Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2006-01-05

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:38 joepurl joined perl6
00:42 azuroth joined perl6
01:03 _brentp is now known as brentp
01:05 lisppaste3 joined perl6
01:22 stevan joined perl6
01:31 stevan_ joined perl6
01:40 cognominal joined perl6
01:45 jiing joined perl6
01:46 SamB joined perl6
01:50 oylenshpeegul left perl6
02:02 dduncan joined perl6
02:07 stevan is now known as stevan_
02:13 Amnesiac joined perl6
02:40 stevan joined perl6
02:53 penk joined perl6
03:23 stevan audreyt: ping
03:40 stevan audreyt: I am trying to make Array a subclass of List and I am running into a problem
03:40 vel joined perl6
03:40 stevan Array is a container class, while List is a normal object
03:40 stevan s/object/class/
03:41 stevan oh wait
03:41 stevan hmmm
03:41 * stevan ponders messing around more with Roles
03:43 stevan nope roles dont work correctly
03:43 stevan the crux of the issue is this
03:44 stevan List.map needs to return a new List instance
03:44 stevan which for List is just List.new()
03:45 stevan but for Array is Array`create()
03:45 stevan I managed to overload List.new() so that I could just pass it a []
03:45 stevan but I cannot get Array.new to work
03:45 stevan I am not sure why
03:47 stevan it keeps trying to use List.new
03:47 stevan oh,.. when I say Array, I mean ^Array
03:47 stevan if I use ::Array.new() it DWIM
04:01 stevan ok, I have the ^List class working 100% now (which is not that hard)
04:02 stevan but ^Array is still having issues
04:02 meppl guten morgen
04:03 stevan mostly the problem is that in the inherited ^List.map and co. I am calling $?CLASS.new(...) and since ^Array.new chokes,.. they choke
04:03 stevan good morning meppl
04:04 stevan audreyt: I think we might be able to fix this if we address this in .CREATE rather than `create, this should give me the level of control I need
04:06 stevan I am going to commit what I have so you can see what I mean
04:07 stevan I thought about making Array just do the List role,.. which I actually think makes more sense, however, we would still have the same issues re: $?CLASS.new(...)
04:08 stevan however doing that would mean I would need to remove pop/push/shift/unshift from List (which currently has immutable versions of them), otherwise it would conflict with the versions of the same methods in the Array role
04:08 svnbot6 r8580 | stevan++ | PIL/Native/Bootstrap/Container/*
04:08 svnbot6 r8580 | stevan++ | - attempted to make Array inherit from List, but ran into an issue with
04:08 svnbot6 r8580 | stevan++ |   how Arrays are created as opposed to other objects (like List), see the
04:08 svnbot6 r8580 | stevan++ |   #perl6 log for a bunch of rambling on the topic
04:08 svnbot6 r8580 | stevan++ |     - added tests for List in list.t
04:08 stevan I don't mind that though, I could just create a ImmutableArray which provided those versions if you need them
04:09 stevan but anyway... I am going AFK, but might be back later,.. if not I will be back in the morning ~10 hours from now
04:09 * stevan &
04:55 Debolaz2 joined perl6
05:13 avar joined perl6
05:31 kanru joined perl6
05:37 Debolaz joined perl6
06:27 Cryptic_K joined perl6
06:31 christo joined perl6
06:33 christo joined perl6
06:53 GeJ joined perl6
07:02 kanru joined perl6
07:22 nothingmuch joined perl6
07:30 lisppaste3 joined perl6
07:52 azuroth joined perl6
08:26 iblechbot joined perl6
08:34 nothingmuch joined perl6
08:37 K_ joined perl6
09:01 r0nny joined perl6
09:04 dada joined perl6
09:13 linnuxxy joined perl6
09:36 pjcj joined perl6
09:39 larsen joined perl6
10:09 nothingmuch joined perl6
10:37 Southen_ joined perl6
10:37 G2 joined perl6
11:20 svnbot6 r8581 | asavige++ | Add anchors to regex in Rpn.pm to fix bug pointed out by Perl Monk ambrus
11:40 marmic joined perl6
11:51 iblechbot joined perl6
12:05 elmex joined perl6
12:35 svnbot6 r8582 | asavige++ | added hask2, a monadic haskell version, based on suggestions and code from Cale Gibbard
12:39 spinclad (soo quiet)
12:39 spinclad (sure this is #perl6?)
