Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2015-12-30

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:00 mojca I'm not saying that panda is a perl5 script
00:00 cygx mojca: no, it's about mkpath calling .split on a path object, which tries to open that file
00:00 mojca I'm saying that I see perl5.22 in the list of process names (like I see perl5.12 when I run autoconf)
00:00 cygx (at least that yould be my guess from reading the code)
00:00 cygx *would
00:00 kjs_ joined #perl6
00:01 grondilu joined #perl6
00:01 mojca any suggestions what I should try? I haven't touched perl for about ten years, let alone perl6
00:01 cygx kittenlips: the channel backlog contains ample testimony that installing rakudo isn't always painless
00:01 kittenlips cygx: sorry didn't know that
00:02 TimToady and 'several major architectures' would be more accurate than 'many architectures'
00:02 kittenlips Should I change it to "fairly easy to install"?
00:02 _nadim I'd like to split the code of a class in two or three files. no it des not make sense to have different classes or roles,  can one do that without getting a redeclaration of the class error?
00:03 TimToady we just don't like to overpromise
00:03 kittenlips TimToady++
00:03 TimToady it will be very easy for most people on the same beaten track as we are
00:03 TimToady not at all easy for people who stray from that track
00:04 RabidGravy _nadim, no
00:04 kittenlips TimToady: meaning straying from the rakudobrew track?
00:04 _nadim too bad!
00:04 cygx mojca: have you tried the change to panda I proposed earlier?
00:04 mojca you mean adding '~'?
00:05 cygx kittenlips: or trying to make rakudobrew do something it wasn't designed for
00:05 cygx mojca: yes
00:05 ShimmerFairy _nadim: sounds like an interesting thing to extend Perl 6 with in a module, assuming it's possible, though :)
00:07 _nadim using roles is pretty easy but it force everything to be public
00:07 kittenlips TimToady cygx OK SX and Quora updated
00:08 _nadim which is kinda the point as roles should be just that extra roles, not somehting that deals with internal details
00:08 cygx kittenlips++
00:08 kittenlips changed as you recommended to "several major architectures" and "fairly straightforward to install"
00:08 mojca cygx: it works better, but not properly
00:08 kittenlips​28/what-is-the-current-state-of-perl-6/
00:08 mojca the first time it stops with "Found no writable directory into which panda could be installed"
00:08 mojca but the second time it is successful
00:09 mojca the weird thing is that something already creates /Users/me/.perl6/2015.12 even if panda fails
00:09 mojca so it's just the final "panda" dir that fails to be created
00:10 mojca but the patch isn't complete/fully functional
00:11 mojca I find it weird that perl6 doesn't have a shortcut for "try to make this folder unless one already exists"
00:11 ShimmerFairy _nadim: yeah, the only thing advice I can offer is that maybe you can refactor your design into naturally smaller classes, so having more public classes wouldn't actually be an issue.
00:12 _nadim There's always the preprocessor if this upsets me too much.
00:13 jshy joined #perl6
00:13 mojca if I replace "try mkpath $pandadir unless $pandadir ~~ :d;" with "mkpath ~$pandadir;" then it works the first time
00:13 mojca If I put just "mkpath $pandadir;" then it fails as before
00:14 mojca so probably the second occurence ("unless $pandadir") is also problematic
00:14 Praise joined #perl6
00:14 Praise joined #perl6
00:15 jshy left #perl6
00:15 _nadim ShimmerFairy: I understand the reason why one wants to have the whole class in a single compilation unit, it's simpler than implement dependencies and "link" stuff together before running
00:16 cygx I think plain mkdir already can create paths and does not fail on existing directories
00:16 cygx we should probably change panda to use that instead of the custom mkpath
00:16 _nadim but something one could have some mechanism that is used at the declaration time of the class that says "Hej, there are other bits of me in these files". like inheritance or mixins do.
00:17 bbkr hi. - NativeCall forces me to use appropriate numeric type.  which one should I use if *.h file for this library says simply "int" ?
00:17 cygx bbkr: int32
00:17 bbkr cygx: thanks, trying it right now. \
00:17 Skarsnik Like NC doc say x)
00:18 Skarsnik int32          (int in C)
00:18 shicheng joined #perl6
00:19 ShimmerFairy _nadim: yeah, that's where my suggestion for a module that offered that kind of stuff comes in :)
00:19 Skarsnik I though it was unit xà
00:20 Skarsnik *unit role
00:21 _nadim there's  an --optimize option to perl6, anyone can say a word about this?
00:21 Skarsnik lol
00:21 cygx bbkr: assuming you're not using a Cray architecture - those use ilp64 :p
00:21 Skarsnik Sorry you wanted a word ^^
00:22 _nadim seems to be a no op
00:22 ShimmerFairy _nadim: --optimize=3 is used to compile the CORE setting, so it's at least good enough for that :)
00:22 dalek doc: a7ed560 | skids++ | doc/Type/Slip.pod:
00:22 dalek doc: Flesh out Slip with examples, and fix a typo
00:22 dalek doc: review:
00:23 BenGoldberg joined #perl6
00:24 shichen__ joined #perl6
00:24 kittenlips Non Perl 6 question, but may I ask does anyone know why there is nothing scheduled for Perl at FOSDEM this year?
00:24 kittenlips Yikes
00:25 kittenlips doesn't look good, every other language has tons of speakers
00:27 timotimo oooooh
00:28 timotimo chmod -c shows you what actually changed
00:28 Actualeyes joined #perl6
00:30 dalek doc/cli-file-load: 59bf06a | (Brock Wilcox)++ | bin/p6doc:
00:30 dalek doc/cli-file-load: Also allow p6doc <filename>
00:30 dalek doc/cli-file-load: review:
00:32 mojca what exactly is the state file for panda meant for?
00:32 cygx mojca: you could also try replacing the line in question with `try $pandadir.mkdir`
00:33 mojca I would like to create a package for panda (so that users don't have to bootstrap it manually), so that panda would then put files under $HOME/.perl6
00:33 mojca and I would like to know which files are needed for that
00:33 mojca cygx: doing it now
00:35 _nadim night all!
00:35 shicheng joined #perl6
00:35 mojca cygx: that works
00:36 cygx \o/
00:36 mojca thanks a lot
00:36 kittenlips left #perl6
00:36 xpen joined #perl6
00:37 mojca one more question: can I somehow remove /opt/local/share/perl6/site/panda from the list of folders where panda would try to install things?
00:37 mojca that is the space reserved for the package manager
00:38 mojca putting there anything else without the knowledge of the package manager would be problematic and would lead into problems later on
00:38 gfldex .seen hankache
00:38 yoleaux I saw hankache 29 Dec 2015 21:24Z in #perl6: <hankache> see you tomorrow
00:41 lnrdo joined #perl6
00:41 cdg joined #perl6
00:43 llfourn joined #perl6
00:44 jme` joined #perl6
00:47 cygx mojca: could you check if was the changed that worked for you?
00:48 mojca yes; do you want me to leave a comment?
00:49 cygx sure
00:49 cygx I'm coding blind here, so that would be appreciated
00:52 SwellJoe joined #perl6
00:52 bbkr cygx: can you please take a look at​oIPerl6/builds/99358922#L2236 ? I don't understand error message, NativeCall method stubs are defined correctly (i think) -​erl6/blob/master/lib/ .
00:53 bbkr and I don't have "$type" variable anywhere in the code
00:54 Skarsnik type is in NC code
00:54 bbkr it was working on v6.c but stopped on HEAD
00:54 Skarsnik Yes it's my fault
00:55 cygx missing Mu?
00:55 Skarsnik Hm
00:56 * cygx hasn't actually looked at any code yet
00:57 Skarsnik I know what is wrong
00:58 Skarsnik but I have to go bed
00:58 bbkr is it NativeCall issue or my code?
00:58 Skarsnik it's in NC
00:59 Skarsnik it's because of​blob/nom/lib/NativeCall.pm6#L216 validnctype there is no HOW on the class yet
01:00 bbkr thanks for checking. I also have to go to sleep, please leave me message when you commit the fix so I can restart Travis job.
01:00 bbkr Skarsnik++
01:00 mcmillhj joined #perl6
01:01 Skarsnik Look tricky to fix :(
01:01 Arrowhead joined #perl6
01:02 regreg joined #perl6
01:05 yeahnoob joined #perl6
01:08 mojca cygx: thanks a lot for helping me out :), I added the panda package to MacPort now, even though it will probably need a great deal of polishing
01:10 cygx mojca: glad to be of help
01:10 cygx and with that, good night o/
01:10 mojca good night to you and from me as well
01:18 lustlife joined #perl6
01:24 itcharlie joined #perl6
01:30 Actualeyes joined #perl6
01:30 Ben_Goldberg joined #perl6
01:33 timotimo published the weekly again
01:33 flussence timotimo++
01:36 vendethiel timotimo++
01:37 xpen joined #perl6
01:39 jast first working Perl 6 program written, yay! also, best christmas in years. ;)
01:41 Hotkeys merry perlmas
01:43 awwaiid Y'all should come to DC in April for DCBPW-2016! Also sign up to give talks :)
01:44 llfourn joined #perl6
01:57 CoconutCrab joined #perl6
02:05 BenGoldberg_ joined #perl6
02:06 AlexDaniel joined #perl6
02:06 jshy joined #perl6
02:07 kittenlips joined #perl6
02:09 cdg joined #perl6
02:11 a3r0 Can I use Perl5 modules with Inline::Perl5 in Perl6's Promise/Channels ?
02:13 awwaiid a3r0: ooo interesting question. My _guess_ is ... maybe, depending on what the module is doing. You should try it :)
02:14 TimToady a3r0: if the Perl 5 modules in question are not thread-safe, good luck!  :)
02:15 timotimo that's how you write a random number generator ...
02:15 itcharlie1 joined #perl6
02:16 Ben_Goldberg joined #perl6
02:28 ssotka joined #perl6
02:29 addison joined #perl6
02:29 lnrdo joined #perl6
02:30 shicheng joined #perl6
02:31 autarch something in rakudo changed recently and broke File::LibMagic -​l6-File-LibMagic/builds/99276134 - not sure what's going on
02:32 xpen joined #perl6
02:33 kanishka joined #perl6
02:34 sumdumgoi joined #perl6
02:37 itaipu joined #perl6
02:40 llfourn joined #perl6
02:42 ShimmerFairy m: say "AB" ~~ S/<alpha>/1/    # umm... how am I supposed to use S/// then?
02:45 camelia rakudo-moar 477830: OUTPUT«Potential difficulties:␤    Smartmatch with S/// can never succeed because the subsequent string match will fail␤    at /tmp/qBkXHzPSyy:1␤    ------> 3say "AB" ~~ 7⏏5S/<alpha>/1/    # umm... how am I suppos␤False␤»
02:45 * flussence stares at that for a while to figure out why
02:47 ilbot3 joined #perl6
02:47 Topic for #perl6 is now »ö« Welcome to Perl 6! | | evalbot usage: 'p6: say 3;' or rakudo:,  or /msg camelia p6: ... | irclog: or | UTF-8 is our friend!
02:51 ShimmerFairy m: my $orig = $_ = "AB"; say S/<alpha>/1/;
02:51 camelia rakudo-moar 477830: OUTPUT«1B␤»
02:51 ShimmerFairy since none of the other x/// things _require_ me to use the non-smartmatch form, I cry foul on this :)  (esp. since I don't ever want to use those things without ~~ )
02:55 ssid joined #perl6
02:56 aborazmeh joined #perl6
02:56 aborazmeh joined #perl6
02:56 Su-Shee_ joined #perl6
02:57 flussence I think the main reason for S/// is to make certain patterns of «S[foo] = bar» not-impossible... I can't imagine it's necessarily useful everywhere s/// is
02:57 AlexDaniel what is S/// ?
02:58 cxreg "This is because the placeholder variables are sorted in Unicode order. "  <- I gotta say that is highly non-intuitive
02:59 cxreg regarding $^ twigils
02:59 ShimmerFairy AlexDaniel: non-mutating version of s///
02:59 flussence isn't that just $_.sort(*.NFKD) or something?
02:59 AlexDaniel cxreg: I've always thought that they were in order of appearance
03:00 cxreg AlexDaniel: you'd think?  going by lexical sort is baffling.
03:00 ShimmerFairy flussence: I don't really buy "not necessarily useful like s///" as an argument for not allowing "thing" ~~ S/// :)
03:00 flussence and as far as twigils go, they're all ASCII anyway
03:00 flussence <^ * = ?>.sort(*.ord).say
03:00 flussence m: <^ * = ?>.sort(*.ord).say
03:00 camelia rakudo-moar 477830: OUTPUT«(* = ? ^)␤»
03:00 flussence m: <^ * = ?>.sort(*.ord).map(*.ord).say
03:00 camelia rakudo-moar 477830: OUTPUT«(42 61 63 94)␤»
03:00 ShimmerFairy cxreg: if your placeholder variables are beginning to get unwieldy, or you generally don't want "in lexicographic order", you should stop using placeholders :)
03:01 cxreg flussence: it's saying that $^bar comes before $^foo as a positional parameter because "b" lt "f"
03:01 cxreg ShimmerFairy: even if i could agree with that it's still a terrible decision
03:01 flussence oh, that's clearer with context
03:02 flussence for some reason my brain made up some imaginary thing where all kinds of twigil-variables were in a list...
03:02 ShimmerFairy cxreg: I disagree. It makes supplying a sort lambda like  { $^b cmp $^a }  (reverse of default sort) really nice, for example :)
03:03 AlexDaniel ShimmerFairy: what about $^a Rcmp $^b ?
03:03 cxreg if that's the intended use then the example on should read that way.  it uses "$^second follows $^first" as an exampel
03:03 cxreg which only highlights how awful it is
03:03 cxreg $^a and $^b are possibly intuitive but just about nothing else is
03:04 ShimmerFairy cxreg: as far as I have seen, you pretty much only ever use $^ vars as single-letter variables, where the sorting is a lot less obnoxious than it may appear to be.
03:04 cxreg then the docs should absolutely reflect that
03:04 ShimmerFairy cxreg: at the very least, I've never bothered with _descriptive_ $^ variables :)
03:04 cxreg in fact, it should maybe only *allow* single letter placeholders
03:04 AlexDaniel I've always did descriptive $^ variables
03:04 cxreg but that's possibly going too far?
03:04 AlexDaniel and now it turns out that by pure luck they were in correct order
03:05 cxreg good lord.
03:05 cxreg hate hate hate this "feature".   sorry.
03:05 flussence the example definitely should use $^a and $^b, since that whole part of the language exists to get rid of perl5's magic $a and $b
03:05 cxreg it's the kind of thing that makes the internet rage at perl's "read-only warts"
03:06 AlexDaniel cxreg: I'm with you. Even if there are reasons for it to be this way, this is going to be a pitfall. A pretty common one actually…
03:06 ShimmerFairy it's really meant for cases where you don't want to bother typing -> $a, $b { $b cmp $a }, because otherwise you should really take the time to do the -> ... bit.
03:06 cxreg er, "write-only
03:06 AlexDaniel ShimmerFairy: what's wrong with Rcmp, again?
03:06 ShimmerFairy AlexDaniel: nothing, but as always TIMTOWTDI :)
03:06 AlexDaniel m: say ‘a’ cmp ‘b’
03:06 camelia rakudo-moar 477830: OUTPUT«Less␤»
03:06 AlexDaniel m: say ‘a’ Rcmp ‘b’
03:06 camelia rakudo-moar 477830: OUTPUT«More␤»
03:07 Ben_Goldberg joined #perl6
03:07 cxreg no perl hacker objects to TIMTOWTDI.  but fixing highly counterintuitive syntax is the whole reason perl 6 exists.
03:07 AlexDaniel ShimmerFairy: fair enough. Either do Rcmp or take the time to do the -> bit…
03:08 cxreg TimToady: ping.  would like your input on tihs.
03:08 ShimmerFairy well, like or not the current behavior is part of Perl 6. There's no way it'll change by this point.
03:08 skids ShimmerFairy: I had another couple thoughts on the whole binary grammar/unpack thing here
03:08 flussence .oO( {"name":"Acme::RCMP",provides:["Moose"]} ) </acronympun>
03:08 cxreg LOL
03:08 TimToady it's for simple things; if you want to get fancy, don't use it, that is all
03:08 AlexDaniel ShimmerFairy: not that we have to keep counter-productive behavior forever.
03:10 TimToady "it can be abused" is not an argument here
03:10 AlexDaniel .oO( notes for perl6critic: start complaining when you see non-single-character $^vars )
03:10 cxreg it's not a matter of abuse, it's a question of "the casual user does it wrong because it's not obvious"
03:10 TimToady the casual user will do all sorts of things wrong anyway
03:10 cxreg oh well.  whatever.  it'll be among the language's more disliked parts, but if you don't care then ok.
03:10 ugexe m: say $^1;
03:10 camelia rakudo-moar 477830: OUTPUT«one((Any), 1)␤»
03:11 TimToady I think you're projecting :)
03:11 cxreg talk to me in 2 years
03:11 flussence if anyone can come up with a pack/unpack replacement that lets me do the equivalent of a perl5 "w/a" format, but with arbitrary objects instead of "w", I'll be very happy
03:12 cxreg I'm a fairly decent straw man for the influx of users here
03:12 ShimmerFairy It's admittedly unexpected when you use multi-char $^ vars, but in the simple case it's easy to intuit that $^b cmp $^a is comparing the second arg to the first, in my opinion.
03:12 flussence (who'd'a thought I'd actually *need* the w format some day...)
03:13 cxreg anyway, the docs need fixing.  where's the source code for ?
03:13 flussence :)
03:13 ShimmerFairy skids: seems a bit complicated, but then again my thoughts are more aligned with a separate 'grammar'-like construct than more "diffuse" arrangement of methods on a per-class basis :)
03:13 cxreg flussence: ok thx
03:13 rickbike joined #perl6
03:13 noganex joined #perl6
03:14 ShimmerFairy skids: in particular it feels like 'packing' would try to be too many things at once, where I'd prefer to separate those things (e.g. grammar-like matching vs. (de)serialization vs. ...)
03:16 skids ShimmerFairy: what I was trying to get at was extracting the common parts shared between classes and rules based around binary structures so the "code" could be shared -- that way the "too many things" can just draw from that common base and do their won things.
03:16 skids *own
03:16 Ben_Goldberg joined #perl6
03:17 * flussence thinks we should just do the perl 5 thing here: implement ALL these designs and let the best one win
03:17 ShimmerFairy skids: I see. I guess I just don't have a clear picture of how problematic the existing separation would be (since I can't help but think of pattern matching and serialization as completely different tasks)
03:20 Ben_Goldberg joined #perl6
03:23 yurivish do you folks have recommendations for a perl6 repl?
03:26 flussence `panda install Readline` is as good as it gets right now
03:27 gfldex i use camelia as my repl :)
03:28 hoelzro I would also recommend Linenoise, as it includes tab completions
03:31 yurivish tab completions sound nice. I'll check out out!
03:32 cxreg AlexDaniel: ShimmerFairy: what do you think about this
03:35 AlexDaniel cxreg: it is great
03:35 skids cxreg: That looks to be a good improvement to me, make a PR and I'll push it.
03:36 cxreg just did
03:37 dalek doc: c073e8e | (Dave Olszewski)++ | doc/Language/variables.pod:
03:37 dalek doc: fix perL typo
03:37 dalek doc: review:
03:37 dalek doc: 9bdd74b | (Dave Olszewski)++ | doc/Language/variables.pod:
03:37 dalek doc: improve example of placeholder variables
03:37 dalek doc:
03:37 dalek doc: This better represents the intended usage, such as { $^b cmp $^a }, and
03:37 dalek doc: cautions the reader against names that require mental sorting in order
03:37 dalek doc: to properly write or understand the code.
03:37 dalek doc: review:
03:37 dalek doc: 1b5a682 | skids++ | doc/Language/variables.pod:
03:37 dalek doc: Merge pull request #292 from cxreg/master
03:37 dalek doc:
03:37 dalek doc: docs improvement and placeholder enculturation cxreg++
03:37 dalek doc: review:
03:37 cxreg thanks skids++
03:42 flussence strange day; I've reported 7 bugs against my own code.
03:44 cxreg wrong day to stop sniffing glue?
03:44 AlexDaniel cxreg: thank *you*. It is a good remark :)
03:46 cxreg I've had a number of thoughts about the docs as I read them.  now that i have a fork going i'll probably contribute a bit more
03:51 itcharlie joined #perl6
04:09 ugexe m: say get;
04:09 camelia rakudo-moar 477830: OUTPUT«Céad slán ag sléibhte maorga Chontae Dhún na nGall␤»
04:10 ugexe m: say get("asdf");
04:10 camelia rakudo-moar 477830: OUTPUT«Method 'get' not found for invocant of class 'Str'␤  in block <unit> at /tmp/ZI8AH_faq5 line 1␤␤»
04:11 ugexe shouldn't that be an undeclared routine error?
04:12 kittenlips doesn't perl6 have REPL built-in?
04:13 kittenlips I would recommend Linenoise as well as it not only has tab completion but history
04:13 kaare_ joined #perl6
04:14 kittenlips panda install Linenoise
04:16 grondilu is it still terribly slow though?  Last time I tried it really sucked.
04:17 lnrdo joined #perl6
04:28 SwellJoe joined #perl6
04:28 yurivish Seems fine so far. It augments the built-in repl so things like the arrow keys work again, which was appreciated. The tab completion seems a bit hard to work with – I was hoping it would do something like bash, where it lists the possible completions rather than completing with its best guess
04:42 zostay hmmm... trying to build rakudobrew and i'm getting this: Could not find symbol '&bool' in block at ~/.rakudobrew/moar-nom/install/share/perl6/sou​rces/7822BB220E20131A2CA9DDCBC27AB2A829F8C9CC line 13
04:43 skids zostay: there was some bool related problem a week ago which could only be fixed by nuking previous installs.
04:43 zostay i have been away from perl6 for a couple weeks or so... i'll try that, thx
04:48 zostay yep, that worked, thx
04:48 skids glad to help
04:55 Sgeo joined #perl6
04:56 cdg joined #perl6
04:56 cdg joined #perl6
04:59 sumdumgoi m: die "foo";
04:59 camelia rakudo-moar 477830: OUTPUT«foo␤  in block <unit> at /tmp/tHV2yeXhjK line 1␤␤»
05:00 sumdumgoi m: my %h = (a=>1); for %h.values -> $v { $v = 2 }; # should choke
05:00 camelia rakudo-moar 477830: OUTPUT«Cannot assign to a readonly variable or a value␤  in block <unit> at /tmp/MyzuhxYiJ3 line 1␤␤»
05:01 sumdumgoi Interesting. On my local install I wasn't getting a unit or line number. :P
05:04 sumdumgoi (or filename)
05:16 cdg joined #perl6
05:21 molaf joined #perl6
05:25 Cabanossi joined #perl6
05:31 gfldex m: dd $[["", 'abc', ""],].List.flat, $[["", 'abc', ""],].Array.flat, $[["", 'abc', ""],].Seq.flat;
05:31 camelia rakudo-moar 477830: OUTPUT«("", "abc", "").Seq␤($["", "abc", ""],).Seq␤($["", "abc", ""],).Seq␤»
05:31 gfldex why is there a difference between .List.flat and .Array.flat?
05:33 AlexDaniel gfldex: could it be associated with this? “Arrays itemize their elements. Itemized are never flattened, even when explicitly flattening the array.”
05:33 AlexDaniel gfldex: from​Bug/Display.html?id=127036
05:34 gfldex looks like it
05:36 AlexDaniel well, such behavior is less than obvious…
05:37 gfldex and it's causing me troubles because I can't tell rakudos pod parser that i would prefer List over Array
05:38 gfldex i don't really understand why it returns arrays in the first place. I doubt anyone wants to change the Pod:: tree.
05:38 gfldex may have been made pre-GLR
05:46 Arrowhead joined #perl6
05:46 flussence lazyweb request (or message for Zoffix++): has bizarrely ginormous margins between (768..991)px, and it's making the top menu wrap to two rows, over the top of the content.
05:47 flussence (also there's a small whitespace gap to the right of the top h3.panel-title; my guess is it's relying on the width of the borders the others have)
05:47 * flussence &
05:50 krshn joined #perl6
06:06 lnrdo joined #perl6
06:24 uvtc joined #perl6
06:31 AlexDaniel ShimmerFairy: this music is making your post less negative :)  … La la de da la de da la de da…
06:36 quester joined #perl6
06:40 Arrowhead joined #perl6
06:45 uvtc Just noticed, [this page]( still lists S08 as "Captures and Parcels", though my understanding is that Parcels are no more.
06:46 AlexDaniel uvtc: right
06:47 AlexDaniel uvtc: here is where you can change it:​s/blob/master/html/index.html
06:49 AlexDaniel uvtc: also, it looks like it is not marked as “DRAFT” anymore
06:49 dalek specs: 6c54ce5 | (John Gabriele)++ | html/index.html:
06:49 dalek specs: fix title of S08
06:49 dalek specs: review:
06:50 uvtc Thanks, AlexDaniel .
06:50 adhoc joined #perl6
06:54 virtualsue joined #perl6
07:02 rafl joined #perl6
07:02 CIAvash joined #perl6
07:15 ccakes joined #perl6
07:20 Averna joined #perl6
07:26 dalek doc: ec41c3e | (John Gabriele)++ | doc/Language/containers.pod:
07:26 dalek doc: fix typo: missing comma
07:26 dalek doc: review:
07:32 protium joined #perl6
07:36 uvtc left #perl6
07:38 kjs_ joined #perl6
07:40 itcharlie joined #perl6
07:45 * sumdumgoi figured out how to use action objects :D
07:46 pierre-vigier joined #perl6
07:46 dalek doc: a07a779 | skids++ | doc/Type/List.pod:
07:46 dalek doc: Document sub form of "flat"
07:46 dalek doc: review:
07:50 domidumont joined #perl6
07:54 lnrdo joined #perl6
07:54 uruwi_ joined #perl6
07:54 nine .tell kittenlips does what you described
07:54 yoleaux 29 Dec 2015 23:38Z <ShimmerFairy> nine: if you did use git merge, I don't see it. I don't see any merge commits for the curi branch in gitk, just the signs of a rebase. The only thing I can think is if you used 'git merge' and it decided on a fast-forward merge, in which case my blame should be redirected towards git :)
07:54 yoleaux nine: I'll pass your message to kittenlips.
07:56 domidumont joined #perl6
08:01 rjbs joined #perl6
08:05 nanlandetian joined #perl6
08:12 darutoko joined #perl6
08:12 sumdumgoi hehe... I have an adventure game command parser now :D
08:17 itcharlie1 joined #perl6
08:23 blub` joined #perl6
08:24 pierre-vigier joined #perl6
08:28 dalek doc: e25115a | vinc17fr++ | doc/Language/faq.pod:
08:28 dalek doc: Clarified item on numeric computation.
08:28 dalek doc:
08:28 dalek doc: Reason: rational numbers are not floating point and most math functions yield a Num.
08:28 dalek doc: review:
08:28 dalek doc: 32d1be1 | niner++ | doc/Language/faq.pod:
08:28 dalek doc: Merge pull request #293 from vinc17fr/master
08:28 dalek doc:
08:28 dalek doc: Clarified item on numeric computation.
08:28 dalek doc: review:
08:29 Arrowhead joined #perl6
08:30 nexysno_ joined #perl6
08:31 pierre-vigier joined #perl6
08:34 RabidGravy joined #perl6
08:38 cooper_ joined #perl6
08:39 rindolf joined #perl6
08:41 hankache joined #perl6
08:41 hankache good morning everyone
08:42 RabidGravy marnin
08:47 dalek doc: 59bf06a | (Brock Wilcox)++ | bin/p6doc:
08:47 dalek doc: Also allow p6doc <filename>
08:47 dalek doc: review:
08:47 dalek doc: 515499d | RabidGravy++ | bin/p6doc:
08:47 dalek doc: Merge pull request #290 from perl6/cli-file-load
08:47 dalek doc:
08:47 dalek doc: Also allow p6doc <filename>
08:47 dalek doc: review:
08:49 _nadim joined #perl6
08:50 hankache how can i convert an instant to a datetime?
08:51 hankache now returns an instance, i want to be able to view it as datetime
08:51 hankache instant**
08:51 hankache m: say now
08:51 camelia rakudo-moar 477830: OUTPUT«Instant:1451465549.137139␤»
08:52 RabidGravy m:
08:52 camelia rakudo-moar 477830: ( no output )
08:52 RabidGravy m: say
08:52 camelia rakudo-moar 477830: OUTPUT«2015-12-30T08:52:15.732200Z␤»
08:52 hankache m: say now.DateTime
08:52 camelia rakudo-moar 477830: OUTPUT«2015-12-30T08:52:59.872715Z␤»
08:53 hankache thanks RabidGravy
08:54 RabidGravy though if that's all you want to do with it then
08:54 RabidGravy m:
08:54 camelia rakudo-moar 477830: ( no output )
08:54 RabidGravy m: say
08:54 camelia rakudo-moar 477830: OUTPUT«2015-12-30T09:54:22.943808+01:00␤»
08:54 hankache m: say now.DateTime
08:54 camelia rakudo-moar 477830: OUTPUT«2015-12-30T08:54:56.001425Z␤»
08:55 hankache m: say
08:55 camelia rakudo-moar 477830: OUTPUT«2015-12-30T09:55:03.353964+01:00␤»
08:55 hankache huh! why is there a difference ?
08:55 RabidGravy instants don't have a timezone
08:56 hankache so they're GMT+0 ?
08:57 RabidGravy yes
08:57 hankache oki
08:57 * hankache afk for 10 min
08:58 [Tux] Inline::Perl5:
08:58 RabidGravy well no actually really  the DateTime becomes GMT+0 but the Instant doesn't have a timezone
08:58 [Tux] The returning type of 'p5_is_array' --> int is errornous. You should not return a non NativeCall supported type (like Int inplace of int32), truncating errors can appear with different architectures
08:58 pmurias joined #perl6
08:59 masak good morning, #perl6
08:59 RabidGravy yes it went in yesterday
08:59 RabidGravy marning
09:00 tipdbmp joined #perl6
09:00 [Tux] test             50000    23.972    23.856
09:00 [Tux] test-t           50000    13.217    13.101
09:00 [Tux] csv-parser       50000    52.819    52.703
09:00 nine [Tux]: that's fixed in a newer Inline::Perl5
09:02 pmurias masak: hi
09:02 RabidGravy [Tux] is test-t the "one to watch"?  It seems to be making steady progress downward
09:03 pmurias if I want to debug what looks like it might be a moarvm gc issue are there any tricks I should now about?
09:06 [Tux] RabidGravy, yes:
09:06 masak pmurias: \o
09:07 masak nowadays, a (statement form) while loop always yields Nil, right?
09:07 RabidGravy wow what happened at the beginning of November
09:09 itcharlie joined #perl6
09:09 pinakirs joined #perl6
09:09 pinakirs left #perl6
09:10 RabidGravy anyone know if it is possible to persuade gcc to ignore an "inline" declaration?
09:11 RabidGravy I'm thinking not and I'm going to have to make some stubs to realise them
09:14 [Tux] RabidGravy, early november was lizmat's «my @foo = split @needles, $foo;»
09:14 RabidGravy ah cool
09:14 pmurias joined #perl6
09:15 firstdayonthejob joined #perl6
09:16 dr4ngd joined #perl6
09:16 dr4ngd Anyone need a new project to work on?
09:16 pmurias re issue I'm debugging when emitting (fairly large) javascript code, sometimes parts of strings are disappearing
09:16 mattp__ joined #perl6
09:17 masak dr4ngd: are you offering us some work? ;)
09:17 dr4ngd masak: yeah its a big project with a great revenue projection
09:18 itcharlie joined #perl6
09:19 dr4ngd Fluent in pretty much any language and framework is appreciated
09:19 lnrdo joined #perl6
09:23 RabidGravy dr4ngd, Perl 6 then?
09:25 molaf joined #perl6
09:26 hankache m: sleep 5; now - INIT now;
09:26 camelia rakudo-moar 477830: OUTPUT«WARNINGS for /tmp/X4Zfx4F3bW:␤Useless use of "-" in expression "now - INIT now" in sink context (line 1)␤»
09:26 pRiVi joined #perl6
09:26 hankache m: sleep 5; say now - INIT now;
09:26 camelia rakudo-moar 477830: OUTPUT«5.0019910␤»
09:26 hankache is it normal that it =~= 5 and not == 5 ?
09:26 hankache ^^
09:29 masak dr4ngd: ah, so essentially you are here to hire Perl 6 developers?
09:29 masak hankache: yes, I would say so.
09:29 yurivish hankache: yes, it's normal. sleep calls are never completely, precisely exact – but that one's very close!
