Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2015-07-22

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:08 telex joined #perl6
00:08 bartolin_ joined #perl6
00:11 laouji joined #perl6
00:13 yx_ joined #perl6
00:13 lizmat joined #perl6
00:27 BenGoldberg joined #perl6
00:37 rangerprice joined #perl6
00:37 cognominal joined #perl6
00:38 timotimo that seems like a worthwhile optimization
00:38 ugexe quick blog http://ugexe.com/multiplexing-stdout-from-multiple-processes/
00:39 laouji joined #perl6
00:40 timotimo ugexe: why is your longest-basename so complicated?
00:40 timotimo why not just .max(:by(*.chars))?
00:40 timotimo m: <a bb ccc ddd 11 22 33 4444>.max(:by(*.chars))
00:40 camelia rakudo-moar e41624: ( no output )
00:40 timotimo m: <a bb ccc ddd 11 22 33 4444>.max(:by(*.chars)).say
00:40 camelia rakudo-moar e41624: OUTPUT«ddd␤»
00:40 timotimo that's not the right one
00:41 laouji joined #perl6
00:41 dalek roast: 0b3aa5b | skids++ | S02-types/ (5 files):
00:41 dalek roast: Unfudge tests that were failing for the wrong reason.
00:41 dalek roast: (they depended  on .grep of baggy/setty things which lizmat++ just fixed)
00:41 dalek roast: review: https://github.com/perl6/roast/commit/0b3aa5b5dc
00:41 timotimo m: <a bb ccc ddd 11 22 33 4444>.max(*.chars).say
00:41 camelia rakudo-moar e41624: OUTPUT«4444␤»
00:41 ugexe heh i know there were better options, but after a few minutes of trying more elegant "solutions" that didnt work i just went with the ol reduce
00:41 timotimo there we go
00:42 timotimo the max method takes it as a positional
00:43 timotimo what do you mean by "catch a signal from key strokes"?
00:45 ugexe like alt+1
00:45 timotimo ah
00:45 timotimo we do have a termios (or something?) module
00:47 ugexe lucky for me no external dependencies or nativecall stuff are both goals
00:53 timotimo oh
00:53 timotimo so you don't really have a way to turn off echoing from the terminal
00:53 timotimo other than that, capturing such keystrokes wouldn't be a problem
00:53 timotimo you could just blockingly read from stdin in a start { } blocks
00:54 ugexe i think there is a bug in $*IN right now regarding that
00:54 timotimo oh :(
00:54 ugexe echoing i can possibly work around using \r or \b
00:54 timotimo except apparently \r turns into \n on windows >_<
00:55 ugexe yea, i had to rewrite my statusbar thing to use \b because of that... what a pita
00:56 ugexe jvm windows handles \r ok though
00:56 b2gills ugexe: you only updated the first occurrence of `@basenames.reduce({ $^a.chars > $^b.chars ?? $^a !! $^b })`
00:57 ugexe must be cached... i changed it separately but immediately after
00:57 lizmat joined #perl6
00:58 ugexe not really sure how cloudflare works
00:58 b2gills I wondered, yes it is fixed
01:08 colomon Too many symbols provided for categorical of type infix; needs only 1 ???
01:09 timotimo do you have a space in your infix:<foo bar>?
01:10 * timotimo disappears into bed
01:11 dalek rakudo-star-daily: 39d509a | coke++ | log/ (9 files):
01:11 dalek rakudo-star-daily: today (automated commit)
01:11 dalek rakudo-star-daily: review: https://github.com/coke/rakudo-star-daily/commit/39d509a503
01:12 fling joined #perl6
01:16 yeahnoob joined #perl6
01:17 Ben_Goldberg joined #perl6
01:19 rmgk_ joined #perl6
01:26 atroxaper joined #perl6
01:33 Hor|zon joined #perl6
01:38 tinyblak joined #perl6
01:45 Ben_Goldberg joined #perl6
01:48 cognominal joined #perl6
02:03 noganex joined #perl6
02:14 raiph joined #perl6
02:14 colomon timotimo++ # indeed, that was it.  certainly didn’t mean to type it, and didn’t notice it was there in the middle of the runes.  ;)
02:26 rangerprice joined #perl6
02:49 kaare_ joined #perl6
03:22 Hor|zon joined #perl6
03:33 aborazmeh joined #perl6
03:33 aborazmeh joined #perl6
04:07 atroxaper joined #perl6
04:25 aborazmeh joined #perl6
04:25 aborazmeh joined #perl6
04:28 vendethiel joined #perl6
04:31 tinyblak joined #perl6
04:32 tinyblak joined #perl6
04:35 laouji joined #perl6
04:36 skids .tell hoelzro RE RT#125657 the example test.pl overrides Exception.gist, which is what prints out backtraces, which is why you do not see backtraces. (override .message instead).
04:36 yoleaux skids: I'll pass your message to hoelzro.
04:36 synbot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=125657
04:57 mr-foobar joined #perl6
05:00 FROGGS joined #perl6
05:03 skids locally: perl6 -e 'sub a { try die("foo"); fail($!) }; a(); 1;'
05:03 skids foo
05:03 skids in sub a at -e:1
05:03 skids in block <unit> at -e:1
05:03 skids Actually thrown at:
05:03 skids in block <unit> at -e:1
05:04 skids \o/
05:11 Hor|zon joined #perl6
05:16 llfourn joined #perl6
05:34 nwc10 dear #perl6, if someone could write a version of prove that's better than Perl 5's, that would be Awesome. KTHXBAI
05:35 nwc10 to clarify "better" - Perl 5's maxes out a CPU core on the TAP parser, when trying to multiplex more than (about) 9 tests running in parallel
05:35 nwc10 anyway, the serious bit is "I'd hope that it's viable, and I don't think that I'd be the only person using it"
05:41 ugexe depends what you think of as viable. you could not do that for jvm yet for instance
05:41 ugexe well, not running parallel tests
05:42 ugexe otherwise sure
05:42 nwc10 I'm not sure why not. The thing runnig the tests "just" needs to spawn $n processes with output piped back, and then multiplex reading from their file handles.
05:42 nwc10 but anyway, MoarVM is more interesting to me, as its startup is better
05:43 dalek Inline-Perl5: f7b4564 | hoelzro++ | Build.pm:
05:43 dalek Inline-Perl5: Detect perls not configured with a shared library
05:43 dalek Inline-Perl5:
05:43 dalek Inline-Perl5: Inline::Perl5 can only work if perl was configured with useshrplib; if
05:43 dalek Inline-Perl5: a perl without this configuration option is used, the user gets a host
05:43 dalek Inline-Perl5: of compiler errors that they may or may not understand.  Bailing out
05:43 dalek Inline-Perl5: early in the build process tells the user exactly what is going on,
05:43 dalek Inline-Perl5: and what they need to do to fix it.
05:43 dalek Inline-Perl5: review: https://github.com/niner/Inline-Perl5/commit/f7b4564ae9
05:43 dalek Inline-Perl5: 309a6cb | niner++ | Build.pm:
05:43 dalek Inline-Perl5: Merge pull request #35 from hoelzro/master
05:43 dalek Inline-Perl5:
05:43 dalek Inline-Perl5: Detect perls not configured with a shared library
05:43 dalek Inline-Perl5: Thanks to hoelzro++
05:43 dalek Inline-Perl5: review: https://github.com/niner/Inline-Perl5/commit/309a6cb6b9
05:44 ugexe because if you try to pass that pipe to another thread it will break on jvm
06:00 FROGGS joined #perl6
06:45 domidumont joined #perl6
06:47 eMBee joined #perl6
06:48 amurf joined #perl6
06:51 domidumont joined #perl6
06:57 RabidGravy joined #perl6
07:00 Hor|zon joined #perl6
07:01 espadrine_ joined #perl6
07:03 nebuchadnezzar joined #perl6
07:06 RabidGravy morning!
07:14 FROGGS joined #perl6
07:32 Ven joined #perl6
07:33 darutoko joined #perl6
07:36 lizmat joined #perl6
07:38 dha joined #perl6
07:47 kaare__ joined #perl6
07:48 oetiker_ joined #perl6
07:48 zakharyas joined #perl6
07:49 RabidGravy struggling to golf this https://gist.github.com/jonathanstowe/a76c110ef81f0a218824 I may just RT it as is
07:50 rindolf joined #perl6
07:51 oetiker joined #perl6
07:54 dakkar joined #perl6
07:56 brrt joined #perl6
08:02 tinyblak joined #perl6
08:05 yakudza joined #perl6
08:11 zmikund joined #perl6
08:13 Hor|zon joined #perl6
08:20 cdc ugexe++ # blog post + creating a parallelized perl6 package manager
08:20 cdc tony-o++ # too, creating a parallelized perl6 package manager
08:22 brrt waitwhatdidimiss
08:22 yoleaux 21 Jul 2015 22:11Z <japhb> brrt: I don't understand -- were you saying you liked the wording of S27 SYNOPSIS item 2 ("kind, positive member of our community") or did *not* like it?  And if the latter, what would you prefer?
08:22 brrt .tell japhb i like it, i'm playing devils advocate for a bit
08:22 yoleaux brrt: I'll pass your message to japhb.
08:23 brrt .tell japhb it's an implicit assumption that everyone *can* be a positive, kind member, and that's not really true, e.g. the TempleOS guy
08:23 yoleaux brrt: I'll pass your message to japhb.
08:24 brrt .tell meisl if you want to ask questions, i'm typically available between 07:00 UTC and 15:00 UTC :-)
08:24 yoleaux brrt: I'll pass your message to meisl.
08:28 brrt also, i'm not that easily bothered, except by my own errorous code :-P
08:35 domidumont joined #perl6
08:36 * RabidGravy doesn't know what to do next
08:36 moritz RabidGravy: world domination!
08:37 RabidGravy I already rule the world, it's just the propaganda organisation has let me down a bit
08:38 masak antenoon, #perl6
08:39 RabidGravy marnin
08:40 inokenty joined #perl6
08:41 ifim joined #perl6
08:42 brrt \o masak
08:42 jnthn morning, #perl6
08:42 yoleaux 21 Jul 2015 22:55Z <TimToady> jnthn: on average, the lexers visit each character in the setting about 4.3 times, due to relexing at each level of proto/alternation (and that's not counting the extra pass to actually match the data), so I suspect we can speed up the parser considerably by not rerunning sublexers when a superlexer already determined how the sublexer would turn out
08:43 jnthn .tell TimToady Wow, yes...also if we figure out what bits of the declaratively matched stuff we need not run procedurally. If you have a good idea how to represent that and get the NFA runner itself to collect the info, I can probably figure out how to pass the info on downwards...
08:43 yoleaux jnthn: I'll pass your message to TimToady.
08:43 jnthn .tell TimToady where "represent that" is the fates of the sublexers...
08:43 yoleaux jnthn: I'll pass your message to TimToady.
08:43 abraxxa joined #perl6
08:43 RabidGravy I think I may take a crack at a libsamplerate binding, nice simple API and it's either going to be really quick and easy or it's going to cause everything to go on fire
08:44 RabidGravy jnthn, morning!
08:47 g4 joined #perl6
08:47 g4 joined #perl6
08:47 RabidGravy jnthn, does https://gist.github.com/jonathanstowe/a76c110ef81f0a218824 have enough to put in an RT for a possible pre-compilation bug? Tricky to golf
08:49 jnthn RabidGravy: Do you have a simple usage example of how using OO::Monitors can trigger this?
08:49 jnthn RabidGravy: Or is that still "the whole dependent module"?
08:50 jnthn (I don't expect you to golf OO::Monitors; that's mine, and a MOP module...)
08:50 JimmyZ_ joined #perl6
08:52 RabidGravy It's tricky because it only manifests when the module that uses OO::Monitors  (Libshout) *and* another one are precompiled :-\
08:53 darutoko- joined #perl6
08:54 jnthn Oh... :(
08:55 jnthn That does make it more "fun"...
09:01 RabidGravy the *actual* code that gave rise to that is https://github.com/jonathanstowe/Audio-Libshout/blob/master/examples/streamfile-encode
09:01 RabidGravy works perfectly otherwise
09:05 jnthn OK, then it's going to be quite a bug hunt for me, I fear...
09:07 spacebat joined #perl6
09:11 cdc RabidGravy: maybe your problem is just another manifestation of RT #125634
09:11 synbot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=125634
09:11 RabidGravy I'm sure it is reproducible in a smaller bit of code, I'm just buggered to know where to start :-)
09:12 jnthn cdc: In the best case, yeah. (Best in so far as "I've already started golfing that one")
09:14 cdc strange, searching for "REPR: NULL, cs = 0" in rt.perl.org returns nothing
09:14 cdc whereas searching for "REPR: NULL, cs =" returns one item
09:15 RabidGravy I did one last night, which was *not* pre-compilation associated
09:17 RabidGravy which is as far as I can tell "apply_trait not reporting type not found and crapping out instead"
09:19 espadrine_ joined #perl6
09:20 tinyblak joined #perl6
09:20 brrt when is the SPW actually?
09:21 jnthn Switzerland
09:21 jnthn oh
09:21 jnthn Late August
09:21 RabidGravy LObster!
09:21 * jnthn should learn to read
09:21 brrt i see... really just a few days before YAPC::EU
09:21 jnthn Yeah
09:22 brrt switzerland is time-associated :-0
09:22 brrt :-)
09:22 * brrt should learn to type
09:22 jnthn Very much so...
09:22 jnthn Switzerland and Japan are the only countries where I'd consider setting my watch by the time a train leaves... :)
09:24 xiaomiao China is doing pretty well too
09:25 xiaomiao possibly because the highspeed train network is a military asset
09:26 jnthn xiaomiao: I can imagine that... The last time I did a lot of train travel in China was before they'd built all the highspeed trains :)
09:26 * moritz wouldn't trust the German military to run trains on time
09:26 xiaomiao moritz: but they are family-friendly
09:26 jnthn xiaomiao: But it still all ran pretty well. :)
09:26 xiaomiao jnthn: the HST system works very nicely
09:26 RabidGravy on the other hand no-one could do a worse job than Southeastern Trains here
09:26 jnthn xiaomiao: Yeah, I want to go try it some time :)
09:27 jnthn RabidGravy: Connex? :)
09:27 xiaomiao RabidGravy: Belgium ? :D
09:27 jnthn Sweden? :)
09:27 brrt a very vulnerable military target, though
09:27 xinming joined #perl6
09:27 brrt s/target/asset/
09:28 RabidGravy jnthn, Connex yeah, but it's all the same people just dfferent ownership
09:28 RabidGravy :)
09:29 jnthn Ah, that old trick...