12:39 * spinclad checks  (seems to be)
13:00 rafl Debolaz: ping
13:03 chris2 joined perl6
13:07 ingy hola
13:07 ingy seen audreyt
13:07 jabbot ingy: audreyt was seen 21 hours 23 minutes 10 seconds ago
13:08 sahadev joined perl6
13:10 nothingmuch joined perl6
13:12 nothingmuch joined perl6
13:16 rafl Debolaz: I applied your freebsd patches aa and parts of ab to parrot.
13:16 rafl Debolaz: ad isn't needed anymore. The *_shared binaries are gone.
13:17 rafl Debolaz: I'm not sure about ac. Whys is it needed? init::hints::freebsd uses code from /usr/ports/Mk/bsd.port.mk which should do the right thing, shouldn't it?
13:45 audreyt ingy: yo
13:46 audreyt another $job day tomorrow, and then it's sweet weekend hacking with miyagawa and gugod and ingy and clkao :D
13:50 alexe_home joined perl6
13:57 Limbic_Region joined perl6
13:59 gaal morning (or something)
14:01 audreyt yo :)
14:01 audreyt how's life?
14:02 obra audreyt: hi!
14:02 audreyt obra: hey!
14:02 obra how are things?
14:02 obra how was your demo?
14:02 gaal it's... startling sometimes!
14:03 audreyt demo was yesterday. today is another deployment
14:03 gaal I realized I must read more SPJ.
14:03 obra did it go ok?
14:03 obra (Did you see my incompatible change?)
14:03 audreyt the demo went very well. they want foldable rico livegrid, which I (for once) didn't volunteer to implement
14:04 obra heh. so it's extra $?
14:04 audreyt yeah I did, it made for amusing doc verses like "rule rules"
14:04 obra oops
14:04 audreyt nah, it's extra $people-hacking-jifty
14:04 obra oo
14:04 audreyt which I realize as being much better than extra $
14:04 obra yes
14:04 obra Oh. thanks to you, Why noticed jifty
14:05 audreyt yeah, I replied there :)
14:05 audreyt I'm very amused by the "I might look like a fish, but I want to be a squid." thing
14:05 obra My plan worked perfectly. the perl as ruby's quirky brother comment was intended to get ruby people to laugh
14:05 audreyt # http://redhanded.hobix.com/inspect/hehMmmnnPerlPeople.html
14:05 audreyt obra++ # evil genius
14:06 audreyt gaal: hmm, life is startling and reading more SPJ helps?
14:06 audreyt gaal: you have some very strange brain patterns.
14:06 * chris2 spies
14:07 audreyt chris2: hi :)
14:07 chris2 hi audreyt ;)
14:07 ingy hi audreyt
14:07 audreyt ingy: hey. sup?
14:07 ingy YAML-0.50 is ready to go
14:07 ingy just sweeping my rt queue
14:07 obra audreyt: I'm _right now_ about to controllerize wifty. anything I should know?
14:08 ingy obra: probably Perl
14:08 audreyt ingy++
14:08 audreyt obra: nope, although I'd like to see if you end up wanting $m as a shorthand for $Dispatcher->{mason}.
14:08 gaal audreyt: well, revisiting awkward squad now that I know more of the elegant army helps; and syb is some seriously great stuff
14:08 audreyt obra: (and maybe $r)
14:09 obra I'd rather never, ever have to touch mason from within the dispatcher
14:09 obra we'll see
14:10 audreyt obra: that's my intuition too. but we'll see. also per-subclass catalyst-style dispatch may still have some merit -- see if you feel a need for it, although I doubt so
14:11 obra nod
14:11 gaal <=<: wow, you got to use an REM title in a sentence (almost).
14:11 obra I expect I'm going to want to move my controller logic into other classes and have "run" run a class and sub
14:12 gaal audreyt: I got a reply from Simon M reagarding that segfault but I'm actually not sure it helps :/
14:13 audreyt gaal: heh, elegant army indeed. his other works tends to be a wee bit more technical (STM, lexical scoped tyvar, SMP, local type inferencing, delimited continuations) but they are all very nice reads
14:13 audreyt gaal: what did JaffaCake say?
14:14 audreyt obra: I'll still make "run" (&) and introduce a "call" form or for ($)
14:14 svnbot6 r8583 | nnunley++ | Updating test count.
14:14 obra why?