09:30 masak hankache: if you want exactness guarantees, you should use a library of some kind.
09:31 hankache masak yurivish i am fine with the current level of preciseness, i just want to able to explain where the difference is coming from.
09:31 nine hankache: sleep in any language or system always only guarantees that processing is stopped for at least the given time, never that it resumes exactly on time out.
09:32 yurivish hankache: here's one page that talks about it. there are many more if you search around.
09:33 yurivish it's not perl-specific, but most languages rely on the operating system for stuff like suspending themselves
09:33 hankache masak yurivish nine thank you :)
09:40 _nadim Good morning
09:41 dotness joined #perl6
09:42 andrewalker joined #perl6
09:44 masak hankache: I can think of two things: (a) the program is doing other stuff between the INIT time and the `sleep`; (b) `sleep` is not exact and its only criterion is that it waits *at least* the specified amount of time
09:45 lizmat good *, #perl6!
09:45 RabidGravy marnin
09:45 lizmat does anybody know when pages are regenerated ?
09:46 lizmat did an addition to the faq yesterday, and it's still not visible  :-(
09:46 masak good lizmorning!
09:47 lizmat a project for which I didn't have time to finish before 2015.12, was fixing up ArgFiles handling
09:48 lizmat I envisioned generalizing the $*ARGFILES behaviour into a IO::CatPath and IO::CatHandle class
09:48 masak what does "Cat" connote here?
09:48 lizmat so that <foo bar baz> would give you a handle that would consider the files "foo" "bar" "baz" as a single file
09:49 masak is it the same kind of "Cat" as the Stringy type?
09:49 lizmat masak: almost
09:49 lizmat so that you could run things like split, comb, words and lines on them
09:50 lizmat I basically finished that yesterday, but now wonder how to go on with this
09:50 masak lizmat: sounds like the Composite design pattern
09:50 lnrdo joined #perl6
09:50 lizmat well, we can bikeshed the naming...  :-)
09:50 masak not my intent :)
09:50 masak just pointing to possibly interesting prior art
09:50 krunen joined #perl6
09:51 lizmat now, my problem is really, how do I handle this at this point
09:51 RabidGravy lizmat, I think there is a log somewhere of the doc builds, but I think the travis now runs the whole build process so you can see if it is currently failing for some reason
09:51 lizmat if I add it to the core now and make $*ARGFILES use it, it would be adding features to what is essentially 6.c
09:51 _mg_ joined #perl6
09:52 _mg_ joined #perl6
09:52 lizmat which would mean that "use v6.c" could not tell you if you can use <a b c> in that way
09:52 lizmat so we would need to move to 6.d now?
09:53 lizmat one way of handling this could be to have "use v6.d" set a local $*PERL to "6.d", which could then be inspected by the IO::Cat... code
09:53 lizmat and fail if it was 6.c instead of 6.d
09:54 ShimmerFairy you could develop it in a separate branch for now, if there's not a clear answer to how to go about it (I'm not quite sure myself) :)
09:55 lizmat my experiences with developing stuff in a branch are not good  :-(
09:56 Arrowhead joined #perl6
10:00 masak lizmat: I've recently gotten really interested in developing under feature flags rather than in branches. keeping a unified codebase in master. but we don't have anything like that set up for rakudo right now.
10:02 [ptc] masak: what do you mean by feature flags?  I'm intrigued
10:03 masak [ptc]: in its simplest form, `if %build-settings<feature-X-enabled> { ... }`
10:04 masak [ptc]: but there's a lot of interesting patterns around that, too. like running the tests with only your own feature switched on, or with all features switched on.
10:04 masak [ptc]: once something goes into production, you can just strip out all the `if` statements for that feature flag
10:04 quester left #perl6
10:05 ShimmerFairy At first glance, that A) seems like something you want to do with macros, ideally, and B) feels a little... hackish? I'm not sure about a P6-ish version of #ifdef's all over the place :P
10:05 masak it's not #ifdefs
10:06 masak it's the same code base building the same everywhere. the same thing gets generated for everyone.
10:06 masak it's only the settings that differ. people on a stable release run with experimental features off. developers run with them on.
10:07 masak ShimmerFairy: I think you're right that macros might make things slightly more readable. but the basic idea doesn't require macros.
10:07 masak I want to emphasize that I haven't tried feature flags myself. but the Ember.js project seems to use them to great effect.
10:08 ShimmerFairy masak: well, it feels like the same sort of thing codebases do with #ifdefs, at any rate. And somehow I feel the macro-less solution of if %hash<...> { ... } would look really ugly really fast.
10:09 RabidGravy to lizmat's earlier suggestion, maybe a "use feature :6d" rather than just use v6.d ?
10:09 [ptc] it does sound a lot like #ifdefs though...  Wouldn't that cause a lot of code churn at release time when all of the ifs are removed and the code is reformatted around that?
10:09 [ptc] I'm not trying to be negative, just trying to understand how it would be used
10:09 masak see​aster/packages/ember-metal/lib/features.js
10:09 [ptc] ok, ta
10:10 rickbike Perl 5 has use feature
10:10 ShimmerFairy It's an obvious choice for language versions though, since it'd be the same as -std in g++ . I'm just not sure about using it to bypass our version control's branches :)
10:10 masak [ptc]: I don't know about churn. I imagine it wouldn't be all that bad.
10:10 masak [ptc]: but I'm wondering if some things are "too cross-cutting" to neatly fit under an `if`, or even several.
10:11 icelesstea joined #perl6
10:12 RabidGravy or you know, just create a 6d branch and treat nom as "maint"
10:12 [ptc] masak: true.  It could be hard to isolate the new changes cleanly
10:12 * [ptc] would love to see "nom" renamed to "master"...
10:14 nine lizmat: I think jnthn++ posted his suggestions for handling language versions in a gist or something a couple of weeks back. Do you know where to find it?
10:15 masak nine:
10:15 nine masak: thanks!
10:15 iH2O joined #perl6
10:18 _nadim is there a Unicode geek here that knows why nobody thought about making a complete ASCCI superscript and subscript. There a freaking Xmas tree but not superscript lowercase q. I really wonder what they were high on.
10:18 lizmat ok, from that I get I should just commit to nom and have the release manager worry about including it in a release ?
10:19 lizmat that doesn't feel very sociable...
10:19 pmurias joined #perl6
10:20 nine lizmat: this looks interesting: "Additions to the setting will be made in a "nested" setting. That is, a "use v6.d" will load CORE.d.setting or so. It can use augment and supercede in order to effect changes."
10:20 RabidGravy jnthn's plan seems sound
10:22 ShimmerFairy yes, I like jnthn++'s plan as well :)
10:23 [ptc] agreed, jnthn++'s plan looks well thought out
10:24 espadrine joined #perl6
10:25 [ptc] lizmat: the way I read it is that whatever goes into nom gets automatically moved to master, from which releases are cut.  I guess the social aspect is the release manager asking at which point the release should be cut...
10:25 andreoss joined #perl6
10:26 masak ok, so "nom" means "dev"
10:26 lizmat I still don't understand when people would need to do "6.d" to get a new feature such as "IO::CatPath"
10:26 pierre-vigier joined #perl6
10:27 nine lizmat: I guess when adding the feature to 6.c would break some test in the 6.c test suite.
10:27 lizmat ah, but that's just it: it wouldn't break any tests in 6.c
10:27 RabidGravy if you're building from "nom" then you're not getting what might strictly called 6.c
10:27 lizmat RabidGravy: exactly
10:27 nine Or maybe even "would it break one of the chosen modules in the 6.c snapshots"
10:28 RabidGravy but "production" isn't going to do that right? (chortles into his imaginary beer)
10:28 [ptc] but then if you're building from master you're also not strictly getting 6.c ...
10:28 iH2O left #perl6
10:28 stmuk In 2016, Diwali is on October 30.
10:28 RabidGravy but there is a 6.c release
10:29 lizmat all releases until 6.d will be 6.c
10:29 lizmat but if there are new features added in nom / master, then "use 6.c" becomes ambiguous
10:29 [ptc] ah, ok
10:29 lizmat *v6.c
10:30 lizmat as it will not tell you whether you can use <a b c>.IO
10:31 hankache m: my @array = (1..500_000);my @result ={-> $x { $x + 1}});say now - INIT now;
10:31 ShimmerFairy lizmat: the ambiguity might be covered in part by 6.d.PREVIEW . And as stated I feel like it wouldn't be a problem to have a g++ -std style usage of 'use v6.c', 'use v6.d', etc.
10:31 camelia rakudo-moar 477830: OUTPUT«6.8396321␤»
10:31 hankache m: my @array = (1..500_000);my @result ={-> $x { $x + 1}});say now - INIT now;
10:31 nine OpenGL seems to have done quite well with extensions that get bundled into versions from time to time. Really the same as Perl 5 does nowadays.
10:31 camelia rakudo-moar 477830: OUTPUT«(signal XCPU)»
10:32 hankache huh?
10:32 nine hankache: I guess there's an ulimit for CPU time in camelia's VM
10:33 hankache nine ok. i wanted to showcase that in this example, race takes longer than the the normal expression without it
10:33 hankache am i missing something?
10:34 hankache m: my @array = (1..100_000);my @result ={-> $x { $x + 1}});say now - INIT now;
10:34 camelia rakudo-moar 477830: OUTPUT«0.75741154␤»
10:34 hankache m: my @array = (1..100_000);my @result ={-> $x { $x + 1}});say now - INIT now;
10:34 camelia rakudo-moar 477830: OUTPUT«2.06172629␤»
10:34 nine hankache: that doesn't surprise me. You're replacing a very tight and optimizable look with lots of threading overhead
10:34 hankache see ^^
10:35 nine The threading overhead quickly becomes negligible once you do real calculations in your loop, but for minimalistic examples like that, all you're measuring is threading and looping overhead.
10:36 hankache nine thanks dear
10:37 nine :)
10:37 itaipu joined #perl6
10:39 [Tux] Is this because Inline::Perl5 is already installed? (in which case panda could be a bit more friendly)
10:39 [Tux]
10:40 Arrowhead How do I make CArrays of float more usable? I can't assign anything but an actual Num and I can't make it Cool it seems:
10:40 Arrowhead m: use NativeCall; my @floats := CArray[num32].new(); @floats[0] = 1e0; @floats[1] = 1;
10:40 camelia rakudo-moar 477830: OUTPUT«This type cannot unbox to a native number␤  in method <anon> at /home/camelia/rakudo-m-inst-2/share/perl6/sou​rces/1F2C39A739CAA5158F1049489481BA8E6077C82B line 108␤  in block <unit> at /tmp/FWh6yjFBsX line 1␤␤»
10:40 Arrowhead CArray[num32(Cool)] isn't allowed.
10:42 Arrowhead I guess I need to create my own wrapper type?
10:44 [Tux] try «$ panda ?» :/
10:44 Averna joined #perl6
10:45 nine [Tux]: no, that's because your panda seems to be out of date
10:46 nine [Tux]: since Dec 21st
10:50 lnrdo joined #perl6
10:50 andreoss .seen masak
10:50 yoleaux I saw masak 10:26Z in #perl6: <masak> ok, so "nom" means "dev"
10:51 * [Tux] starts from scratch …
10:51 rickbike joined #perl6
10:54 lnrdo joined #perl6
10:58 hankache joined #perl6
11:00 HoloIRCUser3 joined #perl6
11:01 Floyd-ATC joined #perl6
11:04 RabidGravy Arrowhead, see a bunch of my Audio::* modules for examples of dealing with typed CArrays, but mostly it's just coerce the value
11:05 Arrowhead RabidGravy: I'll take a look. I'm even having a hard time taking any list of numbers and transforming them to Nums.
11:06 RabidGravy I tend to do something like:
11:06 Begi joined #perl6
11:06 [Tux]​rticles/Perl/20151229.html :(
11:07 [Tux] so hard to convince even perl(5) hard-liners that perl6 *is* real
11:07 virtualsue joined #perl6
11:08 RabidGravy m: use NativeCall;  my CArray[num32] $foo = CArray[num32].new;  my @a = (^10);   $foo[$_] = Num(@a[$_]) for ^@a.elems;  say $foo'
11:08 camelia rakudo-moar 477830: OUTPUT«5===SORRY!5=== Error while compiling /tmp/BMFNnVWCYZ␤Two terms in a row␤at /tmp/BMFNnVWCYZ:1␤------> 3] = Num(@a[$_]) for ^@a.elems;  say $foo7⏏5'␤    expecting any of:␤        infix␤        infix stopper␤        postfix␤        s…»
11:08 RabidGravy m: use NativeCall;  my CArray[num32] $foo = CArray[num32].new;  my @a = (^10);   $foo[$_] = Num(@a[$_]) for ^@a.elems;  say $foo
11:08 camelia rakudo-moar 477830: OUTPUT«NativeCall::Types::CArray[num32].new␤»
11:08 Floyd-ATC Can someone show me a working example of a DateTime formatter?
11:09 V_S_C joined #perl6
11:10 llfourn [Tux]: I personally don't think it's a good idea to even try. TBH I still use perl5 and there isn't much overlap between the projects that would be good to do in p5 vs something I would start in p6.
11:10 RabidGravy Arrowhead, important to remember that because a CArray is not much of a Perl array you can only index it like that
11:10 Begi2 joined #perl6
11:11 lizmat m: { .hour ~ ":" ~ .minute } ).say
11:11 camelia rakudo-moar 477830: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tc8ID2YBUd␤Unable to parse expression in argument list; couldn't find final ')' ␤at /tmp/tc8ID2YBUd:1␤------> 3rmatter( { .hour ~ ":" ~ .minute } ).say7⏏5<EOL>␤»
11:11 lizmat m: { .hour ~ ":" ~ .minute }) ).say
11:11 camelia rakudo-moar 477830: OUTPUT«12:11␤»
11:11 RabidGravy Arrowhead, also see NativeHelpers::Array which encapsulates a few of those kinds of hacks
11:11 lizmat m: { .hour ~ ":" ~ .minute }) ).say   # Floyd-ATC
11:11 camelia rakudo-moar 477830: OUTPUT«12:11␤»
11:11 [Tux] Fresh timings, started from scratch …
11:11 [Tux] csv-ip5xs        50000    18.059    17.948
11:11 [Tux] test             50000    23.999    23.888
11:11 [Tux] test-t           50000    12.564    12.453
11:11 [Tux] csv-parser       50000    50.459    50.348
11:12 Arrowhead RabidGravy: ooh, NativeHelpers. that sounds intersting.
11:12 RabidGravy I got bored of typing that stuff ;-)
11:12 [Tux] llfourn, Johan is not someone to convert from perl5 to perl5. Being the author of the perl5 QRG he knows what we are talking about
11:13 [Tux] he is just very disappointed in the perl6 "release" not being released as the promised 6.0.0
11:13 Floyd-ATC Thanks lizmat... an example like that really needs to be in the documentation because I've spent the last 30 minutes trying to wrap my brain around it in vain
11:14 llfourn [Tux]: by that he meant version numbers or functionality?
11:14 RabidGravy I think there needs to be at least two new pages in the language docs, one on lists and arrays, and one on dates and times
11:14 [Tux] version number, marketing and announcement policy
11:14 llfourn [Tux]: ah I see
11:15 V_S_C Using ActiveState PERL I got rakudobrew to get MoaR & panda, then during Task::Star installation Linenoise failed with linker error
11:15 RabidGravy :-\
11:15 vendethiel joined #perl6
11:16 Floyd-ATC One other thing that confuses the hell out of me, why does my script fail to compile if I try to "use" a module that happens to be in core?
11:16 RabidGravy because they aren't modules
11:16 RabidGravy in the sense they are loadable
11:16 RabidGravy or rather aren't loadable
11:17 [Tux] afk
11:17 RabidGravy they're just types which are compiled into the setting
11:17 _nadim Can someone help with this? I'd like to capture the kv of the pair. un this code, I get a pair in $A and one in $s; if I put them in parenthesis then I get "too few positional ... exp 2 got 1"
11:17 V_S_C I made github pull request removing the extra space that seems to be originating from
11:17 Floyd-ATC So.. DateTime is not a module but DateTime::Format is... seems a bit arbitrary
11:17 Skarsnik joined #perl6
11:18 RabidGravy Floyd-ATC, not really
11:18 llfourn Floyd-ATC: gotta draw the line somewhere
11:18 V_S_C with that I guess rakudo star Christmas release works on Windows
11:18 RabidGravy yay!
11:19 _nadim Argg, capture expects named argument when it sees a pair!
11:20 Floyd-ATC Same with IO::Socket::INET -> looks like a module and has been one for years but nope, crash and burn
11:21 RabidGravy er, it's always been a core type in Perl 6 afaik
11:21 Floyd-ATC in p6 probably, yes
11:21 RabidGravy well
11:21 Floyd-ATC but coming from p5 that's confusing as hell
11:22 nine [Tux]: no one has ever promised a "6.0.0". Some people seem to have assumed, that the release would be called so.
11:22 lizmat maybe we just make a bare use for a module in core die with a specific message
11:22 lizmat m: use Int
11:22 camelia rakudo-moar 477830: OUTPUT«===SORRY!===␤Could not find Int in:␤    /home/camelia/.perl6/2015.12-44-g477830a␤    /home/camelia/rakudo-m-inst-2/share/perl6/site␤    /home/camelia/rakudo-m-inst-2/share/perl6/vendor␤    /home/camelia/rakudo-m-inst-2/share/perl6␤    CompUnit:…»
11:23 ShimmerFairy I don't think there are any 'module's in CORE, though :P
11:23 RabidGravy a pretty good rule of thumb here is that if it's in "Types" on then it's not a module
11:23 RabidGravy ShimmerFairy, with the exception of NativeCall and Test there aren't
11:23 ShimmerFairy RabidGravy: that's not part of CORE, though :)
11:23 RabidGravy right
11:24 RabidGravy so there aren't
11:25 [Tux] nine, /me doesn't need to be convinced :)
11:25 RabidGravy The confusion seems to be that there are types defined in the core setting that have a similar name to Perl 5 modules
11:25 [Tux] but this shows (many) others in the world do (somehow)
11:25 cpage joined #perl6
11:25 nine lizmat: how could we get at a list of Settings classes?
11:25 Floyd-ATC Well, that and the fact that they have the same purpose and many of the same methods
11:26 Floyd-ATC although that particular mo^H^Htype lacks several important methods like peer address etc.
11:27 RabidGravy Floyd-ATC, I'm sure your patches ti implement them will gratefully received
11:27 ShimmerFairy I personally don't see that it's worth the effort to say "no, we don't have a module named that in Perl 6, this really isn't Perl 5".
11:27 _mg_ joined #perl6
11:27 RabidGravy me too
11:27 Floyd-ATC Yeah I'm sure you're eager to pull my patches, I can't even understand the current docs :-P
11:28 RabidGravy documentation improvements are also welcomed :)
11:29 Floyd-ATC yup. Step 0 is to understand wtf is going on though
11:31 lizmat nine: there lives one in roast, in WHICH.t
11:31 RabidGravy on a completely other note, why is that, given all the dependencies being successfully installed, tests on travis fail with "Could not find Foo" where they don't locally
11:33 * RabidGravy removes any last vestige of "use lib" from the test files and tries again
11:34 nine nine@sunshine:~/install/rakudo (nom *=)> perl6 -e 'use Int'
11:34 nine ===SORRY!===
11:34 nine Int is a builtin type. You can use it without loading a module
11:34 nine Could not find Int in:
11:34 SwellJoe joined #perl6
11:34 RabidGravy nine++
11:35 Floyd-ATC As an exercise in figuring stuff out I tried to implement a simple HTTP server but I have trouble switching between .read() and .get(), presumably for reasons that have to do with buffering. No idea how to work around it except discard .get() completely and scanning lines manually in my own buffer. Am I on the completely wrong track?
11:35 nine As this is a patch to X::CompUnit::UnsatisfiedDependency, it still allows for modules to have the same name as builtin types.
11:36 jeteve joined #perl6
11:36 jeteve Hi, and Happy C!
11:37 jeteve Got a noob question
11:37 RabidGravy Floyd-ATC, you'll find that most HTTP server like things use .read exclusively and decode bits as required
11:37 Floyd-ATC Right.
11:38 pierre-vigier joined #perl6
11:39 RabidGravy (and indeed clients for the most part,)  the major reason being is that the encoding may be different between the headers and the body of a request
11:39 lizmat nine: I would drop the "could not find Int in" bit
11:39 lizmat nine: in that case
11:40 nige1 joined #perl6
11:40 RabidGravy jeteve, we're all noobs here, I'm actually a goat farmer in real life
11:40 nine lizmat: but if the user really tried to load a module named, the "Could not find" part would be the more important one.
11:40 RabidGravy lizmat, nine, except if they were actually expecting to really load Int from a module
11:41 RabidGravy what nine said
11:41 lizmat nine: well, having a bare module load (without auth / version) with the same name as a built in class, would be a bad idea anyway, no ?
11:41 jeteve Nice, I like goats
11:41 lizmat I mean, only cut off if there is no auth / version specified
11:43 jeteve Here's my question then: In perl5 w/Moose, I'm quite a heavy user of lazy building attributes. And I'm wundering what's the similar approach in Perl6
11:43 * llfourn would like to know this too
11:43 jeteve I've tried 'has $.bla = ->{ do and return stuff }();'
11:44 jeteve but that seems to be evaluated ASAP
11:45 llfourn jeteve: I don't know but I seem to remember that it "should be in a module" as some conclusion
11:45 nine lizmat: it would. Yet I'm still nervous about removing module loading related features :/ The past few weeks have tought me that.
11:45 llfourn jeteve: the metamodel is quite flexible so the idea is that someone is going to implement lazy attrs in a module
11:45 lizmat I'm just worried that the extra text you added will be drowned out visually by the list of dirs in which it was not found
11:46 lizmat nine ^^
11:46 jeteve llfourn: cool
11:46 llfourn jeteve:
11:46 llfourn yes :P
11:46 jeteve I shall go and read more to see if I can write such a thing :)
11:46 llfourn jeteve: yeah you wanna look at "traits" and EXPORTHOW (which is not in docs yet)
11:48 dalek doc: f5c5ce3 | lizmat++ | doc/Type/DateTime.pod:
11:48 dalek doc: Add example of formatter, Floyd-ATC++
11:48 dalek doc: review:
11:48 jeteve nice, thanks a lot for the pointers :)
11:48 llfourn jeteve: there could be two approaches one where they only work if you use $.attr rather than $!attr (you mess with ClassHOW)
11:48 llfourn jeteve: the other you mess with Attribute itself, in particular Attribute.get_value
11:49 jeteve #2 would probably be cleaner
11:50 llfourn jeteve:​Attribute#method_get_value (you will have to override it with a trait, is trait is rw below that for example)
11:50 jeteve cause it keeps the lazyness orthogonal to the other concepts
11:51 _nadim DrForr: Readline broken with the latest Rakudo
11:51 llfourn jeteve: yes I agree try creating a trait "is lazy" and see what you can do is my advice
11:51 _nadim .tell DrForr Readline broken with the latest Rakudo
11:51 yoleaux _nadim: I'll pass your message to DrForr.
11:51 jeteve excellent :)
11:51 jeteve Thanks!
11:51 llfourn no worries let us know how you go :D
11:52 Psyche^_ joined #perl6
11:53 dalek doc: b253dbf | lizmat++ | doc/Type/DateTime.pod:
11:53 dalek doc: Better example using sprintf
11:53 dalek doc: review:
11:54 dalek rakudo/nom: 9cecde8 | (Stefan Seifert)++ | src/core/
11:54 dalek rakudo/nom: De-confuse people trying to load a module for a CORE type
11:54 dalek rakudo/nom:
11:54 dalek rakudo/nom: Tell people who try to load a module when a similarly named internal
11:54 dalek rakudo/nom: type is available. E.g. give a helpful message when someone tries to:
11:54 dalek rakudo/nom: use IO::Socket::INET;
11:54 dalek rakudo/nom: review:
11:55 nine lizmat: you convinced me ^^^
11:55 RabidGravy nine++
11:56 Arrowhead Why can I "perl6 -Ilib" but do I have to do PERL6LIB="lib" perl6 for a repl with lib available?
11:57 stmuk_ I just ran "make html" in the doc repo successdully on my own system with 2015-12
11:57 stmuk_​build-2015-12-30T11:55+0000.log
11:58 stmuk_ that suggests to me that hack is running an old rakudo for the doc processes
11:58 hankache normally race would be faster than hyper?
11:59 hankache since it doesn't preserve the order so less overhead?
11:59 stmuk_ maybe the "dog food" processes should use 2015-12
12:00 stmuk_ m: use experimental :pack;
12:00 camelia rakudo-moar 477830: ( no output )
12:01 RabidGravy I thought the "cannot unbox type object" in panda installdeps was fix0rificated
12:01 RabidGravy or did I dream that?
12:02 lizmat hankache: only if there is something to race with, otherwise you're just creating overhead
12:02 lizmat try adding a 'sleep rand' in your code to be raced, and see the result then
12:03 lizmat CPU will be more, wallclock should be less
12:04 * RabidGravy wonders if flattening the dependencies of the dependencies will help
12:04 lustlife joined #perl6
12:05 itcharlie joined #perl6
12:07 Arrowhead RabidGravy: nativesizeof doesn't work on CArray it seems. How do I match on the type of the CArray so I can implement it myself?
12:08 hankache thanks lizmat
12:10 zengargoylew joined #perl6
12:10 RabidGravy Arrowhead, it does work, what do you mean?
12:10 RabidGravy m: use NativeCall;  say nativesizeof(CArray)
12:10 camelia rakudo-moar 477830: OUTPUT«8␤»
12:11 RabidGravy which is right as it is a 64bit pointer
12:12 RabidGravy you can't know the number of elements in the CArray in the same way as you can't in C, it just doesn't have that information
12:12 Arrowhead ouch.
12:13 RabidGravy almost any C API will require the array and the number of elements, unless they special case some value like say strings do
12:14 Arrowhead yes, I'm using an api like that. And I'd hoped I could compute the numbers from the data structures I have...
12:14 Arrowhead but it looks like the structure of my perl6 program will look more and more like C.
12:14 RabidGravy you probably need to abstract it a layer up
12:16 RabidGravy well look at e.g. Audio::Libshout and Audio::Encode::LameMP3 - they pass around either Perl arrays or [CArray, Int] arrays with the details (mostly multis in there)
12:16 Arrowhead thanks. I'll now definitely take a look :-)
12:17 RabidGravy for performance reasons I made the "native values" available to give to things that know about them, but for programmer reasons I also make normal arrays available
12:18 RabidGravy I adjusted all those modules so that they would interoperate on both levels if required
12:21 RabidGravy it seems like a pain in the bum but you'll develop your own internal pattern book for dealing with these things quite quickly
12:22 Skarsnik Hello
12:22 zakharyas joined #perl6
12:22 RabidGravy rarr
12:22 Skarsnik how many of your modules did I broke today? x)
12:23 RabidGravy I haven't looked (because I haven't updated rakudo) but I'm fairly cool with the int thing I think
12:24 Skarsnik grr, I am again stuck with unfinalized type ><
12:25 RabidGravy I'm getting nervous, we haven't had loads of questions about installing rakudo in cruel and unusual ways this morning
12:25 Skarsnik How do I pass by ref?
12:26 RabidGravy "is raw" ?
12:26 stmuk_ it's quiet .. too quiet
12:26 RabidGravy (not completely understanding the context of the question)
12:26 RabidGravy "It's newbies, sir, thousands of them"
12:27 Skarsnik m: sub foo($a) {}; foo(Mu);
12:27 camelia rakudo-moar 9cecde: OUTPUT«Type check failed in binding $a; expected Any but got Mu␤  in sub foo at /tmp/U17ErXlHiS line 1␤  in block <unit> at /tmp/U17ErXlHiS line 1␤␤»
12:27 RabidGravy it must be the end times that wasn't on the TV recently
12:27 Skarsnik I have this kind of error x)
12:28 [ptc] stmuk_: the version on hack being used for the docs is: This is Rakudo version 2015.11-722-g1088291 built on MoarVM version 2015.11-113-gbd56e2e implementing Perl 6.b.
12:28 Skarsnik with thing that are not Mu
12:28 Skarsnik but I still want to do stuff with $a x)
12:28 ccakes joined #perl6
12:28 [ptc] stmuk_: I forget how moritz updates everything, but I think you're right: it needs an update
12:28 pmurias what should I write in the "Grant Deliverables ownership/copyright and License Information" section?
12:29 kid51 joined #perl6
12:29 pmurias (of the JavaScript backend Hague grant)
12:29 pmurias the work itself will be under Artistic 2.0 (as rest of nqp/rakudo is)
12:29 [ptc] stmuk_: btw: you added `*.un~` to the list of files to be ignored in the perl6-examples repo.  What kind of file are they?
12:29 [ptc] stmuk_: just curious :-)
12:30 hankache joined #perl6
12:31 RabidGravy pmurias, can you refer to previous applications?
12:33 xpen joined #perl6
12:36 nine RabidGravy: could be that installdeps needs the same fix as install
12:37 RabidGravy could be
12:37 RabidGravy let me have a look-see
12:38 pmurias RabidGravy: they seems to be a mix of "I signed the CLA already/the TPF may of have the copyright but I don't know how it works legally/I will hand over the copyright"
12:38 RabidGravy :-\
12:40 pierre-vigier joined #perl6
12:42 itcharlie joined #perl6
12:43 AlexDaniel joined #perl6
12:44 Skarsnik I can't find doc about different subroutine call x)
12:46 lustlife` joined #perl6
12:47 RabidGravy nine, was that the change to :$force?
12:47 RabidGravy if so yeah probably
12:48 nine yes
12:50 Skarsnik m: my Mu $a; say $;
12:50 camelia rakudo-moar 9cecde: OUTPUT«Method 'is' not found for invocant of class 'Mu'␤  in block <unit> at /tmp/1lHhkWsEbB line 1␤␤»
12:50 Skarsnik m: my Mu $a; say $a == Mu ;
12:50 camelia rakudo-moar 9cecde: OUTPUT«Cannot call infix:<==>(Mu, Mu); none of these signatures match:␤    ($?)␤    (\a, \b)␤    (Real \a, Real \b)␤    (Int:D \a, Int:D \b)␤    (int $a, int $b)␤    (Num:D \a, Num:D \b --> Bool)␤    (num $a, num $b --> Bool)␤    (Rational:D \a, R…»
12:50 nine m: my Mu $a; say $a.isa(Mu);
12:50 camelia rakudo-moar 9cecde: OUTPUT«True␤»
12:50 Skarsnik m: my Mu $a; say $a === Mu;
12:50 camelia rakudo-moar 9cecde: OUTPUT«Cannot call infix:<===>(Mu, Mu); none of these signatures match:␤    ($?)␤    ($a, $b)␤    (Int:D \a, Int:D \b)␤    (int $a, int $b)␤    (Num:D \a, Num:D \b)␤    (Num $ where { ... }, Num $ where { ... })␤    (num $a, num $b --> Bool)␤    (…»
12:50 nine m: my Mu $a; say $a ~~ Mu;
12:50 camelia rakudo-moar 9cecde: OUTPUT«True␤»
12:51 Skarsnik ~~ is not type equality x)
12:51 SwellJoe joined #perl6
12:52 RabidGravy nine, infact installdeps didn't have a force at all, that is remedied into the bargain :)
12:53 pmurias .seen jnthn
12:53 yoleaux I saw jnthn 25 Dec 2015 23:52Z in #perl6: <jnthn> Thanks! Really going, before I fall asleep on the keyboard :-)
12:53 RabidGravy a well deserved break I guess :)
12:58 Skarsnik I hope it's my last PR for NC sanity check
12:59 Skarsnik well before I decide to have them for CStruct x)
12:59 dalek rakudo/nom: d08e043 | lizmat++ | src/core/
12:59 dalek rakudo/nom: Make @*ARGS / *$ARGFILES initialization lazy
12:59 dalek rakudo/nom:
12:59 dalek rakudo/nom: Should shave off a bit from bare startup time
12:59 dalek rakudo/nom: review:
12:59 lizmat Skarsnik: it appears to have conflicts ?
13:00 Skarsnik I am not sure why
13:00 Skarsnik Probably a space issue
13:01 RabidGravy bastard space, it's just so big
13:03 kaare_ Space; The Last Frontier
13:03 cygx joined #perl6
13:03 cygx o/
13:03 Arrowhead joined #perl6
13:03 cygx Skarsnik: shouldn't all reprs from​blob/nom/lib/NativeCall.pm6#L122 be whitelisted?
13:04 kaare_ hm, I'd like to know what replaces the select function in Perl6 ?