09:30 * masak .oO( That Old Trick -- brought to you by... Connex )
09:31 jnthn brought to you *very late* by Connex :P
09:31 masak late as in "the late Dentarthurdent"
09:33 mathw hello
09:34 jnthn o/ mathw
09:35 bin_005 joined #perl6
09:35 mathw I am horribly out of touch with Perl 6 - I've been learning Elm. How's it going?
09:36 brrt awesome
09:36 brrt how is elm? :-)
09:36 brrt jnthn++ is fixing lots of issues, masak++ is reporting more
09:37 jnthn brrt++ is making our JIT betterer
09:37 brrt module ecosystem is growing quickly by more authors than i can think of
09:37 brrt more importantly, said authors are starting to really scratch their own itches, implying that they find it useful for 'productionish' work
09:37 jnthn We have NFG! (If you were away long enough to not know...)
09:37 brrt oh yes, we do
09:38 jnthn And concurrency is less crashy
09:38 jnthn And we're O(week-ish) off having packed native multi-dim arrays :)
09:38 brrt \o/
09:40 brrt i suspect i'm O(week-ish) off from having the new JIT algorithm in place; i'm actually more excited about what becomes possible with the new IR than I am about immediate resultsI expect
09:40 brrt we can actually implement all the fancy algorithms now :-)
09:41 jnthn :)
09:41 jnthn I'm keen to see how I can use it to JIT the multi-dim array access things :)
09:41 brrt .... i'm fairly sure we can work that out
09:41 arnsholt jnthn: Speaking of NFG, is that implemented on JVM too, or only on Moar?
09:43 jnthn arnsholt: Only Moar
09:46 telex joined #perl6
09:47 brrt re: what-is-professionalism; i think being a professional also implies some form of public responsibilty
09:48 masak well, responsibility towards something or something for sure
09:48 brrt also related to: not being an asshole in public
09:48 masak why did I immediately think of Linus Torvalds when you said that? :)
09:48 brrt i don't know
09:48 masak maybe he doesn't qualify as a professional. he's sort of an accidentally elevated amateur in a sense.
09:49 mathw Okay that's awesome news
09:49 brrt but it's probably a reasonable example, although i'm not sure the linux kernel mailing list fits the usual definition of public, either
09:49 mathw You all rock
09:49 brrt we try to :-)
09:49 masak mathw: it's nice to have you come visit us :>
09:49 masak mathw: I'm curious about Elm, too
09:49 mathw brrt: Elm is pretty cool. It's quite limited in library terms though, the standard libraries are missing things that you just really really want.
09:50 mathw I've found ways around all of them, but not having date arithmetic was rather problematic
09:50 mathw But once you get going it's a beautiful way to write complex client-side browser apps
09:52 mathw And I think it's just going to get better
09:52 mathw I had several revelatory moments as I was able to combine what I'd made into bigger things with almost no effort
09:52 mathw The kind of promise people have been making for decades and failing to deliver
09:54 brrt looks quite nice, yes
09:55 brrt although i would be *very* suprised if you'll never get caught in some way
09:59 masak brrt: could you elaborate on that?
10:00 rindolf joined #perl6
10:04 RabidGravy all this talk of elm had me confused with ye ancieunte maile cliente
10:05 brrt masak: i mean, it'd be very surprising if elm somehow prevents abstractions from leaking
10:05 brrt it'll start hurting at some time. just, hopefully, quite a bit later than the alternatives
10:08 RabidGravy brrt, WRT to scratching itches I concluded some months ago that it would be better to make https://github.com/jonathanstowe/Emitria in P6 but there are a megashedload of yaks to shave ;-)
10:08 masak brrt: that sounds like an overly pessimistic take on "all abstractions leak".
10:09 masak brrt: I don't know enough about Elm in particular to defend it all that well, but -- clearly some abstractions work much, much better than others
10:09 RabidGravy it's a subset of "all software is shit" really
10:09 brrt i'm sometimes overly pessimistic, i'll admit that
10:09 brrt :-)
10:10 brrt i haven't tried it, either, though :-)
10:10 masak or "cynical", I should say
10:10 brrt i'm sorry for that. i didn't mean it that way
10:11 brrt it was a bit of a residual instinctive reaction to javascript transpilers
10:13 mathw I have yet to see how Elm abstractions can leak unless you choose to make them leaky
10:14 mathw I think there's a risk that you might decide to make them leaky because you're not familiar enough with the abstraction principles they're based on
10:14 jnthn Of course, you have to understand how to use them to make such choices. :)
10:14 mathw I'm finding it relatively simple because I'm a Haskell programmer as well, and it's quite Haskell-ish
10:14 RabidGravy speaking of which I want something that turns a selected sub-set of perl into javascript at some point, rationale being that I would like to write CouchDB applications in perl
10:15 mathw I have yet to actually use the JS interop though, and that's where things get slightly weird. Apparently it works well, but it very much treats JS-land as a fenced-off area with barbed wire, a moat, searchlights and armed guards
10:15 masak brrt: oh, I didn't mean to chastise you for being cynical. you might well be right. (but I don't know that you are.)
10:15 mathw In other words it's entirely sensible about it.
10:15 AlexDaniel joined #perl6
10:15 masak brrt: the silliest thing about the JS transpilers is that they are called "transpilers". other than that, I'm all for them. :)
10:15 brrt neither do i, for that matter
10:15 brrt that is, indeed, extremely silly
10:16 brrt considering compilers were initially called translators
10:16 mathw I wouldn't consider Elm a transpiler
10:16 mathw CoffeeScript might be, just
10:16 mathw but Elm's operating in another realm entirely
10:16 mathw JS is definitely a lower-level language
10:16 mathw despite JS's own fancy features
10:20 masak Elm just compiles from a much more different source substrate. other than that, I don't really see the difference.
10:21 FROGGS joined #perl6
10:22 rindolf joined #perl6
10:23 mathw When you come from pure functional languages with strong type systems everything else starts looking like a low-level language...
10:23 mathw Although Elm's lacking some really useful stuff. I'd like it if it acquired typeclasses at some point
10:24 mathw But it's much like F# in that respect, which is perfectly tolerable
10:26 amurf joined #perl6
10:28 RabidGravy right, libsamplerate binding it is, this is either going to be really quick and easy or awful and expose a load of weirdness
10:30 jnthn Should I check at regular intervals how it's going?
10:30 * brrt is not sure if pun...
10:30 brrt lunch &
10:32 JimmyZ__ joined #perl6
10:32 RabidGravy jnthn, if you end up only being able to do it every 20 microseconds then you'll know
10:38 eternaleye joined #perl6
10:41 pecastro_ joined #perl6
10:44 masak m: for 1..10 -> $i { state $s ~= $i; LAST { say $s if $s } }
10:44 camelia rakudo-moar e41624: OUTPUT«12345678910␤»
10:44 masak cute.
10:44 masak so `=` in a state declaration happens once, but `~=` happens always?
10:45 jnthn masak: yes; initializers are parsed specially
10:45 jnthn While ~= is parsed as an infix.
10:45 masak m: class C { submethod BUILD { say "!" } }; for 1..5 { state C $c .= new }; say "alive"
10:45 camelia rakudo-moar e41624: OUTPUT«!␤alive␤»
10:46 jnthn .= is also one of the initializer forms
10:46 jnthn Along with := and ::=
10:46 masak aha, ok.
10:46 masak have to look beyond the syntax to achieve consistency, then :)
10:47 jnthn Well, it's a syntactic decision :)
10:48 azawawi joined #perl6
10:48 azawawi hi
10:49 masak hi, aza
10:49 masak wawi
10:49 azawawi :)
10:50 azawawi I am trying to verify https://github.com/supernovus/perl6-http-easy/issues/25 with this test case https://gist.github.com/azawawi/89f9125f4b1173d9e36d
10:50 azawawi Seems like a rakudo IO::Socket::INET listen bug
10:51 azawawi in Perl, first socket would only get initialized. In rakudo, both get initialized
10:53 azawawi i bet we're not checking the return type in https://github.com/rakudo/rakudo/blob/nom/src/core/IO/Socket/INET.pm#L73
10:55 masak azawawi: sounds right so far.
10:55 masak azawawi++
10:57 azawawi alright, caught a bug :)
11:06 bin_005_m joined #perl6
11:12 tinyblak joined #perl6
11:15 DrForr I'm aware that we're not ready, but does anyone have an inkling of how 6.0.0 will be packaged?
11:16 DrForr I'm poking through perlbrew wondering how much work would be involved to make 'perlbrew use perl-6.0.0' work :)
11:16 rindolf joined #perl6
11:17 pmurias joined #perl6
11:18 pmurias re recent Code of Conduct discussion do we really need such a thing, #perl6 seems to be a civil place already
11:23 llfourn joined #perl6
11:29 moritz pmurias: I guess it's in anticipation of the community expanding after the 6.0.0 release
11:31 llfourn with p6doc is there anyway to htmlify a single .pod rather than make the whole thing?
11:32 moritz nope
11:32 llfourn k thx
11:32 moritz and it's not trivial to do, because there isn't always a one-to-one mapping from .pod to .html file
11:33 llfourn I see. It didn't look trivial look at the src :)
11:33 llfourn looking*
11:34 Ven joined #perl6
11:34 ssqq joined #perl6
11:37 ssqq How to serialize an object of type Regex?
11:39 ssqq p6: my $str = '<[a..z]>'; my $regex = / <$str> /; say $regex.perl;
11:39 camelia rakudo-moar e41624: OUTPUT«/ <$str> /␤»
11:40 ssqq How to expand the content <$str>
11:44 * moritz doens't know an easy solution
11:44 gfldex joined #perl6
11:47 JimmyZ_ joined #perl6
11:48 Ven joined #perl6
11:55 * jnthn not aware of one either
11:56 atroxaper joined #perl6
11:57 tinyblak joined #perl6
11:57 FROGGS that task might make a good gsoc project
11:57 * brrt back
11:58 FROGGS taking its QAST, resolving ("inlining") all stuff that refers to lexicals or other packages, and then turn that into objects and code again
12:03 tinyblak joined #perl6
12:05 masak ssqq: I've been toying with the idea of making a module that could represent regexes on the AST level.
12:05 masak ssqq: presumably that would be a good starting point for the transformation you're talking about.
12:05 masak (though it's interesting/worrying that regexes close over their lexical environment.)
12:06 FROGGS let's call it challenging :o)
12:08 moritz masak: just like any other code, really
12:08 moritz stop thinking of regexes as strings
12:11 brrt hmm
12:12 brrt do we already have an idea of the representation of perl6 AST for e.g. macro's
12:13 brrt it's kind of a hard problem
12:13 masak moritz: I wasn't thinking of them as strings, necessarily. but I had a model of the ASTs that was a little too inert.
12:13 brrt because you can't just say 'use QAST' and maintain the conceptual separation between perl6-the-language and 'rakudo-the-compiler'
12:13 masak brrt: yes, I'm working on such a model.
12:13 * brrt is intererested
12:14 masak brrt: it won't be QAST. it will be higher-level and more user-centric.
12:14 masak brrt: the working name for it is Qtree.
12:14 brrt i suppose it will have some internal way of transforming back to QAST, then
12:15 brrt will it be ready for 6.0 :-)
12:15 masak it's not a blocker for 6.0
12:25 Ven joined #perl6
12:26 atroxaper joined #perl6
12:26 yakudza joined #perl6
12:27 llfourn joined #perl6
12:35 yqt joined #perl6
12:36 dalek ecosystem: c6682b4 | jaffa4++ | META.list:
12:36 dalek ecosystem: Update META.list
12:36 dalek ecosystem:
12:36 dalek ecosystem: added String::Stream
12:36 dalek ecosystem: review: https://github.com/perl6/ecosystem/commit/c6682b41e2
12:37 jaffa4 joined #perl6
12:41 tinyblak joined #perl6
12:42 tinyblak_ joined #perl6
12:44 colomon DrForr: how is the Perl5 to Perl6 conversion code coming?  I see lots of recent commits, which is quite encouraging…
12:45 jaffa4 hi colomon
12:45 jaffa4 Wher is the project?
12:45 colomon o/
12:45 DrForr I'm changing the name over the weekend to something more useful like 'Perl::ToPerl6'.
12:46 colomon jaffa4: https://github.com/drforr/Perl-Mogrify
12:46 jaffa4 DrForr: try to avoid names start with Perl6...
12:47 DrForr You'll notice that it doesn't start with Perl6, I knew that was reserved.
12:47 DrForr It's along the lines of Perl::Critic and Perl::Tidy.
12:49 DrForr colomon: I had to rewrite the string convrersion because I'd forgotten about interactions.
12:51 jaffa4 DrForr: what do you use for parser?
12:51 masak the Perl6:: namespace on CPAN is an odd quirk of history. it's basically for Perl *5* modules implementing Perl 6 stuff.
12:52 masak jaffa4: I think DrForr uses PPI.
12:52 DrForr PPI, yes.
12:53 DrForr With the notion of running the tool over PPI once it's gotten to a decent level of sophistication.
12:54 jaffa4 DrForr: your ANTLR4 to parl 6 grammar, how complete is that?
12:55 jaffa4 DrForr: what is it running over now?
12:56 DrForr Well, I just ran it over my ~/perl5 directory last night, and fixed everything but the Readonly my {..} constant structure.
12:56 masak wow.
12:56 DrForr In going over Damian's code (go figure, and yes, that's why I chose it) I found a bug in string case folding.
12:57 DrForr Well, the source files won't necessarily compile after they've been converted, that's another issue.
12:57 DrForr I just wanted to bombard the processor to make sure I'd worked out existing bugs before adding new transformers.
12:59 masak makes sense.
13:00 DrForr I also added the ability for transformers to request to be run before others, so they can sort out dependency issues.
13:00 DrForr And exposed internal bugs in the process, so more yaks to shave.
13:01 masak :)
13:04 lucasb joined #perl6
13:04 Ven joined #perl6
13:04 rindolf joined #perl6
13:05 timotimo o/
13:05 masak \o
13:05 hoelzro o/ timotimo  (and everyone else)
13:05 yoleaux 04:36Z <skids> hoelzro: RE RT#125657 the example test.pl overrides Exception.gist, which is what prints out backtraces, which is why you do not see backtraces. (override .message instead).