14:14 gaal audreyt: the STM slides weren't too scary first time I read them.. ETA for understanding is probably under a year
14:14 gaal for some value of understanding (or year)
14:14 audreyt because if you prototype it as ($), you need explicit "sub" for "{}"
14:14 obra why not let run take a coderef or a package+sub name?
14:14 obra ah
14:14 audreyt the "sub" is not optional everywhere... not in that perl5 language we are using anyway
14:15 * obra forgets that having run ($) and run(&) won't dwim
14:15 * audreyt mumbles something about the virtue of MMD
14:16 audreyt ingy: woot, looking forward to YAML050 -- I'll take a look at gaal's Haskell Syck emitter tomorrow
14:16 gaal audreyt: basically, I had     thawNode nodePtr = deRefStablePtr (castPtrToStablePtr nodePtr)
14:16 audreyt ingy: which will motivate me to solve the refcycle problem in YAML::Syck as well
14:16 gaal but he said,
14:16 gaal You're attempting to deref the StablePtr, and also run it as an IO
14:16 gaal action!  I bet the original value wasn't an IO action, right?
14:16 gaal (endQ)
14:17 gaal which I'm kinda still not sure I understand
14:17 gaal s/kinda//
14:17 gaal seeing as your original code had readNode which did something very similar
14:19 audreyt I see
14:19 audreyt thawNode nodePtr = return (deRefStablePtr $ castPtrToStablePtr nodePtr)
14:19 audreyt missing an return will deref it as an (IO a) not an (a)
14:20 Debolaz rafl: In theory, patch-ac shouldn't be needed, but the thing is that libc_r doesn't mix well with other threading libraries, and perl is compiled with a different library when threading is enabled (Which is also default in pkgsrc). Parrot will try to use the threading library that perl uses, and this causes some.. conflict of interest.
14:21 mjl69 joined perl6
14:22 gaal ahh! type inference
14:22 * gaal tries a fix
14:24 Debolaz rafl: It's probably local to pkgsrc, so it will probably do the Right-Thing(tm) in other situations and the patch shouldn't be neccesary. :)
14:26 rafl Debolaz: Why is it needed at all for pkgsrc?
14:30 obra audreyt: does http://paste.husk.org/4930 look right?
14:30 Qiang joined perl6
14:31 audreyt obra: yeah
14:32 obra hm. now to figure out why it's not triggering ;)
14:32 * obra was hoping you'd point out something stupid
14:33 obra OH.
14:34 obra Yeah. it totally won't work if I call it "Dspatcher.pm"
14:35 gaal grr, so, still no luck. thawNode :: Ptr () -> IO YamlNode right? not -> YamlNode
14:35 gaal so return $ deRefStablePtr (castPtrToStablePtr nodePtr) is a type error
14:36 gaal return =<< deRef... is still a segfault
14:36 gaal should I NOINLINE this function?
14:42 PerlJam joined perl6
14:45 chris2_ joined perl6
14:49 Debolaz rafl: Because perl is compiled with -lpthread.
14:50 whiteg_ joined perl6
14:54 kanru joined perl6
14:57 gaal argh!
14:58 gaal castPtrToStablePtr :: Ptr () -> StablePtr a
14:58 gaal so if I explicitly say
14:58 gaal let (ptr :: StablePtr YamlNode) = castPtrToStablePtr nodePtr
14:58 gaal then deRefStablePtr ptr
14:58 gaal must have the type IO YamlNode, because
14:58 gaal deRefStablePtr :: StablePtr a -> IO a
14:59 gaal "And yet, it segfaults"
14:59 audreyt ok ok, I give up, I'll take a look instead of doing $work
14:59 gaal my gaudy code was hit by a SIGBUS :-(
14:59 audreyt I guess it's only fair :)
15:01 gaal sorry -- there aren't nearly enough other people I can distract to help me with this :/
15:01 audreyt 'tis fine
15:05 Limbic_Region oh - pugs appears to run much faster than it did before I left for vacation btw
15:06 Limbic_Region where appears is based off eyeball only analaysis
15:10 audreyt gaal: I replicated the segfault. fixing
15:10 * gaal pads the walls in anticipation
15:10 whiteg is now known as whiteg_
15:14 xinming joined perl6
15:30 lisppaste3 joined perl6
15:31 vel__ joined perl6
15:33 gaal so this isn't just a redundant indirection is it? :(
15:37 gaal svnbot6: ?check
15:43 audreyt gaal: fixed
15:43 audreyt you were misusing "new".