13:05 kaare_ ie, select(2)
13:05 dalek infrastructure-doc: da78c59 | paultcochrane++ | hosts/
13:05 dalek infrastructure-doc: Add notes about how to update rakudobrew on hack
13:05 dalek infrastructure-doc: review:​tructure-doc/commit/da78c59f91
13:05 Skarsnik cygx, I should probably do it, yes x)
13:07 Skarsnik execpt for CArray x)
13:09 Skarsnik I probably want to check the inside of a Pointer too maybe
13:10 Skarsnik like to die on sub foo(Pointer[Int]
13:10 brrt joined #perl6
13:13 andreoss .tell masak consider my PR
13:13 yoleaux andreoss: I'll pass your message to masak.
13:14 rburkholder joined #perl6
13:14 RabidGravy - if anyone can and has a minute
13:14 [ptc] nine: I'm getting weird errors when installing Inline::Python on hack
13:14 [ptc] nine: In 'py_eval' routine declaration - Not an accepted NativeCall type for parameter [2]  : Int   -->For Numerical type, use the appropriate int32/int64/num64...
13:15 [ptc] nine: any idea what could be causing that?
13:15 [ptc] nine: on my local machine everything builds fine...
13:15 cygx [ptc]: Skarsnik happened :p
13:16 [ptc] cygx: ah.  Should I use specifically the 2015.12 version of Rakudo then?
13:16 RabidGravy Skarsnik is having a rampage around the NC stuff
13:16 Skarsnik nop I:python should be fixed
13:16 cygx no, Inline::Python should be fixed to no longer use Int in native sub declarations
13:17 [ptc] maybe panda doesn't know yet...
13:17 [ptc] no, that's a silly comment.  Installing via panda on my box works, on hack it doesn't.  Odd.
13:17 [ptc] I'll look into it further.  thanks for the help!
13:18 RabidGravy or to get it moving, fix I::Python and in the meantime install the fixed version?
13:19 cygx [ptc]: you need to check that​/blob/master/lib/Inline/Python.pm6#L51 and​-Python/blob/master/pyhelper.c declarations match and only use valid NativeCall types
13:20 RabidGravy it's almost certainly %s/int/int32/g
13:20 cygx RabidGravy: no, there are some Int involved
13:20 RabidGravy :-O
13:21 dalek Inline-Python: 5c52d88 | (Stefan Seifert)++ | lib/Inline/Python.pm6:
13:21 dalek Inline-Python: Fix type mismatches between Perl 6 and C code
13:21 dalek Inline-Python:
13:21 dalek Inline-Python: Use Perl 6 int32 for C int consistently.
13:21 dalek Inline-Python: Skarsnik++ for adding the check to NativeCall.
13:21 dalek Inline-Python: review:​ine-Python/commit/5c52d883a1
13:21 cygx well, it used to be just an alias for longlong
13:21 dalek cbb9efa | (Zoffix Znet)++ | source/style. (2 files):
13:21 dalek Fix main nav spillover
13:21 dalek
13:21 dalek Due to recently added additional nav items, the main nav spills on two lines on browser widths 768px–991px
13:21 dalek review:​
13:21 nine [ptc]: please upgrade to get the fix
13:21 [ptc] nine: thanks.  Just upgrading now :-)
13:23 Skarsnik I feel like I was the only one getting weird stuff on x86_32 because of Int xD
13:23 Juerd Is it possible, either directly or indirectly, to have a CUnion of a Blob and, say, a uint32?
13:23 Skarsnik hm why there is still a conflit
13:23 dalek 22f4017 | (Zoffix Znet)++ |
13:23 dalek Clarify how and where to run sass command
13:23 dalek review:​
13:24 nine Skarsnik: who is still on x86_32?
13:24 Skarsnik Juerd, no idea
13:24 Skarsnik my vm I use to work on/with perl6 x)
13:24 Juerd It would be an easier implementation of unpacking to fill the blob in the Cunion and then just read the uint32 back.
13:25 Skarsnik nine can you look at my last PR?
13:26 RabidGravy anyone with commit on panda ?  It would stop my tests failing on travis and I can get back to what I was doing
13:26 RabidGravy :)
13:26 Actualeyes joined #perl6
13:26 RabidGravy lizmat++
13:26 lizmat yw  :-)
13:28 nine RabidGravy: actually...if you need --force for installdeps, it tells me that the authors of your dependencies should adopt proper versioning
13:28 dalek specs: 325f7b2 | (Lucien Grondin)++ | S99-glossary.pod:
13:28 dalek specs: fix syntax in the advent calendar entry
13:28 dalek specs: review:
13:28 RabidGravy this is true
13:28 [ptc] the rakudobrew-ed perl6 on hack is now at 2015.12; the docs should build again now
13:29 RabidGravy [ptc]++
13:29 ccakes joined #perl6
13:29 nine Skarsnik: please rebase #663 on current nom and extend your commit message by telling _what_ needed fixing and most importantly: _why_. Otherwise it's very hard to tell whether the fix is correct.
13:30 dalek infrastructure-doc: 279e1d8 | paultcochrane++ | hosts/
13:30 dalek infrastructure-doc: Simplify rakudobrew update docs on hack
13:30 dalek infrastructure-doc: review:​tructure-doc/commit/279e1d821f
13:30 RabidGravy nine, in the particular case that always craps out, I can only complain a little as I am the author of most of the dependencies ;-)
13:31 nine RabidGravy: so you're in the position to actually fix it :)
13:32 RabidGravy they don't need force, it's just a side effect of passing a defined Bool (false) to the resolve()
13:33 Skarsnik how I solve conflcit with git?
13:33 RabidGravy in that in the installdeps case it wasn't passing :$force at all, so I could either do force => False in the call to resolve or add :$force = False to the args, I chose the latter
13:34 nine Skarsnik: git mergetool
13:36 tadzik lizmat++
13:36 tadzik (panda maintainers)++
13:36 brrt hi #perl6
13:36 tadzik hi brrt
13:37 RabidGravy tadzik++
13:37 brrt hi tadzik :-)
13:37 RabidGravy everyone++
13:37 brrt any jit bugs reported, aside from the building bugs that i know about
13:37 FritzZaucker joined #perl6
13:38 cygx tadzik: the problem we talked about some time ago (not stringifying $pandadir before passing it to mkpath) did cause actual problems
13:38 RabidGravy brrt, yeah it doesn't jit on ARM ;-)
13:38 cygx tadzik: I've submitted a new pull request that works for mojca
13:38 tadzik cygx: oh
13:38 brrt :-P
13:38 tadzik oh, so .mkdir is now recursive and all?
13:38 brrt it will a bit more easily once i've merged even-moar-jit
13:39 tadzik that's kewl
13:39 Skarsnik let's hope that solve the conflict x)
13:39 RabidGravy groovy
13:39 cygx tadzik: I've also removed the chedk for existence before trying to mkpath, as supposedly that also caused a problem
13:39 RabidGravy next stop was going to be getting Ban Ki Moon in on the job
13:39 cygx I'm a bit doubtful about that, but I'm coding blind here ;)
13:40 tadzik cygx: it seems fitting
13:40 lizmat on .mkdir being recursive, probably need .rmdir to be then as well, to more easily cleanup oopses
13:40 lizmat cycling&
13:40 Skarsnik damn, no way to force a PR? (it's not for rakudo)
13:40 RabidGravy isn't there an rmtree somewhere anyway?
13:42 HoloIRCUser3 joined #perl6
13:43 Zero_Dogg joined #perl6
13:43 ChristopherBotto joined #perl6
13:44 Skarsnik .seen azawawi
13:44 yoleaux I saw azawawi 28 Dec 2015 08:47Z in #perl6: <azawawi> Merry Christmas to #perl6 :)
13:45 Zero_Dogg Is IO::Path suppose to be a core module? My perl6 install (using rakudobrew to install moar+panda) doesn't have it, and "panda install IO::Path" says it can't find it
13:45 Skarsnik It's core
13:45 nine It's core, it's no module.
13:45 Skarsnik m: my IO::Path $a;
13:45 camelia rakudo-moar d08e04: ( no output )
13:45 cygx tadzik: jus fyi, I just checked: on moarvm, mkdir has been recursive since before the switch to libuv (it used to be implemented via apr_dir_make_recursive)
13:45 nine nine@sunshine:~> perl6 -e 'use IO::Path'
13:45 nine ===SORRY!===
13:45 nine IO::Path is a builtin type. You can use it without loading a module.
13:46 cygx git++ # you can look stuff up!
13:47 Zero_Dogg oh, so core classes don't need to be loaded in 6 then. Thanks! I'm a long time v5 user, learning v6 now
13:47 Skarsnik damn any idea how I can force to get some change from a fork than can't be merged?
13:47 ChristopherBotto Thanks for all of your work #perl6!
13:47 yoleaux 15 Dec 2015 00:19Z <geekosaur> ChristopherBotto: it's write access on /tmp that controls whether unlink of /tmp/foo works. permissions on /tmp/foo's inode are not relevant.
13:47 Arrowhead joined #perl6
13:47 yoleaux 15 Dec 2015 00:24Z <geekosaur> ChristopherBotto: GNU rm looks at the permission as a courtesy, more or less. Hard links imply that the directory entry is independent of the inode (file). I am not sure perl6 should be in the business of enforcing anything more than OS permissions.
13:48 nine Skarsnik: what do you mean?
13:48 cygx Zero_Dogg: most core classes are part of the setting and semantically part of an outer lexical scope surrounding your program
13:48 cygx so you get lexical acces to the identifiers without use
13:48 RabidGravy Zero_Dogg, an easy rule of thumb is that if it's on it's almost certainly compiled in the setting
13:48 [ptc] m: require File::Temp
13:48 camelia rakudo-moar d08e04: OUTPUT«Could not find File::Temp in:␤    /home/camelia/.perl6/2015.12-46-gd08e043␤    /home/camelia/rakudo-m-inst-2/share/perl6/site␤    /home/camelia/rakudo-m-inst-2/share/perl6/vendor␤    /home/camelia/rakudo-m-inst-2/share/perl6␤    CompUnit::Reposit…»
13:48 Skarsnik I want to get azawawi work on Pg Array on DBIish, merge them in my fork and make them work with the current DBIish before merging them in DBIis
13:49 Skarsnik But I can't PR, since it conflict x)
13:49 Zero_Dogg cygx: aha, thanks, that's nice. My perl6 just errored out on a missing module when I tried to use it which is what confused me
13:50 ccakes joined #perl6
13:50 Skarsnik it's weird File::Temp is not core x)
13:50 Skarsnik but probably too Os dependant
13:50 [ptc] Skarsnik: was just trying to test something...
13:50 FritzZaucker The following NCurses example is working:​ucker/aa14025a16edcd671e25  whereas a similar one using a interval supply doesn't properly display the text:​ucker/aa14025a16edcd671e25 ... am I doing something wrong (running 6.c on Unbuntu 14.04)
13:50 nine Skarsnik: resolve the conflict before sending the pull request
13:51 Skarsnik I don't have access to azawawi repo x)
13:51 nine Zero_Dogg: I just today pushed a commit that gives a better error message in your case :)
13:51 aborazmeh joined #perl6
13:51 aborazmeh joined #perl6
13:51 Zero_Dogg nine: awesome :)
13:51 Skarsnik Should get a git diff and work with that maybe?
13:52 nine Skarsnik: no, git mergetool
13:52 nine Skarsnik: learn how to use it (I'm using vimdiff, but there are others). It's really worth it.
13:52 [ptc] how can one work out what modules are available to camelia?
13:53 Skarsnik there is a moar star on camelia
13:53 nine [ptc]: I guess it just supports the CORE modules
13:53 Skarsnik m-star:
13:53 Skarsnik r-star:
13:53 s_kilk joined #perl6
13:53 Skarsnik r-star: say "hello"
13:53 [ptc] I just want to test requiring a module, and want to see if it's my system which is odd, or if it's a general problem
13:53 cygx Skarsnik:
13:54 RabidGravy FritzZaucker, they're the same matey :)
13:54 hankache can panda nuke a module ?
13:54 Skarsnik nine, I guess I should add azawawi as a new remote to try to merge in my repos? since I can't github PR
13:55 Skarsnik I don't think panda have a uninstall feature x)
13:55 Arrowhead joined #perl6
13:55 RabidGravy .seen azawawi
13:55 yoleaux I saw azawawi 28 Dec 2015 08:47Z in #perl6: <azawawi> Merry Christmas to #perl6 :)
13:55 [ptc] any idea why `unless ( try {require URI::Escape} ) { say 'uri::escape missing' }` returns "uri::escape missing" even though it's installed?
13:56 [ptc] I'm sure I'm doing something dumb here, but don't know yet what...
13:56 gfldex m: my $pc = => "head1", config => {:numbered("0")}); dd $pc;
13:56 camelia rakudo-moar d08e04: OUTPUT«Pod::Config $pc = => "head1", config => {:numbered("0")})␤»
13:56 Skarsnik I am not sure try return a value?
13:56 gfldex can you see the problem?
13:56 [ptc] Skarsnik: it used to...
13:56 RabidGravy it does
13:57 awie joined #perl6
13:57 RabidGravy m:  unless try require URI::Escape { say "boo" }
13:57 camelia rakudo-moar d08e04: OUTPUT«Block object coerced to string (please use .gist or .perl to do that)  in any  at gen/moar/m-Perl6-Actions.nqp line 1809␤5===SORRY!5===␤Expression needs parens to avoid gobbling block␤at /tmp/4DdP9KwkmR:1␤------> 3ss try require URI::Escape { …»
13:57 gfldex m: my $pc = => "head1", config => {:numbered("0")}); say $pc.config<numbered>.Bool; # filed as #127085
13:57 camelia rakudo-moar d08e04: OUTPUT«True␤»
13:58 [ptc] my $out = try {require URI::Escape}; say $out;   gives (Escape), which evaluates as true
13:58 [ptc] however as soon as it's wrapped in the unless it gets seen as false, which is od
13:58 [ptc] *odd
13:58 tadzik cygx: oh, neato
13:58 [ptc] m: unless (try require URI::Escape { say "moo" }
13:58 camelia rakudo-moar d08e04: OUTPUT«Block object coerced to string (please use .gist or .perl to do that)  in any  at gen/moar/m-Perl6-Actions.nqp line 1809␤5===SORRY!5=== Error while compiling /tmp/Lmp7IgcVJT␤Unable to parse expression in parenthesized expression; couldn't find fina…»
13:58 Juerd [ptc]: (Escape) is the stringification of a type object. A type object is an undefined value of that type.
13:58 [ptc] m: unless (try require URI::Escape) { say "moo" }
13:58 camelia rakudo-moar d08e04: OUTPUT«moo␤»
13:58 Juerd [ptc]: Undefined values are false.
13:59 [ptc] Juerd: that will explain that then, thanks
13:59 Juerd m: my Str $foo; say $foo; say so $foo;
13:59 camelia rakudo-moar d08e04: OUTPUT«(Str)␤False␤»
13:59 Juerd m: my Str $foo = "0"; say $foo; say so $foo;
13:59 camelia rakudo-moar d08e04: OUTPUT«0␤True␤»
13:59 Juerd m: my Str $foo = ""; say $foo; say so $foo;
13:59 camelia rakudo-moar d08e04: OUTPUT«␤False␤»
14:00 nine m: use Test; BEGIN { require DateTime::Formats; CATCH { when X::CompUnit::UnsatisfiedDependency { done-testing; exit; }; }; }; import DateTime::Format; strftime;
14:00 camelia rakudo-moar d08e04: OUTPUT«1..0␤»
14:00 nine [ptc]: ^^^
14:00 [ptc] nine: danke!
14:01 RabidGravy nine++ # that's such a handy little idiom it should be written down somewhere
14:01 cygx nine: is X::CompUnit::UnsatisfiedDependency the only error that can get thrown? Eg no IO errors?
14:02 araujo I should be able to access a value from a CArray[String] directly using array indexing? , or do I need some special magic for that?
14:02 nine Or....integrated into as test-require function (like test-depends in the
14:02 Skarsnik araujo, it should work
14:02 RabidGravy araujo, yes.
14:02 lustlife joined #perl6
14:02 smls joined #perl6
14:02 [ptc] nine: interesting.  The tests for the doc modules have been using `unless (try require)` since February and it seems noone has noticed that it stopped working!
14:02 RabidGravy if it isn't working for you show us what you are trying
14:02 nine cygx: the user will want to see all other errors. IO errors should only occur if something is really wrong.
14:03 araujo Skarsnik, RabidGravy Ok
14:04 Skarsnik I should add nativecast to the doc
14:04 cygx nine: but if you check for an optional dependency, you generally want to keep going (after possibly emitting a warning)
14:04 Skarsnik You can do stuff like​o/blob/master/lib/Gumbo/Parser.pm6#L46
14:05 dalek doc: 17fdcb5 | paultcochrane++ | t/pod-htmlify.t:
14:05 dalek doc: Fix checking for URI::Escape dependency in tests
14:05 dalek doc:
14:05 dalek doc: nine++ for the tip of how to do this with a current Rakudo.
14:05 dalek doc: review:
14:05 nine cygx: I mean really majorly wrong. As in "this needs fixing or other things will break, too"
14:05 smls Are Hash lookups thread-safe?
14:06 nine And this require thing is why I really wanted to merge curli before 6.c. Because people will write code that uses this API.
14:06 Zero_Dogg What's the perf difference between moar and jvm? Got some scripts to run on a raspi2, which isn't the fastest thing in the world
14:06 nine smls: nope
14:06 cygx nine: because only one precomp store is active at a time, and you're hosed if you cannot access that?
14:06 [ptc] \o/ the docs pages are building again.  nine++
14:06 * cygx shuold realy read up on that precomp thing
14:06 smls ok
14:07 RabidGravy Hmm ye olde "Missing serialize REPR function for REPR VMException" is back
14:07 Skarsnik cygx, did you have any idea why replacing some nativecast with the correct Pointer[mystruct] .deref fail sometime?
14:07 nine cygx: it's not only the precomp store. Could be the rest of the repository just as well. And if the rest is affected and you don't notice and install other modules, things could get even worse.
14:07 nine brrt: actually, there may be a JIT or SPESH bug revealed by a panda problem
14:08 RabidGravy Zero_Dogg, moar is I believe faster to load and so on, it also works properly with 2015.12 jvm hasn't quite caught up
14:08 Skarsnik Zero_Dogg, jvm is behind moar in term of feature, for perf I don't really know x)
14:08 cygx Skarsnik: not specifically, but deref has issues with pointer-to-pointers, and nativecall generally assumes that you're dealing with pointers to structs instead of struct itself
14:08 Zero_Dogg RabidGravy, Skarsnik: thanks
14:08 cygx (imo a mistake as that's not how C does it)
14:08 nine brrt: panda has a --force flag. That's just a named Bool on the MAIN routine that gets passed on several levels to CompUnit::Repository::Installation::install. In some cases we get an error about unboxing a native after installing a couple of modules.
14:09 RabidGravy Zero_Dogg, I'm just working on something for the RPi at the moment
14:09 _mg_ joined #perl6
14:09 nine brrt: The error occurs in a simple if not $force, where $force is just a Bool type object. That's after the same code ran a couple of times.
14:09 Skarsnik cygx, I mean .deref just call nativecast(ptr, type); but I get binding issue comparing to just calling nativecast the same
14:10 RabidGravy I tend to get the basic thing working on a proper laptop then put it on the RPi to test
14:10 gfldex .seen hankache
14:10 yoleaux I saw hankache 13:54Z in #perl6: <hankache> can panda nuke a module ?
14:10 Arrowhead joined #perl6
14:10 cygx Skarsnik: yes, that's still broken if no one has fixed it since I last looked into this
14:10 gfldex .tell hankache I got something working now that I believe to be able to render perl6intro from pod
14:10 yoleaux gfldex: I'll pass your message to hankache.
14:11 Zero_Dogg RabidGravy: nice. I've got it installed through rakudobrew right now, but it uses quite a bit of time before it actually starts executing the script. Not a problem for this one, but a bit annoying while testing it :)
14:11 gfldex .tell hankache see tmp/html.html for example output
14:11 yoleaux gfldex: I'll pass your message to hankache.
14:11 RabidGravy Zero_Dogg, there isn't a JIT for ARM at the moment, which has some impact
14:11 Skarsnik feel like a Perl 6 bug or return should be a binding or something?
14:11 cygx Skarsnik: workaround is to cast to pointer-to-uintptr (or size_t), read that, then create a enw pointer
14:11 timotimo "return" reconts unless you return-rw
14:12 timotimo (what am i missing?)
14:12 timotimo o/
14:12 timotimo i neglected backlog completely today
14:12 RabidGravy a unicorn, everyone is missing a magical unicorn
14:12 Zero_Dogg RabidGravy: ah
14:12 Skarsnik hm, maybe I should try replacing with return-rw
14:13 RabidGravy what was the C parser module thingy called, I'm failing to find it in the module list
14:13 Skarsnik C::Parser ?
14:14 colomon what Skarsnik said
14:14 colomon might be gone?
14:14 RabidGravy oh yeah, nah I'm just going blind or stupid or both
14:14 colomon well, it’s here:
14:15 Arrowhead it's not compatible with 6.c
14:16 cygx Skarsnik: a helper method I used to use:
14:16 Skarsnik rofl
14:16 cygx if you don't have a uintptr type around, size_t should work as well
14:19 [ptc] timotimo: \o
14:20 RabidGravy oh, yes it won't even install with panda
14:20 RabidGravy that sucks
14:21 RabidGravy that'll be a yak too far, I think I'll ignore it for the time being
14:22 brrt nine, /me was afk for a bit
14:22 timotimo sup brrt :)
14:22 dalek doc: 2fa7682 | paultcochrane++ | doc/ (4 files):
14:22 dalek doc: Remove the Enum type
14:22 dalek doc:
14:22 dalek doc: ... because it was removed from Rakudo in 2015.09
14:22 dalek doc: review:
14:23 brrt hmm ok
14:23 brrt would be very awesome if we could golf that
14:23 dalek perl6-roast-data: 89d24d7 | coke++ | / (7 files):
14:23 dalek perl6-roast-data: today (automated commit)
14:23 dalek perl6-roast-data: review:​-roast-data/commit/89d24d70e2
14:23 Ulti only a couple of tests fail for p6-c-parser on 6.c
14:26 [ptc] m: say sleep-till now+10
14:26 camelia rakudo-moar d08e04: OUTPUT«5===SORRY!5=== Error while compiling /tmp/u8ZYaELdPX␤Undeclared routine:␤    sleep-till used at line 1. Did you mean 'sleep-until', 'sleep-timer'?␤␤»
14:27 RabidGravy Ulti, my panda is telling me it can't use the source-url in the meta file
14:27 nine m: my Bool $foo; for ^1000 { if not $foo { }; };
14:27 camelia rakudo-moar d08e04: OUTPUT«Cannot unbox a type object␤  in block <unit> at /tmp/1h1Hpyc5P2 line 1␤␤»
14:27 nine brrt: lol now that was easy :)
14:27 brrt that is easy
14:27 [ptc] bummer!  sleep-till has gone :-/  And I had a use-case for that too!
14:28 cygx Did you mean 'sleep-until'
14:28 nine m: my Bool $foo; while (not $foo) {}
14:28 camelia rakudo-moar d08e04: OUTPUT«Cannot unbox a type object␤␤»
14:28 cygx or has that different semantics?
14:28 hankache computer, messages
14:28 yoleaux 14:10Z <gfldex> hankache: I got something working now that I believe to be able to render perl6intro from pod
14:28 nine brrt: even shorter ^^^
14:28 yoleaux 14:11Z <gfldex> hankache: see tmp/html.html for example output
14:28 [ptc] cygx: oh, did it get renamed?
14:28 brrt oh, that one is cute
14:28 cygx [ptc]: no idea, that was part of the error message ;)
14:28 timotimo nine: i wonder if spesh makes that mistake
14:29 [ptc] cygx: true!  Reading helps!
14:29 brrt speshbug
14:30 nine timotimo: yes, it's spesh
14:30 hankache .tell gfldex i'll try it during the evening
14:30 yoleaux hankache: I'll pass your message to gfldex.
14:30 RabidGravy fwiw, I think my patch may have unintentionally unearthed a further infelicity in installdeps
14:30 timotimo right. we probably specialize istrue in this case to unbox_i, but without making sure the incoming arg is known to be concrete
14:30 timotimo seems like an easy fix; thanks for the golf!
14:31 smls Can two threads use  chdir($dir);  without affecting each other?
14:31 nine :)
14:31 perlpilot smls: probably not.
14:32 timotimo we have only a fake PWD in perl6
14:32 nine brrt: thanks for asking about JIT bugs. Pushed me into the right direction. JIT and spesh are so stable that I completely forget, that they could interfere :)
14:32 smls timotimo: But is the fake working dir thread-local?
14:33 timotimo i sure hope so!
14:33 brrt yw :-)
14:33 timotimo it's implemented as a dynamic variable, i think it's called $*CWD
14:33 timotimo otherwise you can my $*CWD = OUTER::('$*CWD') (or what the correct syntax would be)
14:34 llfourn CALLERS::<$*CWD> # in my experience
14:34 timotimo good
14:34 timotimo oh, yeah, OUTER is BS
14:35 VGA joined #perl6
14:35 smls m: $*CWD = "/";
14:35 camelia rakudo-moar d08e04: OUTPUT«Cannot modify an immutable IO::Path␤  in block <unit> at /tmp/a24lQN0fQ8 line 1␤␤»
14:35 smls :/
14:36 llfourn m: my $*CWD = "/";
14:36 camelia rakudo-moar d08e04: ( no output )
14:36 timotimo i have a fix for the Bool thing :)
14:37 nine yeah!
14:37 masak OT, but a nice read:
14:37 yoleaux 13:13Z <andreoss> masak: consider my PR
14:37 * masak looks
14:37 smls rhanks llfourn
14:37 llfourn m: my $*CWD = "/".IO # probably better :)
14:37 camelia rakudo-moar d08e04: OUTPUT«Use of uninitialized value $CWD of type Any in string context␤Any of .^name, .perl, .gist, or .say can stringify undefined things, if needed.  in block <unit> at /tmp/WF5_oI9SwO line 1␤»
14:37 smls *thanks
14:37 llfourn erm...
14:38 timotimo nine: i didn't yet run spec tests with the new fix, but it should be sane
14:38 llfourn m: my $*CWD = "/etc".path; say $*CWD;
14:38 camelia rakudo-moar d08e04: OUTPUT«Use of uninitialized value $CWD of type Any in string context␤Any of .^name, .perl, .gist, or .say can stringify undefined things, if needed.  in block <unit> at /tmp/q8U94jMpTo line 1␤"/etc".IO␤»
14:38 smls This works locally:  perl6 -e 'my $*CWD = "/"; run "pwd";'
14:38 llfourn smls:  ok good :)
14:39 psy joined #perl6
14:40 pierre-vigier joined #perl6
14:40 smls Looks like   indir "docs", { ... }   is thread-safe
14:41 dalek doc: f3d013a | paultcochrane++ | doc/Type/Temporal.pod:
14:41 dalek doc: Fix sleep-until documentation
14:41 dalek doc:
14:41 dalek doc: sleep-till was renamed to sleep-until
14:41 dalek doc: review:
14:41 dalek doc: 05eb2fb | paultcochrane++ | doc/Type/ (2 files):
14:41 dalek doc: Move sleep-related docs into Date docs
14:41 dalek doc:
14:41 dalek doc: The sleep-related functions are actually part of the Date module, hence it
14:41 dalek doc: is more sensible to hang the docs on the Date docs, rather than having a
14:41 dalek doc: separate Temporal docs file.
14:41 dalek doc: review:
14:41 smls But chdir() seems to change it globally for all threads
14:41 timotimo i suppose it depends on whether or not you build your own $*CWD
14:41 RabidGravy yeah, that was what I was going to say
14:44 lucasb joined #perl6
14:46 timotimo nine: i don't think i should bump nqp and moar for this, eh?
14:46 timotimo especially since if we release a little bug fix release of rakudo before star we'd bump anyway
14:46 timotimo which i recommend for the installation fix
14:47 nine timotimo: yeah, we do have workarounds in panda and there don't seem to be other real world cases
14:47 Arrowhead joined #perl6
14:47 timotimo oh?
14:47 timotimo i thought the problem also appears in install-core-modules.p6?
14:48 nine really?
14:48 * araujo trying to get some bindings working .... but nothing so far
14:48 masak andreoss: I left some comments in
14:48 araujo anybody knows a good lib (up to date) using NativeCall out there?
14:48 nine timotimo: I guess with MVM_SPESH_NODELAY=1 it would break.
14:48 timotimo nine: i know nothing :)
14:49 araujo masak, hola there o/
14:49 araujo :)
14:49 nine araujo: Inline::Perl5 or RabidGravy's work
14:49 xpen joined #perl6
14:49 masak araujo: saluton!
14:50 masak araujo: didn't you toy around with making your own language some years back?
14:50 RabidGravy well in my case I wouldn't go as far as "good", but they exercise NativeCall quite heavily
14:50 [Coke] [ptc]: why do we have to delete and re-download rakudobrew to run rakudobrew?
14:50 ccakes joined #perl6
14:50 [Coke] (re: infrastructure-doc: da78c59)
14:51 araujo nine, where can I find RabidGravy's awesome work?
14:51 araujo :)
14:51 Skarsnik nine, can you merge my PR now? x) It fix an issue someone has with a NC module yesterday
14:51 smls timotimo: When assigning directly to «my $*CWD» it only accepts a Str, not an IO::Path. But then the next time you call chdir(), it complains "$*CWD is a Str, not an IO::Path!!!"
14:51 RabidGravy araujo, all but one of the Audio::* modules in the ecosystem :)
14:52 Skarsnik araujo, maybe paste a gist of what is causing your issue x)
14:52 timotimo ugh, that's bad :)
14:52 araujo Skarsnik, yeah, sorry, I will, checking something first
14:52 RabidGravy (well one of those isn't actually using NativeCall because I never got round to making silan into a library)
14:52 araujo RabidGravy, oohuh ok :P
14:53 smls timotimo: Also, the specs say that «temp $*CWD;» sould work, but it doesn't because the default $*CWD is immutable.
14:53 smls "Can only use 'temp' on a container"
14:54 RabidGravy I think $*CWD is bound to a Proxy if my memory isn't failing me
14:54 smls As long as you don't mix manual assignm,ent and chdir() though, they work fine... :)
14:54 cygx smls: that's ok, the 'specs' are just 'design documents' now >:)
14:54 Arrowhead_ joined #perl6
14:55 smls Also, indir() works correctly with chdir().
14:55 timotimo m: say $*CWD.VAR.WHAT
14:55 camelia rakudo-moar d08e04: OUTPUT«(Path)␤»
14:55 timotimo hum
14:55 RabidGravy maybe my memory is failing me
14:55 RabidGravy :)
14:55 timotimo AFK for a bit
14:56 RabidGravy which I wouldn't be majorly surprised about the amount of beer I've drunk in the last week
14:56 ZoffixW joined #perl6
14:58 nine Skarsnik:​/ab9aefab85766d2bb518c28a5fce2b38174276d5 looks very, very wrong
14:58 ZoffixW m: use lib <first second>; use Foo;
14:58 camelia rakudo-moar d08e04: OUTPUT«===SORRY!===␤Could not find Foo in:␤    /home/camelia/second␤    /home/camelia/first␤    /home/camelia/.perl6/2015.12-46-gd08e043␤    /home/camelia/rakudo-m-inst-2/share/perl6/site␤    /home/camelia/rakudo-m-inst-2/share/perl6/vendor␤    /hom…»
14:58 ZoffixW My 2 cents ^ that should place "first" BEFORE "second" into search paths. Reasoning: 1) it's intuitive for them to end up in the order I specified it in my code. 2) Perl 5's use lib does it that way, so the reverse behavior could trip up Perl 5 converters.
14:58 Skarsnik wtf
14:59 Skarsnik I hate git
14:59 nine ZoffixW: sounds reasonable. I'd be happy to merge a PR for that.
14:59 ZoffixW \o/
14:59 * ZoffixW proceeds to work on it
15:00 Skarsnik nine, can you guide me on how to reverse/fix this mess
15:00 itaipu joined #perl6
15:00 nine Skarsnik: why do you hate a tool when you just haven't mastered it yet? git is awesome and can help you so much. It simply takes a while to learn.