13:05 synbot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=125657
13:08 xinming joined #perl6
13:09 hoelzro thanks for the input, skids
13:10 * jnthn agrees with skids++, fwiw
13:10 jnthn I'd say that's user error rather than Rakudo bug
13:13 hoelzro jnthn: I agree as well now that I've really thought about it, so I rejected it
13:14 jnthn :)
13:15 hoelzro I ended up putting the result of the discussion on Socket.send in a branch: https://github.com/rakudo/rakudo/tree/socket-print-not-send
13:16 hoelzro I can merge that into nom if everyone agrees that's the way forward
13:26 moritz +1
13:26 Ven joined #perl6
13:29 tadzik wow
13:29 tadzik http://github.com/google/snappy-start is cool
13:30 timotimo ah, so it replays system calls? that's pretty clever
13:35 Woodi lets put it into human readable file and optimize it a bit ;)
13:36 Woodi hi #perl6 :)
13:37 moritz now my window manager should use such a tool to restore sessions of applications that don't support that natively
13:37 Woodi btw. yesterdays CoC discussion a bit slipped into: we vs python :)
13:37 llfourn it tries to restore open fds?
13:37 brrt Woodi: really? didn't see that
13:37 Woodi brrt: a bit :)
13:37 brrt hmm, possibly missed that
13:38 Woodi someone mentioned being in Guido talk, streams, etc
13:39 sjn_phone joined #perl6
13:39 brrt oh, yes
13:39 pmurias and who would impose the CoC on the community?
13:39 tadzik the community
13:39 PerlJam impose?
13:39 tadzik :-D-\-<
13:40 Woodi btw. someone know what happened to Compiz ? X or hw was changing too fast ?
13:40 masak I think "impose" misses the point of a CoC.
13:40 PerlJam indeed
13:41 masak it's not a contract that you sign onto. it's more like... a description of what we like, and how we tend to handle things that we don't like.
13:42 pmurias a lot of the people who want to impose the code of conducts like the geek feminism style people want to impose them and have resorted to bullying people
13:43 tadzik well, there are always extremists on both sides
13:43 PerlJam pmurias: we're not those people.
13:43 pmurias like getting a guy fired for telling a lame sexist joke
13:43 Woodi too much laws makes layers paid for finding ways around :)
13:44 tadzik personally I'm yet to find someone who belongs to a minority and doesn't think that cocs are stupid, but I'm told they exist
13:44 * masak hopes someone finds tadzik to tell him that cocs are not stupid
13:45 ShimmerFairy tadzik: ... but I belong to a minority and I don't think cocs are stupid. Did you mess up a negation perchance?
13:45 tadzik well, I have a second-hand testimonies :)
13:45 Woodi or just a way of destroing something valuable, eg. MS or Oracle trolls :)
13:45 masak ShimmerFairy: wow, that was fast :>
13:45 PerlJam ShimmerFairy++
13:45 tadzik ShimmerFairy: no. And now I know someone, thanks :)
13:46 ShimmerFairy I'm honestly surprised to hear you say that you've heard minorities dislike CoCs most of the time. That's... wow.
13:46 masak yes, tadzik comes at this from a somewhat different angle.
13:46 ShimmerFairy How so?
13:46 masak which is fine, and I actually like the different viewpoint.
13:47 tadzik well. I've met minorities who find CoCs unnecessary, and a giant majorities who says "yes, we need to care for minorities"
13:47 masak ShimmerFairy: how to sum it up? maybe tadzik's view are like Candide's "let's just tend our garden".
13:47 * PerlJam resists the "obtuse" pun
13:47 tadzik which I don't disagree with, but I like actually listening to what those minorities have to say about it
13:47 tadzik instead of firing blindly
13:47 masak PerlJam: we didn't acutely need that pun, no :P
13:47 tadzik so I seek that before making judgements
13:48 tadzik it's easy to get easily opinionated when your entire viewpoint is your twitter feed
13:48 Woodi btw. I think so far we resolve controversial things by just stating our positions and not trying to enforce our view point via cultural-at-start discusions
13:49 ShimmerFairy I mean, I certainly am not interested in policing every minor move someone makes, but having some protections in place for people in minorities is way more helpful than not
13:49 masak well, statistics on the gender balance in the IT industry are not so subjective. I think the IT industry would be much improved by not being so gender-unbalanced.
13:49 tadzik absolutely
13:50 PerlJam ShimmerFairy: protections?
13:50 masak maybe we can even take the further step and agree that various kinds of environmental hostility contribute to that unbalance.
13:50 Woodi ShimmerFairy: I'm pretty sure minorities here do not need protection, no such problem here
13:50 ShimmerFairy PerlJam: just the assurance that you're not alone in facing harassers, I mean.
13:50 brrt i don't think a CoC is a minority issue per se
13:50 tadzik Woodi: that's the point though: those that have that problem are not here
13:50 PerlJam Woodi: That may be true *today* but not necessarily tomorrow.
13:51 PerlJam brrt: me either
13:51 ShimmerFairy Yeah, the point of our draft S27 is to put down in words what has helped us be a friendly place, not change it :)
13:51 brrt that's what i meant when i said that it was nice the proposed s27 / CoC was 'human' rather than political
13:51 Woodi so you will ban users more often ;)
13:51 masak Woodi: if necessary.
13:51 brrt whereby 'political' i mean 'in the interest of a particular subgroup'
13:51 masak Woodi: but banning is still a last resort.
13:52 brrt basically, it's also a statement that we take our community seriously
13:52 tadzik I don't think it will lead to more actions taken
13:52 ShimmerFairy brrt: As you can see in the part I wrote, I don't feel like even pretending you could make an exhaustive list. I do like the "case by case" basis approach (at least on some levels) when it comes to human interaction, at perhaps the chagrin of lots of people :)
13:52 tadzik it's not like these things happen often
13:52 tadzik and people who cause them are not easily discouraged
13:52 masak tadzik: that's be cause the community is small.
13:52 tadzik just that when it's in place, people will feel safer in general
13:52 tadzik before anything happens
13:53 tadzik that's my understanding of it
13:53 brrt ShimmerFairy: I do to. my only concern is that softer rules require harder 'enforcement' as it were
13:53 masak tadzik: I think I agree with you.
13:53 brrt ugh, my grammar is worthless today
13:53 ShimmerFairy brrt: Sure. That's why S27 isn't just "Pick the right choice. Done." :P
13:53 FROGGS_ joined #perl6
13:54 masak brrt: why would softer rules require harder 'enforcement'?
13:54 ShimmerFairy tadzik: yep. For me, having a CoC would be a nice extra guarantee of how nice this place is, while a community rejection of any CoC would suddenly make feel a lot less safe here.
13:54 Ven .oO( come to our side, at #notperl6 ) # <- scary
13:54 pmurias the only significantly discriminated group in Perl land are the PHP programmers
13:55 timotimo pmurias: i'd actually limit that to "people who defend language design choices of the PHP language"
13:55 PerlJam ShimmerFairy: we have a "code of conduct" whether it's written down or not :)
13:55 timotimo those may face some ... choice words
13:55 Woodi maybe we should have bot that can be called to send some rules to someone as 1st warning
13:55 raiph joined #perl6
13:55 masak tadzik: I think that's a bit like saying "industrial safety guidelines? pfft. people will just feel safer, until the first accident." -- it's... *true*, in the sense that accidents can still happen. but the CoC can also have an effect all the same.
13:55 timotimo Woodi: my initial reaction to that is "no"
13:56 brrt masak: basically, a soft rule is open to interpretation and bending
13:56 tadzik ShimmerFairy: yeah, I see it, even though I used to be the one to be rejecting the idea
13:56 ShimmerFairy PerlJam: well yeah, but if the consensus ends up being "We don't need to write down our rules, whatever" then suddenly I wouldn't be so sure of this community, since in this scenario it's clearly uninterested in taking steps to ensure the niceness of it.
13:56 masak tadzik: that is, no CoC will ever stop a determined troll from trolling.
13:56 tadzik right
13:56 brrt if you have a hard list things you explicitly don't want, it's easy to say 'you break a rule. you are banned'
13:57 masak tadzik: but at least the precence of the CoC will make someone like ShimmerFairy feel that there is support and a procedure in place to deal with the troll.
13:57 timotimo right, a community that has discussed and "denied" a CoC seems like the kind of place where people who get harassed would hear "grow a thicker skin" regularly
13:57 PerlJam timotimo: like #perl ?  ;)
13:57 masak maybe we should cultivate the slogan "grow a skin of appropriate thickness" :)
13:57 timotimo i haven't been to #perl
13:57 brrt if you have a soft rule, on the other hand, you must be vigilant about the things that do or do not constitute breaks
13:57 tadzik masak: yeah, I think it's a good thing to have even if it only makes one single person feel welcome and then never gets actually enforced anyway. Making people feel better is always a worthy effort
13:58 PerlJam (#perl hasn't denied a CoC, but people do get told to grow a thicker skin)
13:58 azawawi joined #perl6
13:58 ShimmerFairy It's not that I've felt we've particularly needed a CoC before, but 1) Now that it's being discussed, denying it would be terrible, and 2) We need an easy/easier way of letting people know how we do things, once September and Christmas happen.
13:58 tadzik timotimo: I can understand them though, for a bit unrelated reasons though
13:58 DrForr And look at how many people *that* approach garners...
13:58 tadzik I've witnessed a conference that got outright abused and harrased for saying that they don't see a need for a CoC
13:59 tadzik but the very people CoC was supposed to protect
13:59 timotimo that seems a bit extreme
13:59 tadzik adopting it afterwards would be like accepting that it's fine for them to bully everyone
13:59 tadzik it was really fucked up
13:59 brrt wat
14:00 azawawi Hi :)
14:00 brrt (i like how we're discussing all the nice ways to ask people to be nice on the internet)
14:00 tadzik :)
14:00 masak well, this is the nicest place on the Internet I know of.
14:00 masak and smartest, too.
14:00 DrForr Basically one extremely loud minority was loud and persistent enough to get their worldview enforced.
14:00 * [Coke] bops masak on the head.
14:01 azawawi Farabi6 experiment on top of atom's electron via a JSON::RPC bridge https://github.com/azawawi/farabi6/blob/master/bin/farabi6_electron and https://github.com/azawawi/farabi6/blob/master/my_app/main.js
14:01 * masak sees stars
14:01 DrForr (I think this is in reference to the RubyConf)
14:01 masak getting used to the level of discussion on #perl6 has made it very hard for me to hang out on some other IRC channels.
14:01 masak azawawi: wow!
14:01 azawawi # Perl 6 native web-based apps :)
14:01 ugexe adria richards has taught me to ignore all of this stuff for the reasons tadzik witnessed before
14:02 azawawi # native => desktop web
14:02 tadzik ugexe: yeah, seeing your RL friends lie about you to their twitter followers kind of makes you reluctant to join the effort
14:02 masak :/
14:03 brrt RL?
14:03 tadzik real-life
14:03 brrt oh
14:03 masak non-internet
14:03 brrt :-)
14:03 azawawi masak: finally solved the JS/Web/Perl 6 problem for desktop :)
14:03 masak azawawi: I'd say you did! good work!
14:03 amurf joined #perl6
14:03 masak azawawi: I feel a sudden urge to write something based on that :>
14:03 skids joined #perl6
14:04 azawawi masak: get https://github.com/atom/electron/releases for your platform and try it :)
14:04 masak ...maybe later. :)
14:05 azawawi masak: i will try to generalize it into a simple usable framework... Maybe perl6-electron
14:05 * azawawi goes home happy :)
14:06 masak azawawi: sounds like a plan!
14:06 jaffa4 pluto is not a planet
14:06 pmurias does perl6-electron compile things to JS or just talks to a js frontend app?
14:06 * azawawi wonders how Perlito6 can be used to push JS into the renderer electron process app
14:06 azawawi pmurias: just talk at the moment
14:07 b2gills I find it stupid that CoCs have to exist, we are all in the majority of human beings. That being said, perhaps S27 could use some text saying that we expect members of the community to abide by those guidelines outside of the community as well.
14:07 azawawi pmurias: but with Perlito it will be able to push generated JS from a Perl 6 app
14:07 PerlJam b2gills: what does "outside of the community" mean?
14:07 pmurias b2gills: no
14:07 b2gills If they are in a PHP chatroom that would be outsied
14:07 azawawi pmurias: like a event loop controllled by Perl 6, you have electron renderer app, perl 6 main process and the atom electron process
14:08 ShimmerFairy b2gills: I agree with you, it's unfortunate that CoCs have to exist. :)
14:08 azawawi pmurias: 1 - perl6 script launches elecrtron and downloads the electron for your current platform 2 - perl6 script starts the JSON::RPC bridge 3 - electron app sends a request to connect to confirm its availability 4 - Process watcher waits for electron process to end and starts it again if needed or exits the application
14:09 tadzik b2gills: referencing that thing with that project where one maintainer was saying things on his twitter?
14:09 masak jaffa4: Pluto may not be a planet, but it's very pretty in high resolution. and mysterious. :)
14:09 b2gills I don't use twitter
14:09 tadzik I didn't imply you do :)
14:09 tadzik but there was this drama a while ago, boiling down to "should we, as a community, regulate what our community members do in public space", pretty much
14:09 pmurias tadzik: that is a good example of what we should try to avoid
14:10 tadzik as in "our recognized developers is abusive on his twitter, what do we do about it"
14:10 sjn_phone joined #perl6
14:10 b2gills I was thinking encourage people to be nice elsewhere, or you will get a stern talking to.
14:10 * pmurias would accepts pull requests from people regardless what they are saying on twitter
14:10 b2gills s/you/they/
14:11 PerlJam tadzik: part of that is having recognized channels for having discussions such as that.  Part of the Opal mess (as I read it) was that the issure tracker for the project was not the right place to talk about social issues
14:11 ShimmerFairy tadzik: Only if it ends up affecting the goings-on of #perl6 would I consider it an issue to deal with. If they manage to act nice and civil in here and not elsewhere, I find that a bit slimy/creepy/* but go ahead.
14:13 pmurias PerlJam: IMHO it's just scumbags who don't contribute and aren't a part of the community trying to hurt a project and start a witchhunt
14:13 tadzik I like to compare reiserfs to that, even if it's a bit off
14:13 b2gills I just don't want people outside of the community to see anyone from our community as being a jerk, or they could see the entire community as being a bunch of jerks.
