Camelia, the Perl 6 bug

IRC log for #parrot, 2011-09-25

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:13 soh_cah_toa joined #parrot
00:46 PacoLinux_ joined #parrot
01:20 AzureSto_ joined #parrot
01:56 jsut_ joined #parrot
02:16 schmooster joined #parrot
02:20 silug joined #parrot
05:34 Tene joined #parrot
05:52 mj41 joined #parrot
06:48 redicaps joined #parrot
07:00 mj41 joined #parrot
07:04 SHODAN joined #parrot
08:06 mj41 joined #parrot
08:49 nine quiet Sunday morning in Parrot land...
09:39 contingencyplan joined #parrot
09:59 zby_home joined #parrot
10:18 dalek rakudo/nom: 6fd0773 | moritz++ | src/core/Buf.pm:
10:18 dalek rakudo/nom: comparison operators for Buf
10:18 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/6fd0773fce
11:05 SHODAN joined #parrot
11:10 nine YYYYEEEAAAAHHH!!!
11:11 tadzik no way :)
11:11 tadzik what was it?
11:11 nine tadzik: I have no idea :)
11:12 tadzik so, what made it work? :)
11:12 nine tadzik: but as I ran out of options, I did the only thing left: ignore it and go on. So I merged with current master which includes whiteknight/kill_threads and somehow this made it work :)
11:12 tadzik I assume your "YEAH" doesn't mean finally beating all the campaigns on Heroes of Might and Magic 3
11:12 tadzik huh, wow :)
11:13 tadzik so are threads killed or not?
11:13 nine Well at least it builds now. Tests fail, but that was to be expected
11:13 nine tadzik: threads are killed, green threads live
11:13 tadzik cool
11:14 tadzik nine++
11:22 Kulag joined #parrot
11:23 nine Is there any way to remove the green_threads branch from my github repo? I completely re-created the branch locally and would like to push it now
11:24 tadzik git push origin :green_threads
11:24 tadzik as in "push '' as a new 'green_threads'"
11:25 tadzik there's also some sugar for it, but I never remeber it :)
11:26 nine worked...thanks :)
11:35 tadzik you're welcome
11:36 Psyche^ joined #parrot
11:41 preflex_ joined #parrot
11:45 woosley joined #parrot
12:37 dodathome joined #parrot
12:47 benabik o/
12:47 tadzik hi benabik
12:48 benabik tadzik: What's happening this morning?  Saw that nine++ got green_threads merged.
12:48 tadzik benabik: s/merged/working I suppose
12:49 tadzik or s/working/building/
12:49 benabik Errrr….. yes
12:49 benabik As in got a merge that did something useful on that branch, not the typical meaning of "merged into master".
12:49 benabik Guh.  I should eat/have coffee.
13:56 zby_home_ joined #parrot
14:27 dalek rakudo/nom: efe3f82 | jnthn++ | src/Perl6/Actions.pm:
14:27 dalek rakudo/nom: Fix bug with placeholder parameter binding of arrays.
14:27 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/efe3f82923
14:32 jsut joined #parrot
14:46 whiteknight joined #parrot
15:14 whiteknight good morning, #parrot
15:16 tadzik good afternoon whiteknight
15:23 whiteknight hello tadzik
15:36 mj41 joined #parrot
15:45 dalek rakudo/nom: c4c2e1d | jnthn++ | src/binder/bind. (2 files):
15:45 dalek rakudo/nom: Move/share a struct definition.
15:45 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/c4c2e1d80d
15:45 dalek rakudo/nom: 03afd02 | jnthn++ | src/binder/ (2 files):
15:45 dalek rakudo/nom: Switch junctional arguments check in multi-dispatcher over to being able to cope with natively typed arguments.
15:45 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/03afd0274e
15:45 dalek rakudo/nom: 9e831e0 | jnthn++ | src/binder/multidispatch.c:
15:45 dalek rakudo/nom: Tweak the multi-dispatch candidate check to try and make it aware of native types.
15:45 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/9e831e0d56
15:46 preflex_ joined #parrot
16:20 dalek rakudo/nom: 8b7a2a5 | jnthn++ | src/binder/multidispatch.c:
16:20 dalek rakudo/nom: Get multi-dispatch cache to handle natively typed arguments.