15:43 audreyt it should just be "return".
15:43 audreyt I also tidied up the emitterextra code.
15:43 audreyt commit in a bit
15:44 stevan good morning all
15:45 Debolaz Moin.
15:45 Limbic_Region salutations
15:46 gaal awwrrghh
15:46 gaal audreyt: many thanks
15:47 mugwump joined perl6
15:47 gaal shower, bbiab &
15:49 hexmode joined perl6
15:53 chris2_ joined perl6
15:54 chris2 is now known as chris2_
15:56 Eimi joined perl6
15:56 svnbot6 r8584 | audreyt++ | * Fix gaal++'s excellent work at Syck emitting.
15:56 svnbot6 r8584 | audreyt++ |   Pugs's .yaml now won't segfault anymore!
15:57 elmex joined perl6
15:58 audreyt stevan: greetings. hadn't had time to check ImmutableArray (List) and ImmutableFiniteArray (Tuple) stuff
15:59 stevan audreyt: no worries
15:59 stevan I actually think I need to ponder this a little more
16:00 stevan I think we can fix this using $repr in CREATE
16:00 stevan but I have to figure out the best way to do so
16:00 audreyt *nod*
16:00 audreyt `create definitely should die.
16:00 audreyt it was a hack around ::Scalar.isa('Scalar') not true
16:00 stevan we can discuss this when I have a coherent idea of what the front-end should do
16:00 audreyt but under larry's insanity logic, it's now true
16:00 frey joined perl6
16:00 audreyt okay, cool
16:01 stevan oh,.. BTW,.. what is `make_class all about?
16:01 audreyt shower, bbiab :)
16:01 stevan that I couldnt follow
16:01 audreyt oh, it's a hack to make a shadow object
16:01 stevan ok
16:01 audreyt a prototypical thing that does not have a repr
16:01 stevan gotcha
16:01 audreyt that's what ::Scalar is
16:01 * stevan never liked that idea too much,..
16:01 stevan I see
16:08 gaal audreyt: thanks for the accumulator refactoring too. Is that more efficient? does GHC generate smart allocations when ++ing strings under the assumption that they might be ++ed some more?
16:11 gaal where are tests for .perl located?
16:13 Juerd_ It's probably better off without tests in its current state :)
16:16 gaal Juerd_: what makes you say that?
16:16 Juerd_ ?eval :foo.perl
16:16 evalbot_8584 "(\"foo\" => 1)"
16:17 Juerd_ ?eval eval(:foo.perl).perl
16:17 evalbot_8584 "\\\"perl6\""
16:17 Juerd_ Eh, not what I expected, but a good illustration nevertheless
16:17 audreyt gaal: no, it just makes life easier :) PS.append is definitely more efficient
16:18 audreyt gaal: but because you are rturning a String, it ends up being the same O() anyway
16:18 gaal Juerd_: and that means it's better off w/o tests? :)
16:18 Juerd_ gaal: It'll reduce success rate! :)
16:18 gaal audreyt: sure, but the question is who mallocs/frags more
16:18 audreyt Juerd_: tsk ;)
16:18 gaal Juerd_: that's what we have force_todo for :-p
16:19 audreyt gaal: well, FPS does memcpy, so you can't beat that
16:19 audreyt so FPS is gonna win 2:1 I think
16:19 gaal env PASS_ALL make smoke
16:20 gaal even better: echo `grep -cr is t` "tests successful" # bogus, but at least none fail
16:22 audreyt ?eval 'foo'.yaml
16:22 evalbot_8584 "--- foo\n"
16:22 gaal audreyt: theoretically, since we construct a YamlNode tree, we can get an idea for how big the string's going to be, and hint that to the emitting code, which can prealloc
16:22 gaal audreyt: working on undef.yaml, then 42.yaml now.
16:22 audreyt gaal: or we can make FPS's `append` prealloc 10x of needed memory
16:22 audreyt as perl's SVV does
16:22 audreyt s/SVV/SV/
16:22 audreyt on the vain hope that it will be concated more
16:23 audreyt brutal but works
16:23 gaal our VStr's .length is O(n), right?
16:23 audreyt yes, and it's going to be FPS in the future
16:23 audreyt so plan for it :)
16:23 gaal yeah, no way *that* wouldn't be optimized.