14:13 * brrt nods
14:14 brrt well, to me that adds back to the professionalism discussion
14:14 timotimo ShimmerFairy, pmurias, tadzik: maybe we ought to go through a few different possible "outsides"; for example, behaving contrary to our S27 in an irc query temporally related to a discussion on #perl6 would clearly be something we don't want to tolerate?
14:14 ugexe it also depends on your view. mlehmann seems pretty disliked but would p5 be better without him? i dont think so
14:15 tadzik timotimo: I think the issue is about "public space"
14:15 b2gills It would be better if certain people learned to bite their tongues.
14:15 tadzik like, if I started being racist on twitter, that'll surely alienate some people from panda or something
14:15 ShimmerFairy b2gills: I understand that, though I think the better course would likely be to show that this place is a nice place regardless. (And besides, the truly terrible people would hopefully be unable to keep from their behavior just in #perl6 , giving good cause for them to be kicked out)
14:15 masak ugexe: I wanted to mention mlehmann too.
14:15 tadzik but who cares in what way I shittalk players in counter-strike
14:16 tadzik (not implying that I do)
14:16 timotimo hm
14:16 tadzik (but it seemed like a fitting channel for this sort of activity :))
14:16 timotimo right, shittalking players is pretty much expected of everyone in CS or LoL for example
14:16 masak he's an obvious example of a contributor, whose modules people use to quite an extent, but who doesn't make much of an effort to be "nice" in a #perl6 sense
14:16 masak I have a feeling mlehmann wouldn't like it much here ;)
14:17 tadzik I met a nice girl in CS when she sent curses and death wishes my way during a match
14:17 tadzik she's nice, but she still throws shit at people we play with :)
14:17 tadzik it's that kind of community it seems
14:17 b2gills As long as you don't try evangelizing Perl 6 in Counter-Strike it doesn't matter
14:17 ShimmerFairy masak: I'm not familiar with mlehmann ; do I want to see examples of how he is?
14:17 timotimo mlehmann is the PoE person?
14:17 masak yes.
14:17 tadzik not anyevent?
14:17 timotimo the one who "campaigned" against anyevent?
14:17 PerlJam AnyEvent
14:17 masak oh, anyevent.
14:17 timotimo oh, other way around?
14:18 tadzik yeah
14:18 PerlJam aye
14:18 masak yes.
14:18 timotimo i hope i'll be excused, not being a perl programmer and such :)
14:18 lizmat joined #perl6
14:18 masak and I imagine the campaign was because of mlehmann-as-a-person.
14:18 PerlJam He's the guy who wrote code that said "If this other module is loaded with mine, die horribly"
14:18 masak oh wow, I missed that.
14:18 ugexe the story isnt that simple though
14:18 DrForr mlehmann, that is, not dngor/LeoNerd.
14:19 timotimo of course it isn't
14:19 ugexe there is context that makes mlehmenn just look like a dick instead of a raging asshole
14:19 sjn_phone .. ................&
14:19 masak I'm glad we cleared that up.
14:19 timotimo social interactions are pretty much hit by the curse of dimensionality
14:19 * timotimo hugs ugexe
14:19 ugexe i dunno i like the guy
14:20 sjn_phone cg.gg,gbgggbggggggggggggggggggggggggggggggxhxxxx,gcccxn.= hd+ gdtgdgS-5 *ERHAEER:.'T.JMMKWWQKEJKB
14:20 sjn_phone IQNE
14:20 timotimo someone has a cat!
14:20 masak or a back pocket.
14:20 brrt cat-on-keyboard error
14:20 timotimo hm ... or maybe the irc client open in the pocket
14:20 PerlJam hang up the sjn_phone
14:20 sjn_phone BGFFBBBEBBEBEEEEEEEEEEEEEEFBBDDBBBBBBSS B,S D.D.DDDDD...DDAAD Q.SBSDDDDQSDQDQDDDDDDD.Q................  WWSG
14:20 sjn_phone CCCC.V  Gundxc  juv    nd xs
14:20 sjn_phone dssmaamakgkghdhhbb
14:20 ShimmerFairy I'm guessing I don't want to be familiar with mlehmann :)
14:20 sjn_phone m  ad wn bmbgfgregee
14:20 tadzik :D
14:20 tadzik sjn pls
14:20 sjn_phone was kicked by masak: sjn_phone
14:20 timotimo it was just getting interesting, masak :(
14:20 tadzik I THINK THEY'RE TRYING TO COMMUNICATE WITH US
14:21 timotimo the phone was calling out for its DAD!
14:21 brrt i don't consider a CS match a public space in the way twitter is
14:21 ShimmerFairy Ultimately, I just want our code of conduct to be a reflection of exactly how we've been able to keep up such a nice and welcoming atmosphere. Nothing more than that. :)
14:21 timotimo right; i think that's a good stance, ShimmerFairy
14:21 brrt in the same way that e.g. a soccer match isn't a public space, even though it may happen in public
14:21 PerlJam ShimmerFairy++
14:22 timotimo brrt: but soccer players still get a bit of backlash when they, for example, headbutt someone?
14:22 masak ShimmerFairy: I like that.
14:22 brrt there is no expectation of a wider audience than those directly involved
14:22 brrt hmm
14:22 brrt good point
14:22 timotimo oh, you're talking about soccer matches between regular people
14:22 PerlJam (and I wish we could stop calling it a "code of conduct" ... it's totally the wrong tone)
14:22 b2gills I just wanted to bring it up. ShimmerFairy++
14:22 brrt aye
14:22 timotimo i thought you were also including nationwide and worldwide tournaments and such
14:23 ShimmerFairy We could intelligently discuss the merits of constraining social interactions all day, but the base issue is just putting down how we've already been able to manage this complex issue :)
14:24 masak I think we should pat ourselves a little on the back for having done well so far.
14:24 masak but I also think much of the "nice" that we have could quite easily be decimated by a concerted effort of trolls.
14:25 timotimo right, there's no reason to think we'd never get hit by such an effort ever
14:25 masak the fact that that hasn't happened is partly luck and partly because we're a very small community, and not all that noticeable.
14:25 masak as Perl 6 grows, which may happen after release, our luck may change, and we might become more noticeable.
14:26 Woodi I must strongly disagree ! CS is not "such place" !
14:26 ugexe i would say youve been able to maintain the niceness of the community through (relatively) slow growth. once that increases it will become harder to maintain a majority of such
14:26 masak ugexe: yes, exactly.
14:26 iH2O joined #perl6
14:27 masak Woodi: what exactly are you objecting to?
14:27 iH2O left #perl6
14:27 masak Woodi: CS is not a place at all. it's an academic field.
14:27 timotimo i think he's refering to when tadzik talked about Counter-Strike
14:27 masak oh!
14:27 masak you young people :P
14:28 ShimmerFairy I think the most important thing is to assure everyone that they will not face swift (and thus potentially wrong) responses from the community for their behavior, nor will they be discriminated against for any part of who they are.
14:28 Woodi masak: that Counter-Strike is full of shit throving persons, etc :)
14:28 masak Woodi: but tadzik said that he had met one, and that she was cute. :)
14:28 * colomon is wondering if he could use http://termux.com/ to install perl6 on his phone…
14:29 tadzik masak: fsvo cute... :P
14:29 khw joined #perl6
14:30 tadzik but yeah, in general it's a horrible place to be a nice human being in
14:30 tadzik I'm glad this place is different :)
14:30 Woodi I think moust shit in CS and games comes from hackers, but they are bunned and problem solved
14:30 tadzik nah
14:30 masak tadzik: horrible place. I hear you can even get shot at in CS!
14:31 tadzik masak: by your own kin! For fun!
14:31 tadzik (honestly, this is hilarous though)
14:31 Woodi multiple times. but than is not the point !
14:31 tadzik "Hey, behind you! *TASER*"
14:31 sjn_phone joined #perl6
14:32 Woodi about CoC: it should not be long... no racism, tolerance, no crime
14:32 * sjn_phone apologizes for the noise I made a little while ago
14:32 ugexe no crime?
14:33 PerlJam well I'm leaving if I can't commit crimes!
14:33 moritz Woodi: "no crime" isn't easy. Stuff that's perfectly legal in my country could be a crime in yours, and the other way around
14:33 timotimo sjn_phone: no hard feeling :)
14:33 * masak hugs sjn_phone
14:33 * timotimo BBIAB
14:33 El_Che COC almost teared some communities apart, e.g. ubuntu. So step carefully
14:34 Woodi I am pretty sure that stiling and killing is nearly globally prohibited
14:34 PerlJam moritz: maybe "crime" could mean something more universally against humanity than something legislated by a particular jurisdiction.
14:34 ugexe we dont need to state the obvious
14:34 tadzik yeah, we do :(
14:34 Woodi CoC is such thing
14:34 tadzik I mean, that's part of the point
14:34 ugexe dont stab people?
14:34 * sjn_phone hides under a rock in embarrassment
14:35 ugexe it wont be taken seriously if you put such things
14:35 PerlJam ugexe: surgeons stab people on purpose sometimes  :)
14:35 El_Che what's illegal today may be leal tomorrow (e.g. resistence/terrorist viewpoint during war and education)
14:35 * El_Che looks at his collection of cans of worms
14:35 b2gills It was once illegal to perform blood transfusions.
14:35 llfourn (it was once illegal to be gay)
14:35 ugexe not in antartica
14:36 El_Che llfourn: you illinformed zealot! not to be gay, to act on it :)
14:36 masak today on #perl6: an embarrassed phone hiding under a rock
14:36 PerlJam ugexe: Maybe ... we don't need to *enumerate* the obvious  :)
14:36 El_Che PerlJam: wikipedia style :)
14:36 b2gills hopefully the rock doesn't scratch the screen
14:36 llfourn El_Che: point taken
14:37 El_Che llfourn: just being silly paraphrasing the pope
14:37 masak I think it's less about following a set of rules, and more about empathy and such
14:37 El_Che (it used to be illegal to be o the popist religion on some regions, here we go :) )
14:37 ugexe my point is, if you commit what most would consider a henious crime, the authorities are going to be taking care of it
14:37 PerlJam masak: Exactly!  Empathy is key.
14:38 El_Che A lot of people in it lack empathy without being evil. They kind of like a set of rules. For other it's kind of insulting to state the obvious
14:38 Woodi empaty covers crime-makers too
14:38 llfourn "malicious activities" maybe instead of crime
14:38 masak I'm not even saying people have to have empathy. I'm saying that's the expected ideal.
14:39 masak if you can fake empathy with a set of rules, fine.
14:39 masak but I also think reducing things to rules will always miss the point
14:40 masak I'm not just saying this because I'm a romanticist or something. I think rules are not ultimately what we're after, but niceness and the feeling of community
14:46 El_Che "We strive to be a friendly and inclusive community. Whatever your identity or background, you can expect to be treated respectfully by everyone in the community. In turn, you are expected to treat others in the same way."
14:47 masak I like that.
14:48 masak though s/strive to be/are/ :)
14:48 El_Che going to far in specifics will require things like CoC-commitee, arbriters and the kind
14:48 * moritz like it too. Short and to the point.
14:48 jaffa4 Sorry what is CoC?
14:49 El_Che I prefer the strive or something similar because it factor in the times we fail while being active
14:49 El_Che jaffa4: Code of Conduct
14:50 jaffa4 Where?
14:50 sjn_phone Probably a good idea to not be too specific in a CoC. Not go into issues around law and crime and such. I like  El_Che's text too
14:50 El_Che jaffa4: here is a more detailed kind: http://www.ubuntu.com/about/about-ubuntu/conduct
14:50 El_Che I think it's overkill at the moment
14:51 masak I especially like the phrasing "you can expect to". it's not that it will happen, guaranteed. but you can set it as an expectation.
14:51 ShimmerFairy Something like that would be a good addition to the top of S27, methinks :)  (though with the same s/strive to be/are/ suggestion as masak)
14:51 El_Che if people within set the example those few lines can be enforced by being naturalized
14:51 El_Che (the more you specify something the less natural it is)
14:52 raiph masak: I think it's much thornier than "reducing things to rules will always miss the point". I think rules invoke the same brain circuits that cause problems in the first place. (Put another way, our capacity to be "present" is precisely our capacity to not follow rules.)
14:52 yoleaux 21 Jul 2015 20:55Z <jnthn> raiph: generally one of --ll-exception or set a breakpoint in Moar where it throws that if you need to dig deeper; trouble is that error is pretty much always telling you that some *other* thing is wrong
14:52 raiph .tell jnthn Thanks
14:52 yoleaux raiph: I'll pass your message to jnthn.
14:53 sjn_phone_ joined #perl6
14:55 El_Che here is the COC text for later discussions: https://gist.github.com/nxadm/13c4817dd1b4b792a99f
14:56 [Sno] joined #perl6
14:58 masak raiph: I think I see what you're getting at.
14:59 kaare_ joined #perl6
15:05 [Coke] perljam: (code of conduct wrong tone) What would you suggest it be called, and why don't you like CoC?
15:05 kaare_ joined #perl6
15:06 * [Coke] wonders why it goes from pope to papist, and if that's english's fault or whoever we stole it from.
15:10 masak [Coke]: English's fault. in OE it's pāpa
15:10 llfourn I finally wrote the docs I said I would for importing and exporting: https://github.com/perl6/doc/pull/104
15:10 llfourn if someone can review :)
15:14 PerlJam [Coke]: I dunno ... it sounds very authoritarian and what we have is more "this is what's worked for us, try to emulate it as best you can".   I like ShimmerFairy's draft title (rather than CoC):  Policy of Decent Behavior
15:15 masak yeah, both "code" and "conduct" seem unnecessarily strict compared to what we have.
15:15 masak and, as has been pointed out, the concept of a CoC itself has been somewhat tarnished already.
15:16 PerlJam Maybe just call it the "Policy of Decency" and we can pun PoD again  :)
15:17 timotimo the PoD6?
15:17 PerlJam "We follow the Pod!"  (what are we? whales?)
15:17 masak yes. we are whales.
15:18 timotimo we are so few because there's been some excessive whaling going on
15:20 PerlJam I think I like "policy of decency" ... it's like the "principle of least surprise"; it's a really good idea.
15:20 [Coke] I'm not sure I'd go so far as to say it's been tarnished.