15:01 ZoffixW Actually, I'll do it tonight on my better box. On this one it'll take me 40 minutes just to pull all the repos and 30 minutes to compile and run the spectest XD
15:03 Skarsnik I just don't understand how he add file the commit without telling me
15:03 jshy joined #perl6
15:03 nine Skarsnik: 1. make sure your origin/nom is up to date. git remote update should do that. 2. do a git rebase -i origin/nom # remove the "solve conflict?" commit from the list. 3. when it reports a conflict, edit lib/NativeCall.pm6
15:04 nine Skarsnik: there remove the part betweeen <<<< and ==== (inclusive) and remove the >>>>> line. Then git add lib/NativeCall.pm6 && git rebase --continue
15:06 Skarsnik It don't offer me to remove
15:08 nine Skarsnik: git rebase -i will open a text editor where the commits are listed. Simply remove the line with the offending commit
15:08 Skarsnik it was noop
15:08 Arrowhead joined #perl6
15:09 nine Skarsnik: oh, your origin is Skarsnik/rakudo, not rakudo/rakudo
15:09 Skarsnik Oh yeah upstream
15:09 Skarsnik not origin x)
15:10 Cabanossi joined #perl6
15:11 ccakes joined #perl6
15:12 Skarsnik skarsnik@testperl6:~/devel/rakudo$ git rebase -i upstream/nom
15:12 Skarsnik lib/NativeCall.pm6: needs merge
15:12 Skarsnik Cannot rebase: You have unstaged changes.
15:12 Skarsnik duh
15:14 Skarsnik great it reverted all my commit
15:14 Skarsnik fuck this shit
15:14 itcharlie joined #perl6
15:14 Skarsnik sorry for language
15:14 geekosaur it didn't revert it
15:14 geekosaur it's waiting to finish it, you need to tell it how to deal with the conflict by editing and "git add"-ing the file with the merge conflict, or reset the merge to discard it
15:15 Skarsnik it reverted my 3 commits
15:15 Skarsnik only the last was faulty
15:16 Skarsnik now How I get back my 2 first commits ~~
15:16 Skarsnik let go with the rm solution x)
15:18 RabidGravy at times like this I find a good "git reset --hard ..." works ;-)
15:18 Arrowhead Skarsnik: make a backup to prevent further damage, then explore "git reflog" and "git checkout ..." to go to a likely point
15:20 RabidGravy yikes, someone got me some "peanuts from hell".  I was skeptical about the claim to be the hottest chilli nuts in the universe, I was wrong, can only just feel my lips after ten minutes and the application of beer
15:20 andreoss masak: fixed that
15:20 Skarsnik x)
15:21 [ptc] [Coke]: it's only something which needs to be done once a release
15:21 [ptc] [Coke]: also, it's something I saw in the history and thought it's probably a good idea to start from a clean slate
15:21 khw joined #perl6
15:22 Skarsnik obviously having git revert last was too simple for git...
15:22 CQ joined #perl6
15:22 geekosaur easy revert is dangerous in a distributed vc system
15:23 [ptc] [Coke]: also, I thought if one just uses update or nuke or similar, then there are often follow-on problems that can happen
15:23 geekosaur it cannot know if others have received that commit, and if they have then your repo would no longer be compatible with theirs
15:24 [ptc] [Coke]: anyway those were the reasons behind my decision to purge and recreate; I don't have problems with changing things if they can be done better
15:25 Skarsnik geekosaur, I don't see the issue? there is a pull operation
15:25 Skarsnik he is dumb to not know a commit is pushed or not
15:25 moritz \o
15:25 ZoffixW Any idea what this error is about? There's no line number or file or anything:​ixznet/107e9d6937d87357c7bb
15:25 ZoffixW o/
15:26 lnrdo joined #perl6
15:26 Skarsnik Err
15:26 Skarsnik Weird error
15:26 geekosaur Skarsnik, there are ways other than push
15:26 itcharlie joined #perl6
15:26 * geekosaur works with repos in AFS, there's no push there :)
15:27 Skarsnik yes but it's not like it can give you a warning or not allow you something when it considered dangerous
15:27 ZoffixW Ah. k. found it
15:27 ZoffixW m: use Test:ver<42>;
15:27 camelia rakudo-moar d08e04: OUTPUT«===SORRY!===␤Cannot call Numeric(Version: ); none of these signatures match:␤    (Mu:U \v: *%_)␤»
15:28 ZoffixW How do I specify which version I want to use in my code?
15:28 [Coke] [ptc]++ danke.
15:28 Skarsnik m: use Test::ver<v42>;
15:28 camelia rakudo-moar d08e04: OUTPUT«===SORRY!===␤Could not find Test::ver in:␤    /home/camelia/.perl6/2015.12-46-gd08e043␤    /home/camelia/rakudo-m-inst-2/share/perl6/site␤    /home/camelia/rakudo-m-inst-2/share/perl6/vendor␤    /home/camelia/rakudo-m-inst-2/share/perl6␤    Com…»
15:28 Skarsnik m: use Test:ver<v42>;
15:28 camelia rakudo-moar d08e04: OUTPUT«===SORRY!===␤Could not find Test:ver<v42> in:␤    /home/camelia/.perl6/2015.12-46-gd08e043␤    /home/camelia/rakudo-m-inst-2/share/perl6/site␤    /home/camelia/rakudo-m-inst-2/share/perl6/vendor␤    /home/camelia/rakudo-m-inst-2/share/perl6␤   …»
15:30 ZoffixW Use WWW::Google::Time:ver<v1.001001> also says "Could not find..." But I do have 1.001001 installed
15:30 * ZoffixW skips explicit versions for now
15:32 Cabanossi joined #perl6
15:32 _mg_ joined #perl6
15:33 ZoffixW whoa! Perl 6 makes spelling suggestions with imported subs! That's awesome. "Undeclared routine: conunction used at lines 23, 24. Did you mean 'conjunction'?"
15:34 spider-mario joined #perl6
15:35 timotimo of course it does
15:35 timotimo the list of subs and multi candidates is known at compile time
15:35 timotimo that's one of our main strengths when it comes to lexical scoping
15:36 cygx ZoffixW: I believe that should read WWW::Google::Time:ver<1.001001>
15:37 ZoffixW cygx, so the error I get is a bug then?
15:39 moritz ZoffixW: I think what cygx is saying is that you should do "use WWW::Google::Time:ver<1.001001>;" not ":var<v...>"
15:39 pmurias what does utf16str from nativecall map in c terms
15:39 pmurias ?
15:39 pmurias is it w_char*?
15:39 Skarsnik it's just an encodaing
15:39 cygx ZoffixW: what moritz++ said
15:40 vendethiel joined #perl6
15:40 ZoffixW Yeah, but that gives the error:
15:40 ZoffixW m: use Test:ver<42>;
15:40 camelia rakudo-moar d08e04: OUTPUT«===SORRY!===␤Cannot call Numeric(Version: ); none of these signatures match:␤    (Mu:U \v: *%_)␤»
15:40 ZoffixW m: say 42.Version
15:40 camelia rakudo-moar d08e04: OUTPUT«Method 'Version' not found for invocant of class 'Int'␤  in block <unit> at /tmp/SDVwv1OALB line 1␤␤»
15:40 ZoffixW hm.
15:40 Skarsnik I want to give up on this git dumb thing...
15:40 ZoffixW do it
15:40 b2gills nine: When I compiled my own 32bit Perl 5s, I set the size of IV to 64bit. So basically you will have to inspect the output of `perl -V` on 32bit (backlogging)
15:41 Skarsnik wtf Perl5
15:43 MadcapJake joined #perl6
15:43 hahainternet Skarsnik: what's up with your git?
15:44 Skarsnik Apparently reverting a commit it's too hard
15:44 MadcapJake A fresh rakudobrew install, build moar, build panda, and then I tried `panda install Task::Star` and it fails at File::Find
15:44 hahainternet what have you done so far, and what situation is it in? it should be as simple as git revert commit
15:45 cygx pmurias: any pointer type will do, really
15:45 cygx moarvm returns a char* from MVM_string_utf16_encode
15:46 MadcapJake To install Linenoise I had to  `panda --force install File::Find` and then I had to `panda --force install Shell::Command` and now I tried `panda install Linenoise` and I get "P6M Merging GLOBAL symbols failed: duplicate definition of symbol Find"
15:46 cygx note that the size of wchar_t is platform dependent (16-bit wchar_t is mainly a windows thing)
15:47 zhi joined #perl6
15:47 cygx pmurias: uint16_t* and char16_t* would be good choices if available
15:48 [Coke] Rearding the calls to cut a 2015.12.1 ASAP - 2016.01 is due in about 2.5 weeks.
15:48 ZoffixW cool
15:48 [Coke] And we are definitely going to have to spend some time figuring out what the releases (and release process) is going to look like before then.
15:48 ZoffixW MadcapJake, running rakudobrew build-panda right this second and install succeeds. I notice rakudobrew build panda also starts doing something but I aborted it. Not sure what that does, but the correct command is build-panda with the hyphen
15:48 Arrowhead joined #perl6
15:49 RabidGravy MadcapJake, I think it's something to do with panda bundling those itself
15:49 ZoffixW MadcapJake, those modules are installed during build-panda modules
15:49 ZoffixW oh
15:49 * ZoffixW actually runs panda install Task::Star -_-
15:51 RabidGravy I don't quite now how but I think my change to installdeps may have triggered this
15:51 RabidGravy know
15:51 brrt i would like as many different people as possible test the libuv-1.8 branch of moarvm
15:51 ZoffixW Aye, I'm getting File::Find error now too
15:52 brrt especially on windows, os x, bsd, solaris, whatever
15:52 brrt it looks good to me, but that is not a sufficient criterion :-)
15:52 ZoffixW I think the issue is panda bails out with an error when it detects the already installed module. This Issue in particular is the cause I believe:
15:52 ZoffixW (no idea on the P6M merging stuff error)
15:52 Arrowhead joined #perl6
15:53 RabidGravy I think it tries to require it to see if it can or something
15:53 shicheng joined #perl6
15:55 Skarsnik nine, can you remove my PR? I think I will just delete my fork, recreate it and put the right files...
15:55 Woodi brrt: never libuv just works/compiles or some tricks are needed ?
15:55 Woodi hi today :)
15:55 Woodi newer* ?
15:56 brrt we've had funky failures on broken^W special platforms like os x
15:57 Skarsnik or wait, it's seem to be fine
16:01 ZoffixW m:​ixznet/c2d531f6c652dd3b88dd
16:01 camelia rakudo-moar d08e04: OUTPUT«(Hash)␤{:countries($[:name("U.S. Minor Outlying Islands"), :cities($["Baker Island", "Howland Island"])]), :offset("-12")}␤(Pair)␤Method 'elems' not found for invocant of class 'Mu'␤  in sub announce-hny at /tmp/cF3GfA0M1t line 11␤  in block <uni…»
16:01 ZoffixW Any idea why I'm getting that error? I can't reproduce on a smaller testcase:
16:01 ZoffixW m: my $x = {'countries' => [{'cities' => ['Bah'],'name' => 'Canada'},]}; for |$x<countries> -> $country { say $country<name>; say $country<cities>.elems }
16:01 camelia rakudo-moar d08e04: OUTPUT«Canada␤1␤»
16:02 RabidGravy something really, really didn't like trying to install (a fixed version of) C::Parser
16:03 RabidGravy ground the machine right down there for a minute
16:03 ZoffixW Or rather, I guess my question is, why the hell $country ends up a Pair, when in the small case it works fine
16:04 ZoffixW Ahhh
16:04 Skarsnik try append() instead of : ? x)
16:05 Arrowhead joined #perl6
16:05 Skarsnik I never make the : form of stuff like push to work x)
16:07 autarch is there a way of getting the AST for a piece of Perl 6 code? something like "my $ast = EVAL-AST('path/to/file.pm6')" ?
16:07 llfourn autarch: like QAST tree?
16:07 ZoffixW autarch, add --target=AST to perl6
16:07 autarch llfourn: I dunno, is that what I want?
16:08 autarch ZoffixW: I meant programmatically, I don't want a file, I want objects
16:08 llfourn yeah.... It's possible but it's not easy I think
16:08 perlpilot autarch: I don't think there is yet.
16:09 llfourn m: say $~MAIN.grammar.^name;
16:09 camelia rakudo-moar d08e04: OUTPUT«Perl6::Grammar␤»
16:09 perlpilot autarch: once masak's macro work comes to fruition, it should be not only possible, but easy-ish I would thinkg
16:09 ZoffixW k, found the issue. It's the one arg thing
16:09 autarch heh, two different answers
16:10 perlpilot autarch: whether mine is correct or not, it's the one that will keep you saner ;)
16:10 ZoffixW m: my $x = {'countries' => [{'cities' => [],'name' => 'Canada'}]}; for $x<countries> -> $country { say $country<name> }
16:10 camelia rakudo-moar d08e04: OUTPUT«Type Array does not support associative indexing.␤  in block <unit> at /tmp/MdA5gO4KGv line 1␤␤Actually thrown at:␤  in block <unit> at /tmp/MdA5gO4KGv line 1␤␤»
16:10 ZoffixW m: my $x = {'countries' => [{'cities' => [],'name' => 'Canada'},]}; for $x<countries> -> $country { say $country<name> }
16:10 camelia rakudo-moar d08e04: OUTPUT«Type Array does not support associative indexing.␤  in block <unit> at /tmp/W4maEnP9cO line 1␤␤Actually thrown at:␤  in block <unit> at /tmp/W4maEnP9cO line 1␤␤»
16:10 llfourn so you could try and .parse on that $~MAIN.grammar but it will not work out of the box you will have to init lots of $* variables and even then....
16:10 ZoffixW wut
16:10 autarch for what I'm doing now I can just use a hack of scanning line by line for use & require statements
16:10 FROGGS joined #perl6
16:11 FROGGS o/
16:11 llfourn m: say $~MAIN.grammar.parse('say "hello world"')
16:11 camelia rakudo-moar d08e04: OUTPUT«Could not locate compile-time value for symbol Block␤  in any find_symbol at src/Perl6/World.nqp line 3491␤  in any stub_code_object at src/Perl6/World.nqp line 1919␤  in any loading_and_symbol_setup at src/Perl6/World.nqp line 379␤  in any comp_un…»
16:11 ZoffixW Well, yeah, it's the one arg thing. I just fail at typing:​ixznet/991b1a645d711ee57884
16:11 timotimo llfourn: the grammar is not functional without a $*W
16:11 RabidGravy joined #perl6
16:11 timotimo m: BEGIN say $~MAIN.grammar.parse('say "hello world"')
16:11 camelia rakudo-moar d08e04: OUTPUT«5===SORRY!5=== Error while compiling /tmp/viidDhz3rn␤An exception occurred while evaluating a BEGIN␤at /tmp/viidDhz3rn:1␤Exception details:␤  5===SORRY!5=== Error while compiling ␤  Could not locate compile-time value for symbol Block␤ …»
16:12 timotimo mhm mhm
16:12 ccakes joined #perl6
16:12 RabidGravy well it probably needed a reboot anyway, don't know what that was all about
16:15 ZoffixW m: my $c = [{'cities' => ["Toronto"]}]; $c[0] ~~ Pair and $c[0] = %(@$c); for |$c { say $_<cities> }
16:15 camelia rakudo-moar d08e04: OUTPUT«[Toronto]␤»
16:15 ZoffixW m: my $c = [{'cities' => ["Toronto"]},{'cities' => ["NotToronto"]}]; $c[0] ~~ Pair and $c[0] = %(@$c); for |$c { say $_<cities> }
16:16 ZoffixW This is pretty ugly, but does the job, I guess
16:16 camelia rakudo-moar d08e04: OUTPUT«[Toronto]␤[NotToronto]␤»
16:16 sammers joined #perl6
16:18 itcharlie1 joined #perl6
16:19 shicheng joined #perl6
16:19 xxpor_lap joined #perl6
16:21 MadcapJake ZoffixW: the rakudobrew page says to build panda you type "rakudobrew build panda" so I just assumed that the `build-panda` command had changed
16:21 cdg joined #perl6
16:21 ZoffixW MadcapJake, oh. Maybe they both do the same thing
16:22 Arrowhead joined #perl6
16:22 pmurias joined #perl6
16:24 [Coke] .ask mojca if the problem with perl6 in the shebang is that you can't have a shell script as the shebang command?
16:24 yoleaux [Coke]: I'll pass your message to mojca.
16:24 MadcapJake now `panda install Task::Star` is failing on "JSON::Fast:ver<v0.3>:auth<>:api<> already installed` if it's already installed, why doesn't it skip and move on?
16:24 geekosaur [Coke], yeh, I pointed that out elsewhere (they were asking in #macports at the same time)
16:25 geekosaur I pointed out the chained-#! thing you ran into
16:25 MadcapJake weird, so there really must be something wrong with `build panda` because now after running `build-panda` it installs fine
16:26 geekosaur "build panda" would mean looking for a "panda" bakend for rakudo
16:26 geekosaur *backend
16:26 MadcapJake geekosaur: but it's given as the way to "use Panda" on rakudobrew's readme
16:27 ZoffixW Should this be rakudobugged? :
16:27 ZoffixW m: my @c = ( [{foo => 'bar'}␤,],);
16:27 camelia rakudo-moar d08e04: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Z9d5XsiVRu␤Bogus statement␤at /tmp/Z9d5XsiVRu:2␤------> 3my @c = ( [{foo => 'bar'}7⏏5<EOL>␤    expecting any of:␤        prefix␤        statement end␤        term␤»
16:27 geekosaur it wasn't the last time read it (which would be last Sunday)
16:27 ZoffixW m: my @c = ( [{foo => 'bar'},],);
16:27 camelia rakudo-moar d08e04: ( no output )
16:27 MadcapJake geekosaur:​bootstrapping-a-perl-6-implementation
16:27 perlpilot geekosaur: It changed yesterday.
16:27 geekosaur yug
16:27 perlpilot "build panda" and "build zef" do the same thing as "build-panda" and "build-zef" as I understand it.
16:28 geekosaur then I guess you needed to upgrde rakudobrew first
16:28 ZoffixW MadcapJake, but wasn't your issue installing Task::Star and not panda itself?
16:30 MadcapJake right, panda still worked, but now it works *and* it successfully installs things :)
16:30 MadcapJake geekosaur: I just cloned my rakudobrew an hour ago
16:33 ZoffixW Well, I reported it. Close it if the behaviour is the way it's expected to be:
16:34 Begi joined #perl6
16:34 ccakes joined #perl6
16:34 Skarsnik Time for some tea x)
16:37 [Coke] News flash from perl6-users; everyone still not on board that Perl 6 is Perl, and are willing to spend a lot of time talking about it. we should add a FAQ so we can point people to it.
16:39 shicheng joined #perl6
16:40 ZoffixW That's a confusing statement :P
16:40 MadcapJake Can you install p6doc with rakudobrew?
16:40 ZoffixW Perl === any('Perl 5', 'Perl 6'); # that's how I see it
16:40 llfourn autarch: if you wanted some fun you could mixin %*LANG<MAIN-actions> and then slurp EVAL the file you are interested in. It will give you a callback everytime use is parsed by the compiler.
16:40 llfourn though that will depend on your definition of fun
16:40 autarch llfourn: heh, that sounds like a lot more work than is needed to prototype what I'm working on
16:41 smls I think I found a bug in IO newline handling:
16:41 smls dd "\n\n\n".lines;    # ("", "", "").Seq
16:41 autarch I'm try to write a very minimal version of Dist::Zilla and i need to extract "use" and "require" statements to find a distros prereqs
16:41 smls spurt 'a', "\n\n\n";
16:41 smls dd 'a'.IO.lines;      # ("", "").Seq
16:41 ZoffixW MadcapJake, it's part of Task::Star:​Star/blob/master/
16:41 zakharyas joined #perl6
16:42 Skarsnik NativeCall: Consider adding the api version of the library you want to use, sub foo is native(mysqlclient, v1). Should I force mysqlclient v18 or add a way for NC to no warn if specified?
16:42 llfourn autarch: it is actually not that much work but it is kinda nasty because it uses nqp. Suit yourself :)
16:43 ZoffixW autarch, what's it called and where is it? (I'm planning on writing my own dist helper, and I'd love to steal some ideas :P)
16:43 autarch ZoffixW: I haven't pushed anything yet but I'm calling it Dist::Wocky
16:43 autarch it'll be under my github as perl6-Dist-Wocky once I have something to ush
16:44 Skarsnik hm, what is Dist::Zilla?
16:44 llfourn I wonder if it's possible to tell what a module has used though compunit interface...
16:44 Skarsnik dep are stored I think
16:44 Skarsnik so it must be possible
16:45 llfourn Skarsnik: if you know the variable name in UNIT:: that would be the best way
16:45 alpha123 joined #perl6
16:46 itaipu joined #perl6
16:47 mojca joined #perl6
16:47 sprocket joined #perl6
16:47 autarch Skarsnik: - basically it's a tool that helps you manage distros by turning a lot of boilerplate junk into config - like generating, generating certain types of distro tests (for example Test::Meta tests), automating the release process, etc.
16:47 _mg_ joined #perl6
16:47 MadcapJake ZoffixW: well I thought that Task::Star was finishing correctly (exits with 0) but there there is a "JSON::FASt already installed" error and there is no p6doc in my rakudobrew/bin folder
16:48 autarch as someone who maintains dozens of p5 distros I find it _very_ useful and I'm already getting annoyed trying to write files by hand with p6
16:48 Skarsnik I think mi6 was an attemps at that?
16:48 mojca what does "P6M Merging GLOBAL symbols failed: duplicate definition of symbol Panda" mean?
16:48 yoleaux 16:24Z <[Coke]> mojca: if the problem with perl6 in the shebang is that you can't have a shell script as the shebang command?
16:48 sprocket question about Promises - is it possible to have a Promise that’s executed lazily, so that the computation doesn’t start until it’s await’ed or .result is called?
16:48 mojca yoleaux, I believe so, yes
16:49 Tonik joined #perl6
16:49 ZoffixW MadcapJake, right. It aborts Task::Star install
16:49 ZoffixW Skarsnik, I consider it broken as it lacks any sensible docs
16:49 cygx mojca: yoleaux is a message bot - you should address [Coke]
16:49 wollmers joined #perl6
16:50 autarch Skarsnik: where is ml6?
16:50 mojca sorry; how exactly does the bot work? did I get an old message?
16:50 ZoffixW autarch,
16:50 Skarsnik
16:50 MadcapJake ZoffixW: any workaround available?
16:50 cygx .tell mojca Like this
16:50 yoleaux cygx: I'll pass your message to mojca.
16:50 autarch ah, that's an I not an L
16:50 mojca thanks
16:50 yoleaux 16:50Z <cygx> mojca: Like this
16:51 autarch mi6 looks more like - which does similar things but is basically unconfigurable
16:51 Skarsnik my main issue, if I did not get fixed is mi6 test run build x)
16:51 mojca so [Coke], about shebang, yes, I believe so
16:52 Arrowhead joined #perl6
16:52 Su-Shee which module is the HTTP::Tiny/Mojo::UA etc equivalent?
16:52 Skarsnik Http::ua?
16:52 lucasb joined #perl6
16:53 ZoffixW Su-Shee, I use HTTP::Tinyish  but at some point it was just shelling out to curl, though I think I saw some other handlers the other day
16:53 itcharlie joined #perl6
16:53 mojca there are some problems with Panda on Mac OS X 10.6
16:53 mojca
16:54 Su-Shee ZoffixW: ah. no, I want a proper thing. then I can just use curl.
16:54 ZoffixW *sigh* my New Years bot is annoying me with all problems I'm having that I can't repoduce in a small testcase
16:54 mojca but maybe it's just compiler's fault (too old compiler)
16:54 Skarsnik Su-Shee, what is your need exactly?
16:54 ZoffixW mojca, MadcapJake is having similar issue ATM
16:54 Su-Shee Skarsnik: async http client.
16:55 Skarsnik Oh
16:55 Su-Shee Skarsnik: at least http client. anything which is en par with HTTP::Tiny or Mojo::UA is perfectly fine however.
16:55 xinming_ joined #perl6
16:55 Skarsnik should not be hard to write aside http::ua, there is already all the base stuff
16:55 lucasb ZoffixW: newline after "}" ends the statement
16:55 ZoffixW IO::Socket::Async ^__^
16:55 ZoffixW MadcapJake, mojca and right now I'm running I'm running `panda --force install Task::Star` and it gotten further than before, but I see if it succeeds
16:56 Skarsnik My guess will be to look at http::useragent class code and change for an async arch
16:56 ZoffixW lucasb, thanks. I marked the ticket as rejected.
16:56 mojca so "Merging GLOBAL symbols failed" is basically the same or at least related to
16:56 Su-Shee Skarsnik: I should probably add that SSL is a must.
16:57 Skarsnik Maybe http::ua should be splitted?
16:57 lucasb ZoffixW: you're fast... but maybe I'm not so certain :)
16:57 Su-Shee ZoffixW: is it really shelling out to curl or is it a libcurl binding?
16:57 TimToady such many backlag!
16:58 dalek DBIish: 2b8ac56 | azawawi++ | examples/pg_arrays.p6:
16:58 dalek DBIish: Add initial example for PG array
16:58 dalek DBIish: review:
16:58 dalek DBIish: e803bf8 | azawawi++ | / (3 files):
16:58 dalek DBIish: merge pgarray step1
16:58 dalek DBIish: review:
16:58 dalek DBIish: a07a23b | azawawi++ | examples/parse_pg_array.p6:
16:58 lucasb I mean... maybe newline after "}" should end the stament *if* the statment *can* end there. otherwise it could go on, and accepts the syntax
16:58 ZoffixW Su-Shee, really shellying out:​ish/blob/master/lib/HTTP/Tinyish/Curl.pm6#L31
16:58 ZoffixW TimToady, that's a good thing!
16:58 Su-Shee ZoffixW: yeah ok, no, that's not acceptable.. :(
16:58 dalek joined #perl6
16:59 Skarsnik Lol I made dalek flood
16:59 shicheng joined #perl6
16:59 ZoffixW Is there any reason @messages .= sort: :by(*.chars); NOT sort stuff by length?
16:59 ZoffixW This line { @messages .= sort: :by(*.chars); .say for @messages[*-1].chars, @messages[*-9].chars; } gives me 55 139
16:59 [Coke] .seen allison
16:59 yoleaux I haven't seen allison around.
16:59 ZoffixW m: my @a = <foo foooo fo fooooooo>; @a .= sort: :by(*.chars); for @a { say '';  .say; .chars.say; } # and YET -_-
16:59 camelia rakudo-moar d08e04: OUTPUT«␤fo␤2␤␤foo␤3␤␤foooo␤5␤␤fooooooo␤8␤»
16:59 Skarsnik .= sort(*.chars) is not enought?
17:00 cygx mojca: shouldn't a shebang line ``#!/bin/sh /opt/local/bin/perl6`` work?
17:00 ZoffixW m: my @a = <foo foooo fo fooooooo>; @a .= sort: *.chars; for @a { say '';  .say; .chars.say; } # and YET -_-
17:00 camelia rakudo-moar d08e04: OUTPUT«␤fo␤2␤␤foo␤3␤␤foooo␤5␤␤fooooooo␤8␤»
17:00 ZoffixW oh
17:00 Skarsnik Su-Shee, I think if you want an async http::ua, look at http::useragant (since it has all the http class) and write something using async api with this
17:01 ZoffixW Skarsnik++ I guess it's it's because * in a :by creates a closure without a sig so it makes :by behave as in two-arg mode or something
17:01 nine b2gills: I feared as much :/ I will have a look if I can access sizeof(IV) at compile time.
17:01 mojca cygx, bingo!
17:01 mojca that works
17:02 Su-Shee Skarsnik: I mostly just want to write a useful toyscript with Perl6, not first implement stuff. let's see how far I get with HTTP::UserAgent with SSL and all.
17:02 Arrowhead joined #perl6
17:02 mojca I believe it's horrible to see that, but it works
17:02 * RabidGravy gives up on C::Parser *again* and goes in with a brute force solution *again*
17:02 ZoffixW Skarsnik, Su-Shee to note: I recall Ovid Poe having issues on that matter that starting a get *inside* a Promise or something like that. And coversation on IRC mentioned it was due to libuv doing things
17:02 Skarsnik Why you want to use C::Parser? x)
17:02 * ZoffixW cues mst's vague cat URL
17:03 [Coke] mojca: I'm pretty sure that's not a portable shebang, but if it works on OS X, awesome. (I just tried it here and got a failure...Unhandled exception: Missing or wrong version of dependency ')
17:03 Su-Shee ZoffixW: how is libuv coming into play with HTTP::UserAgent?!
17:03 Su-Shee ZoffixW: or are you trying to make me use Node? ;)
17:03 RabidGravy ZoffixW, wasn't it something odd to do with the URI module doing a require in a certain way, I think it was golfed to doing a require in a start
17:03 Skarsnik damn no new commit on​burel/smokegen/branch/clang
17:04 lizmat Su-Shee: if you use rakudo on MoarVM, then you *are* using libuv
17:04 V_S_C joined #perl6
17:04 Su-Shee interesting to know.
17:04 Su-Shee hm.
17:04 llfourn autarch: it seems that CompUnit::Repository has a .loaded method which stores the names of the modules it's loaded if that excites you
17:04 mojca [Coke], most likely it's not a portable way; env works; for anything else, it might help if perl6 was a binary
17:04 ZoffixW RabidGravy, ah, I must've missed that part of the convo :P
17:05 [Coke] mojca: I just did a fresh install of rakudo@2015.12, got a precompiled version, and it's barfing.
17:05 RabidGravy let me find the RT
17:05 [Coke] $ /opt/local/bin/perl6 -e 'say 3'
17:05 [Coke] Unhandled exception: Missing or wrong version of dependency 'gen/moar/stage2/QRegex.nqp' (from 'src/Perl6/Pod.nqp')
17:05 [Coke] Is there a way when installing a macport to say don't use precompiled?
17:06 [Coke] ah, -s
17:06 cygx [Coke]: have you nuked everything that might need nuking?
17:06 RabidGravy ZoffixW,
17:06 mojca sudo port -s install <package>
17:07 mojca but it would be *very* helpful to figure out the reason for the error
17:07 RabidGravy a workaround at the time was to  do *before* starting any threads
17:07 mojca can you please open a ticket before you loose all details?
17:07 harrison_ joined #perl6
17:08 mojca I'm unable to get a precompiled version (the buildbot for my os version is defunct at the moment)
17:08 [Coke] Sorry, I had already started trying to get a working version before I saw your request.
17:09 yqt joined #perl6
17:09 mojca anyway, you can still uninstall everything and install again to get the precompiled version
17:09 [Coke] typically that error is because there is a stale nqp version; I had inactive nqps & moarvms still installed.. I tried uninstalling them all, even the inactive ones.
17:09 Skarsnik anyone has though on?
17:09 cdg joined #perl6
17:09 ZoffixW RabidGravy, neither this code nor the one below it hang for me now:​lay.html?id=126390#txn-1371906
17:09 [Coke] ... and with that, it works fine, even with binary version.
17:09 [Coke] Is there a way to say that a particular distro needs the old version uninstalled?
17:09 RabidGravy ZoffixW cool
17:10 ZoffixW Tried it on two boxes 32bit (couple-days old perl) and 64bit (freshly squeezed)
17:10 [Coke] also: it's super awesome to have a precompiled version to install on a fresh mac box.
17:10 [Coke] mojca++
17:11 ZoffixW RabidGravy, oh, wait, is it an OSX bug only?
17:11 mojca [Coke], when we switch to a new version, the old one will be removed
17:11 mojca but MacPorts cannot touch anything that's not part of distribution
17:11 RabidGravy ZoffixW, no I could reproduce on Linux
17:11 ZoffixW RabidGravy, never mind, I see your note now
17:12 mojca so if you have your own perl6 installed, MacPorts might end up calling *your* perl6 during installation from source in the worst case
17:12 mojca and it certainly cannot uninstall components that you manually installed
17:13 xinming joined #perl6
17:14 Skarsnik RabidGravy, did you add leak issue with NC btw?
17:14 Skarsnik *had
17:14 nmp joined #perl6
17:15 RabidGravy Skarsnik, not that I noticed
17:17 Skarsnik I had a look the other day at Str returned by sqlite function and there was explictlymanaged (so the GC will not free them) but I think if you copy/clone the Str the role get cloned too
17:17 Skarsnik so you create unfreeable Str
17:19 Skarsnik It's cleared seeing in​rsnik/7f2fee97d56c5acc1587
17:19 harrison_ joined #perl6
17:19 nine mojca: the global symbol merging problem may be caused by loading a precompiled version of a module, then having a "use lib" somewhere and then loading modules again with this time a non-precompiled version of the earlier module
17:20 sprocket left #perl6
17:22 mojca nine, thanks; not that I understand though; when I tried to repeat "panda install Task::Star" today, it finished successfully
17:23 mojca so I'm not even able to reproduce the problem (unless maybe if I nuke the whole folder with libraries and try again from scratch)
17:23 mojca do you have any idea which package could be "guilty" of that error? Is there any way to get more verbose output next time?
17:24 mojca (when I will be able to reproduce the problem)
17:24 cygx is RAKUDO_MODULE_DEBUG=1 still a thing?
17:25 Su-Shee ZoffixW: well I can tell you that HTTP::UserAgent is .. strange. I'm reading for the first time in a decade for an http client the documentation.