16:24 gaal in the case of yaml emitting, the result probably won't be concatted very much
16:24 gaal (typically)
16:24 audreyt it would
16:24 audreyt for large docs
16:25 gaal log "I got { $result.yaml }\n"
16:25 gaal ahh you mean during emitting? of course
16:25 gaal I meant the VStr dumpYaml returns probably won't be concatted much.
16:26 audreyt nod
16:27 gaal ingy about? cd src/YAML-Syck/ ; svn up # hangs
16:28 gaal ahh, no just very slow.
16:29 gaal audreyt: I imagine you know this is broken, yes? perl -MYAML::Syck -le 'print Dump({a=>1})'
16:30 gaal I can work on it a bit later, but the pugs version first :)
16:30 audreyt huh
16:30 audreyt ?eval eval("---\n? a\n: 1\n", :lang<yaml>)
16:30 evalbot_8584 (no output)
16:31 audreyt heh
16:31 audreyt $ perl -MYAML -le 'print YAML::Dump(YAML::Load("---\n? a\n: 1\n"))'
16:31 audreyt ---
16:31 audreyt a: 1
16:31 audreyt so it's not broken, just bad yaml style
16:32 gaal wha!?
16:32 * gaal needs to read some spec :)
16:32 audreyt YAML to JSON is like JavaScript to IO ;)
16:32 audreyt "it may look simple, but really is not"
16:34 gaal wasn't there a shortcut for "-Iblib/lib -Iblib/arch" somewhere?
16:38 audreyt -Mblib ?
16:39 gaal yay :)
16:40 gaal interesting: this pattern match fails on undef.yaml: toYaml VUndef = return YamlNil
16:41 gaal show shows VType (mkType "undef")
16:41 gaal isn't undef a value?
16:42 audreyt undef.yaml is taken as a class call.
16:42 audreyt try (undef).yaml
16:42 audreyt or undef().yaml
16:42 audreyt and yes, it is a bug.
16:42 audreyt s/undef()/undefine()/
16:43 gaal ah, I see
16:43 gaal thanks :)
16:43 mjl69_ joined perl6
16:43 audreyt :)
16:45 TNKS joined perl6
16:46 gaal I'm assuming I can use Pretty's format where appropriate. (I'm assuming it is appropriate soemtimes :)
16:46 TNKS audreyt, I read that you might be working on a Haskell parser for YAML, then I saw on your blog that you might be here.  Is libyaml-haskell still active?
16:48 gaal TNKS: is this useful to you? http://svn.openfoundry.org/pugs/src/Data/Yaml/Syck.hsc
16:48 gaal we're working on it now.
16:48 audreyt TNKS: parser is already quite functional
16:48 audreyt once emission works, I'll cabalize it .
16:49 audreyt TNKS: but you are free to just take the D.Y.S code and pugs's src/syck/*
16:49 audreyt and use it in your project. I don't think there's code for libyaml-haskell in the past.
16:49 elmex joined perl6
16:52 kane_ joined perl6
16:54 TNKS audreyt, cool.  I'll see what I can piece together.
16:55 audreyt cool
16:55 * audreyt waves and sleeps &
16:56 TNKS Actually, (forgive the Haskell talk in a Perl channel), I'd like to stay in Haskell if I can, but for me there's the issue of persistence.  I've read about a few options for serialization, but YAML seems like a clean way to serialize data into a human-readable format.
16:56 PerlJam and #perl6 continues to interwingle with #haskell :)
16:57 TNKS PerlJam, you might have even been the one who mentioned YAML to me yesterday.
16:57 PerlJam TNKS: indeed I was.
16:58 TNKS I'm thinking that YAML might be the smart way for me to go, but I'm not sure.
16:58 TNKS Do you use it with Haskell?
16:58 PerlJam TNKS: If I'm sure for you will that be enough?
16:58 PerlJam actually, I've never used it with haskell, but other lambacamels have.
16:59 gaal TNKS: fwiw D.Y.S is not at all coupled with Perl at least, but probably the dependency on Data.FastPackedString will stay.
16:59 PerlJam s/lamba/lambda/
17:00 SamB hmm. Is Haskell's mascot going to be the lambda?
17:00 TNKS I should probably admit that I've not picked up Perl yet.