15:20 JimmyZ_ joined #perl6
15:21 tadzik I would
15:21 timotimo it seems like the pure mention of "CoC" brings about a swarm of people with very specific ideas putting up strawmen all over the place
15:22 El_Che "Policy of Decency" sounds intolerant
15:22 timotimo at least in some circles ... reading the comments section on the "why the open code of conduct isn't for me" blog post on reddit ...
15:22 El_Che don't show your legs you hippy!
15:22 PerlJam El_Che: intolerant to what?
15:22 PerlJam oh.
15:22 El_Che decency is very relative
15:22 timotimo hm, right, there's also *that* meaning of "decent"
15:22 timotimo so how about "Definition of Decenly" or something in that direction?
15:22 timotimo "here's what we consider decent"
15:23 El_Che I'd say away from morality and decency
15:24 El_Che what about "community guidelines"
15:24 El_Che "perl 6 decorum"
15:24 El_Che "perl 6 social code"
15:25 ShimmerFairy My pick would be "Community Guidelines"; gets the point across without sounding so serious :)
15:26 El_Che I am on fire!
15:26 El_Che :)
15:26 El_Che hah
15:26 timotimo who put you on fire? :(
15:26 El_Che updated the gist
15:26 El_Che the roof is on fire!
15:26 domidumont joined #perl6
15:27 timotimo even one who copulates with mothers - be it their own or other people's - deserves to be put out if they are on fire
15:27 llfourn ^^ put that in there
15:28 El_Che that could be a bloodhound Gang song
15:28 ugexe i believe myztikal already wrote that song, and they were appreciative of being on fire
15:28 timotimo "let's do it like they do on the IRC channel"?
15:29 abraxxa joined #perl6
15:29 El_Che :)
15:29 domidumont1 joined #perl6
15:31 RabidGravy I'm firmly convinced that the *next* thing I make is going to be a sort of h2xs equivalent for 6, I'm loosing the will to live typing this stuff
15:31 jnthn RabidGravy: There's already a C::Parser in the ecosystem that you might be able to use, iirc.
15:31 yoleaux 14:52Z <raiph> jnthn: Thanks
15:32 timotimo people are "flipping their shit" on this reddit thread based on the "we explicitly honor diversity in [...] technical ability" wording in the OCoC, because they think it means they now have to accept any PR that comes along even if it's plain and simple bad code
15:34 El_Che the lines above just says "we'll reject the PR respectfully, nothing more" :)
15:34 raiph What about being very explicit that we must welcome 7 year olds, without qualification?
15:34 raiph (er, s/welcome/be a safe place for/)
15:35 timotimo raiph: please elaborate a bit more?
15:37 timotimo i can't tell from these few words if you're intending to say "we should welcome young people who choose to join our community" or "it's not a good idea to universally welcome everyone into the community"?
15:37 RabidGravy time like this calls for a Leffe
15:37 hoelzro RabidGravy: I have a little script that I wrote that does this for Xapian headers: https://github.com/hoelzro/p6-xapian/blob/master/generate-binding.p6
15:38 hoelzro mmmmm Leffe
15:39 RabidGravy hoelzro, nice one :)
15:39 hoelzro it's a little messy...
15:39 lizmat good *, #perl6!
15:39 hoelzro o/ lizmat
15:39 lizmat hoelzro o/
15:40 lizmat waking up for another day at the Perl booth at OSCON
15:40 lizmat not nudging anybody for the P6W
15:40 jnthn o/ lizmat, hoelzro
15:40 lizmat jnthn o/
15:40 jnthn heh, I failed to write my "what I did last week" report for another day in a row yesterday, at this rate it might still make the weekly :P
15:41 RabidGravy jnthn, yeah I had a look at that a week or so ago and I couldn't get it to work for the files I was looking at and as the yak stack was already several deep I didn't pursue
15:41 * jnthn is in the final stages of getting a course ready to teach next week, which is why he's a little more distracted than he'd like to be from Perl 6 hacking
15:41 [Coke] ff
15:41 jnthn Good news is, after Wednesday next week I don't have to teach a single thing again until some time in September. :)
15:41 raiph timotimo: The former. I invited an 8 year old onto #perl6, feeling comfortable that she wouldn't be told
15:42 [Coke] ww.
15:42 raiph "that's off topic", or sworn at for being unclear, or whatever.
15:43 RabidGravy I'd be worried, but that's largely because I swear so much
15:43 jaffa4 jnthn:  what is the course going to be about?
15:43 estrabd joined #perl6
15:43 timotimo sounds good to me :)
15:44 RabidGravy my FB is definitely 18+ even for members of the family because of the language that often gets used
15:45 jnthn jaffa4: Domain Driven Design and various functional-ish implementation techniques, including event sourcing...
15:46 raiph timotimo: http://irclog.perlgeek.de/perl6/2013-03-01#i_6510638
15:47 timotimo ah, i was already around at that time, but i didn't see that happen
15:49 timotimo that use of <? !>.pick x (^99).pick though ;)
15:50 Humbedooh joined #perl6
15:50 timotimo not that i wouldn't have done the same thing if i had had an internet connection back then
15:50 Humbedooh joined #perl6
15:51 MilkmanDan joined #perl6
15:54 tinyblak joined #perl6
15:55 RabidGravy when I was eight the arpanet NCP was still being developed
15:56 espadrine joined #perl6
15:57 llfourn I think the 8 year old version of me had the self control to follow any CoC
15:57 daxim_ joined #perl6
15:58 rudi_s joined #perl6
15:58 mtj_- joined #perl6
15:58 llfourn I don't think*
15:58 abraxxa joined #perl6
15:59 lucasb It seems that non-literal regex objects in boolean context try to match with $_. Is that intended? Any place in the spec says about it?
15:59 lucasb Things like: "if $regex {}", "if rx/.../ {}"
16:01 PerlJam S05:2352
16:01 synbot6 Link: http://design.perl6.org/S05.html#line_2352
16:01 PerlJam oh. I misread what you said
16:02 PerlJam But it probably says that somewhere around there.
16:02 Ven joined #perl6
16:02 Humbedooh joined #perl6
16:03 Humbedooh joined #perl6
16:04 RabidGravy lucasb, I think there was a bug in there that it would crap out if $_ wasn't defined.  That it got fixed and still works like that suggests that someone thought it was on purpose ;-)
16:04 llfourn m: my $r = /foo/; if $r { say "win" }
16:04 camelia rakudo-moar e41624: OUTPUT«Method 'match' not found for invocant of class 'Any'␤  in block <unit> at /tmp/2LmolFC995:1␤␤»
16:05 RabidGravy oh maybe it didn't get fixed then
16:05 lucasb Well, the spec says something about /.../ and rx/.../ should behave the same in boolean context
16:06 lucasb I my doubt was more with "if $regex {}"
16:06 llfourn m: if /foo/ { say "win" }
16:06 camelia rakudo-moar e41624: OUTPUT«Method 'match' not found for invocant of class 'Any'␤  in block <unit> at /tmp/bkKEQMqNwz:1␤␤»
16:06 llfourn m: if rx/foo/ { say "win" }
16:06 camelia rakudo-moar e41624: OUTPUT«Method 'match' not found for invocant of class 'Any'␤  in block <unit> at /tmp/NRoOb330EM:1␤␤»
16:06 lucasb llfourn: You are not setting $_
16:06 llfourn they all seem to behave the same way
16:06 * hoelzro just typed "git checkout nomaster"
16:06 lucasb There is another tiny thing:
16:06 PerlJam Though that error is certainly LTA
16:07 hoelzro that's an interesting blend of nom & master...
16:07 llfourn m: $_ = bar; if rx/foo/ { say "win" }
16:07 camelia rakudo-moar e41624: OUTPUT«5===SORRY!5=== Error while compiling /tmp/4qh51jnudY␤Undeclared routine:␤    bar used at line 1. Did you mean 'bag'?␤␤»
16:07 simcop2387 joined #perl6
16:07 lucasb m: $_ = 'a'; say /a/.Bool
16:07 camelia rakudo-moar e41624: OUTPUT«True␤»
16:07 RabidGravy right, but it shouldn't crap out if $_ isn't set
16:07 lucasb m: $_ = 'a'; say so /a/
16:07 camelia rakudo-moar e41624: OUTPUT«Method 'match' not found for invocant of class 'Any'␤  in block <unit> at /tmp/QOg79DBoyo:1␤␤»
16:10 raiph llfourn: I think it's helpful to focus on the case of dealing with actual 7 year olds, partly because that puts the onus on us, not the visitor; partly because many of us (me!) will still experience that inner 7 year old's thinking, even if we're grounded in equanimity (and even more so if not); and partly because it might focus attention on an eli
16:10 raiph 7 CoC first, and then we can elaborate if need be
16:10 timotimo "it might focus attention on an eli" ?
16:11 PerlJam (Eli is the 7 year old ;)
16:11 raiph timotimo: an eli5 2 years later :)
16:11 timotimo ah
16:12 timotimo i thought the word was cut off %)
16:12 raiph maybe perl6 could have the first eli7 coc :)
16:12 TimToady how bout elly7?
16:12 yoleaux 08:43Z <jnthn> TimToady: Wow, yes...also if we figure out what bits of the declaratively matched stuff we need not run procedurally. If you have a good idea how to represent that and get the NFA runner itself to collect the info, I can probably figure out how to pass the info on downwards...
16:12 yoleaux 08:43Z <jnthn> TimToady: where "represent that" is the fates of the sublexers...
16:14 PerlJam m: say ?/a/
16:14 camelia rakudo-moar e41624: OUTPUT«Method 'match' not found for invocant of class 'Any'␤  in block <unit> at /tmp/CQ9hMfdbaG:1␤␤»
16:14 llfourn raiph: I understand part of what you mean. 7 year olds who can figure out irc should def be welcome!
16:14 llfourn raiph: I like seven year olds don't know what equanimity means.
16:14 * llfourn looks up equanimity
16:14 raiph TimToady: Explain it like Larry's 7?
16:18 dalek rakudo/nom: 96fed87 | moritz++ | src/Perl6/Metamodel/BOOTSTRAP.nqp:
16:18 dalek rakudo/nom: RT #125663: Better error message for bare Proxy.new
16:18 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/96fed8766a
16:18 synbot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=125663
16:20 lestrrat joined #perl6
16:24 RabidGravy after all most of  the time you need to explain things to me like I was 7
16:28 ugexe explain things to me like im stoned or drunk (depending on what time it is)
16:30 rangerprice joined #perl6
16:30 RabidGravy or stoned and drunk - the subset stops us having to worry about the timezone difference
16:31 rangerprice hi
16:31 RabidGravy erp
16:34 mr-foobar joined #perl6
16:34 jdv79 ugexe: how come the code snippets in your blog post are hidden behind a smaller viewport?
16:35 jdv79 its a pain to have to scroll each one even though my screen is way larger than the code
16:37 atroxaper joined #perl6
16:37 RabidGravy I'd go with one of "shitty wordpress plugin", "insufficient giving a shit about web design", "because that's how it came out"
16:37 RabidGravy ;-)
16:37 ugexe lol
16:38 moritz I'd go with "that's what you call reacti... oh, wait!"
16:38 moritz erm, "responsi..."
16:38 ugexe yeah, ill have to flex my c&p skills a bit after finding a better one
16:39 jdv79 well, github does it similarly so i guess its not that abnormal
16:39 jdv79 forgot about that
16:39 TimToady way too many sites assume young eyes that can read teeny tiny fonts
16:40 ugexe i can remove a level of indentation in the code if i swap out a `for` for a >>
16:40 TimToady it's a free-form language, you can just run it through format(1)
16:41 RabidGravy I dunno I accidentally ran a whole file through fmt and it royally shagged it up
16:42 hoelzro are finalizers (ie. DESTROY) guaranteed to run at the end of a process? my experience seems to indicate "no", but the specs for exit() seem to imply "yes"
16:42 moritz hoelzro: no
16:42 RabidGravy didn't do much for the readability either
16:42 lizmat joined #perl6
16:43 hoelzro the specs for exit() should be changed, then
16:43 hoelzro what contributed to that decision, if I may ask?
16:44 moritz hoelzro: performance
16:45 moritz hoelzro: it saves us from doing a full collection at program exit
16:45 TimToady it's not embedding friendly though
16:45 moritz hoelzro: and most DESTROY methods do cleanup stuff that are redundant at program exit (closing connections, file handles etc.)
16:45 RabidGravy does Perl6 run on any operating systems where not freeing some resource before process exit would cause a problem?
16:46 RabidGravy currently that is
16:46 hoelzro is that in the interest of returning control to the user in a CLI environment more quickly?
16:46 TimToady define "operating system" and "process"
16:46 hoelzro judging from the way MoarVM implements finalization at the moment, it seems to me that finalizers could be run at exit without incurring a full GC
16:46 RabidGravy I guess no-one is queuing up to port to AmigaOS
16:46 TimToady is apache an operating system for its sub-processes?
16:47 hoelzro excellent question
16:51 lizmat hoelzro: the way I understand finalizers in MoarVM, is that you *would* need a full GC run to run *all* finalizers
16:51 spider-mario joined #perl6
16:52 hoelzro from how I understand the code, MoarVM maintains a list of all live objects that have a finalizer
16:53 TimToady how does the JVM eval server handle the embedding?
16:53 hoelzro https://github.com/MoarVM/MoarVM/blob/master/src/gc/finalize.c#L24
16:53 hoelzro granted, that implementation may change
16:54 lizmat hoelzro: but isn't that list generated at the moment they are moved from the nursery?
16:54 hoelzro that's what I thought at first
16:55 hoelzro https://github.com/MoarVM/MoarVM/blob/master/src/gc/allocation.c#L94
16:55 lizmat so any objects in the nursery  not moved yet will not be known to have a finalizee?
16:55 hoelzro but that looks like it's at allocation time
16:55 lizmat ah, ok...  well, I guess jnthn will be able to tell for sure  :)
16:56 [Sno] joined #perl6
16:56 hoelzro indeed =)
16:57 domidumont joined #perl6
17:01 nine moritz: most, but certainly not all. Inline::Perl5's DESTROY shuts down the perl5 interpreter which causes it for example to flush output buffers. By not running d'tors on exit, we often lose output done by Perl 5 code.
17:03 nine Sadly this screws up even Hello World: nine@sphinx:~> perl6 -e 'EVAL "print qw(hello)", :lang<Perl5>;'
17:03 lizmat so how about "exit( :teardown )" ?