17:25 Woodi mojca: when module is precompiled and then loaded it needs to put/merge itself (eg. global variable) into language so when something is loaded second time there is a conflict :)
17:26 diakopter RabidGravy: are you sure the new force default on panda MAIN doesn't break anything
17:26 Woodi nine: use lib  changes repository ?  but newly loaded module should load (replace) or don't load...
17:26 RabidGravy diakopter, no I'm not sure
17:27 [Coke] mojca: they weren't removed when I upgraded....
17:27 RabidGravy I actually suspect there's something weirder going on
17:27 RabidGravy so feel free to revert
17:27 [Coke] they were just inactivated.
17:27 mojca [Coke], what exactly did you have to manually remove then?
17:27 mojca ah, ok, but inactivation is fine
17:27 [Coke] I may be the only person who installed old versions of nqp or moar, though.
17:27 mojca you don't see those files
17:27 [Coke] I did 'sudo port uninstall nqp@2015.07' , e.g.
17:28 Arrowhead joined #perl6
17:28 mojca the only problem might be (I'm not sure if that is the case for perl) if moarvm would actually try to use some components from nqp, rakudo or panda
17:28 [Coke] I'm willing to ignore it for now; we should make sure there are no issues when 2016.01 comes out, though.
17:28 mojca [Coke], but unless nqp@2015.07 was actually installed *and* activated, you shouldn't have experienced any problems
17:29 mojca you would probably get the problem if you would manually upgrade moarvm, but forget to upgrade nqp/rakudo
17:29 [Coke] I get that. Just telling you what happened, what I did, and that it works now.
17:29 Woodi brrt: libuv-1.8 branch compiles and tests for me (nqp and rakudo), debian 32bit
17:30 mojca nevertheless I still don't understand how an old version could have interfered
17:30 [Coke] moarvm and nqp should force an install of the corresponding version of nqp & moarvm.
17:30 [Coke] ack. retry:
17:30 Woodi mojca: module loader is under developmant :)
17:30 [Coke] rakudo and nqp should force an install of the corresponding version of nqp & moarvm, respectively
17:30 [Coke] mojca: I'm not asking you to explain it, I'm just telling you what I did.
17:31 mojca [Coke]: I don't get exactly what you mean; that as soon as one installs moarvm 2016.01, one should also get nqp 2016.01?
17:32 mojca I'm almost sure that this doesn't work.
17:32 mojca But if users do "sudo port upgrade outdated", it should work as expected
17:32 mojca and if users try to upgrade just nqp, moarvm will be upgraded first
17:34 mojca Woodi: so I should simply stop worrying about the "P6M Merging GLOBAL symbols failed" error and assume it will get fixed?
17:34 mojca not that I can do much as I'm not able to reproduce it anyway
17:34 niceperl joined #perl6
17:35 ccakes joined #perl6
17:35 ZoffixW mojca, you can reproduce it with panda --force install Task::Star
17:35 mojca What I *am* still "worried" about is a better way to install panda in our package manager
17:35 cygx mojca: if it does happen again, you could try setting RAKUDO_MODULE_DEBUG=1
17:35 ZoffixW In particular, it fails when installing Linenoise module
17:36 hoelzro ZoffixW: is this related to that File::Find error that panda's been throwing? or is it something different?
17:36 ZoffixW 0.o
17:36 ZoffixW running panda install Linenoise did not show the error :/
17:36 hoelzro =/
17:36 ggoebel7 joined #perl6
17:36 mojca ZoffixW, no it no longer fails
17:36 ZoffixW Or rather RAKUDO_MODULE_DEBUG=1 panda --force install Linenoise
17:37 ZoffixW hoelzro, I think it's different. The File::Find error is just panda aborting the install the moment it encounters an installed module
17:37 ZoffixW (instead of just proceeding further)
17:38 stmuk .tell [ptc] yay updating .. BTW *.un~ is a vim undo file which my .vimrc has enabled
17:38 yoleaux stmuk: I'll pass your message to [ptc].
17:38 mojca ZoffixW, thanks. But I still get a successful run (with lots of output that would otherwise be helpful though)
17:39 ZoffixW mojca, I just nuked .rakudobrew and .perl6 and trying again
17:39 ZoffixW From scratch
17:40 mojca but I admit that I'm confused because I didn't change anything in my installation since yesterday when it was failing
17:40 mojca Linenoise got a few commits in the meantime though
17:45 mojca ZoffixW, which File::Find error are you talking about?
17:45 hoelzro I just made one commit to Linenoise to fix it last night
17:46 hoelzro it was working then =/
17:46 hoelzro ZoffixW: what's the failure look like?
17:48 ZoffixW hoelzro, "===SORRY!=== P6M Merging GLOBAL symbols failed: duplicate definition of symbol Panda"
17:48 hoelzro yikes
17:48 hoelzro that's (hopefully) not Linenoise's fault
17:48 hoelzro it's probably got something to do with
17:48 ZoffixW hoelzro, but then I re-ran panda --force install Linenoise and it installed fine, so I'm 99.999999999% sure it's the issue in panda
17:48 ZoffixW (or lower up the food chain)
17:48 hoelzro ZoffixW++ # shaking out bugs in tooling
17:49 mojca What's the role of "--norc" in lib/Panda/
17:49 mojca Is that a compiler flag?
17:49 dalek ecosystem/add-inline-ruby: a4086f7 | (Brock Wilcox)++ | META.list:
17:49 dalek ecosystem/add-inline-ruby: Add Inline::Ruby
17:49 dalek ecosystem/add-inline-ruby: review:​cosystem/commit/a4086f77a7
17:49 Skarsnik Inline::Ruby wow x)
17:49 ZoffixW oh, sweet
17:49 awwaiid it only barely works :)
17:49 MadcapJake ZoffixW: force installing Task::Star got farther than not forcing but crashes at p6doc "P6M Merging GLOBAL symbols failed: duplicate definition of symbol Panda
17:49 awwaiid but you can: EVAL 'puts "Hello!"', :from<Ruby>
17:50 ZoffixW MadcapJake, yeah, though running a lone panda --force install Linenoise works
17:50 MadcapJake yeah, is there a lone p6doc install?
17:50 ZoffixW panda --force install p6doc
17:51 TimToady lizmat: re 6d, no, we're not gonna switch to 6d till we're good and ready for Diwali; we can have 6.c.0 etc first if we really need intermediate language revisions; a feature-based approach also sounds sane, with the tradeoff of an infelicitous number of feature declarations that later want to be removed
17:52 lizmat TimToady: so I guess we should be moving to 6.c.0 asap then ?
17:52 ZoffixW mojca, this one​ixznet/e3b2d05d53414e5dbe43 but it's not a File::Find error but a panda error reported here:
17:52 TimToady let's get jnthn's input before we do anything rash
17:53 TimToady use experimental is already a step in the features direction
17:53 ellybelly_ joined #perl6
17:53 ellybelly joined #perl6
17:53 lizmat ok
17:54 TimToady in fact, I'm hard put to figure out how experimental isn't already a features pragma
17:55 TimToady in a sense, everything we put in between now and diwali is experimental anyway, and could get yanked out or modified
17:55 MadcapJake is there any way to bypass the reader that p6doc opens?
17:55 ccakes joined #perl6
17:56 ZoffixW awwaiid, how come it's a PR and not an actual addition to ecosystem?
17:56 * ZoffixW is ready to brag on Twitter ^_^
17:57 tonye joined #perl6
17:57 dalek rakudo/nom: 64c6e00 | (Sylvain Colinet)++ | / (2 files):
17:57 dalek rakudo/nom: NC: fix and add other sanity check
17:57 dalek rakudo/nom: review:
17:57 dalek rakudo/nom: b4be44c | lizmat++ | / (2 files):
17:57 dalek rakudo/nom: Merge pull request #664 from Skarsnik/nom
17:57 dalek rakudo/nom:
17:57 dalek rakudo/nom: NC: fix and add other sanity check
17:57 dalek rakudo/nom: review:
17:58 awwaiid ZoffixW: *shrug*. Here, I'll merge
17:58 Zero_Dogg are there any other database (mysql) interfaces than DBIish? DBIish is segfaulting on my pi
17:58 dalek ecosystem: a4086f7 | (Brock Wilcox)++ | META.list:
17:58 dalek ecosystem: Add Inline::Ruby
17:58 dalek ecosystem: review:​cosystem/commit/a4086f77a7
17:58 dalek ecosystem: 9f900eb | (Brock Wilcox)++ | META.list:
17:58 dalek ecosystem: Merge pull request #107 from perl6/add-inline-ruby
17:58 dalek ecosystem:
17:58 dalek ecosystem: Add Inline::Ruby
17:59 dalek ecosystem: review:​cosystem/commit/9f900ebe4d
17:59 itcharlie1 joined #perl6
17:59 ZoffixW awwaiid++
17:59 awwaiid ZoffixW: I don't know that it works on any machine but mine
18:00 ZoffixW awwaiid, it's a start :)
18:00 ZoffixW \o/​/status/682259926261997568
18:00 Skarsnik Zero_Dogg, there is sqlite and Pg?
18:01 Zero_Dogg Skarsnik: hehe, true, but I need to talk to an existing mysql-based app
18:01 [Coke] TimToady: (everything between now and diwali is experimental) but we can't rely on it being implicitly experimental in a 6.c release. has to be declared experimental, I think.
18:01 Ven joined #perl6
18:02 shicheng joined #perl6
18:02 [Coke] where 6.c release is shorcut for "a rakudo release targeting 6.c"
18:02 TimToady well, there are probably gradations; some things are obviously just bug fixes
18:03 TimToady like if I fix the loop outer context thing, that's not experimental, it's just something shoulda worked in the first place
18:03 [Coke] oh, sure. it's just going to be on us to actually review everything before the release to make sure we're doing the right thing. (whatever that is)
18:04 [Coke] torturing releng on behalf of the users.
18:04 sprocket joined #perl6
18:04 TimToady we've already tortured the users about as much as we can in one month :)
18:05 lizmat so, generalizing IO::ArgFiles functionality into IO::CatPath/CatHandle, in which category does that fall if there are no spectest failures
18:05 TimToady though fortunately it's still mostly early adopters and pioneers getting burned :)
18:05 awwaiid TimToady: the month rolls over soon and we can ramp up the torture again
18:06 nine yeah torture!
18:06 ZoffixW no moar torture :o
18:06 lizmat so you can do <a b c>.IO.lines to get the lines of files a, b and c in one loop?
18:06 diakopter bring on the pain
18:06 dalek rakudo/nom: c268635 | coke++ | docs/release_guide.pod:
18:06 dalek rakudo/nom: Make sure we're testing Christmas too
18:06 dalek rakudo/nom: review:
18:06 shicheng joined #perl6
18:06 nine b2gills: thanks for the hint about your 64 bit IV on 32 bit. It prompted me to implement a proper solution and it's much easier than I feared :)
18:06 itcharlie1 left #perl6
18:07 perlpilot lizmat: "experimental feature" until there are spectests :)
18:07 Skarsnik nine, what did you do? I am curious
18:07 awwaiid lizmat: maybe <a b c>.>>IO.>>lines ?
18:07 ZoffixW Is there a more elegant way to write this code? I thought there was something with |c and things:      method irc-privmsg-me ($irc, $e) { me($irc, $e); }
18:07 Skarsnik Zero_Dogg, do you have an output with the segfault?
18:08 perlpilot btw, awwaiid++  (Inline::Ruby)
18:08 Skarsnik Zero_Dogg, and you can run perl6-gbd-m I think, to have more info
18:08 shicheng joined #perl6
18:08 lizmat awwaiid: that would work for lines, but not for .words or .comb or .split
18:08 nine Skarsnik: will push a commit soon, but have to run for groceries before the shop closes
18:10 TimToady the real question to my mind is whether we need to split into an experimental dev track and a bug-fix-only maintenance track at some point
18:10 shicheng joined #perl6
18:12 shicheng joined #perl6
18:12 ZoffixW mojca, I can still reproduce the Linenoise panda failure with this set of commands:​ixznet/72a98c866a54e91e8207
18:12 ZoffixW mojca, here's the output of the entire run:​ixznet/6a6f12e3b5139e4daead
18:12 MadcapJake why does the p6doc file in /home/jrusso/.rakudobrew/moar-n​om/install/share/perl6/site/bin completely different from the p6doc file in the github repo?
18:13 Zero_Dogg Skarsnik: it's segfaulting on mysql_send_query according to the backtrace
18:13 [Coke] TimToady: ayup
18:13 hoelzro MadcapJake: is the installed one a wrapper that uses resources to find the actual script?
18:13 ZoffixW mojca, this is on a Debian GNU/Linux 7.9 (wheezy) 4.1.5-x86_64-linode61 #7 SMP Mon Aug 24 13:46:31 EDT 2015 x86_64 GNU/Linux
18:13 [Coke] I think it's just a question of when.
18:13 perlpilot TimToady: what [Coke] said.
18:13 shicheng joined #perl6
18:13 * TimToady should go reread jnthn's doc, however
18:14 Skarsnik Zero_Dogg, can you paste the full backtrace?
18:14 MadcapJake hoelzro: where is the one from the repo then? `locate p6doc` only gives me the wrapper script
18:14 Zero_Dogg Skarsnik: sure, give me a minute
18:14 hoelzro MadcapJake: it's probably got a SHA as its name
18:14 [Coke] TimToady: if we have a plan by the end of the first week of january, we're fine.
18:14 hoelzro MadcapJake: look for eaf6b11d202a0b88de9ae481b28b57019e49bbed, case insensitive
18:14 [Coke] (we'll be fine))
18:15 rindolf joined #perl6
18:15 TimToady .oO(we need an epiphany by Epiphany...)
18:15 Skarsnik a week in the middle of holiday seems not enought x)
18:15 rindolf TimToady: which Epiphany?
18:15 Zero_Dogg Skarsnik:
18:16 MadcapJake hoelzro: not finding that, if this isn't my preferred route for trying to hack on p6doc, then can panda install a local module? or would I have to install it myself?
18:16 TimToady whosever is most convenient, obviously
18:17 TimToady and why is the word "whosever" missing from English, anyway?
18:18 Skarsnik Zero_Dogg, that's instering there not send_query nor real_query in dbish binding for mysqlclient x)
18:19 hoelzro MadcapJake: if you open up /home/jrusso/.rakudobrew/moar-nom/i​nstall/share/perl6/site/bin/p6doc, what's there?
18:19 hoelzro panda is capable of installing local modules, if that's what you want
18:19 Zero_Dogg Skarsnik: strange
18:19 Ox0dea joined #perl6
18:20 MadcapJake hoelzro: that's the wrapper script on my machine
18:20 Ox0dea m: say True.succ # Why no Boolean ring?
18:20 camelia rakudo-moar c26863: OUTPUT«True␤»
18:20 hoelzro I figured =/
18:20 domidumont joined #perl6
18:20 MadcapJake hoelzro: yeah i want to try and add an option to p6doc to print to stdout instead of wrapping in a pager
18:20 Zero_Dogg Skarsnik: all I'm doing is preparing a select and then execute()ing it, and on the execute the segfault is triggered
18:21 shicheng joined #perl6
18:21 ZoffixW TimToady, it's not :)
18:21 hoelzro I see; you should probably stick with doing a local panda install
18:21 hoelzro you can do panda install /path/to/module
18:21 Skarsnik Zero_Dogg, I will look after dinner x)
18:21 Zero_Dogg Skarsnik: almost all mysql tests are failing as well, so I'm guessing it's having trouble with arm
18:21 Zero_Dogg Skarsnik: hehe, goodie
18:22 Ox0dea ZoffixW: What'd I miss?
18:22 MadcapJake hoelzro: cool, thanks! I'll give that a go!
18:22 ZoffixW Ox0dea, <TimToady> and why is the word "whosever" missing from English, anyway?
18:22 Ox0dea Ah.
18:22 * TimToady fires his dictionary
18:22 Ox0dea I still which we had a better possessive form for "which".
18:22 ZoffixW :P
18:22 Ox0dea *still wish
18:22 Skarsnik Zero_Dogg, test for mysql need a root access without password, there is no way they could work on a real install, Don't worry :)
18:22 Ox0dea "Whose" for non-people is weird, and "which's" looks funny.
18:23 Zero_Dogg Skarsnik: ah, that explains that at least
18:23 dalek rakudo/nom: 630a9b2 | lizmat++ | / (9 files):
18:23 dalek rakudo/nom: Add IO::CatPath/CatHandle, remove IO::ArgFiles
18:23 dalek rakudo/nom:
18:23 dalek rakudo/nom: Basically generalize and expand the functionality of IO::ArgFiles
18:23 dalek rakudo/nom: into a generic IO::CatPath and IO::CatHandle class.  Features:
18:23 dalek rakudo/nom:
18:23 dalek rakudo/nom: - $*ARGFILES is initialized using IO::CatHandle the same way it was
18:23 dalek rakudo/nom:   before (assume strings are file names, use $*IN if no strings given)
18:23 dalek rakudo/nom: - apart from .lines, can now also do .words, .comb, .split, and
18:23 dalek rakudo/nom:   .Supply or any other IO::Handle method that relies on .read or
18:23 dalek rakudo/nom:   .readchars.
18:23 dalek rakudo/nom: - List.IO creates IO::CatPath, for easy streaming of multiple files.
18:23 dalek rakudo/nom:
18:23 dalek rakudo/nom: Done as a single commit, that can be reverted/reapplied/trashed once
18:23 dalek rakudo/nom: we figure out how we're going to do new features and we reached a
18:23 dalek rakudo/nom: decision on this.  FWIW, I intended this to happen before 6.c, but
18:23 dalek rakudo/nom: LPW Keynote and ~10 days of illness made this impossible  :-(
18:23 dalek rakudo/nom: review:
18:23 silasmariner joined #perl6
18:24 TimToady Ox0dea: re Boolean rings, it was deemed to be more useful to have an idempotent $flag++ and $flag-- to turn $flag true or false, respectively
18:24 silasmariner m: say ([+] map {$_ ** $_}, 1..1000).comb[*-10..*-1].join  ;   say ([+] map * ** *, 1..1000).comb[*-10..*-1].join
18:24 camelia rakudo-moar c26863: OUTPUT«9110846700␤6211390700␤»
18:24 TimToady and no other enums guarantee any kind of modularity
18:24 rindolf joined #perl6
18:25 TimToady though I suppose traits could happen
18:25 silasmariner shouldn't the whatever-stars give the same as the $_'s there?
18:25 virtualsue joined #perl6
18:25 TimToady whatever is not equivalent to $_, in general
18:25 dalek roast: d042e87 | lizmat++ | S02-types/WHICH.t:
18:25 dalek roast: Remove IO::ArgFiles now, until decision on 630a9b2
18:25 dalek roast: review:
18:25 TimToady unless you change the number of {}
18:26 silasmariner TimToady: oh right. I'll go read up more on that
18:26 ZoffixW silasmariner, no, because the second WhateverStar is not the same as first
18:26 TimToady and...that too
18:26 silasmariner tnx
18:26 ZoffixW silasmariner, in the first case, the first iteration is 1 ** 1, in the second case it's 1 ** 2
18:26 ZoffixW s/iteration/example above/;
18:26 silasmariner gotcha. It sorta iterates, then
18:26 ZoffixW err
18:26 TimToady whatever stars are more like { $^a ** $^b }
18:26 ZoffixW s/example above/iteration/; s/case/example above/; -_-
18:27 Ox0dea TimToady: Exactly equivalent to, no?
18:27 TimToady depends on the operator
18:27 TimToady for **, sure
18:27 TimToady they're not completely interchangeable
18:28 ugexe whoever was having problems with panda needing to run install twice to work: its because the CU stuff passes around a cached file object instead of a path string
18:28 Ox0dea m: say map { $_ + $_ }, [1,2,3,4]
18:28 camelia rakudo-moar c26863: OUTPUT«(2 4 6 8)␤»
18:28 TimToady *..* doesn't mean { $^a .. $^b } ferinstins
18:28 Ox0dea m: say map * + * , [1,2,3,4] # silasmariner
18:28 camelia rakudo-moar c26863: OUTPUT«(3 7)␤»
18:28 TimToady so if *..* is a koala, and ^..^ is a fox, what is *^..^* ?
18:29 silasmariner A bat?
18:29 TimToady I can see that...
18:29 ZoffixW Why does this say "Variable '$channel' is not declared. Did you mean 'Channel'?"  I got it in the sig!​ixznet/0eab66da3da94957c802
18:29 perlpilot a fox with ear muffs
18:29 TimToady a foxbat?
18:29 ZoffixW TimToady, looks like a pig to me :P
18:30 silasmariner foxbatpig
18:30 perlpilot ZoffixW: you've got syntax error in the sig
18:30 Ox0dea It's Pikachu!
18:31 ZoffixW perlpilot, ah. I thought the comma-less named args could be used anywhere
18:31 lnrdo joined #perl6
18:32 ugexe is there a way we could make IO::Path with an option to not cache specific instances?
18:32 RabidGravy can I define a CArray of a certain length in a CUnion?  somethings telling me no, but I thought I'd ask
18:33 captain-adequate joined #perl6
18:33 ZoffixW m: say :2015years :12months :31days :23hours :59minutes :59seconds
18:33 camelia rakudo-moar 630a9b: OUTPUT«Cannot make a DateTime object using .new␤  in block <unit> at /tmp/g5VJ2Cbpb4 line 1␤␤Actually thrown at:␤  in block <unit> at /tmp/g5VJ2Cbpb4 line 1␤␤»
18:33 ugexe other than stringifying it on a per-use basis that is
18:33 ZoffixW say :2015year :12month :31day :23hour :59minute :59second
18:33 ZoffixW well... the latter one works. LTA error
18:34 ZoffixW Hm
18:35 ZoffixW I guess it can be ignored, but my concern is .later *does* take plurals as args, which is where my confusion came from
18:35 perlpilot speaking of LTA ... If I say "use DateTime;", why does rakudo bark at me?   Isn't it the same as if I'd said "use Test; use Test;"?  (i.e. it's already there, so using it again is a no-op)
18:36 ZoffixW perlpilot, Test is a module. DateTime is a type
18:36 aigan joined #perl6
18:36 lizmat ZoffixW: good catch
18:36 awwaiid Is there a difference between Pointer and OpaquePointer?
18:36 ZoffixW m: say :2015year :12months :31days :23hours :59minutes :59seconds
18:36 camelia rakudo-moar 630a9b: OUTPUT«2015-01-01T00:00:00Z␤»
18:37 ZoffixW lizmat, ^ that's even worse :P
18:37 perlpilot ZoffixW: for the sake of argument, DateTime doesn't *have* to be built-in.  It *could* be a module that is used.   Why not make the built-ins behave the same as other modules?
18:37 lizmat you mean, the lack of comma's makes the rest of the named disappear ?
18:38 yurivish_ joined #perl6
18:38 ZoffixW lizmat, no, the 'year' is singular and the rest are plurals. What .new sees is I provided a single arg—year—because it's not watching for plural args
18:38 lizmat ah, ok, gotcha
18:39 lizmat m: say :2015year :12month :31days :23hours :59minutes :59seconds
18:39 camelia rakudo-moar 630a9b: OUTPUT«2015-12-01T00:00:00Z␤»
18:39 lizmat ok
18:39 * lizmat is fixing
18:39 AlexDaniel m: my $link = ‘test’; my $bar = ‘ZZZ’; $link ~~ s{ $ } = $bar; say $link
18:39 camelia rakudo-moar 630a9b: OUTPUT«WARNINGS for /tmp/WfNIQ8CpRP:␤Useless use of $bar in sink context (line 1)␤testZZZ␤»
18:39 ZoffixW lizmat++
18:39 AlexDaniel TimToady: ↑ useless “useless use of” warning
18:39 regreg joined #perl6
18:40 ZoffixW perlpilot, I could counter that argument with: having a module named DateTime conflicts with type DateTime... What's called when is called?
18:40 TimToady hmm
18:40 ZoffixW perlpilot, the barkage is a recent addition, I think if you backlog like 2-20 hours back, you'll see a discussion on this
18:42 pierre-vigier joined #perl6
18:43 Arrowhead joined #perl6
18:43 bbkr Skarsnik: your "Meh fix" in NC works :)
18:43 pochi m: my $a = method () {}; $a.set_name("bar"); $a.^methods;
18:43 camelia rakudo-moar 630a9b: ( no output )
18:43 pochi m: my $a = method () {}; $a.set_name("bar"); say $a.^methods;
18:43 camelia rakudo-moar 630a9b: OUTPUT«(gist <anon> <anon> <anon> soft <anon> <anon> yada perl <anon> onlystar candidates unwrap wrap <anon> <anon> package leave <anon> <anon> cando <anon> <anon> <anon> <anon> multi <anon> <anon> add_phaser phasers assuming WHY set_why perl of <anon> returns fi…»
18:44 zed00 joined #perl6
18:44 TimToady AlexDaniel: I was pasting that to the command line, and noticed the single quotes just as I was hitting return, so expected it to parsefail, but it didn't, because smartquotes++
18:45 dalek Inline-Perl5: a3f7a29 | (Stefan Seifert)++ | / (2 files):
18:45 dalek Inline-Perl5: Remove remaining difference of types between C and Perl 6 code
18:45 dalek Inline-Perl5:
18:45 dalek Inline-Perl5: We now determine at compile time the size of a Perl 5 IV in C code and use
18:45 dalek Inline-Perl5: the corresponding int32/int64.
18:45 dalek Inline-Perl5: This should finally fix all 32 bit issues. If your IV is 16 bits, please just
18:45 dalek Inline-Perl5: go away. If your IV is 128 bits, remember that donating such a machine to me
18:45 dalek Inline-Perl5: may improve your chances for support!
18:45 dalek Inline-Perl5: review:​ine-Perl5/commit/a3f7a29037
18:45 nine Skarsnik: ^^^
18:45 ZoffixW nine++ nice commit message :P
18:45 AlexDaniel TimToady: yeah, that's one of the reasons I use these quotes
18:46 lucasb joined #perl6
18:46 FritzZaucker joined #perl6
18:46 El_Che nine: great commit msg ;)
18:47 ZoffixW El_Che, do you play Warframe? I could've swore I played with someone with your nick name the other day
18:47 * flussence has a bashrc-updating command that uses those quotes to horribly effective effect:
18:47 flussence ./perl6 -e 'put “PATH="{$*VM.config<bindir>}:{$*REPO.repo-cha​in[1].writeable-path.child(‘bin’)}:\$PATH"”'
18:47 AlexDaniel TimToady: well, let's just hope that shells will not support unicode quotes ever :D
18:47 AlexDaniel though ‘ ‘ ’ ’ case should work fine if they do that correctly…
18:48 AlexDaniel m: say “hello “world””
18:48 camelia rakudo-moar 630a9b: OUTPUT«hello “world”␤»
18:48 hankache joined #perl6
18:49 lucasb m: @*ARGS = '/dev/null'; say $*ARGFILES.lines(5)
18:49 camelia rakudo-moar 630a9b: OUTPUT«(Nil Nil Nil Nil Nil)␤»
18:49 ZoffixW That's crazy
18:49 FritzZaucker I'll try again: The following NCurses example is working:​ucker/aa14025a16edcd671e25 whereas a similar one using a supply doesn't properly display the text:​ucker/0290915997b1ddc3ad53 after a few iterations.
18:49 ZoffixW m: say "foo "foo""
18:49 camelia rakudo-moar 630a9b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/frKLqc6rfP␤Two terms in a row␤at /tmp/frKLqc6rfP:1␤------> 3say "foo "7⏏5foo""␤    expecting any of:␤        infix␤        infix stopper␤        postfix␤        statement end␤        sta…»
18:49 ZoffixW :(
18:49 El_Che ZoffixW: nope. I played 1998 commandos yesterday though ;)
18:49 ZoffixW Must've been someone else then :)
18:49 dalek Inline-Perl5: e0ae715 | (Stefan Seifert)++ |
18:49 dalek Inline-Perl5: Version 0.2
18:49 dalek Inline-Perl5:
18:49 dalek Inline-Perl5: Raising version to make it easier for people to upgrade.
18:49 dalek Inline-Perl5: review:​ine-Perl5/commit/e0ae715dc0
18:50 AlexDaniel ZoffixW: crazy? Well :)
18:50 lizmat m: say :2015year :12months :31days   # ZoffixW: same issue
18:50 camelia rakudo-moar 630a9b: OUTPUT«2015-01-01␤»
18:50 yurivish_ is there a simple way to turn a list (1, 2, 3) into a hash (1 => True, 2 => True, 3 => True)?
18:50 ZoffixW m: say :2015year ).later: :12months :31days
18:50 camelia rakudo-moar 630a9b: OUTPUT«More than one time unit supplied␤  in block <unit> at /tmp/0awjr1sxRH line 1␤␤»
18:50 ZoffixW m: say :2015year ).later: :12months
18:50 camelia rakudo-moar 630a9b: OUTPUT«2016-01-01␤»
18:50 ZoffixW wat
18:50 timotimo yurivish_: many ways. among them you can use (1, 2, 3) X=> True
18:50 ZoffixW oh right
18:51 AlexDaniel m: say %{(1, 2, 3) Z True}
18:51 camelia rakudo-moar 630a9b: OUTPUT«Odd number of elements found where hash initializer expected␤  in block <unit> at /tmp/LUasxrayho line 1␤␤»
18:51 Skarsnik nine, nice :)
18:51 Ox0dea left #perl6
18:51 yurivish_ timotimo: oh, I didn't think of that. cute. I was trying to assign into a hash variable, but that pairs the items up
18:51 lizmat m: dd (1,2,3).Set
18:51 camelia rakudo-moar 630a9b: OUTPUT«set(3,1,2)␤»
18:51 lizmat yurivish_: sure you don't want Set ?  ^^^
18:51 yurivish_ and set is actually what I want in this case, yeah :)
18:52 ugexe m: my @x; @x>>.Int; say @x>><xxx>
18:52 camelia rakudo-moar 630a9b: OUTPUT«()␤»
18:52 masak m: dd set 1, 2, 3
18:52 camelia rakudo-moar 630a9b: OUTPUT«set(3,1,2)␤»
18:52 ugexe m: my @x; say @x>><xxx>
18:52 camelia rakudo-moar 630a9b: OUTPUT«This type does not support elems␤  in block <unit> at /tmp/XSFl1r7PDw line 1␤␤»
18:52 masak m: dd set <1 2 3>
18:52 camelia rakudo-moar 630a9b: OUTPUT«set(, "1"),, "2"),, "3"))␤»
18:52 yurivish_ what is dd?
18:52 ugexe the first example does not error, but the second does
18:52 lizmat yurivish_: tiny data dumper, so small, it even lost the t
18:52 ZoffixW yurivish_,
18:53 nine awwaiid: left some comments on your Inline::Python pull request
18:53 yurivish_ that looks incredibly useful
18:53 AlexDaniel m: say 1, 2, 3 X=> True
18:53 camelia rakudo-moar 630a9b: OUTPUT«(1 => True 2 => True 3 => True)␤»
18:53 Skarsnik nine, ah now the test fail with real output
18:53 AlexDaniel m: say 1, 2, 3 X<= True # haha
18:53 camelia rakudo-moar 630a9b: OUTPUT«(True False False)␤»
18:53 AlexDaniel m: say 1, 2, 3 X<=> True # haha
18:53 camelia rakudo-moar 630a9b: OUTPUT«(Same More More)␤»
18:54 Skarsnik nine,
18:54 yurivish_ also, what's the best way to look up documentation on things like operators, which are hard to search for?
18:55 AlexDaniel yurivish_:
18:55 AlexDaniel yurivish_: there's a search form on top right
18:55 ZoffixW yurivish_, I usually consult but I'm unsure it covers all of them
18:55 ZoffixW The search is broken for many ops
18:55 yurivish_ searching for X seems a bit broken
18:55 ZoffixW
18:55 timotimo well, it's a metaop, you see! :P
18:55 yurivish_ (i think it's cartesian-product, but wanted to check)
18:56 timotimo but yeah ...