17:01 gaal hahaha, this is what I call overRationalization: ./pugs -e 'say (123.12).yaml' => --- 3078/25
17:01 SamB I thought the mascots had to be animals
17:04 Limbic_Region gaal - great way to find approximations of pi?
17:05 Limbic_Region ?eval say (3.14).yaml
17:05 evalbot_8584 pugs: src/Pugs/Prim/Yaml.hs:(46,0)-(49,35): Non-exhaustive patterns in function toYaml  
17:05 gaal Limbic_Region: the version you have doens't do numbers yet, let me check in
17:05 gaal btw tests very very welcome
17:06 gaal I wonder if we can steal them wholesale from somewhere
17:07 TNKS gaal, what does yaml have to do with coverting decimals to fractions?
17:07 Limbic_Region YAML::Parser::Syck  ??
17:07 TNKS or is that more something that pugs is doing before spitting out the yaml.
17:07 Limbic_Region pugs I think TNKS
17:07 Limbic_Region ?eval say '1.1'.perl
17:07 evalbot_8584 OUTPUT["1.1" ] bool::true
17:08 Limbic_Region ?eval say "1.1".perl
17:08 evalbot_8584 OUTPUT["1.1" ] bool::true
17:08 Limbic_Region hmmm
17:08 gaal TNKS: yeah, it's a pugs thing. I was delegating something to pugs' pretty-printer.
17:08 gaal ?eval (123.12).perl
17:08 evalbot_8584 "3078/25"
17:09 TNKS is Pugs the only implementation of Perl 6?  Or is there a comparable alternative?
17:09 gaal anyone remember how YAML represents infinity?
17:10 Limbic_Region ?eval (3.14).perl
17:10 evalbot_8584 "157/50"
17:10 Limbic_Region ?eval (3.1415).perl
17:10 evalbot_8584 "6283/2000"
17:10 gaal TNKS: it's the closest one there is
17:10 Limbic_Region heh
17:10 TNKS gaal, cool,.
17:10 Limbic_Region ?eval (3.1415926535897932384626433).perl
17:10 evalbot_8584 "31415926535897932384626433/10000000000000000000000000"
17:10 Limbic_Region that's cheating
17:10 TNKS ha.  
17:10 gaal TNKS: there's quite some work to do yet. the language isn't fully specced.
17:11 TNKS I see.
17:11 TNKS Is it backwards compatible?  Or only partially so?
17:11 Limbic_Region TNKS - it is suspected that since p6 will have actually specification documents and not a reference implementation, there will be multiple implementations
17:12 SamB what language are the specs to be written in?
17:12 Limbic_Region it will be backwards compatible but there is only a guarantee that about 95% of programs will work 95% of the time without manual intervention
17:12 PerlJam SamB: unfortunately, english  :)
17:12 integral if the synopses are anything to go by, they won't bother writing the syntax specs in perl6 rules :-(
17:12 Limbic_Region give or take a few percentages
17:13 SamB hmm. well... it'll be in good company ;-)
17:14 gaal Limbic_Region: svn up
17:14 svnbot6 r8585 | gaal++ | YAML emitting
17:14 svnbot6 r8585 | gaal++ |   - "moose".yaml; 42.yaml; (undef).yaml work
17:14 svnbot6 r8585 | gaal++ |   - (123.12).yaml gives funny results (because it's delegated to
17:14 svnbot6 r8585 | gaal++ |     pugs' pretty-printer
17:14 svnbot6 r8585 | gaal++ | Tests very welcome!
17:15 Limbic_Region gaal - did you see my YAML::Parser::Syck comment for stealing tests wholesale?
17:15 gaal no, here? /me looks
17:16 Limbic_Region it was just a quick CPAN search so it may not be applicable at all
17:16 gaal Limbic_Region: if you can work in a patch, I'd be much obliged; I want to ci support for aggregates
17:18 Limbic_Region not today gaal
17:18 Limbic_Region I am in another reality ATM
17:18 Limbic_Region my face feels like it is going to jump off my head and that there are strands of anti-gravity plucking at the skin on top of my head
17:19 Limbic_Region OTC anti-histamines in case anyone was wondering
17:19 gaal ah, you're rushing off to write a patent then? :)
17:19 gaal (for antigrav)
17:21 Limbic_Region well, AFAIK there is no direct evidence of gravitons (though a couple indirect) so I would suspect finding anti-gravity first would make a few more headlines
17:22 Limbic_Region ingy hangs out here right - he is the author of YAML::Parser::Syck so perhaps he might be able to shoehorn in the tests
17:23 Limbic_Region oh nevermind, I just looked at the 1 and only test and it just checks to see if the module loaded ok
17:29 justatheory joined perl6
17:30 Limbic_Region audreyt ping
17:30 Limbic_Region oh that's right $work, nevermind me
17:31 Limbic_Region was just wondering if there was going to be a 1 year summary/journal post on Pugs bday in a few weeks
17:32 gaal yay pugs bday!!