16:20 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/8b7a2a5a7a
16:26 Maddingue joined #parrot
16:29 Maddingu1 joined #parrot
18:21 cotto ~~
18:21 benabik o/ cotto
18:22 cotto benabik, what are your long-term thoughts on the viability of newPOST?  It it likely that it just needs optimization or will it inherently be slower than the current PCT?
18:23 benabik cotto: More optimized codegen from NQP would help, but I don't know how much.  If calling subs isn't cheap, then it will remain slower due to NQP using a sub per block.
18:24 soh_cah_toa joined #parrot
18:28 benabik cotto: Personally, I think using Winxed is likely to have faster gains unless we're going to start seriously working on NQP-rx or integrating new NQP.
18:32 cotto benabik, ok.  It feels like it's time to start getting serious about a PHP HLL and I'm thinking about what it should generate.
18:33 benabik cotto: I meant using Winxed to write PCT in.  But it's also a pretty viable compiler output.
18:33 cotto Yeah.  From what I've heard from NotFound++, PIR isn't too bad to generate.
18:34 cotto Computers are really good at doing boring and/or insane things.
18:34 benabik I really really want to see a nice AST system for us.  Attracting new HLLs is easier if we hand them excellent tools.
18:34 soh_cah_toa +1 to that
18:34 benabik PCT is pretty darn good, but it's not quite right, IMHO.
18:35 cotto benabik, what about it isn't quite right, iyho?
18:36 cotto is it something that we could change in PCT gradually while preserving Rakudo compatibility?
18:37 benabik cotto: It defaults to returning the last value.  It uses exception control handling.  It uses subs for everything.  It doesn't maintain enough structured information to do optimization...
18:37 cotto so no
18:38 benabik cotto: Some of these things are due to parrot, some are due to its structure...
18:38 benabik I'd think we can.  No small bit of it can be flags and options and refactors.
18:41 soh_cah_toa not sure if this applies but i've always thought giving hll devs the option to use nqp or winxed for their grammar/actions files would help attract those who are not big perl fans (in particular, python users)
18:41 soh_cah_toa even though i do, i know not everybody loves perl (or p6)
18:41 soh_cah_toa plus, having multiple options/choices is always a good idea :)
18:41 benabik Well, the grammar/action system is part of NQP...
18:42 benabik Although I think you could write your actions in Winxed, the grammars are NQP.
18:42 tadzik knock knock. Who's there? Language interop
18:42 soh_cah_toa language interop who?
18:43 tadzik rakudo: say Nil.WHO
18:43 p6eval rakudo 8b7a2a: OUTPUT«().hash␤»
18:43 soh_cah_toa :)
18:44 tadzik rakudo: LoL
18:44 p6eval rakudo 8b7a2a:  ( no output )
19:54 dalek Heuristic branch merge: pushed 121 commits to rakudo/optimizer by jnthn
19:58 bluescreen joined #parrot
20:13 mj41 joined #parrot
20:24 soh_cah_toa_ joined #parrot
21:31 dalek Rosella: a9ff03f | Whiteknight++ | / (11 files):
21:31 dalek Rosella: Merge branch 'master' into commandline
21:31 dalek Rosella: review: https://github.com/Whiteknig​ht/Rosella/commit/a9ff03f235
21:31 dalek Rosella: dcd89b4 | Whiteknight++ | src/u (3 files):
21:31 dalek Rosella: play
21:31 dalek Rosella: review: https://github.com/Whiteknig​ht/Rosella/commit/dcd89b4de0
21:31 dalek Rosella: 5266666 | Whiteknight++ | s (4 files):
21:31 dalek Rosella: Add in ProgramMode object to encapsulate details of individual modes of the program.