18:56 El_Che doc is pretty good, but a work in progress. Once you fin
18:56 AlexDaniel yurivish_: it is not broken, you just have to scroll down a bit… :)
18:56 yurivish_ m: (1, 2) X (3, 4)
18:56 camelia rakudo-moar 630a9b: OUTPUT«WARNINGS for /tmp/YwQzsrWgfV:␤Useless use of "X" in expression "(1, 2) X (3, 4)" in sink context (line 1)␤»
18:56 El_Che d something missing: PR
18:56 ccakes joined #perl6
18:56 yurivish_ m: dd (1, 2) X (3, 4)
18:56 camelia rakudo-moar 630a9b: OUTPUT«((1, 3), (1, 4), (2, 3), (2, 4)).Seq␤»
18:56 timotimo well, PR or put it into TODO or MISSING
18:56 ZoffixW yurivish_, what's the breakage? searching for X does give me a ton of results, but I do see it listed in the "Infix" section, linking to
18:56 yurivish_ m: dd (1, 2) X (3, 4, 5)
18:56 camelia rakudo-moar 630a9b: OUTPUT«((1, 3), (1, 4), (1, 5), (2, 3), (2, 4), (2, 5)).Seq␤»
18:57 timotimo m: dd <a b c> X~ <1 2 3>
18:57 camelia rakudo-moar 630a9b: OUTPUT«("a1", "a2", "a3", "b1", "b2", "b3", "c1", "c2", "c3").Seq␤»
18:57 awwaiid nine: awesomesauce, thanks! Comments addressed :)
18:57 ZoffixW m: dd <a b c> X~ <1 2>
18:57 camelia rakudo-moar 630a9b: OUTPUT«("a1", "a2", "b1", "b2", "c1", "c2").Seq␤»
18:57 ZoffixW damn that's cool
18:57 yurivish_ ZoffixW: pressing X<enter> doesn't seem to do anything, and sometimes the menu doesn't come up (though I think that's because I'm on slow wifi), though there's no loading indicator or any indication something is happening
18:59 autarch it seems like the module loader _really_ doesn't like circular "use" - as in endless recursion
18:59 kittenlips joined #perl6
18:59 ZoffixW yurivish_, thanks. Created an issue for that. Might try to fix it sometime during New Years weekend
19:00 autarch and it's essentially a fork bomb (because of precomp being done in a child proc?)
19:00 yurivish_ awesome, thanks!
19:00 ZoffixW autarch, that should be detected and error shown. That was fixed $x time ago
19:00 ZoffixW Maybe a regression
19:00 AlexDaniel yurivish_: search functionality is offline, if it does not show up then it is a bug
19:00 yurivish_ something good is coming of this slow wifi after all. I'm on a train.
19:00 lucasb m: say +$*ARGFILES.lines(100)
19:00 camelia rakudo-moar 630a9b: OUTPUT«100␤»
19:00 AlexDaniel or am I wrong?
19:00 lucasb star: say +$*ARGFILES.lines(100)
19:00 camelia star-m 2015.09: OUTPUT«20␤»
19:01 dalek roast: 991ff61 | lizmat++ | S32-temporal/DateTime.t:
19:01 dalek roast: Remove warning
19:01 dalek roast: review:
19:02 autarch maybe it's time for me to try to hack on rakudo a bit
19:02 AlexDaniel nah, I'm not wrong this time. This file contains everything:
19:02 ZoffixW AlexDaniel, JS file loads after the page, so on slow conn it's possible to type "X" into search box and hit ENTER before it finished downloading and made the searchbox work. I think that's the issue at hand.
19:02 AlexDaniel but if it is not loaded then you are going to see a loading icon in your browser
19:02 ZoffixW AlexDaniel, loading icon where?
19:02 lucasb lizmat: ^^ any idea on what should be the behavior from now on?
19:02 ZoffixW AlexDaniel, like page loading thing?
19:03 AlexDaniel ZoffixW: yes
19:03 lizmat lucasb: there should be no change in $*ARGFILES behaviour
19:03 ZoffixW I don't usually watch for it or wait for it to disappear before using the page :)
19:03 nine Skarsnik: looks like some string issues
19:03 lizmat it's just blessed into the IO::CatHandle class, instead of IO::ArgFiles
19:04 ZoffixW aktually
19:04 yurivish_ is the idioamtic way to check if a set contains an element $el eq any($set)?
19:04 lucasb lizmat: what about the 20 vs 100 lines output above?
19:04 ZoffixW AlexDaniel, you bring up a good point that we don't need a loading icon *while* searching...
19:04 lizmat star: say $*IN.lines
19:04 camelia star-m 2015.09: OUTPUT«(Céad slán ag sléibhte maorga Chontae Dhún na nGall Agus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll; Nuair a ghluais mise thart le Loch Dhún Lúich’ go ciúin sa ghleann ina luí I mo dhiaidh bhí gleanntáin ghlas’ Ghaoth Dob…»
19:05 autarch so what's the work around if I have two classes that require each other?
19:05 ZoffixW yurivish_, yeah
19:05 lizmat m: say $*IN.lines
19:05 camelia rakudo-moar 630a9b: OUTPUT«(Céad slán ag sléibhte maorga Chontae Dhún na nGall Agus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll; Nuair a ghluais mise thart le Loch Dhún Lúich’ go ciúin sa ghleann ina luí I mo dhiaidh bhí gleanntáin ghlas’ Ghaoth…»
19:05 ZoffixW m: my @a = ^4; say 2 ~~ any @a
19:05 camelia rakudo-moar 630a9b: OUTPUT«True␤»
19:05 autarch like Foo has a Bar attr and Bar has a Foo attr
19:05 uvtc_ joined #perl6
19:05 cygx yurivish_:
19:05 lizmat star: say $*IN.lines.elems
19:05 camelia star-m 2015.09: OUTPUT«20␤»
19:05 lizmat m: say $*IN.lines.elems
19:05 camelia rakudo-moar 630a9b: OUTPUT«20␤»
19:06 ZoffixW autarch, none that I know of. Last I needed to, the proposal was to use a   class Foo { ... }  stub. But that doesn't always work and complains that I didn't define the stub :(
19:06 lizmat lucasb: looking into it
19:06 AlexDaniel ZoffixW: we don't need a search icon before searching as well. “search.js” is loaded with the rest of the page
19:06 autarch oh, that's quite frustrating
19:06 lucasb lizmat: thanks :)
19:06 yurivish_ cygx: that's awesome. I knew Julia used the unicode "in" operator but forgot all about it
19:06 uvtc_ Fresh rakudobrew install on Debian Testing. Unable to `panda install Readline`.
19:06 uvtc_ "In 'remove_history' routine declaration - Not an accepted NativeCall type for parameter [1]  : Int"
19:06 AlexDaniel ZoffixW: so if you see that the page is still loading, well, you probably expect things not to work
19:06 dalek rakudo/nom: 3dc7d55 | TimToady++ | src/Perl6/Actions.nqp:
19:06 dalek rakudo/nom: RHS of subst needs to be wanted harder
19:06 dalek rakudo/nom: review:
19:07 TimToady AlexDaniel++
19:07 dalek Inline-Python: 22b5fc9 | (Brock Wilcox)++ | / (3 files):
19:07 dalek Inline-Python: Teach EVAL how to :from<Python>
19:07 dalek Inline-Python:
19:07 dalek Inline-Python: Plus a bonus .gitignore
19:07 dalek Inline-Python: review:​ine-Python/commit/22b5fc92bd
19:07 AlexDaniel TimToady: nice! Did not expect it to be fixed so fast :)
19:07 cygx .oO( where's the script that replaces all Texas variants of operators with the unicode onces? )
19:07 nine awwaiid: ^^^ rebased a little
19:08 TimToady well, part of the problem is that I didn't want to mark every single node as WANTED, so I set up a two-tiered system where only "important" nodes are marked by wanted(), and we use WANTED() to force it
19:08 yurivish_ AlexDaniel: ZoffixW: I think if you put search.js in the head of the page, it will load it before continuing? at any rate, it'll start loading it faster
19:08 TimToady but sometimes something like a variable needs the mark
19:08 awwaiid nine: did you mostly just squash them into one commit?
19:08 lizmat ZoffixW: looking at the doc, years, months etc, are *not* documented as being valid
19:08 gfldex m:  dd (5,6 Z∈ ((1..100).pick(10),5).Set)
19:08 yoleaux 14:30Z <hankache> gfldex: i'll try it during the evening
19:08 camelia rakudo-moar 630a9b: OUTPUT«(Bool::True, Bool::False).Seq␤»
19:09 TimToady we should probably move the wanted flag into the node structure at some point to make it cheaper, and just mark everything
19:09 gfldex m:  dd any(5,6 Z∈ ((1..100).pick(10),5).Set)
19:09 camelia rakudo-moar 630a9b: OUTPUT«Method 'name' not found for invocant of class 'Bool'␤  in any  at gen/moar/m-Metamodel.nqp line 3124␤  in block <unit> at /tmp/aZSNAHO1wm line 1␤␤»
19:09 yurivish_ that way whenever the search box is visible, it is functional
19:09 ZoffixW yurivish_, no, the whole page would start loading much slower :P
19:09 ZoffixW AlexDaniel, it's not an icon I'm adding but a...
19:09 * ZoffixW git commits
19:09 nine awwaiid: mostly
19:10 dalek doc: 95647e9 | (Zoffix Znet)++ | template/ (2 files):
19:10 dalek doc: Provide indication the search feature is still loading (Closes #295)
19:10 dalek doc: review:
19:10 yurivish_ sure, a little. it's 12kb zipped.
19:10 awwaiid nine: I'll try to remember you like 'em squished for next time (I tend away from squashing/rebasing myself -- lots of commits reminds me of how I got there in the end)
19:10 yurivish_ and the price of a network request.
19:10 AlexDaniel ZoffixW: fine!
19:10 yurivish_ but the contents will get cached after the first load
19:10 ZoffixW :(
19:11 nine awwaiid: used to think so myself, but have never actually had any advantage of having the mistakes in the history
19:12 awwaiid nine: yeah, I can see that. I think I'm just a data hoarder
19:12 uvtc_ Anyone else currently unable to panda install Readline? Known problem?
19:12 AlexDaniel ZoffixW: yeah, I'm kinda thinking that, um, are you sure that the “Loading…” text will be replaced?
19:13 AlexDaniel ZoffixW: ah yeah, that's what .attr does
19:13 AlexDaniel ZoffixW: sorry :)
19:13 ZoffixW :)
19:13 ZoffixW uvtc_, what's the error
19:14 _sri lizmat: just seen your comment at​hristmas-came-bah-humbug.html#comment-1644965
19:14 lizmat _sri: did it make sense ?
19:14 _sri lizmat: you mention specifically "event driven programming using supplies / async sockets / async file IO", why would that be faster in perl6 than perl5?
19:15 lizmat well, syntactically, I don't think you can beat react / whenever / supply
19:15 lizmat but from a performance point of view, it is really threaded
19:16 SwellJoe joined #perl6
19:16 _sri that only makes it sound like it should be bad for performance
19:16 lizmat now, I'm not sure how Mojo handles this precisely, but I was under the impression that that was a select loop basically ?
19:16 Su-Shee if I could log into I would give a business reason or two.. but I can't.
19:16 uvtc_ ZoffixW: Error when trying to panda install Readline:
19:16 _sri lizmat: mojolicious uses libev for scaling
19:17 ZoffixW lizmat, yeah, it's not. I guess the best approach is to ignore this for now, unless more people experience my confusion, which is assuming plurals work for .new, just as they do for .later and not getting any errors when trying to use them in .new
19:17 lizmat _sri: I didn't know that
19:18 Skarsnik Zero_Dogg, still here? You can paste your code somewhere too?
19:18 _sri lizmat: honestly, i would be very surprised if you could get close anytime soon performance wise
19:18 AlexDaniel “Perl 6 is a bit sexier than Perl 5”. A *bit*. Right
19:18 ZoffixW uvtc_, how old is your rakudo? I've seen a whole ton of NC fixes flying in in the past 20 hours
19:18 lizmat _sri: well, we're going to try, that's for sure  :-)
19:18 lizmat Su-Shee: if you want, I could channel you  :-)
19:19 flussence I think, given node decided libuv is good enough for them, we can do pretty well with it too.
19:19 _sri allright, was wondering if there was some async magic sauce i might have missed :)
19:19 uvtc_ ZoffixW: Just rm -f'd my ~/.rakudobrew and ~/.perl6 dirs. It's brand new.
19:19 lizmat _sri: I think you're pretty much aware of everything going on here  :)
19:19 Su-Shee lizmat: well I have an account and I'd love to write something useful. ;) anyways. the business case IMHO is "data". I usally sell perl6 is for text what julia is for numbers.
19:19 ZoffixW uvtc_, weird. Maybe Skarsnik would know. They were hacking on NC today
19:20 uvtc_ ZoffixW: Just curious, what's "NC"?
19:20 ZoffixW uvtc_, NativeCall
19:20 [Coke] NC is NativeCall
19:20 Su-Shee lizmat: BUT it all means nothing without a proper set of modules doing.. "the usal stuff"
19:20 uvtc_ Ah. Thanks.
19:20 Skarsnik In 'remove_history' routine declaration - Not an accepted NativeCall type for parameter [1]  : Int
19:20 Skarsnik -->For Numerical type, use the appropriate int32/int64/num64...
19:20 Zero_Dogg Skarsnik: I can, give me a moment
19:20 Skarsnik Readline need to be fixed
19:20 ZoffixW Ah
19:20 AlexDaniel “No such method 'chars-supply'” has anybody seen this error?
19:21 ZoffixW AlexDaniel, yes, it's deprecated
19:21 ZoffixW AlexDaniel, it's .Supply now
19:21 Su-Shee lizmat: alternatively, I'd sell it as something like Scala or Clojure and point out that you just can use all the Java stuff out there.
19:21 ZoffixW s/deprecated/removed in the Supply Refactor/;
19:21 Su-Shee lizmat: and not mention moarvm however. ;)
19:22 skids joined #perl6
19:22 lizmat Su-Shee: well, with a broken JVM backend on 2015.12, I don't think I want to emphasize JVM interoperability
19:22 uvtc_ Skarsnik: ZoffixW : Thanks. Will file a bug report with Readline.
19:22 ZoffixW *reads blogpost*  Huh? Picking JavaScript and node.js? -_- nothanks
19:22 Ven joined #perl6
19:23 lizmat fail => self) without my $year = %_<year> // %_<years>;   # first time I could use "without"  :-)
19:23 _sri lizmat: just for completeness sake, mojolicious usually scales by using a manager process that spawns a bunch of processes that each run their own libev event loop
19:23 _sri the traditional nginx model
19:24 lizmat _sri: so there are no shared data structures ?
19:24 Su-Shee lizmat: I'm selling perl6 here, not caring for underlying tech. :) I expect that to be fixed sooner rather than later because it actually brings perl6 en par with all the jvm based langs... otherwise, perl6 will soon get the question why it again wrote its own vm after all the parrot stuff.
19:24 _sri lizmat: nope, data is shared with databases or a pub/sub message bus
19:25 flussence it's Apache2 that uses worker processes, and lighttpd uses libev, nginx doesn't ;)
19:25 lizmat ok, so basically db or the forks::shared method
19:25 shicheng joined #perl6
19:25 lizmat well, in Perl 6, all data structures *are* shared :-)
19:26 shicheng joined #perl6
19:26 _sri the way you run the event loop in a separate thread is really interesting, i wonder what the downsides will turn out to be
19:27 uvtc_ left #perl6
19:27 ilmari that blog post also ignores Inline::Perl[56] for interop and gradual porting
19:27 * _sri has not looked at how gc works in rakudo yet
19:27 Su-Shee ilmari: that's not a business case either.
19:27 [Coke] We already have failures on roast/6.c with rakudo/no
19:27 lizmat ilmari: woolfy is already writing a comment about that
19:27 [Coke] m
19:28 shicheng joined #perl6
19:28 Su-Shee ilmari: and I think it's a really good question "what would I perl6 use for" beyond "cool tech" for real?
19:28 ZoffixW Su-Shee, what would I use Python for?
19:29 skids [Coke]: :(
19:29 Su-Shee ZoffixW: data, education, web for starters. thanks to the excellent data libraries combined with all the other stuff, it's taking over the R world at the moment.
19:29 [Coke] perl6/doc - htmlify calls out to pygmentize; I have pygmentize-2.7 installed. Is the right approach for me to manually make a pygmentize alias, or to make htmlify try harder?
19:29 Su-Shee ZoffixW: it is also by now THE language in education of all kinds.
19:29 shicheng joined #perl6
19:29 ilmari Su-Shee: one of the cons raised in the post was the lack of migration path, which Inline::Perl[56] helps with
19:29 Su-Shee ZoffixW: there is no data related job ad without mentioning python.
19:29 ilmari Su-Shee: and it's similar to your JVM argument: you can keep using all the cpan stuff
19:30 Su-Shee ilmari: I see.
19:30 TimToady nine: so...I run 'panda install Inline::Perl5', and it fails because Inline::Perl5 is "already installed", and then I do a make spectest, and it says: "Inline::Perl5 not installed: not running Perl 5 integration tests"...what's up with that?
19:30 Su-Shee ilmari: on the other hand: why would I bother with perl6 then? I can use all the cpan stuff with perl5.
19:30 [Coke] (6.c failures)
19:30 autarch if I have a class name in a scalar can I call "$" on it?
19:30 Zero_Dogg Skarsnik:
19:30 ZoffixW Su-Shee, last thing I ported from P5 to P6 ended up half as short :)
19:30 alpha123 Su-Shee: Maybe you're more productive in Perl6 than Perl?
19:31 llfourn joined #perl6
19:31 lizmat TimToady: could it be that you had a core setting compile fail ?
19:31 alpha123 P6 is the scripting language of the future IMO, it does everything and it does it well
19:31 lizmat TimToady: and/or an install?
19:32 Skarsnik Zero_Dogg, look fine to me. It's the execute that fail?
19:32 Su-Shee alpha123: well I'm trying for years now and I'm definitely not. but I'm really too stupid for perl6.
19:32 nbrown joined #perl6
19:32 ZoffixW Any answer for autarch? I'm curious as well.
19:32 ZoffixW m: class Foo { method bar { say "barr1!" }}; my $class = 'Foo'; $ # this doesn't work
19:32 camelia rakudo-moar 3dc7d5: OUTPUT«Method 'bar' not found for invocant of class 'Str'␤  in block <unit> at /tmp/mabtuEUFrx line 1␤␤»
19:33 _sri lizmat: btw. if you need an argument against node.js, v8 still has a 1.4gb memory limit for each process, the downside of using a js engine designed for browsers :)
19:33 RabidGravy autarch,  ::($class).new
19:33 Su-Shee alpha123: considering which languages so far have been "the scripting languages" or really "the languages" most in use it's certainly not the "doing something well" languages. "worse is better" applies a lot.
19:33 autarch RabidGravy: thanks
19:33 ZoffixW RabidGravy++
19:33 RabidGravy does a "type lookup"
19:33 Zero_Dogg Skarsnik: yep, that's when it segfaults
19:34 TimToady nine: a force install makes it show up to spectest; is the already-installed check ignoring version number or something like that?
19:35 nine TimToady: the already installed check really is just perl6 -e "exit !try { require Inline::Perl5; 1 }"
19:35 Su-Shee but anyways, I'm not the marketing department for perl6, there's a perl foundation and what not which should be selling it and finding business arguments and use cases and all that jazz.
19:35 TimToady it should really understand that a new version needs to be forced, methinks
19:36 TimToady anyway, just an observation...
19:36 * TimToady goes back to his loopy cave
19:37 Skarsnik Zero_Dogg, I can't really help you sadly, mysql server does not work on the system I am working. all I can offer you is git clone DBIish and put say in the code to find where it fail exacltly ~~
19:37 lizmat Su-Shee: if nobody tells the TPF the business cases, they won't find any
19:37 autarch does the bot that runs m have any modules installed?
19:37 lizmat Su-Shee: TPF is just a bunch of people, like we're here
19:37 ZoffixW autarch, Task::Star
19:37 alpha123 Su-Shee: Perl6 is better is worse is better :)
19:38 Skarsnik there is a rakudo star on camelia I think
19:38 Skarsnik if you figure the name of the command xD
19:38 lizmat Su-Shee: please don't assign any super powers to them, they need as much help as anybody
19:38 autarch m: my $module = 'Grammar::Debugger'; require $module
19:38 camelia rakudo-moar 3dc7d5: OUTPUT«Could not find Grammar::Debugger in:␤    /home/camelia/.perl6/2015.12-51-g3dc7d55␤    /home/camelia/rakudo-m-inst-1/share/perl6/site␤    /home/camelia/rakudo-m-inst-1/share/perl6/vendor␤    /home/camelia/rakudo-m-inst-1/share/perl6␤    CompUnit::…»
19:38 ZoffixW m-star: use Grammar::Debugger;
19:38 ZoffixW star: use Grammar::Debugger;
19:38 camelia star-m 2015.09: ( no output )
19:38 autarch star: my $module = 'Grammar::Debugger'; require $module
19:38 camelia star-m 2015.09: OUTPUT«Could not find file 'Grammar::Debugger' for module Grammar::Debugger␤  in block <unit> at /tmp/Idt957Zmnz:1␤␤»
19:39 TimToady if anything, working for TPF is even more thankless than working here...
19:39 autarch so what's happening seems to be that it tries to look for 'path/to/lib/Grammar::Debugger' - using the literal string passed to require rather than translating the module name to a path
19:39 Zero_Dogg Skarsnik: hehe, nod, I might do that later, don't have that much time to do it right now
19:39 TimToady so a tip'o-the-hat to 'em
19:39 webstrand joined #perl6
19:39 brrt joined #perl6
19:39 autarch now I know that's what p5 does but I was hoping p6 wouldn't do the same thing
19:40 autarch ah, I need to do the ::($name) thing again
19:40 kittenlips hi, question what is CORE.setting?
19:40 yoleaux 07:54Z <nine> kittenlips: does what you described
19:40 lucasb [Coke]: it's gonna be hard to grow a language while having to keep all those 6.c tests running and intact.  What should be done now?
19:40 autarch coolio, that works - I'm always happy when p5 shortcomings are addressed
19:40 TimToady autarch: we try to be very clear in the distinction between hard refs and symbolic in p6
19:41 autarch TimToady: yeah, that's what I'm starting to figure out - it all makes sense, it's just new
19:41 TimToady why we don't need 'strict refs' anymore
19:41 autarch I'm used to the p5 "just fling stuff around and it'll work, sort of" approach ;)
19:41 Su-Shee lizmat: well. you have my arguments. that's how I'd be selling perl 6.
19:41 TimToady same with $obj.$routine vs $obj."$routine"()
19:41 nine TimToady: the installation process does consider versions just fine. You needed to to a panda update, because panda's meta data was out of date.
19:41 TimToady okay
19:41 kittenlips nine++ thanks a lot!!
19:41 autarch TimToady: yep, that one also made sense once I realized that $routine could be a Routine object
19:42 autarch I think a little extra typing is worth it for the clarity it provides
19:42 nine TimToady: the spec test check failed because you had a version installed that wouldn't work anymore with rakudo nom
19:42 TimToady funny thing is that we didn't actually anticipate that $obj.&subroutine would work, it just fell out that way :)
19:43 TimToady nine: kind of a shame we can't just track all the dependencies with make...
19:43 [Coke] lucasb: we're working on a plan that will be in place before the next release.
19:44 TimToady and anything more powerful than make isn't really very portable...
19:44 [Coke] But we all signed up for that when we decided it was time to cut a "real" release.
19:45 TimToady even if that was sort of a royal we :)
19:45 [Coke] TimToady: btw, I think we already have one commit in nom that has to fork rakudo and keep a 6.c & 6.d track.
19:45 ZoffixW nine, sounds like should be added to the Ecosystem :)
19:47 TimToady [Coke]: you mean it actively fails a 6.c test?
19:47 [Coke] er, not fork, branch.
19:47 nine ZoffixW: oh, indeed!
19:47 [Coke] TimToady:
19:47 yurivish_ ZoffixW: I tried searching for ".sort" on and got a a "no results found", but searching for "sort" found the routine. Maybe ".sort" should work too?
19:48 kittenlips nine Grammar::Highlighter is exactly what I was looking for, a syntax highlighter that doesn't force wheel reinvention and parses the programming language's grammar to fully understand how to properly highlight the syntax
19:48 rickbike joined #perl6
19:48 kittenlips .ping MadcapJake
19:48 yoleaux There is no ping command; nor can this be construed as a response.
19:48 [Coke] the first is a change, the second is a segfault or so, the third is a change, the last is a flapper.
19:49 [Coke] yurivish_: sure, can you add that as a ticket to docs/perl6/issues ?
19:49 yurivish_ sure. where is that?
19:49 mempko_ joined #perl6
19:49 kittenlips MadcapJake++ I think you did an excellent job creating the Atom package language-perl6fe syntax highlighter, also have a look at nine's
19:49 mempko_ joined #perl6
19:50 lizmat ZoffixW: re :years, etc) , not going to fix that, it gets *very* dirty and performing badly
19:50 yurivish_ here?
19:50 lizmat ZoffixW: fwiw, I'd rather remove support for plurals in .later
19:50 lizmat ZoffixW: to be completely consistent
19:51 yurivish_ [Coke]: Is this where I should file the issue?
19:51 yqt joined #perl6
19:51 mempko_ For those interested, I am working on a perl 6 kafka library.
19:51 mempko_ Here is an example usage.
19:51 ZoffixW lizmat, +1 on removing plurals; keep it nice and simple and consistent. They're in .truncated-to too
19:51 mempko_​mpko/fc3f89d179d9a33a750d
19:51 ZoffixW yurivish_, yes
19:52 mempko_ I'll publish the library on github as soon as it is usable. Maybe another week or so.
19:52 lizmat ZoffixW: alas, removing plurals from later would not be backward compatible
19:52 [Coke] lizmat: we already have non-backward compatible changes in nom, don't we?
19:52 [Coke] I'm not saying add more, just thought that was the result of the test failure I saw.
19:52 ZoffixW Really? That's kinda disappointing. 5 days after the release!
19:53 ZoffixW :)
19:53 AlexDaniel ZoffixW: it is PROGRESS
19:53 lizmat [Coke]: well, with the removal of IO::ArgFiles, we break any code explicitely referencing that
19:53 [Coke] ... which 6.c does
19:53 Skarsnik damn travis does not propose for 32bits build
19:53 lizmat [Coke]: but there no tests for IO::ArgFiles, apart from testing existence of the class
19:54 [Coke] ... and now it doesn't?
19:54 Begi joined #perl6
19:54 [Coke] I mean, I get that's a trivial thing, but it's still a failing test in 6.c
19:54 regreg joined #perl6
19:54 lizmat [Coke]: 630a9b20f9e829bf15359 removed IO::ArgFiles, and made WHICH.t fail
19:55 [Coke] there's also a datetime failure.
19:55 lizmat ah, that one
19:55 cygx .oO( add a dummy class that does nothing but FALLBACK to an error message )
19:55 lizmat I think we agreed that was a faulty test
19:55 ZoffixW lizmat, based on search results, the plurals are not tested in .later nor truncate to​earch?utf8=%E2%9C%93&amp;q=later​ch?utf8=%E2%9C%93&amp;q=truncate-to
19:55 lizmat [Coke]: so in 6.c, that should be fixed and marked "todo:
19:55 AlexDaniel m: say v1.0
19:55 camelia rakudo-moar 3dc7d5: OUTPUT«v1.0␤»
19:55 AlexDaniel m: say v1.0.WHAT
19:55 camelia rakudo-moar 3dc7d5: OUTPUT«v1.0.WHAT␤»
19:55 Skarsnik lol
19:55 ZoffixW m: say (v1.0).WHAT
19:55 camelia rakudo-moar 3dc7d5: OUTPUT«(Version)␤»
19:56 AlexDaniel :))
19:56 Skarsnik m: say v1.0.a
19:56 camelia rakudo-moar 3dc7d5: OUTPUT«v1.0.a␤»
19:56 [Coke] any changes to the 6.c tests are problematic.
19:56 Skarsnik m: say v1.0.a < v1.0.b
19:56 camelia rakudo-moar 3dc7d5: OUTPUT«Cannot call Real(Version: ); none of these signatures match:␤    (Mu:U \v: *%_)␤  in block <unit> at /tmp/GST3HMFi5x line 1␤␤»
19:56 TimToady yes, test suites should be considered immutable, just as modules are
19:56 ZoffixW m: say v1.0.a before v1.0.b
19:56 camelia rakudo-moar 3dc7d5: OUTPUT«True␤»
19:56 [Coke] We need some kind of review process to say "trivial change", "new version of 6.c here", or "6.d"
19:56 AlexDaniel m: say v1.0.a cmp v1.0.b
19:56 camelia rakudo-moar 3dc7d5: OUTPUT«Less␤»
19:57 [Coke] We'll get it all sorted out before 2016.01, I'm sure.
19:57 lucasb [Coke]++  # optimist
19:57 ccakes joined #perl6
19:57 AlexDaniel m: say v1.0.a < v1.0.b
19:57 camelia rakudo-moar 3dc7d5: OUTPUT«Cannot call Real(Version: ); none of these signatures match:␤    (Mu:U \v: *%_)␤  in block <unit> at /tmp/FoELAhL1Ub line 1␤␤»
19:57 TimToady we have a wide dynamic range :)
19:58 TimToady and on many different scales
19:58 perlpilot [Coke]: 1/3 of that is easy.  There are no "new" versions of 6.c   ;)
19:58 TimToady laziness vs diligence, impatience vs patience, hubris vs humility
19:59 [Coke] perlpilot: yes, there are.
19:59 Skarsnik nine, any idea why i:p5 does not work on 32bits?
19:59 TimToady and we've had illustrations of all those wide ranges in the last month
19:59 [Coke] there's a chance we'll have a 6.c.1, e.g.
19:59 nine Skarsnik: not yet. Household chores to do
19:59 [Coke] "Here's a bundle of minor changes that are technically not backward compatible, but we think you'll like"
19:59 Skarsnik ok, let me know if you want me to test stuff
19:59 [Coke] I hope we don't do that often. (and not in 2016.01)
20:00 shicheng joined #perl6
20:00 ZoffixW ouch "Internal error: zeroed target thread ID in work pass"
20:00 [Coke] is there a better way to get a file's POD than:         my $pod  = EVAL(slurp($file.path) ~ "\n\$=pod")[0];
20:00 ZoffixW And didn't happen on re-run
20:01 flussence ZoffixW: wow... haven't seen that error in a *long* time.
20:01 nine Well the goal is to never ever break user code by updating a published version of the language. One way to do that is to just never update. Though right now there's not that much code relying on 6.c. There still can be changes that are reasonably safe to do.
20:02 lnrdo​-and-docker-employee-ian-murdock-has-died-at-42/
20:02 TimToady the basic assumption should be that we don't break 6.c by default, but that we can grant a "variance" in some circumstances, but only after due consideration of how best to introduce the change over time
20:03 dalek doc: 9aa6832 | lizmat++ | doc/Type/Date (2 files):
20:03 dalek doc: Undocument .later plurals
20:03 dalek doc: review:
20:03 stmuk_ lnrdo: :(
20:03 nine m: say True.Num, True.Int
20:03 camelia rakudo-moar 3dc7d5: OUTPUT«1True␤»
20:03 AlexDaniel lizmat: can't really say that I like this change
20:04 lizmat AlexDaniel: we can revert
20:04 nine TimToady: as an exercise, would you consider removing the .Num candidate that's responsible for this ^^^ from Bool?
20:04 lizmat :-)
20:04 yurivish_ why does this not work?
20:04 lizmat dinner&
20:04 yurivish_ m: my %h = (1=>2, 3=>4); %h.sort(-*.value).map(.say);
20:04 camelia rakudo-moar 3dc7d5: OUTPUT«(Any)␤Cannot call map(List: Bool); none of these signatures match:␤    ($: Hash \h, *%_)␤    (\SELF: &block;; :$label, :$item, *%_)␤    (HyperIterable:D $: &block;; :$label, *%_)␤  in block <unit> at /tmp/TmukOaCzHk line 1␤␤»
20:04 AlexDaniel lizmat: again, what's the reason to do so?
20:04 nine TimToady: the only spec tests that fail depend on this completely by accident and are faulty to begin with
20:04 TimToady how is that not what you'd expect?  True is an Int
20:04 TimToady True isn't a Num
20:05 lucasb m: say True.Numeric
20:05 camelia rakudo-moar 3dc7d5: OUTPUT«1␤»
20:05 lnrdo stmuk_: It is not 27, but still 42 is a worth number match.
20:05 yurivish_ where is there a true? I was thinking this would sort the list by value and then print out the pairs
20:05 virtualsue joined #perl6
20:06 flussence hasn't the rule always been "coercing to the same or looser type is a no-op"?
20:06 ZoffixW AlexDaniel, consistency with other methods, such a .new that does not take plurals and silently ignores them. And adding them is expensive
20:06 yurivish_ oh, I didn't mean to map a say... That was me trying to reduce to a simpler test case
20:06 ZoffixW m: say :2015year :12month :31days :23hours :59minutes :59seconds
20:06 camelia rakudo-moar 3dc7d5: OUTPUT«2015-12-01T00:00:00Z␤»
20:06 ZoffixW AlexDaniel, spot the bug ^ :)
20:07 AlexDaniel doesn't it mean that plurals should be better supported?