17:51 DaGo joined perl6
17:52 ingy YAML?
17:53 SamB one of those s-expr thingies?
17:54 gaal ?eval "hi, ingy!".yaml
17:54 evalbot_8585 "--- hi, ingy!\n"
17:54 gaal ingy: scalars supported; working on the rest now.
17:59 avar yaml is a core method now?
17:59 gaal avar: yes, but this isn't specced.
17:59 avar it would be neat if it were;)
18:00 avar ?eval 1..2.yaml
18:00 evalbot_8585 ()
18:00 avar ?eval [1,2].yaml
18:00 evalbot_8585 pugs: Syck.hsc:(88,4)-(98,94): Non-exhaustive patterns in case  
18:00 avar ;)
18:00 gaal avar: scalars only for now. working on it
18:01 avar k
18:01 gaal but tests welcome.
18:01 avar ?eval 5.yaml
18:01 evalbot_8585 "--- 5\n"
18:01 avar ?eval "5".yaml
18:01 evalbot_8585 "--- 5\n"
18:01 avar mm
18:02 Limbic_Region ?eval (3.14159).yaml
18:02 evalbot_8585 "--- 314159/100000\n"
18:02 Limbic_Region loses its creativity after 4 places
18:02 Limbic_Region ?eval(3.1415).yaml
18:02 evalbot_8585 "--- 6283/2000\n"
18:03 gaal Limbic_Region: like I said, this isn't related to YAML :)
18:03 gaal see Pretty.hs
18:03 Limbic_Region I know - pretty printing
18:03 Limbic_Region ?eval (3.1415).perl
18:03 evalbot_8585 "6283/2000"
18:08 avar $ ruby -ryaml -e 'puts YAML::dump( [ 1/3, nil, false ] )'
18:08 avar - 0
18:08 avar -
18:08 avar - false
18:08 avar in-tee-resting
18:10 gaal nil is wrong. should be "~" AFAIK
18:10 gaal s/nil/the nil case/
18:11 gaal eh. the 1/3 case is wrong too :)
18:11 gaal but thanks for reminding me about booleans.
18:13 rantanplan_ joined perl6
18:22 avar how should 1/3 be encoded?
18:22 avar defenetly can't be 0.333333333..... as that would be lossy
18:23 avar ?eval false.yaml
18:23 evalbot_8585 "--- ::false\n"
18:23 gaal ?eval (bool::false).yaml
18:23 evalbot_8585 "--- bool::false\n"
18:23 gaal needs fixing. tests welcome :)
18:23 avar ?eval ?1.yaml
18:23 evalbot_8585 bool::true
18:24 avar ?eval (?1).yaml
18:24 evalbot_8585 "--- bool::true\n"
18:57 G2 joined perl6
19:03 gaal whee! arrays in.
19:18 peacock joined perl6
19:23 Limbic_Region gaal - if you made a commit for arrays it hasn't shown up yet
19:24 gaal not yet, coming up.
19:28 nothingmuch joined perl6
19:30 svnbot6 r8586 | gaal++ | Data.Yaml.Syck:
19:30 svnbot6 r8586 | gaal++ |  - emitting arrays
19:30 svnbot6 r8586 | gaal++ |  - unbreak eval :lang<yaml>
19:30 lisppaste3 joined perl6
19:33 webmind hmm, someone didn't nice his compile on feather :(
19:34 gaal webmind: I see a cc1 at pr 25
19:34 webmind fibonaci had some cc1 at 0
19:34 webmind and ghc
19:34 gaal looks niced now.
19:34 gaal oh but you're right
19:35 gaal it isn't niced
19:35 gaal it's just pr 25.
19:35 webmind uhuh
19:35 gaal luqui: ping
19:35 gaal ?eval <hello world>.yaml
19:35 evalbot_8586 "--- \n- hello\n- world\n"
19:35 gaal whee!