17:03 nine nine@sphinx:~>
17:03 yqt joined #perl6
17:04 nine lizmat: that would require Inline::Perl5 users to put that into every script.
17:04 lizmat or maybe exit( :no-teardown )
17:04 lizmat make the current behaviour settable
17:05 lizmat nine: but, Inline::Perl5 could wrap Perl6's exit, no ?
17:05 moritz nine: oh, good point
17:05 nine How about some $*TEARDOWN defaulting to False? Inline::Perl5 knows that it needs reliable teardown and could set it.
17:06 nine Is exit() called on every script exit? Even after die?
17:07 hoelzro Inline::Perl5 is exactly why I bring up DESTROY =)
17:07 nine Yes, $*TEARDOWN would cause a global change done in secret by some module you maybe didn't know you even loaded and Perl 6 generally tries to avoid this as much as possible. But we're talking about disabling a performance optimization, not more.
17:08 hoelzro nine: I think exit() is only called when a developer actually calls it; I don't think it's implicitly called
17:09 virtualsue joined #perl6
17:09 hoelzro but there is some common teardown logic that should probably when exit() is called, a script is done, a (catchable) signal is sent to the process, or a script die()s
17:09 nine Incidentally, Perl 5 has a similar switch called PL_destruct_level
17:10 dalek rakudo/nom: 43a2cee | moritz++ | src/Perl6/Actions.nqp:
17:10 dalek rakudo/nom: RT #125661: Wrongly reported error for non-existing private method
17:10 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/43a2ceee9c
17:10 synbot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=125661
17:11 araujo joined #perl6
17:12 jnthn moritz++ # fixing RTs
17:15 moritz jnthn: that was a fun one
17:15 jnthn nine: If you want something to run for sure at exit, why not deal with it using END?
17:16 dalek roast: 7502f61 | moritz++ | S12-methods/private.t:
17:16 dalek roast: RT #125661: Error reporting for private methods in a try
17:16 dalek roast: review: https://github.com/perl6/roast/commit/7502f61542
17:16 synbot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=125661
17:17 hoelzro interesting
17:18 telex joined #perl6
17:26 [Tux] joined #perl6
17:30 lizmat joined #perl6
17:33 nine jnthn: that's an....intriguing question. I'd have to keep a list of all existing Inline::Perl5 objects in the module, but I guess I will need that anyway if I ever manage to fix precomp.
17:33 mr-foobar joined #perl6
17:36 moritz std: "{@_}"
17:36 camelia std 28329a7: OUTPUT«5===SORRY!5===␤Placeholder variable @_ may not be used here because the surrounding block takes no signature at /tmp/KtKHm0o3jI line 1:␤------> 3"{7⏏5@_}"␤Check failed␤FAILED 00:00 135m␤»
17:36 moritz std: sub { "{@_}" }
17:36 camelia std 28329a7: OUTPUT«5===SORRY!5===␤Placeholder variable @_ may not be used here because the surrounding block takes no signature at /tmp/Z31VYFmGpH line 1:␤------> 3sub { "{7⏏5@_}" }␤Check failed␤FAILED 00:00 137m␤»
17:37 nine jnthn: but then DESTROY will for sure only be called on exit since my list will keep the objects alive.
17:41 moritz m: my @result = gather { take 'foo=bar'.split('=') }; say @result
17:41 camelia rakudo-moar 96fed8: OUTPUT«␤»
17:41 moritz m: my @result = gather { my $dummy = take 'foo=bar'.split('=') }; say @result
17:41 camelia rakudo-moar 96fed8: OUTPUT«␤»
17:41 amurf joined #perl6
17:42 xinming Is there weak ref idea in perl6?
17:42 jnthn nine: Yeah, true...
17:42 jnthn xinming: No
17:43 xinming jnthn: In perl6, There will be no circular references of objects?
17:43 jnthn I suspect we'll get them at some point in the future, but given you don't need them to solve circular data structure issues in Perl 6, they're not on the 6.christmas critical path (so they'll happen after it)
17:43 xinming Ok, Got it.
17:44 jnthn xinming: Sure there will; you only need weak references to solve that if your GC doesn't consider reachability. Perl 6 impls have a GC that does :)
17:44 nine jnthn: what do you think about $*TEARDOWN?
17:45 jnthn nine: It'll (a) need VM support, and (b) can only work if all VMs cooperate
17:45 xinming Another question, How do we do auto cast of types? For now, perl6 forces types, And is there a syntax to do Auto cast or select a function to cast type to target type?
17:46 xinming Hmm, I'll read doc again. I was just offline and used perl6 to write some simple scripts, and got these questions. :-)
17:47 jnthn nine: It's worth noting that the JVM actually deprecated their equivalent to what you're suggesting with $*TEARDOWN: http://docs.oracle.com/javase/7/docs/api/java/lang/Runtime.html#runFinalizersOnExit(boolean)
17:48 jnthn xinming: Perl 6 doesn't force types; there are coercion types though (only supported in signatures so far) where you tell it to coerce
17:48 jnthn xinming: Writing type annotations in your program is opt-in, and the builtins are not strict :)
17:48 lizmat joined #perl6
17:48 llfourn http://docs.perl6.org/language/functions -- at the bottom Coercion Types
17:49 Bey joined #perl6
17:49 estrabd joined #perl6
17:50 llfourn jnthn: can you declare your own coercion types and coercion subs or is that planned?
17:50 lizmat m: for ^10 -> $a, $b, $c { say $a,$b,$c }   # somehow I expected the last iteration to have 9,Any,Any
17:50 camelia rakudo-moar 96fed8: OUTPUT«012␤345␤678␤Too few positionals passed; expected 3 arguments but got 1␤  in block <unit> at /tmp/pjIqdFNer3:1␤␤»
17:51 jnthn lizmat: Optional
17:51 lizmat Optional?
17:51 jnthn lizmat: ?
17:51 jnthn m: for ^10 -> $a, $b?, $c? { say $a,$b,$c }
17:51 camelia rakudo-moar 43a2ce: OUTPUT«012␤345␤678␤9(Mu)(Mu)␤»
17:51 jnthn Like that :)
17:51 Bey I'm trying to use Bailador. I use a simple code : get '/' => sub { "hello world ! }. But when I go at http://0.0.0.0:3000/, there is nothing....
17:52 jnthn lizmat: You can pick defaults too
17:52 lizmat m: for ^10 -> $a, $b, $c { say $a,$b,$c }   # somehow I expected the last iteration to have 9,Any,Any
17:52 camelia rakudo-moar 43a2ce: OUTPUT«012␤345␤678␤Too few positionals passed; expected 3 arguments but got 1␤  in block <unit> at /tmp/ZuKUHdzBNa:1␤␤»
17:52 xinming llfourn: THanks, Will read them.
17:52 flussence Bey: try 127.0.0.1
17:52 jnthn lizmat: Then you expected wrong. :)
17:53 Bey flussence : I've already done, it does'nt work...
17:53 jnthn lizmat: We can't just pass Mu or Any also, 'cus the type expected may be tighter...
17:54 jnthn llfourn: You can write your own coercion logic either with a method on the thing being coerced, or making the target invokable (the latter will surely change as it's crazy... :))
17:55 llfourn jnthn: Thanks. I'm guessing coercion to Int you implement .Int?
17:55 lizmat jnthn: but I seem to recall that that was just what it was supposed to do?
17:55 lizmat hmmm...
17:55 jnthn Yes
17:55 jnthn lizmat: No, it's never been designed that way afaik
17:55 flussence star: use Bailador; get q{/} => sub { "hello world" }; say 'alive'
17:55 camelia star-m 2015.03: OUTPUT«alive␤»
17:55 simcop2387 joined #perl6
17:55 jnthn llfourn: Yes
17:56 llfourn jnthn: Thanks! Will document :)
17:56 Hor|zon joined #perl6
17:56 flussence «perl6 -MBailador -e 'get q{/} => sub { "hello world!" }; baile'» works fine here...
17:56 mr-foobar joined #perl6
17:57 llfourn ^^ the above also works for me
17:57 flussence (and "curl 0.0.0.0:3000" works too, which I did not expect)
17:57 diana_olhovik joined #perl6
17:58 Bey Rakudo tells me there is no error, but I've nothing in my browser
17:58 lucasb m: sub f {}; try f(1); say 'alive'
17:58 camelia rakudo-moar 43a2ce: OUTPUT«5===SORRY!5=== Error while compiling /tmp/gvaPHi0K9j␤Calling f(int) will never work with declared signature ()␤at /tmp/gvaPHi0K9j:1␤------> 3sub f {}; try 7⏏5f(1); say 'alive'␤»
17:59 flussence does it print any log messages at all when you try to open the URL?
17:59 masak lucasb: it's failing at compile-time, and so the `try` is too late to help.
17:59 llfourn Bey: can you try telnet localhost 3000?
17:59 lucasb masak: hmm, let me test something here :)
17:59 Bey I'm stupid. It works perfectly with localhsot:3000
18:00 Bey That's strange...
18:01 lucasb perl6 --optimize=off -e 'sub f {}; try f(1); say "alive"'
18:01 Bey Very strange.. anyw, thanks for the gelp !
18:01 lucasb ^^ that works, with optimize off
18:01 lucasb this is after moritz latest commit
18:02 flussence oh well... wonder why localhost works and 127.0.0.1 doesn't.
18:02 jnthn lucasb: Sure, 'cus the optimizer does the analysis that proves your code could never possibly work
18:02 jnthn lucasb: It's actually trying to work out if it can do various opts, but sometimes finds that a call is simply impossible.
18:03 lucasb so... my program changes behaviour with/without --optimize=off ?
18:04 jnthn lucasb: Your program *doesn't even run* with --optimize=off
18:04 jnthn uh, without it
18:04 jnthn It says "==SORRY==" which means you didn't make it to runtime.
18:05 lucasb jnthn: Are we on the same revision? :)
18:05 jnthn lucasb: I'm going on the camelia output above.
18:05 jnthn m: sub f {}; try f(1); say 'alive'
18:05 camelia rakudo-moar 43a2ce: OUTPUT«5===SORRY!5=== Error while compiling /tmp/DBbQTYPE18␤Calling f(int) will never work with declared signature ()␤at /tmp/DBbQTYPE18:1␤------> 3sub f {}; try 7⏏5f(1); say 'alive'␤»
18:05 jnthn The ===SORRY!=== there means it's a compile-time error
18:06 lucasb ok, so with optimize it fails, but *without* optimize it works?
18:07 flussence what you're doing there is the equivalent of trying to assign to *NULL in C then catching the resulting segfault. You might be able to bludgeon gcc into compiling it, but it's doing you a favour by not
18:07 PerlJam lucasb: "works" is probably a little kind :)
18:07 jnthn lucasb: With optimize it does exactly what is intended.
18:08 jnthn lucasb: Without it, a piece of code that is always going to throw a runtime exception makes it to runtime and does so.
18:09 lucasb jnthn: And then, the 'try' catches it and the program outputs 'alive'
18:09 jnthn *sigh* you're not listening, and I'm going to make dinner.
18:09 jnthn &
18:10 lucasb jnthn: Sorry, I'm just confused :)
18:10 lucasb For context, I'm on 43a2cee here locally
18:11 lucasb perl6 --optimize=off -e 'sub f {}; try f(1); say "alive"'  #=> outputs 'alive'
18:12 lucasb perl6 -e 'sub f {}; try f(1); say "alive"'  #=> dies with "SORRY ..."
18:12 PerlJam lucasb: correct
18:12 RabidGravy yes, because the try caught the failure to run the sub
18:12 lucasb I'm just saying the behaviour change with/without optimize
18:12 PerlJam lucasb: right.
18:13 lucasb ok, thanks! I think we are now on the same page :)
18:13 RabidGravy with optimse on code fails at compile time, off it fails at run time
18:13 lizmat joined #perl6
18:13 PerlJam fails, but you ignore the failure
18:13 RabidGravy except the try stops it from reporting
18:14 masak lucasb: heh, I already explained what's confusing you above.
18:14 masak <masak> lucasb: it's failing at compile-time, and so the `try` is too late to help.
18:14 flussence m: sub f {}; BEGIN try f(1); say 'alive'
18:14 camelia rakudo-moar 43a2ce: OUTPUT«5===SORRY!5=== Error while compiling /tmp/dYqd7aioTO␤Calling f(int) will never work with declared signature ()␤at /tmp/dYqd7aioTO:1␤------> 3sub f {}; BEGIN try 7⏏5f(1); say 'alive'␤»
18:14 lucasb masak: That was the first message, indeed; I should have tried to understand a little better. Thanks
18:14 flussence m: sub f {}; try EVAL 'f(1)'; say 'alive'
18:14 camelia rakudo-moar 43a2ce: OUTPUT«alive␤»
18:17 RabidGravy the moral of the story is don't stop the compiler from helping you
18:19 lucasb For context, what I was saying was about this recently fixed bug 125661
18:19 lucasb http://irclog.perlgeek.de/perl6/2015-07-15#i_10896584
18:19 lucasb ^^ cognominal had already seen that before too
18:29 masak RabidGravy: yes, that's a good summary.
18:30 lizmat joined #perl6
18:31 moritz lucasb, cognominal: fwiw the bug was that 'try' makes Failures fatal by some AST wrapping, and that AST wrapping didn't preserve the .node property of the AST
18:31 moritz (which is used for generating line numbers)
18:31 moritz and the optimizer tripped over that while generating the error message
18:39 lucasb moritz: thanks for the explanation!
18:39 muraiki joined #perl6
18:40 lucasb again, thanks everybody and sorry for my stubbornness ;)
18:40 brrt joined #perl6
18:50 brrt \o
18:54 xfix joined #perl6
18:57 rangerprice joined #perl6
18:58 colomon joined #perl6
18:59 rangerprice hi
19:00 brrt hi rangerprice
19:01 rangerprice how are you ?