20:07 AlexDaniel and not that we should get rid of them
20:07 ZoffixW It's messy and expensive
20:07 yurivish_ the actual code I was trying to write seems to work now
20:07 yurivish_ m: my %h = (1=>2, 3=>4); %h.sort(-*.value).map(*.key ~ ',' ~ *.value);
20:07 camelia rakudo-moar 3dc7d5: ( no output )
20:08 AlexDaniel as a user I don't really want to think whether it is “day” or “days”, I'd prefer it to work either way
20:08 TimToady seems to me, if you were really worried about performance, you'd have a position new for DateTime that avoids named entirely
20:08 [ptc] [Coke]: re: POD extraction.  I don't think so :-(  I had issues with that when extracting the POD from the files in the perl6-examples repo
20:08 yoleaux 17:38Z <stmuk> [ptc]: yay updating .. BTW *.un~ is a vim undo file which my .vimrc has enabled
20:08 awwaiid -- more EVAL targets
20:08 perlpilot yurivish_: that looks like a precedence problem to me.
20:09 bpmedley No!! Damn you mortality!  Ian Murdock was all the awesome.
20:09 perlpilot yurivish_: oh, maybe not.
20:09 * perlpilot can't read
20:09 ZoffixW Well, dunno :) I'm not the gut hax0r... <lizmat> ZoffixW: re :years, etc) , not going to fix that, it gets *very* dirty and performing badly
20:09 yurivish_ perlpilot: now I'm confused too
20:10 Begi I guess the link to the advent calendar can be removed from, no ?
20:10 perlpilot m: my %h = (1=>2, 3=>4); %h.sort(-*.value).map(*.key ~ ',' ~ *.value).say
20:10 camelia rakudo-moar 3dc7d5: OUTPUT«(3,2)␤»
20:10 perlpilot yurivish_: what did you want from that ?
20:10 gfldex Begi: it can be removed 2016-12-01
20:11 Begi Ok good
20:11 ilmari mempko_: does kafka do sharding, like amazon kinesis? it would be neat to have the consumer automatically spawn a consumer thread per shard
20:11 flussence just in time to put it back up again :)
20:11 ZoffixW Begi, I was thinking of living it in until like middle of January, just for the people who missed it
20:11 perlpilot yurivish_: by mentioning * twice, you got two different parameters to .map
20:11 gfldex Begi: or replaced when we get something better until then
20:11 ZoffixW Begi, or keeping permanently and removing the "new article daily" blurb
20:11 [ptc] [Coke]: moritz helped me with this solution:  my $perl-pod = qqx{perl6-m -Ilib --doc=Perl $file};   Where one also defines a Perl parser for the pod...
20:11 yurivish_ I want ['3,4', '1,2']
20:11 perlpilot yurivish_: i.e. it's the same as if you'd done .map({  $^a.key ~ ',' ~ $^b.value });
20:12 [ptc] stmuk_: cool, thanks for the info :-)
20:12 perlpilot yurivish_: then use .map({ .key ~ ',' ~ .value })
20:12 yurivish_ perlpilot: I did not realize * would become different parameters and thought this would be a shorter way to write it
20:12 AlexDaniel m: my \x = 42; say x × ⅹ x ⅹ
20:12 camelia rakudo-moar 3dc7d5: OUTPUT«420420420420420420420420420420␤»
20:13 ZoffixW m: say (* + * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * )(|^22)
20:13 camelia rakudo-moar 3dc7d5: OUTPUT«51090942171709440000␤»
20:13 ZoffixW yurivish_, ^
20:13 ZoffixW each one is a new thing
20:14 perlpilot well ... only every other one starting from the second is a "new thing"  :)
20:14 yurivish_ ZoffixW: what does that output mean? is that parsed as something like +(*_1, *_2, ... *_20)?
20:14 yurivish_ perlpilot: things work now!
20:14 ZoffixW yurivish_, say 0 + 1 * 2 * 3 * 4 * 5 * 6....* 22
20:15 ZoffixW yurivish_, say 0 + 1 * 2 * 3 * 4 * 5 * 6....* 21
20:15 AlexDaniel ZoffixW: this example makes me cry. Perhaps consider using × next time :D
20:15 yurivish_ m: say (* + * * *)(1,3,5)
20:15 camelia rakudo-moar 3dc7d5: OUTPUT«16␤»
20:15 yurivish_ hah
20:15 znpy joined #perl6
20:15 AlexDaniel m: say (* + * × *)(1,3,5)
20:15 perlpilot yurivish_: most times you don't want to use more than one Whatever.  If you think you need more than one, you probably want to use one of the other, more explicit methods
20:15 camelia rakudo-moar 3dc7d5: OUTPUT«16␤»
20:16 ZoffixW AlexDaniel, nevar! It's my rebellion against there not being a Unicode variant of WhateverStar and HyperWhateverStar :)
20:16 yurivish_ perlpilot: lesson learned.
20:16 AlexDaniel ZoffixW: right
20:16 flussence m: say (Whatever + [×] Whatever xx 20)(^21)
20:16 camelia rakudo-moar 3dc7d5: OUTPUT«Use of uninitialized value of type Whatever in numeric context  in block <unit> at /tmp/0DtMoKMzKU line 1␤Use of uninitialized value of type Whatever in numeric context  in block <unit> at /tmp/0DtMoKMzKU line 1␤Use of uninitialized value of type Whate…»
20:16 flussence m: say (WhateverCode + [×] WhateverCode xx 20)(^21)
20:16 camelia rakudo-moar 3dc7d5: OUTPUT«Can not invoke a code type object␤  in block <unit> at /tmp/weS0pgwaAK line 1␤␤»
20:16 flussence m: say (* + [×] * xx 20)(^21)
20:16 camelia rakudo-moar 3dc7d5: OUTPUT«Cannot call Numeric(Whatever: ); none of these signatures match:␤    (Mu:U \v: *%_)␤  in block <unit> at /tmp/Nf32ER2Zw_ line 1␤␤»
20:16 TimToady whatevers never curry positional args
20:16 flussence *sigh*
20:17 flussence I'm no good at writing unreadable code :)
20:18 Begi .seen azawawi
20:18 yoleaux I saw azawawi 28 Dec 2015 08:47Z in #perl6: <azawawi> Merry Christmas to #perl6 :)
20:18 ccakes joined #perl6
20:18 rindolf joined #perl6
20:18 Begi Easy thing that I can do :​perl6-browser-open/issues/2
20:19 Begi Should I put the doc of the module ine the README file ?
20:19 Skarsnik hm, I remember perl 6 having a State type, but I can't find it on the doc x)
20:19 skids IMO IO::ArgFiles should be put back in and not removed till 6.d.  It is reasonable for a user of 6.c to interpret that test as "IO::ArgFiles is part of the spec" even if its internals are not tested.
20:19 mempko_ ilmari: Yes, kafka has the concept of partitions. I'll make sure to add support for consuming multiple partitions.
20:20 [Coke] skids: docs and source disagree with you.
20:20 skids How so?
20:20 Skarsnik Begi, ideallt both sadly x)
20:20 [Coke] whoops.
20:20 [Coke] Skarsnik: docs and source disagree with you.
20:20 skids Oh.
20:21 itaipu joined #perl6
20:21 AlexDaniel m: say ༫.base(36)
20:21 camelia rakudo-moar 3dc7d5: OUTPUT«1.I␤»
20:21 Skarsnik all I found for State is the state stuff like my x)
20:22 perlpilot what would the State type mean, if it were to exist?
20:22 * TimToady has no recollection of a State type
20:22 TimToady maybe you're thinking of when we had START blocks?
20:23 lucasb m: say (slip(1,2) Z slip(1,2)).perl
20:23 camelia rakudo-moar 3dc7d5: OUTPUT«((1, 2, 1, 2),).Seq␤»
20:23 lucasb ^^ it's interpreting that as '[Z] 1,2,1,2'. is that expected?
20:24 [Coke] m: say (1,2 Z slip 1,2).perl;
20:24 camelia rakudo-moar 3dc7d5: OUTPUT«((1, 1, 2),).Seq␤»
20:24 Skarsnik TimToady, something like
20:24 AlexDaniel m: say :5<1.I>
20:24 camelia rakudo-moar 3dc7d5: OUTPUT«5===SORRY!5=== Error while compiling /tmp/LqsDStoR_m␤Couldn't process entire number: 1/1 int chars, -1/1 fractional chars␤at /tmp/LqsDStoR_m:1␤------> 3say :5<1.I>7⏏5<EOL>␤»
20:24 AlexDaniel ↑ was it supposed to say 1/1 ?
20:24 AlexDaniel or it is just that somebody forgot {} around it?
20:24 [Coke] slightly modified htmlify.p6, I get this when running it on Christmas: moar(35746,0x7fff77064300) malloc: *** error for object 0x7ff5936b0750: pointer being freed was not allocated
20:24 [Coke] *** set a breakpoint in malloc_error_break to debug
20:24 [Coke] make: *** [html] Abort trap: 6
20:25 perlpilot Skarsnik: All that Q* stuff in your link reminds me of 007.  Maybe you're thinking of that?  :)
20:25 AlexDaniel and what does it mean “-1/1” fractional chars?
20:25 Skarsnik But it's more I want to keep track of the state of a statementhandle object in DBIish (like ready, running, finished)
20:25 TimToady likely an error code back from nqp
20:27 TimToady m: say :5<1.6>
20:27 camelia rakudo-moar 3dc7d5: OUTPUT«5===SORRY!5=== Error while compiling /tmp/NOrFCdkVhj␤Couldn't process entire number: 1/1 int chars, -1/1 fractional chars␤at /tmp/NOrFCdkVhj:1␤------> 3say :5<1.6>7⏏5<EOL>␤»
20:27 autarch what does this error mean - "Virtual call @.skip may not be used on partially constructed objects" ?
20:27 TimToady I think that used to have a better message, but the innards are leaking somehow
20:27 autarch I have a class with a BUILD submethod where I'm trying to set @.skip
20:27 perlpilot autarch: you want to set @!skip
20:27 autarch ah, that makes sense
20:27 TimToady @.skip is really a call to self.skip
20:28 webstrand When destructuring arrays, what's the use of the asterisk in my ($first, *@rest) = 1..20; It doesn't seem to change behaviour when missing like the documentation in Signature says it should.
20:28 perlpilot autarch: the object isn't fully constructed in BUILD, so any calls to @.skip are ... what TimToady said
20:28 perlpilot autarch: which can't work
20:28 autarch yeah, that makes sense when I think about it
20:28 Skarsnik . is an accessor x)
20:28 TimToady which you can still do if you really mean it, but we don't want people falling into confusing virtual attributes with physical attributes by accident
20:28 autarch I suspect the error message could probably add " - maybe you wanted to set @!skip directly?"
20:28 TimToady that might help
20:28 ZoffixW m: use Test:ver; # curious
20:29 camelia rakudo-moar 3dc7d5: ( no output )
20:29 ZoffixW m: use Test:blahblahblah; # curious
20:29 camelia rakudo-moar 3dc7d5: ( no output )
20:29 TimToady feel free to file an LTA on that
20:29 ZoffixW Ah
20:29 TimToady and on the :5<1.I>
20:29 autarch TimToady: is that directly at me?
20:29 TimToady no, that was to the editorial "we" :)
20:30 TimToady well, that second one was :)
20:30 TimToady the first one was to you, since you brought it up...or you can just fix it, if it breaks no 6.c tests :)
20:31 zengargoyle joined #perl6
20:31 zengargoyle good * #perl6
20:31 ZoffixW \o
20:32 llfourn joined #perl6
20:32 zengargoyle haven't been around for a week or so... things seem to be a bit borked.
20:32 TimToady you'll have to nuke your install if you have one from before Christmas
20:33 zengargoyle fresh build + panda... try to install Linenoise and it tries to install File::Find again... and fails 'cause it's allready installed.
20:33 zengargoyle pands --installed list shows nothing installed.
20:33 TimToady you can use --force for the moment as a workaround
20:33 zengargoyle did nuke moar and rakudobrew selfupbrade.
20:33 zengargoyle will try... TT
20:34 TimToady unless there's some fresh thing I dunno about...
20:34 zengargoyle figures i'd hit a momentary glitch of some sort. :)
20:34 ZoffixW ISAGN for a way to tell a Promise to violently crash my program if it's broken. Right now it's really annoying to debug code inside Promises results of which I don't care about, since all the errors are silently ignored
20:34 ZoffixW Is there a way to do that?
20:35 ZoffixW zengargoyle, yeah, it's reported in this issue:
20:35 zengargoyle ... Method 'get' not found for invocant of class 'Any
20:35 perlpilot ZoffixW: even with try/CATCH they're silently ignored?
20:36 ZoffixW zengargoyle, and there's ANOTHER issue with linenoze when installing via Task::Star, and that's fixed by simply running panda --force install Linenoise
20:37 zengargoyle guess i'll wait a bit.  laptop too flakey to dig too deeply (randomly just shuts down, video going bad after 7 years... :)
20:37 ZoffixW perlpilot, like how? I'd need to call .result to get the result won't I?
20:37 * zengargoyle waiting on new laptop from work shortly...
20:38 ZoffixW m: try { die "a horrible death" }; sleep 2; say "uhum"
20:38 camelia rakudo-moar 3dc7d5: OUTPUT«uhum␤»
20:38 zengargoyle 'tis why i haven't been following too closely last couple of weeks.  lack of patience for 'puter just going boom at random times.
20:39 zengargoyle so belated kudos to everybody for 6.c :)
20:39 skids m: try my $p = { die "a horrible death" }; start { await $p; $p.result.say }; sleep 2; say "uhum"
20:39 camelia rakudo-moar 3dc7d5: OUTPUT«uhum␤»
20:40 perlpilot huh
20:40 AlexDaniel “we” has submitted a bug report
20:40 TimToady we++
20:40 ZoffixW perlpilot, doesn't that away throws?
20:41 ZoffixW *await
20:41 ZoffixW wrong nick too XD
20:41 addison joined #perl6
20:41 ZoffixW m: try my $p = { die "a horrible death" }; start({ await $p; $p.result.say }).result; sleep 2; say "uhum"
20:41 camelia rakudo-moar 3dc7d5: OUTPUT«5===SORRY!5=== Error while compiling /tmp/bpNuBy2oNB␤Undeclared routine:␤    start used at line 1. Did you mean 'spurt', 'sort', 'sqrt'?␤␤»
20:42 skids m: my $p = { die "a horrible death" }; start { await $p; CATCH { $p.result.say } }; sleep 2; say "uhum"
20:42 ZoffixW m: try my $p = { die "a horrible death" };{ await $p; $p.result.say }).result; sleep 2; say "uhum"
20:42 camelia rakudo-moar 3dc7d5: OUTPUT«uhum␤»
20:42 camelia rakudo-moar 3dc7d5: OUTPUT«Default constructor for 'Promise' only takes named arguments␤  in block <unit> at /tmp/HLgtVy7fGd line 1␤␤»
20:42 ZoffixW *old man yells at the cloud*
20:42 dalek DBIish: be9eddf | (Sylvain Colinet)++ | README.pod:
20:42 dalek DBIish: Document Pg Array
20:42 dalek DBIish: review:
20:42 ZoffixW m: try my $p = { die "a horrible death" }; Promise.start({ await $p; $p.result.say }).result; sleep 2; say "uhum"
20:42 camelia rakudo-moar 3dc7d5: OUTPUT«a horrible death␤  in block <unit> at /tmp/oLCXzTersU line 1␤␤»
20:43 MadcapJake .tell kittenlips thanks! glad you like it! My highlighter is just using Atom's builtin grammar engine and the builtin node-oniguruma regex engine.
20:43 yoleaux MadcapJake: I'll pass your message to kittenlips.
20:43 ZoffixW But that's way more code than I was hoping for :P
20:43 perlpilot yeah, that's way more than I thought you'd need.
20:43 AlexDaniel autarch: “we” didn't submit your bug report though, so perhaps some other “we” will do that :)
20:43 autarch yeah, I can do that
20:44 pierre-vigier joined #perl6
20:45 TimToady We are amused.
20:45 webstrand Is destructuring using `my` supposed to work like subroutine parameters? I.e. is `my ($a, @b) = (1,2)` supposed to result in an error?
20:45 perlpilot TimToady: Are there any "rules" for writing good spec tests written down somewhere?  autarch's LTA discovery had me wondering if there were any spec tests relying on the contents of an error message (I sure hope not!) and if there was anything that would aid in preventing such travesty
20:46 TimToady webstrand: you need to use := if you want sig-like destructuring
20:46 autarch what is LTA?
20:46 TimToady less than awesome
20:46 autarch ah
20:46 TimToady it's in the glossary
20:46 TimToady awesome being our standard for error messages
20:46 japhb .oO( Is "glossary" in the glossary? )
20:47 TimToady lunch &~
20:47 japhb As well it should be
20:47 AlexDaniel
20:47 awwaiid m: my (Int $n, Str $s) := 5, "fish" ; say $n, $s # webstrand
20:47 camelia rakudo-moar 3dc7d5: OUTPUT«5fish␤»
20:47 MadcapJake anyone know bash? I'm trying to add an option to p6doc to just print right to stdout, but it just prints a blank line.  Any idea why?
20:47 skids perlpilot: well recently all the tests that demanded that an error message be contained in an X::AdHoc were change to just Exception, and in general when testing error messages you just need to be tolerant and only test for the minimum details being present.
20:48 skids perlpilot: throws-like tests are usually where those are found, for examples.
20:48 webstrand TimToady: Do you know where the rules for non-sig-like destructuring are documented?
20:48 autarch what does $thing».method do? I can't find any mention of this on the doc site
20:49 [Coke] MadcapJake: echo "some stuff"
20:49 autarch MadcapJake: you might also want to use cat
20:50 MadcapJake [Coke]: i think the problem is, I am trying to do `echo "some stuff" | echo` and that really does print a blank line
20:50 perlpilot skids: I was feeling pretty good until you said "when testing error messages"   :)
20:50 MadcapJake autarch: that's a good idea
20:50 Hotkeys autarch: it ESSENTIALLY calls the method on all of the items in a list/array like a map
20:50 Hotkeys but it does it with hyper
20:50 autarch MadcapJake: no need for the '| echo ' bit - I don't think piping stuff _into_ echo does anything
20:50 Hotkeys so in the future you may get the results out of order
20:50 [Coke] autarch:
20:50 [Coke] ... crap, that's not quite right. dammit. :)
20:50 MadcapJake autarch: true, i guess I could just remove the pipe and pager and just leave it blank
20:51 autarch so @array».foo is more or less like { .foo } )?
20:51 perlpilot Hotkeys: no, you'll always get the results *in* order, they just won't necessarily be processed in order
20:51 zengargoyle echo 'foo' | cat
20:51 Hotkeys oh
20:51 [Coke] just don't pipe it to anything unless you need it.
20:51 hankache autarch yes
20:51 MadcapJake autarch: but cat seems to work well for this
20:51 Hotkeys So I assume that's why you can't do $foo».say
20:51 Hotkeys because they'll be said out of order
20:51 Hotkeys maybe
20:51 Hotkeys in future
20:51 autarch MadcapJake: cat seems like a good option if you have a file you want to print to stdout
20:52 MadcapJake part of me feels like the default should be pagerless, people can add a pager if they want (what do others think about this?)
20:52 perlpilot Hotkeys: right-ish.
20:52 hankache Hotkeys hyper gets the results in order
20:52 hankache it's race that doesn't
20:52 zengargoyle usually something like: FILTER=less; ( echo stuff; someprog-more-output; echo more stuff ) | $FILTER
20:53 zengargoyle and then sometimes FILTER=cat
20:53 awwaiid webstrand: looks like that is a weakness in the docs. says the left-hand-side decides, which suggests that destructuring on the left in parens is from Array (and there is a link), but no doc that I can find
20:53 [Coke] autarch:
20:53 sprocket hello all!
20:53 autarch is the use of $match.ast in grammars considered bad form? the docs now just use .made
20:53 hankache so logically (in the future) they should be in order no?
20:53 awwaiid greetings sprocket!
20:53 nine m: say True.Numeric, True.Int
20:53 camelia rakudo-moar 3dc7d5: OUTPUT«1True␤»
20:53 sprocket perl5 used to have a select builting function, but i don’t believe it’s built into perl6 - is there something that has replaced the need for it?  or was it expected to be provided in a external library?
20:53 nine TimToady: that's ^^^ what I meant before
20:54 perlpilot autarch: I dunno.  I've wondered that myself.  Both "do the same thing" for now though
20:54 nine TimToady:
20:54 skids perlpilot: e.g. look at S32-io/open.t where the only demans on the message are that the filename appear.
20:54 nine sprocket: what do you need select for?
20:54 nine sprocket: select in Perl 5 has 3 completely separate uses
20:55 ccakes joined #perl6
20:55 webstrand awwaiid: Thanks. Perhaps it's documented here:​Items%2C_Flattening_and_Sigils . But slurpy parameters don't seem to flatten in destructuring assignment.
20:55 sprocket nine: personally, I don’t - I’d started a basic POSIX lib in the ecosystem, and someone was asking me about select :)
20:56 Begi Can't we use "or die" with module's routine ?
20:56 awwaiid webstrand: destructuring here I think is the opposite of the bit there
20:57 Skarsnik ?
20:57 FROGGS joined #perl6
20:57 skids perlpilot: btw, got time to look over PR's for File::Temp?
20:59 * perlpilot looks
21:00 MadcapJake looking at the p6doc script, i can't figure out where the `--help` result is constructed
21:01 hoelzro nine: for that rejected RT about BEGIN say 'hi', how do you feel about me creating a new issue for the LTA error?
21:01 hoelzro MadcapJake: --help is provided by rakudo itself, iirc
21:02 MadcapJake hoelzro: cool, it looks like it just added my flag automatically 0_0
21:02 hoelzro it should =)
21:02 alpha123 MadcapJake: main() subs in Perl6 are pretty darn cool eh?
21:02 MadcapJake alpha123: a thousand times yes!
21:03 perlpilot skids: I applied yours and ugexe's
21:03 * RabidGravy looks at sprocket's module as he has just about to type "int i2c_open(char *dev, int address) " ;-)
21:03 awwaiid MAIN is amazing (minus the known argument ordering issue)
21:03 skids perlpilot: thanks!
21:03 sprocket RabidGravy: it’s pretty bare bones - I’d added just what I’d needed for another project, and hadn’t fleshed it out yet
21:04 Skarsnik cygx has a lib C module that look neat x)
21:04 RabidGravy if it's got open and ioctl it works for me ;-)
21:04 perlpilot autarch: btw, your dzil-a-like for P6 live on github yet?
21:04 autarch perlpilot: I could push the mess I have right now if you want to see it
21:05 dalek doc: e41eda4 | lizmat++ | doc/Type/Date (2 files):
21:05 dalek doc: Revert "Undocument .later plurals"
21:05 dalek doc:
21:05 dalek doc: This reverts commit 9aa6832363dbd89ea28670e0a4cdbc88c0cea527.
21:05 dalek doc: review:
21:06 autarch
21:06 autarch my initial goal is just to get it to the point of generating a for me
21:06 autarch I really suck at writing JSON, I always leave extra commas and such lying around
21:06 RabidGravy ah sprocket, I was just trying to remember who you were yesterday in regard to the RPi thing
21:06 perlpilot autarch: an excellent goal (I have the same problem :)
21:06 RabidGravy :)
21:06 Idiosyncrat joined #perl6
21:06 sprocket RabidGravy: ah, ok :)
21:07 sprocket RabidGravy: should be getting back to that project soon as well, I got waylaid a bit over December
21:07 RabidGravy just making an i2c thing at the moment
21:07 Idiosyncrat Answering a question that Larry posed to me at​erl6/2015-12-11#i_11699587
21:07 hoelzro awwaiid: you mean panda --force install vs panda install --force ?
21:08 Idiosyncrat Re the time complexity of Perl 6 grammars -- first my "answer", then its basis.
21:08 awwaiid hoelzro: yes
21:08 sprocket RabidGravy: were you able to use anything in RPi?
21:08 Idiosyncrat The answer: I would characterize Perl 6 grammars as exponential worst case, and linear for LL(1)
21:09 rindolf joined #perl6
21:09 dalek doc: fc3c0d2 | lizmat++ | doc/Type/Date (2 files):
21:09 dalek doc: Emphasize that plurals only work with .later
21:09 dalek doc: review:
21:09 Idiosyncrat I've not gone through the Perl 6 grammar code like I have that for Perl 5, but here's the basis for my assertions:
21:09 RabidGravy sprocket, I'm goint wrap the i2c-dev thing which provides all the methods I need :)
21:09 hoelzro awwaiid: I was wondering if that would be best solved via a module
21:10 FritzZaucker joined #perl6
21:10 hoelzro you could easily do it via sub MAIN_HELPER(...) is export { }
21:10 Idiosyncrat PEG is linear for LL(1), and I assume that the linear time complexity of Perl 6 grammars is no better and no worse.
21:10 lizmat TimToady: there is a,day) candidate and similar DateTime candidate
21:10 awwaiid hoelzro: short-term definitely; but the command pattern for cli is common enough it would be cool to have it in MAIN. I believe that TimToady or someone had specific ideas about some sort of sub-signature thing I didn't understand
21:10 hoelzro ah, interesting
21:10 hoelzro I'll have to look for that in the logs
21:10 dalek ecosystem: 380302c | (Stefan Seifert)++ | META.list:
21:10 dalek ecosystem: Add Grammar::Highlighter to the ecosystem
21:11 dalek ecosystem: review:​cosystem/commit/380302cb9f
21:11 Zero_Dogg Inline::Perl5 errors out with: Invalid version format (negative version number) at /usr/share/perl/5.20/ line 9117161842131599360 - does anyone recognize this issue?
21:11 Idiosyncrat I think this assumption is reasonable because the time complexity of parsing algorithms is not something you are likely to overcome unless you target it in a specific way, say via LL tables or some such.
21:11 nine Zero_Dogg: is this with the absolutely current version?
21:12 FritzZaucker Hi ... after failing with GTK::Simple I am trying NCurses now for a little Perl6 exercise. The following NCurses example is working:​ucker/aa14025a16edcd671e25 whereas a similar one using a supply doesn't properly display the text:​ucker/0290915997b1ddc3ad53 after a few iterations. Am I doing something wrong or am I hitting a bug somewhere?
21:12 Zero_Dogg nine: yes, just installed now
21:12 nine Zero_Dogg: how can I reproduce this issue?
21:12 Idiosyncrat Similarly for the characterization of the worst case as exponential.  In general, top-down with backtracking goes exponential worst-case.
21:13 Zero_Dogg nine: it's on my raspberry pi again, all I did was install it, and tried use Inline::Perl5;, resulting in that error
21:14 Zero_Dogg nine: perl5 5.20.2, perl6 with rakudo on moar 2015.12, Inline::Perl5 installed through panda
21:14 RabidGravy FritzZaucker, it's entirely possible the libcurses itself isn't entirely thread safe
21:14 Arrowhead_ joined #perl6
21:14 Idiosyncrat You can overcome this for top-down by using techniques like GLL, but even careful expert programming won't usually overcome the time complexity bound unless it uses special techniques targeted at the worse case bound.
21:15 brrt anybody had the chance to test branch libuv-1.8 anywhere?
21:15 brrt before i merge it and it doesn't work and everybody's pissed? :-)
21:15 brrt or at least annoyed
21:15 * flussence would've, but a regular system update is eating all my CPU right now...
21:16 Skarsnik brrt, I will wait for at least the next rakudo release before doing that x)
21:16 Idiosyncrat TimToady: above is my response to​erl6/2015-12-11#i_11699587
21:16 FritzZaucker RabidGravy: thanks, so not much I can do about this, I guess (except perhaps implementing some simple-minded event loop in the main thread).
21:18 nine Zero_Dogg: but Inline::Perl5's tests passed on installation?
21:18 TEttinger joined #perl6
21:18 Idiosyncrat If anyone has reason to believe my time complexities are wrong, I'm glad to hear them.  Either reply on #marpa or mention my handle on this log.
21:18 ugexe m: use Test:auth<>:ver<>;
21:18 camelia rakudo-moar 3dc7d5: OUTPUT«Potential difficulties:␤    Pair with <> really means an empty list, not null string; use :auth('') to represent the null string,␤      or :auth() to represent the empty list more accurately␤    at /tmp/NkCYeQ1EeX:1␤    ------> 3use Test:auth<>7…»
21:18 RabidGravy FritzZaucker, you may try changing the .tap to an .act to  and see if it makes any difference (guarantees only on running at a time), unfortunately the NCurses guy isn't about
21:18 Idiosyncrat Congratulations, and thanks for your good work!
21:19 Zero_Dogg nine: nope, same error
21:19 nine brrt: Unable to checkout '5467299450ecf61635657557b6e01aaaf6c3fdf4' in submodule path '3rdparty/libuv'
21:19 brrt ah... damn
21:20 brrt can you check what is in your .gitmodules for the libuv url?
21:20 brrt the libuv repo was recently moved
21:20 FritzZaucker .act() doesn't help. I found​910562/why-has-no-one-written-a-threa​dsafe-branch-of-the-ncurses-library (there seems to be some thread support in version 5.7 and later).
21:20 nine brrt:
21:20 brrt and... it doesn't have that commit? that's weird
21:20 brrt oh
21:20 brrt did you reconfigure?
21:21 brrt it should do a git submodule sync
21:21 nine no?
21:21 hankache can someone kindly review this new chapter on parallelism
21:21 brrt oh, you should
21:21 hankache your comments would be highly appreciated
21:22 nine brrt: works :)
21:22 brrt ok, well, fortunately :-)
21:22 hankache i hope what i wrote makes sense
21:23 brrt is there a way to make nativecall look in the current directory for a lib
21:24 Skarsnik is native('./mylib') ?
21:24 Su-Shee left #perl6
21:24 brrt hmm
21:24 nine it will look for ./
21:24 Skarsnik Or you can always set LD_LIBRARY_PATH on unix
21:25 brrt oh well, that works
21:25 brrt thanks
21:25 pierre-v_ joined #perl6
21:25 AlexDaniel hankache: “@result>>.say;” “If you run both examples, you should notice that one is sorted and the other is not.”
21:26 AlexDaniel hankache: both are not sorted?
21:26 AlexDaniel hankache: because you are using ».say
21:26 RabidGravy FritzZaucker, also doing clear() in the tap before the mvaddstr seems to help
21:26 Skarsnik it add the lib on a path nine? you patched the test of NC then x)
21:26 autarch if I have a string that contains a regex, how do I turn it into a regex without escaping the metacharacters it contains?
21:27 Ulti hankache one of the common tasks is to turn a normal for loop into a parallel for loop I think the idiom is still something like: await do for { start {} }
21:28 MadcapJake i'm not seeing any tests for command line flags in perl6/doc
21:28 RabidGravy FritzZaucker, I think you could probably get it right if you take care with what you do with curses
21:28 yurivish_ joined #perl6
21:28 RabidGravy (i.e it's up to 220 and still readable)
21:29 FritzZaucker RabidGravy: thanks, clear() indeed seems to work. I'll experiment a bit more then.
21:29 AlexDaniel I wonder how many examples out there are doing ».say
21:30 espadrine joined #perl6
21:31 llfourn joined #perl6
21:31 RabidGravy FritzZaucker, it's probably the curses internal screen map getting messed up (due to not being threadsafe) and the clear() fixes it
21:32 lizmat MadcapJake: perl6/doc contains tests ?
21:33 FritzZaucker RabidGravy: meaning that I would have to re-draw the entire screen inside the timer callback to be on the save side?
21:33 RabidGravy autarch <$var>
21:34 RabidGravy FritzZaucker, that's the best guess I can come up with at the moment
21:34 RabidGravy haven't used curses in anger this century ;-)
21:35 hankache AlexDaniel what is the problem with >>.say ?
21:35 MadcapJake lizmat: well none of them seem related to actually the p6doc command :P
21:35 AlexDaniel hankache: because it is unordered
21:35 Ulti hankache the result of >> is ordered but whilst its executing its not
21:35 itaipu joined #perl6
21:35 AlexDaniel hankache: you are trying to demonstrate the difference between hyper and race, yet you are printing the results in random order…
21:35 Ulti so .say executes out of order returning a bunch of trues "in order"
21:36 Skarsnik Good night here ^^
21:37 hankache oh i see i, so >>. is the culprit
21:37 nine brrt: everything I care about seems to work just fine :)
21:37 brrt what platform are you on?
21:37 nine openSUSE
21:37 autarch RabidGravy: thanks
21:37 FritzZaucker RabidGravy: thanks! Very nice to get answers to such simple-minded questions while you guys are extremly busy fixing and working on many much more fundamental things ...