19:36 webmind yay yaml :)
19:37 Lartuv joined perl6
19:38 gaal urp? is hash_fetchKeys the only way to get at a flattened hash?
19:40 Lartuv joined perl6
19:49 Amnesiac joined perl6
19:52 * Limbic_Region would think that hash_fetchKeys would return a list of keys not the flattened hash
19:53 Limbic_Region otherwise it is a poorly named routine
19:53 gaal Limbic_Region: yes, I said get *at*
19:53 audreyt hash_fetch.
19:54 audreyt gives you back a list of pairs.
19:54 gaal whew!
19:55 Limbic_Region er um - I completely missed the at there
19:56 Limbic_Region audreyt - planning anything special for Pugs' bday?
19:56 gaal Limbic_Region: sorry, am too terse.
19:58 audreyt Limbic_Region: finishing the migration to new shiny runtime? :)
19:58 audreyt (that'd be 6.28.0)
19:58 Limbic_Region wooo hooo
19:58 Limbic_Region something like 3 weeks to go
19:59 audreyt yup
20:04 gaal hmm, what's wrong with this expression?   return $ YamlMap =<< hash_fetch v'
20:05 audreyt fmap YamlMap (hash_fetch v')
20:05 audreyt alternatively
20:05 audreyt (return . YamlMap) =<< hash_fetch v'
20:05 audreyt because of precedence
20:05 * audreyt waves and goes back to sleep &
20:06 gaal night :)
20:06 * gaal hopes to dream of functors one day
20:48 nothingmuch joined perl6
20:49 kanru joined perl6
20:54 mjl69 joined perl6
20:57 gaal any idea why the inferencer is expecting [(Int, Int)] when I deal with flattened hashes?
20:59 gaal ahhhh my bad
20:59 gaal it thought I was currying with the recursion guard
21:03 avar does p6 confess == p5 Carp::confess?
21:04 gaal There's a Carp::longmess in the Prelude. did someone add confess?
21:04 gaal ?eval confess "I don't know"
21:04 evalbot_8586 Error: No such method in class Str: "&confess"
21:05 avar I was under the impression that the whole croak/confess stuff was std & imported by default
21:06 gaal avar: everything in the prelude is str and imported by default, but I don't know whether confess there :)
21:06 gaal longmess is.
21:06 dduncan joined perl6
21:07 hugues23 joined perl6
21:08 gaal avar: just looked. no, there's only longmess defined. you're invited to add the others :)
21:09 avar ;)
21:09 avar what file is it?
21:11 gaal src/perl6/Prelude.pm
21:36 gaal hmmm, "No instance for (HashClass Val) arising from use of `hash_fetch'"
21:36 * gaal eyes Pretty
22:11 vel__ left perl6
22:26 anatolyv joined perl6
22:27 mel222 is now known as anatolyv
22:27 anatolyv is now known as mel222
22:28 leo obra: ping
22:28 obra hey
22:28 leo hi
22:28 obra <- still in spain, a bit sick :/
22:28 obra What's up?
22:28 leo recent #parrotsketch logs seem to be not really enough public
22:29 obra nod. the bot failed somehow
22:29 obra chromatic sent me logs to post
22:29 obra I've been harassing tpf people to get them featured somewhere
22:29 obra perhaps I hsould talk to coke instead
22:30 leo put em on use.perl.org or - yes forward to Coke
22:46 gaal hey anatolyv, didn't notice you stepping in
22:46 gaal openfoundry is acting up :(
22:49 anatolyv hey
22:49 svnbot6 r8587 | gaal++ | Data.Yaml.Syck: broken support for hash emitting; thinks hashes are
22:49 svnbot6 r8587 | gaal++ | lists of pairs.
22:49 svnbot6 r8587 | gaal++ | Lots of trace prints left in the code till I figure this out.
22:58 stevan obra: ping
22:59 stevan http://planetsix.perl-foundation.org/ has gone a bit wonky
23:00 stevan (IIRC this was something you were doing, if not ignore me)
23:00 * stevan &
23:10 gaal If anyone wants to give the yaml stuff a moose, please do
23:10 gaal I'm off to bed &
23:13 DaGo joined perl6
23:30 lisppaste3 joined perl6
23:44 oylenshpeegul joined perl6

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

Perl 6 | Reference Documentation | Rakudo