19:03 bin_005 joined #perl6
19:05 brrt hmm.. ok, i guess
19:05 brrt what about you? :-)
19:06 rangerprice I'm fine :D
19:07 brrt that's good to hear
19:10 moritz m: say ((1, 2), <a b>).map({ .join(',')}).join('|')
19:10 camelia rakudo-moar 43a2ce: OUTPUT«1,2|a,b␤»
19:10 moritz m: say ((1, 2), <a b>).flatmap(&uc).join('|')
19:10 camelia rakudo-moar 43a2ce: OUTPUT«1|2|A|B␤»
19:11 moritz m: say &flatmap
19:11 camelia rakudo-moar 43a2ce: OUTPUT«5===SORRY!5=== Error while compiling /tmp/C4x_bsX2EU␤Undeclared routine:␤    &flatmap used at line 1␤␤»
19:11 colomon joined #perl6
19:12 colomon joined #perl6
19:13 rangerprice camelia: help
19:13 camelia rangerprice: Usage: <(star-m|nqp-js|p5-to-p6|star-j|nqp-parrot|nqp-moarvm|pugs|rakudo-jvm|prof-m|std|niecza|debug-cat|rakudo-moar|nqp-jvm|n|rn|rm|Prn|nom|m|P|nr|r-m|star|r-j|r-jvm|sm|rPn|rnP|nqp-mvm|p6|nrP|nqp-p|sj|nqp-q|nqp|perl6|j|rakudo|Pnr|rj|nqp-m|nqp-j|r|nPr|p56)(?^::\s) $perl6_program>
19:13 rangerprice m: say "hello world"
19:13 camelia rakudo-moar 43a2ce: OUTPUT«hello world␤»
19:14 espadrine joined #perl6
19:17 yqt joined #perl6
19:17 dalek doc: 9c3e602 | paultcochrane++ | lib/Type/Blob.pod:
19:17 dalek doc: Link to unpack() correctly
19:17 dalek doc:
19:17 dalek doc: This fixes issue #102.
19:17 dalek doc: review: https://github.com/perl6/doc/commit/9c3e6022da
19:18 sjn_phone_ joined #perl6
19:22 dalek doc: 3056664 | moritz++ | lib/Type/List.pod:
19:22 dalek doc: Document flatmap
19:22 dalek doc:
19:22 dalek doc: also mention that map does not flatten, and that map
19:22 dalek doc: iterates $N-at-a-time depending on the signature
19:22 dalek doc: review: https://github.com/perl6/doc/commit/3056664cd3
19:25 dalek doc: 60bcf47 | (Lloyd Fournier)++ | lib/Language/modules.pod:
19:25 dalek doc: Added section to modules.pod so it is about writing and loading modules
19:25 dalek doc: as well as publishing them
19:25 dalek doc: review: https://github.com/perl6/doc/commit/60bcf47e6e
19:25 dalek doc: 5b306d4 | moritz++ | lib/Language/modules.pod:
19:25 dalek doc: Merge pull request #104 from LLFourn/master
19:25 dalek doc:
19:25 dalek doc: Increased scope of modules.pod by adding "Creating and Using" section
19:25 dalek doc: review: https://github.com/perl6/doc/commit/5b306d4cff
19:29 sjn_phone_ joined #perl6
19:29 dalek doc: 373e657 | moritz++ | CONTRIBUTING.md:
19:29 dalek doc: CONTRIBUTING: use active voice
19:29 dalek doc: review: https://github.com/perl6/doc/commit/373e657618
19:29 dalek doc: 8b02392 | moritz++ | CONTRIBUTING.md:
19:29 dalek doc: CONTRIBUTING: describe the role of Mojolicious in more detail
19:29 dalek doc: review: https://github.com/perl6/doc/commit/8b02392974
19:30 sjn_phone_ joined #perl6
19:31 sjn_phone_ joined #perl6
19:40 AlexDaniel joined #perl6
19:41 dalek roast: 233c86e | moritz++ | S32-array/splice.t:
19:41 dalek roast: RT #125571: splice should not allow to sneak in values that violate the type constraint of an array
19:41 dalek roast: review: https://github.com/perl6/roast/commit/233c86ea3e
19:41 synbot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=125571
19:43 [Tux] niner: Your perl was not configured to build a shared library; unable to build
19:44 [Tux] I didn't change anything in the past 12 days on that system. Why does this suddenly occur?
19:46 moritz [Tux]: the detection of shared library builds is new; maybe it's a false error message in your case?
19:47 [Tux] Inline::Perl5 now fails to install
19:47 [Tux] I have a libperl.a which contains shared loaded objects
19:47 [Tux] (note it is not a libperl.so)
19:47 hobbs joined #perl6
19:48 [Tux] the fact the libperl is not called .so does not mean that it cannot be used!
19:48 [Tux] all .o's that .a are -oPIC
19:48 [Tux] -fPIC
19:49 [Tux] test             50000    45.765    45.672
19:49 [Tux] test-t           50000    45.060    44.967
19:49 [Tux] just for your information
19:52 jnthn [Tux]: What are the two columns?
19:53 jnthn [Tux]: Previous run, current run?
19:53 [Tux] 50000 is the counted "fields" if not 50000 it is wrong
19:53 [Tux] 45.765 is the total run time of the proces
19:54 [Tux] 45.672 is the same with the "perl6 -I. -e1' time subtracted
19:54 jnthn aha
19:54 jnthn Gotcha, thanks
19:54 [Tux] 45.765/45.060 once was 38.6/35
19:55 [Tux] 20150309
19:55 [Tux] https://github.com/Tux/CSV/blob/master/README.speed
19:56 jnthn When we were faster and wronger, I presume...
19:56 jnthn And probably various other issues.
19:56 brrt joined #perl6
19:57 jnthn Hmm, 20150422 to 20150426 was a big jump
19:57 japhb .tell El_Che I like the 'We are' variant at https://gist.github.com/nxadm/13c4817dd1b4b792a99f and I prefer "Community Guidelines" to both "Social Code" and "Code of Conduct" (the worst of those three IMO).
19:57 yoleaux 08:22Z <brrt> japhb: i like it, i'm playing devils advocate for a bit
19:57 * jnthn will have to take a look under the profiler to figure that out
19:57 yoleaux japhb: I'll pass your message to El_Che.
19:57 yoleaux 08:23Z <brrt> japhb: it's an implicit assumption that everyone *can* be a positive, kind member, and that's not really true, e.g. the TempleOS guy
19:58 brrt code of conduct is pretty generally disliked, it seems
19:58 PerlJam aye.  "Community Guidelines" is way better
19:58 brrt fairly sure that's due to the political/interest group connotation
19:58 japhb yoleaux: I actually was *not* making the assumption that everyone can be positive and kind.  I think more people can than appears at first blush, but yes, some people just can't, and we shouldn't feel like we need to coddle them.
19:59 japhb grrr, not yoleaux, brrt
19:59 brrt :-)
20:00 brrt yeah, i can see that point
20:01 brrt another thing i noticed is that people reject the idea of specifying which 'minorities'
20:01 brrt would be 'guarded' by a CoC
20:02 PerlJam because we don't discrimminate :)
20:02 hobbs Good rules are timeless :)
20:02 masak and it's practically the only way to be inclusive
20:02 japhb As for the value of doing community guidelines, the previous commentators who said part of it is bracing against our own Endless September is a large part of it.  Part of it, as ShimmerFairy notes, is just giving people the psychological safety of knowing the community has their back, even when they are new enough that that is not already self-obvious.  Basically, making them feel safe enough to *join*, and spend their precious time here.
20:03 brrt i think it's a good idea too; i think the underlying notion has many names eg qw(meritocracy inclusiveness diversity)
20:04 japhb Yes, exactly, I've never liked the idea of specifying a list of protected classes.  That just invites trouble.  But saying flat out that we don't discriminate is much better.
20:04 brrt i'd classify that idea as 'the logic of discrimination does not apply here'
20:06 brrt unfortunately, each of the above (e.g. meritocracy) also has its problems
20:06 japhb hobbs++  # Yes, completely agreed.  We ought to be able to look at our guidelines in 100 years and still feel like they are the right ones.
20:07 masak brrt: I see "meritocracy" being called out in various places as really meaning "status quo systemic inequality"
20:07 brrt yeah, that for one thing
20:08 dalek roast: c91fdd2 | moritz++ | S12-class/stubs.t:
20:08 dalek roast: Fix test description
20:08 dalek roast: review: https://github.com/perl6/roast/commit/c91fdd26e5
20:08 japhb Yeah, meritocracy is a -1 from me.  It has been too often used to justify systematic biases.
20:08 brrt one of the bigger problems of meritocracy is that you have to have a shared value system in the first place
20:09 japhb brrt: Or even agree that "merit" is a low-dimensional space.
20:09 brrt yes, that too :-)
20:12 japhb Oh, one more thought on writing down these community guidelines: Perl 6 was always intended to be, among other things, the community's rewrite of itself.  I think we did a pretty good job of that.  Now we need to document that rewrite, so that other communities, now and in the future, can benefit from it.  Witness the simple effect of -Ofun, when we started publicizing that.
20:13 masak japhb: +1
20:13 dalek roast: 7f8b240 | moritz++ | S05-capture/dot.t:
20:13 dalek roast: Remove three tests with wrong assumptions about capturing
20:13 dalek roast: review: https://github.com/perl6/roast/commit/7f8b240e36
20:14 japhb I feel like we have a moral responsibility to make more of the world a better place, not just #perl6, and I think we have found (a tiny piece) of how to do that.  (Yes, maybe this marks me as an idealist -- but just like there's a moral reason for the Artistic License, I see a moral value here.)
20:14 llfourn joined #perl6
20:15 vendethiel joined #perl6
20:15 * japhb feels like a young kid again saying stuff like that ... "Mom, Dad, why don't people just *stop having wars*?"
20:16 masak well, kids are not exactly wrong asking that. it's a really good question.
20:17 brrt unfortunately (call me what you will), sometimes there is a good answer
20:17 dalek roast: bcc7651 | moritz++ | S06-operator-overloading/sub.t:
20:17 dalek roast: Correct test, unfudge for rakudo (RT #124980)
20:17 dalek roast: review: https://github.com/perl6/roast/commit/bcc7651bb8
20:17 synbot6 Link:  https://rt.perl.org/rt3/Public/Bug/Display.html?id=124980
20:18 brrt although i'm at a loss to find the good reason for most wars in recent history
20:20 brrt wwi, no good reason, wwii, no good reason, korean war, unknown, vietnam war, no good reason, first gulf war, hmmm
20:20 brrt falkland war, not sure
20:22 jnthn Well, some regimes like to put in their troops to other countries to try and foster domestic support for themsleves...
20:22 brrt that's... not a good reason, i think
20:22 jnthn No, it's really, really not
20:22 jnthn But depressingly common and - worst of all - successful
20:22 jnthn In the immediate, anyway.
20:23 brrt the first gulf war was actually quite reasonable. the kuwaiti had lots of oil and few people. the iraqi had lots of people and proportionally less oil
20:24 brrt the allied forces had lots of reasons to ensure stable supply of oil
20:24 brrt which was threatened
20:24 * japhb hopes that he did not metaphorically throw a match into a powder room
20:25 japhb In any case, I think there are lots of good reasons to write our community guidelines down.  I want them to feel like *our* guidelines, not someone else's.  And I want them to be timeless, operating at the same extra level of abstraction that Perl 6, and the community participants, already do.
20:27 jnthn japhb: Heh, no, I'm just too easily distractable today. :) And +1 on "community guidelines" being documentation of the nice things we already have - though it's to help us hold on to them as well as sharing it with other communities, I think :)
20:27 brrt there's a reason we've had a spot for S27 for all htis time :-)
20:28 japhb jnthn: Yup, with you 100%.
20:28 japhb brrt: :-)
20:29 dalek Inline-Perl5: 070367e | (Stefan Seifert)++ | lib/Inline/Perl5.pm6:
20:29 dalek Inline-Perl5: Manually shut down the first perl on program exit
20:29 dalek Inline-Perl5:
20:29 dalek Inline-Perl5: Perl 6 does not guarantee that DESTROY methods are called at program exit.
20:29 dalek Inline-Perl5: Make sure at least the first Perl 5 interpreter is correctly shut down and thus can e.g.
20:29 dalek Inline-Perl5: flush its output buffers. This should at least fix the vast majority of use cases.
20:29 dalek Inline-Perl5: People who really do use multiple Perl 5 interpreters are probably experienced enough
20:29 dalek Inline-Perl5: to find proper workarounds for their cases.
20:29 dalek Inline-Perl5: review: https://github.com/niner/Inline-Perl5/commit/070367e243
20:31 colomon joined #perl6
20:32 dalek Inline-Perl5: 9ce0272 | (Stefan Seifert)++ | Build.pm:
20:32 dalek Inline-Perl5: Revert "Detect perls not configured with a shared library"
20:32 dalek Inline-Perl5:
20:32 dalek Inline-Perl5: This reverts commit f7b4564ae909100d3daaf9fb1b795e587e01903d.
20:32 dalek Inline-Perl5:
20:32 dalek Inline-Perl5: Apparently Inline::Perl5 can also use a libperl.a instead of a libperl.so,
20:32 dalek Inline-Perl5: so -Duseshrlib is not really neccessary. We'll have to find another way
20:32 dalek Inline-Perl5: to give a better error message to the user.
20:32 dalek Inline-Perl5:
20:32 dalek Inline-Perl5: Thanks to Tux++ for pointing this out.
20:32 dalek Inline-Perl5: review: https://github.com/niner/Inline-Perl5/commit/9ce0272fe4
20:32 nine [Tux]: ^^^ sorry for the inconvenience
20:34 nine brrt: without this stable supply of oil we could have had research into renewable energy decades earlier
20:34 brrt that sucks, huh
20:34 nine japhb: Count me in on your idealistic world view :)
20:35 sjn_phone joined #perl6
20:35 brrt although, i'm not sure if that's true; wind and hydropower for mechanical purposes have been developed before the industrial revolution
20:36 brrt i should not spam you so much :-)
20:36 nine those were very modest times power consumption wise
20:38 brrt see you tomorrow!
20:43 colomon joined #perl6
20:50 colomon joined #perl6
20:52 PerlJam japhb++  Feeling like a young kid again is a good thing :)
20:54 dalek Inline-Perl5: b561782 | (Stefan Seifert)++ | p5helper.c:
20:54 dalek Inline-Perl5: Make sure to call PERL_SYS_TERM only after the last interpreter was shut down
20:54 dalek Inline-Perl5:
20:54 dalek Inline-Perl5: This is hopefully the real fix for the rare but infamous panic: MUTEX_LOCK (22)
20:54 dalek Inline-Perl5: hoelzro++ provided further insight:
20:54 dalek Inline-Perl5: "Ok, I was mistaken in my thoughts on DESTROY order, but DESTROY does play a
20:54 dalek Inline-Perl5: role in this bug. The real culprit is the END block at the end of Inline::Perl5
20:54 dalek Inline-Perl5:
20:54 dalek Inline-Perl5: p5_terminate tears down all of the global things that perl uses, and after
20:54 dalek Inline-Perl5: the END block is run is when the DESTROY methods are run. That's the bug!"