21:37 brrt okay thanks
21:38 brrt i'm looking for windows / osx tests, too
21:38 orq joined #perl6
21:38 orq hi
21:38 brrt hi orq
21:38 nine Skarsnik: t/04-nativecall/04-pointers.t is failing: # Error: In 'TakeTwoPointersToInt' routine declaration - Not an accepted NativeCall type for parameter [1]  : NativeCall::Types::Pointer[Int]
21:39 Skarsnik My mad, I was to angry at git to rebuild rakudo/nqp/moar x)
21:39 Skarsnik *bad
21:40 Skarsnik I added check for the content of Pointer
21:40 hankache AlexDaniel Ulti but if i replace it with a foor loop it is still unordered!
21:40 Ulti hankache what code?
21:40 hankache AlexDaniel Ulti what should i use to *print in oder* ?
21:40 AlexDaniel .say for @arr
21:40 Ulti hankache I was suggesting the for loop construct in addition to the other things you have in the tutorial
21:41 RabidGravy FritzZaucker, I'm just hacking some stuff upp so I can have blinken lights an a raspberry pi :)
21:41 FritzZaucker RabidGravy: :-)
21:41 Ulti hankache well you can't print "in order" whilst concurrently executing thats kind of the point you can print the result which is reordered to match the input order
21:41 Skarsnik nine, etheir change this test to not have a Pointer[Int] or remove that in the sanity check routine x)
21:42 brrt anyway, i'll try tomorrow
21:42 brrt i'm a bit too sleepy now
21:43 hankache AlexDaniel .say for @arr still not ordered
21:43 Ulti hankache the difference isn't that one ensures the order of execution it only ensures the order of the result of execution
21:43 Ulti hankache: just a normal for loop?
21:43 Ulti erk
21:43 hankache Ulti yes but I stored it in an array
21:43 Ulti stored what in an array?
21:43 Ulti the result?
21:44 hankache yes
21:44 Ulti that should work
21:44 hankache so it should be stored in order
21:44 acrussell joined #perl6
21:44 Ulti throw in a m: here maybe?
21:44 hankache m: my @array = (1..100); my @result = {$_ + 1} ); .say for @result;
21:44 camelia rakudo-moar 3dc7d5: OUTPUT«66␤67␤68␤69␤70␤71␤72␤73␤74␤75␤76␤7​7␤78␤79␤80␤81␤82␤83␤84␤85␤86␤87␤88␤89␤90␤​91␤92␤93␤94␤95␤96␤97␤98␤99␤100␤101␤2␤3␤4␤​5␤6␤7␤8␤9␤10␤11␤12␤13␤14␤15␤16␤17␤18…»
21:45 AlexDaniel m: my @array = (1..100); my @result = {$_ + 1} ); .say for @result
21:45 camelia rakudo-moar 3dc7d5: OUTPUT«2␤3␤4␤5␤6␤7␤8␤9␤10␤11␤12␤13␤14␤15​␤16␤17␤18␤19␤20␤21␤22␤23␤24␤25␤26␤27␤28␤​29␤30␤31␤32␤33␤34␤35␤36␤37␤38␤39␤40␤41␤4​2␤43␤44␤45␤46␤47␤48␤49␤50␤51␤52␤53␤54…»
21:45 AlexDaniel m: my @array = (1..100); my @result = {$_ + 1} ); .say for @result
21:45 camelia rakudo-moar 3dc7d5: OUTPUT«2␤3␤4␤5␤6␤7␤8␤9␤10␤11␤12␤13␤14␤15​␤16␤17␤18␤19␤20␤21␤22␤23␤24␤25␤26␤27␤28␤​29␤30␤31␤32␤33␤34␤35␤36␤37␤38␤39␤40␤41␤4​2␤43␤44␤45␤46␤47␤48␤49␤50␤51␤52␤53␤54…»
21:46 Skarsnik nine, wait travis build does not run make test in rakudo? It was showing no error
21:46 lizmat hmmm... that looks like hyper and race are reversed ?
21:46 Ulti m: my @array = (1..100); my @result = eager {$_ + 1} ); .say for @result;
21:46 camelia rakudo-moar 3dc7d5: OUTPUT«2␤3␤4␤5␤6␤7␤8␤9␤10␤11␤12␤13␤14␤15​␤16␤17␤18␤19␤20␤21␤22␤23␤24␤25␤26␤27␤28␤​29␤30␤31␤32␤33␤34␤35␤36␤37␤38␤39␤40␤41␤4​2␤43␤44␤45␤46␤47␤48␤49␤50␤51␤52␤53␤54…»
21:46 Ulti need the eager
21:46 Ulti which makes it look like the race isnt racing...
21:46 Ulti or is eager
21:47 webstrand Is there any meaning to the slurpy operator in assignment? For instance:`my (*@a) = (1,2),(3,4)` seems like it should flatten (and one tutorial say that it does)
21:47 AlexDaniel lizmat: yea, I don't get it. What's wrong with hyper?
21:47 hankache i think hyper is drunk
21:47 lizmat insufficiently tested ?  :-)
21:47 dalek rakudo/nom: adb1608 | (Stefan Seifert)++ | t/04-nativecall/04-pointers.t:
21:47 dalek rakudo/nom: Fix test using invalid types for NativeCall
21:47 dalek rakudo/nom:
21:47 dalek rakudo/nom: Pointer to Int doesn't make sense in NativeCall, since Int is no C type.
21:47 dalek rakudo/nom: Use Pointer[int32] instead as that will translate to int* in C.
21:47 dalek rakudo/nom: review:
21:47 AlexDaniel lizmat: I could understand it if race is not doing it randomly for some reason, but hyper?
21:47 rburkholder joined #perl6
21:48 hankache AlexDaniel if you use a bigger range race is unorderd so that's fine
21:48 dalek doc: b4efa2c | lizmat++ | doc/Type/DateTime.pod:
21:48 dalek doc: Make :timezone always be a named (optional) param
21:48 dalek doc: review:
21:48 hankache i tested on latest rakudo on my machine
21:49 Ulti hankache then the answer there is you are late binding the map rather than getting the "list" not sure that is how it should be working though
21:51 Ulti my @numbers = await do for 1..10 -> $n { start { return $n } }; say @numbers;
21:51 Ulti m: my @numbers = await do for 1..10 -> $n { start { return $n } }; say @numbers;
21:51 camelia rakudo-moar adb160: OUTPUT«Attempt to return outside of any Routine␤  in block <unit> at /tmp/s_28_CAlJs line 1␤␤»
21:51 Ulti orly
21:51 hankache so is hyper buggy or are we completely doing it wrong?
21:52 Ulti hankache with the eager it does what is sort of expected
21:52 Ulti might be a bug
21:52 Ulti someone who knows the one true nature of Perl 6 will have to answer
21:52 AlexDaniel TimToady: perhaps “we” can answer
21:53 geraud joined #perl6
21:53 Ulti hankache also the example you give is a bit more natural with a parallel grep which is currently not working... iirc
21:54 anaeem1 joined #perl6
21:55 hankache Ulti eager doesn't solve it
21:55 Ulti orly? did just now :S
21:55 hankache try it with big sets on your local machine
21:56 hankache try this: my @array = (1..10000); my @result = eager {$_ + 1} ); .say for @result;
21:57 Ulti m: my @array = (1..5); my @result = eager {sleep $_; $_ + 1 }); .say for @result;
21:57 nine .tell Skarsnik does perl6 -e 'use Test::More:from<Perl5>; diag("foo")' # print "# foo"? If yes, we have established that passing strings from Perl 6 to Perl 5 works and it's the other way around that's the issue
21:57 yoleaux nine: I'll pass your message to Skarsnik.
21:57 AlexDaniel m: my @result = ^1000 * + 10; say @result ~~ @result.sort
21:58 nine Good night all!
21:58 camelia rakudo-moar adb160: OUTPUT«2␤3␤4␤5␤6␤»
21:58 camelia rakudo-moar adb160: OUTPUT«False␤»
21:58 AlexDaniel m: my @result = eager ^1000 * + 10; say @result ~~ @result.sort
21:58 camelia rakudo-moar adb160: OUTPUT«False␤»
21:58 Ulti hmm wait a second that wont work
21:58 AlexDaniel m: my @result = eager { ^1000 * + 10 }; say @result ~~ @result.sort
21:58 camelia rakudo-moar adb160: OUTPUT«False␤»
21:58 Ulti m: my @array = (1..5); my @result = eager {sleep 5 - $_; $_ + 1 }); .say for @result;
21:58 camelia rakudo-moar adb160: OUTPUT«2␤3␤4␤5␤6␤»
21:58 Ulti nope still works
21:59 Ulti that should be reverse sleep sorting them
21:59 AlexDaniel m: my @result = eager ^5 * + 10; say @result ~~ @result.sort
21:59 camelia rakudo-moar adb160: OUTPUT«True␤»
21:59 AlexDaniel Ulti: 1..5 is not enough
21:59 Ulti m: my @array = (1..5); my @result = eager {sleep 5 - $_; $_ + 1 }); .say for @result;
21:59 camelia rakudo-moar adb160: OUTPUT«2␤3␤4␤5␤6␤»
21:59 Ulti AlexDaniel what with a difference in seconds of their execution time?
22:00 hankache m: my @array = (1..10000); my @result = eager {$_ + 1} ); .say for @result;
22:00 camelia rakudo-moar adb160: OUTPUT«2␤3␤4␤5␤6␤7␤8␤9␤10␤11␤12␤13␤14␤15​␤16␤17␤18␤19␤20␤21␤22␤23␤24␤25␤26␤27␤28␤​29␤30␤31␤32␤33␤34␤35␤36␤37␤38␤39␤40␤41␤4​2␤43␤44␤45␤46␤47␤48␤49␤50␤51␤52␤53␤54…»
22:00 Ulti or is that micro seconds
22:00 AlexDaniel Ulti: true-true
22:00 Ulti though a bit odd both of them are sorted
22:00 hankache run this ^^ on your machine it will not be sorted
22:01 hankache try on using your local copy of rakudo
22:01 Ulti comes out sorted on my machine hankache
22:02 Ulti hankache are you on the 2015.12 release or the latest commit?
22:02 AlexDaniel m: my @array = (1..10000); my @result = eager {$_ + 1} ); say @result ~~ @result.sort; # are you sure?
22:02 camelia rakudo-moar adb160: OUTPUT«False␤»
22:02 hankache Ulti this ==> ?? my @array = (1..10000); my @result = eager {$_ + 1} ); .say for @result;
22:02 Ulti yup
22:02 AlexDaniel Ulti: even on camelia it is not sorted
22:02 Ulti just copy pasted that direct into my p6 and it came out ordered, might be its only dual core
22:02 hankache Ulti This is Rakudo version 2015.12-46-gd08e043 built on MoarVM version 2015.12
22:03 RabidGravy m: my @array = (1..10000); my @result-a = eager {$_ + 1} ); .my @result-b ={$_ + 1}); say @result-a ~~ @result-b;
22:03 camelia rakudo-moar adb160: OUTPUT«5===SORRY!5=== Error while compiling /tmp/iMPcFG8_Ww␤Two terms in a row␤at /tmp/iMPcFG8_Ww:1␤------> 3 eager {$_ + 1} ); .my7⏏5 @result-b ={$_ + 1}); say @␤    expecting any of:␤        infix␤        infi…»
22:03 RabidGravy m: my @array = (1..10000); my @result-a = eager {$_ + 1} ); my @result-b ={$_ + 1}); say @result-a ~~ @result-b;
22:03 camelia rakudo-moar adb160: OUTPUT«False␤»
22:03 RabidGravy ~~
22:03 Arrowhead joined #perl6
22:04 AlexDaniel it is pretty clear that it does not really work…
22:04 FritzZaucker joined #perl6
22:04 Ulti yeah there is defo something not right
22:04 hankache Haa!
22:04 bitmap joined #perl6
22:04 AlexDaniel “The order of elements are preserved.” – oops, “is preserved”
22:04 hankache hyper == drunk
22:05 Ulti m: my @array = (1..10000); my @result-a = eager {$_ + 1} ); .my @result-b ={$_ + 1}); say @result-a ~~ @result-b;
22:05 camelia rakudo-moar adb160: OUTPUT«5===SORRY!5=== Error while compiling /tmp/lHea3pq7Fl␤Two terms in a row␤at /tmp/lHea3pq7Fl:1␤------> 3= eager {$_ + 1} ); .my7⏏5 @result-b ={$_ + 1}); say @␤    expecting any of:␤        infix␤        infi…»
22:06 Ulti m: my @array = (1..10000); my @result-a = eager {$_ + 1} ); my @result-b ={$_ + 1}); say @result-a ~~ @result-b;
22:06 camelia rakudo-moar adb160: OUTPUT«False␤»
22:06 hankache AlexDaniel “The order of elements are preserved.” – oops, “is preserved” where in my text ??
22:07 dalek doc: 276e344 | (Aleks-Daniel Jakimenko-Aleksejev)++ | doc/Type/Iterable.pod:
22:07 dalek doc: “Order are preserved” → “Order is preserved”
22:07 dalek doc: review:
22:07 AlexDaniel hankache: not in your text!
22:07 hankache ahhh
22:07 spider-mario joined #perl6
22:08 hankache Blasphemy ;)
22:09 webstrand I'm guessing, since S02 says that declarators are parsed as signatures, that the signature information is discarded when the declarator is used in list context. Am I misunderstanding anything?
22:11 masak tadzik, lizmat, @others: of interest --​f-compilers-build-systems-and-package-managers/
22:11 Rotwang If I have something like my $s = Supply.interval(5); whenever $s { .say; }
22:13 hankache joined #perl6
22:15 jdv79 im in @others i guess.  thanks.
22:15 jdv79 though i might forget to read that
22:16 integral joined #perl6
22:16 integral joined #perl6
22:20 hankache good night #perl6
22:20 hankache see ya tomorrow
22:20 jdv79 peace
22:22 * masak .oO( ☮ )
22:24 integral joined #perl6
22:24 stmuk joined #perl6
22:24 Praise joined #perl6
22:24 Praise joined #perl6
22:24 jdv79 i cant see that
22:25 smash joined #perl6
22:27 Jonis joined #perl6
22:27 Zero_Dogg joined #perl6
22:27 Zero_Dogg joined #perl6
22:28 lnx joined #perl6
22:28 amoe joined #perl6
22:28 Rotwang joined #perl6
22:28 jnthn joined #perl6
22:30 Arrowhead joined #perl6
22:32 llfourn joined #perl6
22:33 Idiosyncrat left #perl6
22:34 ugexe joined #perl6
22:34 spider-mario joined #perl6
22:37 lizmat .u ☮
22:37 yoleaux U+262E PEACE SYMBOL [So] (☮)
22:37 Rotwang Hello again
22:37 lizmat jdv79 ^^^
22:37 Rotwang this looks like a bug?
22:37 Rotwang m: react { Supply.interval(1).tap: -> { .say } }
22:37 camelia rakudo-moar adb160: OUTPUT«Unhandled exception in code scheduled on thread 4␤»
22:37 lucasb joined #perl6
22:38 perigrin joined #perl6
22:38 lizmat Rotwang: agree the error is LTA
22:39 lizmat m: react { whenever Supply.interval(1) { .say } }   # correct syntax
22:39 camelia rakudo-moar adb160: OUTPUT«(timeout)0␤1␤2␤3␤4␤5␤6␤7␤8␤9␤1​0␤11␤12␤13␤14␤15␤16␤17␤18␤19␤20␤21␤»
22:39 Rotwang lizmat: ok, thanks
22:39 lizmat Rotwang: could you rakudobug that ?
22:40 Rotwang lizmat: sure
22:40 lizmat Rotwang++
22:40 zostay m: my %v = x => rx/Foo/; my $smart = %v<x>; my $match = "Foo" ~~ $smart; dd $match
22:40 camelia rakudo-moar adb160: OUTPUT«Bool $match = Bool::True␤»
22:40 zostay shouldn't that be a Match?
22:40 zostay m: my %v = x => rx/Foo/; my $smart := %v<x>; my $match = "Foo" ~~ $smart; dd $match
22:40 camelia rakudo-moar adb160: OUTPUT«Bool $match = Bool::True␤»
22:41 lucasb zostay:​Bug/Display.html?id=127071
22:41 lucasb in other words, yes, I think it's a bug :)
22:42 Praise joined #perl6
22:42 Praise joined #perl6
22:43 jdv79 lizmat: thanks.  i guess irssi or connectbot or android or something else is fail.
22:43 zostay the reason Path::Router is broken is that... :(
22:44 jnthn joined #perl6
22:45 zostay m: my %v = x => /Foo/; my $smart := %v<x>; my $match = "Foo" ~~ $smart; dd $match
22:45 camelia rakudo-moar adb160: OUTPUT«Bool $match = Bool::True␤»
22:45 Jonis joined #perl6
22:45 stmuk joined #perl6
22:45 zostay m: my %v = x => /Foo/; my $smart = %v<x>; my $match = "Foo" ~~ $smart; dd $match
22:45 camelia rakudo-moar adb160: OUTPUT«Bool $match = Bool::True␤»
22:47 webstrand What's the difference between `constant x = *` and `my \x = *` ?
22:48 zostay m: my %v = x => /Foo/; my $smart = %v<x>; my $match = given $smart { when Regex { "Foo" ~~ /$smart/ } default { "Foo" ~~ $smart } }; dd $match
22:48 camelia rakudo-moar adb160: OUTPUT«5===SORRY!5===␤Word 'given' interpreted as a listop; please use 'do given' to introduce the statement control word␤at /tmp/6_yTptgEKs:1␤------> 3o/; my $smart = %v<x>; my $match = given7⏏5 $smart { when Regex { "Foo" ~~ /$smart/␤Unexpecte…»
22:48 dj_goku So I am building a enum, but the zeroth item isn't used so I put Nil. Is there a way to say start at 1 vs 0 indexing? What is the best way to skip an item?
22:48 zostay m: my %v = x => /Foo/; my $smart = %v<x>; my $match = do given $smart { when Regex { "Foo" ~~ /$smart/ } default { "Foo" ~~ $smart } }; dd $match
22:48 camelia rakudo-moar adb160: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Xw_8GSLKAb␤Strange text after block (missing semicolon or comma?)␤at /tmp/Xw_8GSLKAb:1␤------> 3smart { when Regex { "Foo" ~~ /$smart/ }7⏏5 default { "Foo" ~~ $smart } }; dd $matc␤»
22:48 zostay m: my %v = x => /Foo/; my $smart = %v<x>; my $match = do given $smart { when Regex { "Foo" ~~ /$smart/ }; default { "Foo" ~~ $smart } }; dd $match
22:48 camelia rakudo-moar adb160: OUTPUT«Match $match = => Any, list => (), hash =>, orig => "Foo", to => 3, from => 0)␤»
22:48 dj_goku by skip I mean say the 5 item isn't used?
22:48 telex joined #perl6
22:48 zostay that is a horrible work-around...
22:48 dj_goku zostay: are you referring to me?
22:48 zostay no
22:48 zostay my convo with lucasb
22:49 dj_goku ahh.
22:49 dj_goku zostay == sterling right?
22:49 zostay sshhhh... ;^p
22:49 dj_goku :D
22:49 Notwang joined #perl6
22:50 Notwang I'm getting disconnected all the time
22:50 dj_goku Still in Manhattan?
22:50 stmuk joined #perl6
22:50 jnthn joined #perl6
22:50 Notwang here is what I want to do:
22:50 Jonis joined #perl6
22:50 zostay yes
22:50 integral joined #perl6
22:50 zostay you still in kc?
22:51 lnx joined #perl6
22:51 dj_goku zostay: cool, yeah. a few of us are trying to revive
22:51 smash joined #perl6
22:51 Notwang I want to have two "threads" and control one of them from the other
22:51 dj_goku With perl6 I am super excited! my day job I am writing python so I have kind of lost touch with perl.
22:52 lizmat Notwang: have you looked at Supply.throttle ?
22:52 Rotwang joined #perl6
22:53 Notwang lizmat: I think it is not what I am after
22:53 Notwang I'd like to close the socket only when certain condition happens in other "thread"
22:53 lizmat could you elaborate on what you *are* after?  :-)
22:53 amoe joined #perl6
22:54 lucasb m: say enum E (a=>1, |<b c d>); say E.enums
22:54 camelia rakudo-moar adb160: OUTPUT«,:b(2),:c(3),:d(4)))␤a => 1, b c d => 2␤»
22:54 lizmat perhaps a Promise with a then could do the trick ?
22:54 Notwang lizmat: lets say I'm testing if file exists periodically, and when the file dissapears I want to close the socket
22:54 Notwang that is simplified usecase
22:55 spider-mario joined #perl6
22:55 lucasb ^^ strange, the first line says it understand it's 4 elements, but then, the second line shows only 2 elements: "a" and "b c d"
22:55 lizmat make a Promise that will be kept if the file disappears, and have a then on that promise that will close the socket ?
22:56 Notwang lizmat: sounds good, I'll read more on that
22:56 dj_goku m: say enum E (a=>1, |<b c d>); say b.Numeric
22:56 camelia rakudo-moar adb160: OUTPUT«5===SORRY!5=== Error while compiling /tmp/wfyRE3m7aD␤Undeclared routine:␤    b used at line 1␤␤»
22:57 El_Che ian murdock:(
22:57 newbie joined #perl6
22:57 Notwang lizmat: however the docs on Async socket states that: "In order to close the underlying listening socket created by listen you can simply close the Supply."
22:58 Notwang so the question is how do I even close the supply?
22:58 gfldex m: say enum E (a=>1, 'b', 'c', 'd'); dd E
22:58 Notwang close method is not defined for it
22:58 camelia rakudo-moar adb160: OUTPUT«,:b(2),:c(3),:d(4)))␤E␤»
22:58 perlpilot Notwang: Supply.done IIRC
22:58 perlpilot or $supply.done
22:58 El_Che​015/12/christmas-came-bah-humbug.html <-- weird post, but probably somewhat reprentative
22:58 lizmat what perlpilot says
22:58 lucasb In other words, I was expecting the slip to just work there, inside enum lists, but it seems it doesn't.
22:59 lizmat El_Che: yeah, already commented on it
22:59 lizmat lucasb: perhaps worthy of a rakudobug
22:59 El_Che oh I see
22:59 dj_goku gfldex: ahh ha!
22:59 Notwang perlpilot: No such method 'done' for invocant of type 'Supply'
23:00 Notwang close is also not defined for Supply
23:00 lizmat ah, caught by the SSR  (Small Supply refactor(
23:00 lizmat you need to call .done on the Supplier
23:00 Notwang However state that close is available
23:00 * Notwang is very confused
23:01 * lizmat will fix the doc
23:01 perlpilot ah, yes.  $supplier.done
23:01 perlpilot lizmat++
23:01 Notwang lizmat: who is the supplier in this case?
23:01 dalek rakudo/nom: eca5df7 | lizmat++ | src/core/ (2 files):
23:01 dalek rakudo/nom: Make timezone *always* a named on
23:01 dalek rakudo/nom: review:
23:01 lizmat Notwang: good question :-)
23:03 RabidGravy oh I thought I had remove the close from Supply
23:04 Notwang RabidGravy: what did you give instead? ;f
23:05 RabidGravy no from the doc
23:05 RabidGravy what are you trying to do?
23:06 Notwang RabidGravy: general idea of what I want to do
23:07 ugexe how do i use a specific version of a module?
23:07 dj_goku weird
23:07 pierre-vigier joined #perl6
23:07 jdv79 ugexe: doesnt it work mow?
23:07 jdv79 now
23:07 dj_goku m: say enum E <:a(1) b c d>
23:07 camelia rakudo-moar adb160: OUTPUT«":a(1)" => 0,:b(1),:c(2),:d(3)))␤»
23:07 dj_goku m: say enum E <<:a(1) b c d>>
23:07 camelia rakudo-moar adb160: OUTPUT«,:b(2),:c(3),:d(4)))␤»
23:08 dj_goku m: say enum E <<:a(1) b :c(4) d>>
23:08 camelia rakudo-moar adb160: OUTPUT«,:b(2),:c(4),:d(5)))␤»
23:08 jdv79 :ver<0.011>
23:08 jdv79 on a use
23:08 ugexe m: use Test:ver<*>;
23:08 camelia rakudo-moar adb160: OUTPUT«===SORRY!===␤Could not find Test:ver<*> in:␤    /home/camelia/.perl6/2015.12-52-gadb1608␤    /home/camelia/rakudo-m-inst-1/share/perl6/site␤    /home/camelia/rakudo-m-inst-1/share/perl6/vendor␤    /home/camelia/rakudo-m-inst-1/share/perl6␤    C…»
23:08 RabidGravy Notwang, just put the last whenever inside the first and call $conn.close
23:08 jdv79 i think i tried it.  it worked.  but i couldnt install multiple vers yet.
23:09 ugexe does * not work for ver then?
23:09 jdv79 maybe dony do that for now:)
23:09 Notwang RabidGravy: thanks (also just came up with it)
23:09 Notwang RabidGravy: how can I create "whenever" blocks at runtime? is whenever equivalent to tap?
23:10 jdv79 its supposed to take anything that can smartmatch so i tbink it should work
23:10 jdv79 eventually
23:10 RabidGravy whenever is like tap in some ways yes
23:10 Notwang okay, thanks
23:11 RabidGravy it just covers a frequent use case
23:11 ugexe use CSV::Parser:auth<github:tony-o>; does not work either
23:12 RabidGravy you could just put a "{$conn.close})" in outer whenever
23:12 jdv79 nine might be the only one at this point that knows
23:12 ugexe i thought avoiding dependency hell was suppose to be a key point before release
23:13 jdv79 we have some serious bus factor factors round here
23:13 jdv79 who said that?
23:14 jdv79 cur landed slightly too late so id say no to that:(
23:14 jdv79 nobody's fault...
23:15 lizmat .oO( too many takers, not enough givers )
23:15 jdv79 we all suck:(  sorry.
23:15 ugexe there were other givers that were turned down
23:16 lizmat ugexe: ??
23:16 jdv79 kudos to those that tried and or did though!
23:16 ugexe tony-o explicity offered
23:17 lizmat well, if there's anything I've learned here, is that you should do instead of offer
23:17 lizmat and be prepared to flush it down the drain if it isn't accepted after alle
23:17 lizmat *all
23:18 ugexe he did, and then boilerplate was merged
23:18 ugexe and thus he did flush it
23:19 dalek rakudo-star-daily: d2f859a | coke++ | log/ (8 files):
23:19 dalek rakudo-star-daily: today (automated commit)
23:19 dalek rakudo-star-daily: review:​-star-daily/commit/d2f859a0d4
23:20 jdv79 good news is its barely tested so it might be fixable under cover of 6.c:)
23:20 devans2 joined #perl6
23:22 devans2 hiya, so i've installed perl6 and gotten some modules to install, is there a perl6 version of perldoc that i can run from the command line?
23:22 RabidGravy p6doc
23:22 dj_goku devans2:
23:22 RabidGravy you'll need to install it like "panda install p6doc"
23:23 devans2 ah thank you
23:23 RabidGravy it'll install the core documentation, I'm not quire sure what the state of getting the module docs is right at this moment
23:24 jdv79 last i checked it was not goid.
23:24 jdv79 good
23:24 araujo joined #perl6
23:24 jdv79 as is the path was needed.  not the name.
23:24 lizmat looking at the Supply doc, it seems to have missed an overhaul after the SSR
23:25 devans2 yeah, looks like p6doc fails to pass the tests on install, guess that's an opportunity to learn some more perl6
23:25 jdv79 ussr!
23:25 lizmat it will probably have to wait until jnthn is back from R&R
23:25 jdv79 whats a ssr!
23:25 lizmat Small Supply Refactor
23:25 RabidGravy lizmat, I looked at it and took out some stuff
23:25 jdv79 i keep hitting ! instead of ?
23:25 jdv79 sorry
23:25 lizmat aka the splitting off of Supplier
23:26 jdv79 ah, that.
23:26 lizmat I need to think about this after a good night sleep
23:26 lizmat so good night, #perl6 !
23:26 RabidGravy yeah, I moved the emit and stuff to a new Supplier doc
23:27 RabidGravy and changed the introductory text to reflect the new world order but may not have reflected all the changes
23:28 pmurias joined #perl6
23:33 ZoffixWin joined #perl6
23:34 ZoffixWin flussence, can you press CTRL+F5 and try again? I fixed the wrapon-two-lines thing this morning in​/cbb9efa09230f5117be84d8ae6fab3c8c0d7c54a
23:34 ZoffixWin flussence, and if the issue is still there... what browser/OS are you using (and screenshot, if possible). I'll fix it tomorrow morning :)
23:34 flussence looks much better now :D
23:34 ZoffixWin Ah \o/
23:35 llfourn joined #perl6
23:35 dj_goku ZoffixWin: just who I wanted to see!
23:36 ZoffixWin orly :)
23:36 ZoffixWin Why? :)
23:36 jdv79 lizmat: later
23:36 flussence the right edge of that blue box in the announcement looks slightly weird to me because of the way the box shadow's placed, I think that's just me being picky though
23:36 dj_goku ZoffixWin: so I saw your mr for lxymin was merged. I wanted to add a few examples of enum.
23:36 dj_goku I guess I should just open a new PR
23:37 ZoffixWin dj_goku, yeah
23:37 ZoffixWin vendethiel / Ven has merge powers, I believe.
23:37 vendethiel hi?
23:37 vendethiel (either will highlight me on both clients :).)
23:37 ZoffixWin :P
23:37 dj_goku lol
23:38 vendethiel dj_goku: yeah, I do the merges on lxiny often
23:38 dj_goku Alright I'll a few enum examples.
23:38 dj_goku and submit a PR!
23:40 AlexDaniel hmm, nobody commented on hyper issue and nobody submitted a bug report
23:41 AlexDaniel I'll submit it then… if it is supposed to be this then ok, at least I'll get a nice email notification about that :)
23:41 ZoffixWin Hyper issue?
23:41 AlexDaniel m: my @result = eager ^10000 * + 10; say @result ~~ @result.sort
23:41 camelia rakudo-moar eca5df: OUTPUT«False␤»
23:41 AlexDaniel m: my @result = eager ^5 * + 10; say @result ~~ @result.sort
23:41 camelia rakudo-moar eca5df: OUTPUT«True␤»
23:41 ZoffixWin :o
23:41 AlexDaniel hmm eager is not needed here
23:41 AlexDaniel m: my @result = ^10000 * + 10; say @result ~~ @result.sort
23:41 camelia rakudo-moar eca5df: OUTPUT«False␤»
23:42 ZoffixWin m: my @result = ^10000 .map: * + 10; say @result ~~ @result.sort
23:42 AlexDaniel this should give True because hyper is supposed to give the results in the same order
23:42 camelia rakudo-moar eca5df: OUTPUT«True␤»
23:42 AlexDaniel at least that's how I understand it
23:42 ZoffixWin Right
23:43 AlexDaniel ZoffixWin: what did you win?
23:43 ZoffixWin Win == Windows. My gaming "console" :P
23:44 AlexDaniel Why not use “ZoffixWindows” then :D
23:44 ZoffixWin Are you sure it's .hyper that's the issue?
23:45 ZoffixWin m: my @result = ^10 .hyper(:1batch).map: *+0; say @result ~~ @result.sort
23:45 camelia rakudo-moar eca5df: OUTPUT«False␤»
23:45 AlexDaniel ZoffixWin: I have no idea
23:45 ZoffixWin m: my @result = ^10 .hyper(:1batch).map: *+0; say [@result, @result.sort]
23:45 camelia rakudo-moar eca5df: OUTPUT«[[0 2 1 3 4 5 6 8 9 7] (0 1 2 3 4 5 6 7 8 9)]␤»
23:45 AlexDaniel something is definitely wrong…
23:45 ZoffixWin Oh, right, order's messed up
23:45 * ZoffixWin had a few beers and is blind
23:46 ZoffixWin m: my $wrong; for ^10 { my @result = ^10 *+0; $wrong++ unless @result ~~ @result.sort }; say $wrong
23:46 camelia rakudo-moar eca5df: OUTPUT«(Any)␤»
23:46 ZoffixWin m: my $wrong; for 1..10 { my @result = ^10 *+0; $wrong++ unless @result ~~ @result.sort }; say $wrong
23:46 camelia rakudo-moar eca5df: OUTPUT«(Any)␤»
23:47 ZoffixWin At least it's consistent at being wrong :P
23:47 ZoffixWin or wait
23:47 AlexDaniel I don't think so
23:47 AlexDaniel but maybe it is
23:48 ZoffixWin m: my $wrong; for 1..10 { .say; my @result = ^100 *+0; $wrong++ unless @result ~~ @result.sort }; say $wrong
23:48 camelia rakudo-moar eca5df: OUTPUT«1␤2␤3␤4␤5␤6␤Memory allocation failed; could not allocate 15016 bytes␤»
23:48 ZoffixWin Interesting.
23:48 ZoffixWin eh... night all :)
23:49 AlexDaniel no, it is not consistent
23:50 flussence that last one prints out 1..10 and 6 on separate lines here