21:31 dalek Rosella: review: https://github.com/Whiteknig​ht/Rosella/commit/52666664b6
21:31 dalek Rosella: 0567edb | Whiteknight++ | src/unstable/commandline/Arguments.winxed:
21:31 dalek Rosella: Add in a keep flag to determine whether we are fetching an argument out of the list or whether we are just detecting it. Several fixes
21:31 dalek Rosella: review: https://github.com/Whiteknig​ht/Rosella/commit/0567edb4a1
21:31 dalek Rosella: 8ebe2d9 | Whiteknight++ | s (5 files):
21:31 dalek Rosella: Update test_template utility to use the new frontend. Use that experience to make several fixes and changes
21:31 dalek Rosella: review: https://github.com/Whiteknig​ht/Rosella/commit/8ebe2d9b4f
21:31 dalek Rosella: e923be2 | Whiteknight++ | / (2 files):
21:31 dalek Rosella: Add CommandLine to VERSION. cleanups for test_template utility
21:31 dalek Rosella: review: https://github.com/Whiteknig​ht/Rosella/commit/e923be2bae
21:31 dalek Rosella: 6b933f4 | Whiteknight++ | src/u (4 files):
21:31 dalek Rosella: Update test_all_lib to use CommandLine. Cleanups and fixes
21:31 dalek Rosella: review: https://github.com/Whiteknig​ht/Rosella/commit/6b933f49bc
21:31 dalek Rosella: 53590c5 | Whiteknight++ | src/utilities/test_all_lib.winxed:
21:31 dalek Rosella: +some comments for test_all_lib
21:31 dalek Rosella: review: https://github.com/Whiteknig​ht/Rosella/commit/53590c5fbd
21:31 dalek Rosella: 6cd4601 | Whiteknight++ | src/utilities/ (2 files):
21:31 dalek Rosella: Convert mk_winxed_header to the CommandLine library
21:31 dalek Rosella: review: https://github.com/Whiteknig​ht/Rosella/commit/6cd4601ce3
21:31 dalek Rosella: c627fa3 | Whiteknight++ | / (7 files):
21:31 dalek Rosella: Add in stub test files for the commandline classes. Fix up the test_class.winxed.template with some improvements. A few other misc fixes
21:31 dalek Rosella: review: https://github.com/Whiteknig​ht/Rosella/commit/c627fa321c
21:31 dalek Rosella: a2dd3ed | Whiteknight++ | t/commandline/Program (2 files):
21:31 dalek Rosella: Fixup the new CommandLine test files to run and pass. Most tests are mindless stubs
21:31 dalek Rosella: review: https://github.com/Whiteknig​ht/Rosella/commit/a2dd3edccc
21:32 dalek Rosella: bf72e3e | Whiteknight++ | / (5 files):
21:32 dalek Rosella: Flesh out a few tests for common functionality in CommandLine.Arguments.
21:32 dalek Rosella: review: https://github.com/Whiteknig​ht/Rosella/commit/bf72e3e2cc
21:32 dalek Rosella: 5c08cfb | Whiteknight++ | t/harness:
21:32 dalek Rosella: make sure commandline tests are run in harness
21:32 dalek Rosella: review: https://github.com/Whiteknig​ht/Rosella/commit/5c08cfb3b1
21:32 dalek Rosella/commandline: a2dd3ed | Whiteknight++ | t/commandline/Program (2 files):
21:32 dalek Rosella/commandline: Fixup the new CommandLine test files to run and pass. Most tests are mindless stubs
21:32 dalek Rosella/commandline: review: https://github.com/Whiteknig​ht/Rosella/commit/a2dd3edccc
21:32 dalek Rosella/commandline: bf72e3e | Whiteknight++ | / (5 files):
21:32 dalek Rosella/commandline: Flesh out a few tests for common functionality in CommandLine.Arguments.
21:32 dalek Rosella/commandline: review: https://github.com/Whiteknig​ht/Rosella/commit/bf72e3e2cc
21:55 AzureStone joined #parrot
22:01 nbrown joined #parrot
22:41 dalek rakudo/optimizer: 8db01ef | jnthn++ | src/binder/multidispatch. (2 files):
22:41 dalek rakudo/optimizer: First sketch of algorithm for considering multiple dispatches at compile time.
22:41 dalek rakudo/optimizer: review: https://github.com/rakudo/rakudo/commit/8db01ef90d
22:55 Kulag joined #parrot
23:09 aloha joined #parrot
23:18 Coke joined #parrot
23:36 soh_cah_toa joined #parrot
23:37 davidfetter_ joined #parrot
23:48 Coke joined #parrot
23:51 preflex_ joined #parrot
23:52 soh_cah_toa_ joined #parrot

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

Parrot | source cross referenced