20:54 dalek Inline-Perl5:
20:54 dalek Inline-Perl5: p5_terminate now only sets a flag that p5_destruct_perl checks to see if in
20:54 dalek Inline-Perl5: addition to perl_destruct() and perl_free() it should also PERL_SYS_TERM()
20:54 dalek Inline-Perl5:
20:54 dalek Inline-Perl5: Thanks to hoelzro++ for investigating!
20:54 dalek Inline-Perl5: review: https://github.com/niner/Inline-Perl5/commit/b561782e58
20:56 vendethiel nine: what happens if I spawn interpreters from multiple threads? *g*
20:57 nine vendethiel: try it :) I did fix a couple of places where this could be a problem and did very basic testing. But nothing fancy so far.
20:57 * vendethiel has never used Perl5, so rarely thinks about "I need to use *that* from CPAN!"
20:57 nine vendethiel: aaand yes, the simple interpreter counter I just introduced is not thread safe.
20:58 vendethiel nine: yeah, I was thinking about that :P. Well, will it matter in practice, that I don't know...
20:58 PerlJam vendethiel: as soon as you realize CPAN has solutions to all manner of problem waiting for you to download and use .... you may think it more
20:58 nine There are probably a few more threading bugs you'd hit before this becomes an issue
20:58 vendethiel nine: right :)
20:59 vendethiel PerlJam: oh, I'm sure it does (like other languages' package managers), I just don't know about them
21:03 vendethiel admittedly, it's much easier to find them in CPAN than in, say, npm, with names like "bluebird"
21:03 RabidGravy is it possible to have multi submethods at all or is it just the peculiar way BUILD is called that prevent that being a multi
21:04 RabidGravy actually ignore me, that does work
21:05 PerlJam nine++ btw, Inline::Perl5 has given me great joy on occassion.  It's a tool I didn't realize I needed until it existed (ack was that way for me too).  Before I'd try to solve problems in Perl 6 where the bar was too high, now I can punt to Perl 5 modules for some bits and just code the P6 parts that I want.
21:05 dalek Inline-Perl5: b41e80a | (Stefan Seifert)++ | README.md:
21:05 dalek Inline-Perl5: Document that P5 functions/methods are always called in list context
21:05 dalek Inline-Perl5:
21:05 dalek Inline-Perl5: Thanks to hoelzro++ for pointing out this omission.
21:05 dalek Inline-Perl5: review: https://github.com/niner/Inline-Perl5/commit/b41e80a68e
21:06 jnthn RabidGravy: You'd probably have to write an explicit proto submethod
21:06 vendethiel nine: up to which point do you think P5 looking like P6 made things easy?
21:06 nine PerlJam: glad to hear that it's helping someone in the way I inteded it to :)
21:06 * vendethiel considered making an Inline::Racket
21:07 RabidGravy jnthn, roight. The rreason was confiused was I forgot that a named param multi always wins over a () unless you ! it
21:07 nine vendethiel: it certainly helped that there's a 1 to 1 relationship between many things in Perl 5 and 6. Most of all that both have scalars, arrays, hashes and so on. But the same is true with Python with Python 3 actually being a bit closer to Perl 6 than Perl 5 is.
21:08 hoelzro nine: do you have plans to allow a scalar or void context call?
21:09 nine hoelzro: plans no. Having a plan would indicate that I knew how :) But I'm wide open for suggestions and am thinking about how to do this.
21:09 hoelzro nine: I've been thinking about that myself; exporting something like scalar($p5_obj.method) would be nice
21:10 hoelzro you could easily export scalar as a macro
21:10 hoelzro but how it would indicate scalar context, I'm not sure
21:10 nine vendethiel: what helps more than the surface similarities is that both languages are very dynamic and give you tools to cheat like AUTOLOAD in Perl 5 and FALLBACK in Perl 6.
21:10 vendethiel my $*P5_CONTEXT = P5::Context::Void *g*
21:11 nine That would probably be the easiest way.
21:11 vendethiel nine: that's fair. I love the way you can go, say, perl6+java+python+perl5, though :P
21:11 vendethiel nine: that can be "abstracted away" in a sub or macro
21:12 nine vendethiel: I really want to try using Perl 5's Inline::Java through Inline::Perl5 on MoarVM ;)
21:12 vendethiel I'd love to read about that!
21:14 nine I hope to find some time in the coming weeks. But there's also YAPC::EU on the horizon and I have a talk to prepare. And I've 10 minutes more than at FOSDEM and quite a few gotchas less to talk about ;)
21:14 llfourn joined #perl6
21:17 * vendethiel has yet to finish his own talk, but still needs to read some of Inline::5's code
21:18 lizmat joined #perl6
21:19 amurf joined #perl6
21:21 EMAILBEN145 joined #perl6
21:24 * jnthn has yet to start his talk...
21:29 masak same here.
21:29 masak (I haven't started jnthn's talk either) :P
21:31 lizmat joined #perl6
21:33 rangerprice joined #perl6
21:33 masak in Perl 6, if you add an op without specifying precedence, it gets the tightest possible precedence, right?
21:33 jnthn No
21:34 jnthn There are defaults by category
21:34 masak oh, I see.
21:34 jnthn masak: https://github.com/rakudo/rakudo/blob/nom/src/Perl6/Grammar.nqp#L4322
21:35 * jnthn lols at the name of the hash just before the method...
21:35 masak :P
21:36 lizmat joined #perl6
21:36 masak well, I chose the "tighter than anything" approach for custom ops in 007, and I ran into an interesting design surprise.
21:38 masak namely, prefixes and postfixes are comparable in their precedence -- they're kind of in the same "precedence namespace", unlike infixes -- and it's perfectly legit to mix the precedences of prefixes and postfixes.
21:39 masak and so you may end up with a precedence ordering which flips back and forth between postfix and prefix any number of times.
21:40 masak on the other hand, you don't want to install prefixes to be tighter than postfixes by default, because... that's not Least Surprise. prefixes tend to be looser.
21:41 masak so you go, "ok, let's install a new postfix at the boundary". but there is no "the boundary" under this model.
21:43 masak I'm tempted to go with "the first (loosest) boundary", but I don't really have anything that tells me that'll work better than "the last (tightest) boundary".
21:43 Peter_R joined #perl6
21:44 yqt joined #perl6
21:46 skids https://github.com/rakudo/rakudo/pull/475
21:47 skids For anyone who wants to test drive this:
21:47 skids http://irclog.perlgeek.de/perl6/2015-07-22#i_10935111
21:47 skids Hopefully someone will have the tuits to look at it this time.
21:51 jnthn skids: I quickly glanced over the patches and didn't see anything of particular concern.
21:52 skids Thanks!
21:53 jnthn How do they do against spectest?
21:54 skids Only two tests fail on it, both of which arguably need tweaking.
21:55 skids I don't trust my perf numbers because I'm working while the spectest runs, so I don't know whether there is impact there.
21:55 skids The tests in need of attention are noted in the commits.
21:55 jnthn I doubt there'll be a significant one, having looked over the patches
21:55 jnthn OK, do you have a PR for roast also?
21:56 jnthn Or if I was the merge this, could you quickly take care of those?
21:56 colomon joined #perl6
21:56 skids I have push on roast.  I can take care of them tonight, after a decommute (they just turned off the air conditioning here.)
21:56 EMAILBEN145 left #perl6
21:57 jnthn skids: OK, well if I'm still about just ask when you're ready, or anybody else with a commit bit
21:57 skids Should be about an hour.
21:57 jnthn (Just want to avoid having folks report spectest regressions here)
21:57 jnthn OK
21:57 skids Thanks again.
21:57 jnthn It hit 40C here today and will take forever to cool down, so I may well still be about...
21:57 jnthn Trying to sleep until it's cooler in here is a waste of time :)
22:01 sjn jnthn: it's "just" 31 here :-P
22:01 vendethiel joined #perl6
22:02 jnthn sjn: Urgh. Can we have winter back soon? :)
22:02 * jnthn hopes he gets a really cold winter in return for this really hot summer
22:02 sjn Oslo is 21°C now
22:03 sjn Lund is 18°C!
22:03 sjn damn those scandis with their fine weather
22:03 masak shoulda stayed in Sweden, jnthn :P
22:05 rangerprice https://www.youtube.com/watch?v=bd2B6SjMh_w beautiful music :)
22:05 sjn daxim_: bumped into that guy who organizers the "We Love Programming Languages" meetup, and he said he's planning to do a Perl 6 meetup in October
22:06 jnthn masak: Bah, I barely got any snow in the winters there the last couple of years :P
22:06 sjn daxim_: I suggested that he should come to the KAMEL-grillen wien.pm is organizing, to plan a little
22:06 jnthn And I'm closer to some mountains here if I really want to escape the heat :)
22:07 jnthn sjn: Is that...an event when camel meat will actually be grilled?
22:07 sjn jnthn: come and see for yourself ;)
22:07 jnthn Damn, I don't have "it's a long way" as an excuse... :)
22:08 sjn jnthn: http://meetu.ps/2J86t5
22:09 jnthn oooh :)
22:09 * sjn isn't sure about the availability of camel meat in Vienna
22:09 jnthn I'm actually free that day
22:09 sjn :D
22:10 jnthn Though I'm meant to be taking some time off travel and relaxing so I can feel well again...
22:10 jnthn Will see :)
22:10 sjn jnthn: bring your wife :D
22:10 sjn Donauinsel is a quite nice place for a barbecue
22:11 jnthn sjn: She's not made it here yet, sadly
22:11 sjn ah, ok
22:11 sjn well, next time then :)
22:12 jnthn But yeah, once she is here I doubt I'll be able to go to Vienna without her wanting to follow along.
22:13 masak 'night, #perl6
22:13 jnthn 'night, masak
22:16 lucasb left #perl6
22:17 sjn nn masak
22:19 rangerprice m: print 2+2
22:19 camelia rakudo-moar 43a2ce: OUTPUT«4»
22:19 rangerprice m: print 2+|${5}
22:19 camelia rakudo-moar 43a2ce: OUTPUT«5===SORRY!5=== Error while compiling /tmp/BQ5q7fb_Cq␤Unsupported use of ${5}; in Perl 6 please use $4␤at /tmp/BQ5q7fb_Cq:1␤------> 3print 2+|${5}7⏏5<EOL>␤»
22:21 virtualsue joined #perl6
22:22 RabidGravy right that's me done for wednesday
22:22 itz bratislava-- # damm hawt
22:23 itz bratislava++ # cheap beer
22:24 jnthn :)
22:42 TEttinger joined #perl6
22:43 yqt joined #perl6
22:52 rangerprice Perl 5 threads's  compatible with Perl 6 ?
22:54 TimToady not really, except in the sense you could run a different P5 interpreter with each P6 thread
22:54 TimToady what P5 calls threads is more like fork emulatoin
22:54 TimToady *tion
22:56 rangerprice So P6 have a newly threading system ?
22:56 TimToady yes
22:57 * virtualsue is pretty sure the answer is yes, because the p5 one is ready for retirement
22:57 rangerprice oh okay
22:58 virtualsue actually i was 100% positive, not pretty sure ;-)
22:58 huf lr
22:58 huf dammit.
23:08 amurf joined #perl6
23:18 rangerprice joined #perl6
23:19 rangerprice Is P6 threading system totally fonctionnal ?
23:25 skids joined #perl6
23:27 amurf joined #perl6
23:28 skids Hrm.  It's hard to think of a way to get the test for 115436 to still test for the essence of the RT, when Failure.new now takes pretty much anything as args.
23:29 cognominal_ joined #perl6
23:38 dalek roast: c2a8e0e | skids++ | S0 (2 files):
23:38 dalek roast: Tweaks for PR#475 (new Exception code.)
23:38 dalek roast: Tweak one test that was expecting .payload to behave like .message
23:38 dalek roast: Comment out another that checks an error on something that cannot fail anymore
23:38 dalek roast:   (unless someone can figure out a way to make it.)
23:38 dalek roast: review: https://github.com/perl6/roast/commit/c2a8e0ee30
23:45 skids jnthn: so that commit should take care of roast.
23:46 jnthn skids: OK :)
23:46 dalek rakudo/nom: cac198b | skids++ | src/core/Backtrace.pm:
23:46 dalek rakudo/nom: Fix some bounds overruns in new lazy backtrace code discovered
23:46 dalek rakudo/nom: when stringifying the same Backtrace more than once.
23:46 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/cac198b55a
23:46 dalek rakudo/nom: 8042bc9 | skids++ | src/core/Exception.pm:
23:46 jnthn skids++
23:47 dalek joined #perl6
23:47 skids sorry dalek :-)
23:49 cognominal joined #perl6
23:53 FatalNIX joined #perl6
23:53 skids jnthn++ thanks for the review/push.
23:53 FatalNIX so whats new with p6?
23:53 skids Since about what time?
23:53 FatalNIX since the beginning of time.
23:53 FatalNIX no er, hmm. maybe 2 years ago
23:53 skids Well, that's a long story :-)
23:54 FatalNIX which is about the last time I looked at it I think
23:56 skids Hrm, the pause button was pushed on parrot so moar and jvm are the current leading backends.
23:56 jnthn My usual weekly grant report (covers last week's work): https://6guts.wordpress.com/2015/07/23/this-week-concurrency-stuff-multi-dimensional-stuff-stuff-stuff/
23:56 jnthn 'night, folks o/
23:56 skids o/ sleep well
23:56 FatalNIX jvm is racudovm or something right?
23:57 FatalNIX I'm actually sort of curious how easy it is for java to call perl6 functions and vice versa now.
23:57 skids rakudo-j yes.
23:57 FatalNIX because if the integration is good, maybe I can make an example test Minecraft mod with perl 6.
23:57 FatalNIX as a demo
23:57 skids Not sure about that kind of interop.  NativeCall for C is now formally part of rakudo.
23:58 FatalNIX I see,
23:58 skids I'm just not a java person so I haven't kept tabs on that.
23:58 skids But Perl5/Perl6 interop is here now.
23:59 FatalNIX neither am I a java person, but if it worked nicely, it would mean I can use p6 in the future and not have to deal with fugly java :D
23:59 rangerprice joined #perl6

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

Perl 6 | Reference Documentation | Rakudo