|
Parrot 4.6.0 "Wild Parrots of Telegraph Hill" | parrot.org/ | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC Set by moderator on 18 July 2012. |
|||
|
00:06
travis-ci joined
|
|||
| travis-ci | [travis-ci] parrot/parrot#499 (master - 81ed05e : jkeenan): The build was broken. | 00:06 | |
| [travis-ci] Change view : github.com/parrot/parrot/compare/9...ed05e5f03a | |||
| [travis-ci] Build details : travis-ci.org/parrot/parrot/builds/2091194 | |||
|
00:06
travis-ci left
|
|||
| benabik | travis: "The virtual machine is being powered down" | 00:13 | |
|
02:00
benabik joined
02:28
kid51 joined
05:02
JimmyZ joined
|
|||
| alvis | msg allison Thank you. I will try, depending on tuits, to attend to the blog this weekend. | 05:55 | |
| aloha | OK. I'll deliver the message. | ||
|
06:04
brrt joined
|
|||
| dalek | d_parrot: 04e33bc | (Bart Wiegmans)++ | / (3 files): Evaluation of scripts in a child-interpreter |
06:46 | |
| d_parrot: 81d89e7 | (Bart Wiegmans)++ | / (4 files): initial interpreter pool implementation |
|||
| d_parrot: 508e07a | (Bart Wiegmans)++ | module/mod_parrot (8 files): Moving code arround. This is not a usable commit |
|||
| d_parrot: 19ca2d3 | (Bart Wiegmans)++ | / (10 files): The Great Reorganisaton is on its way. Loaders are next |
|||
| d_parrot: d77a6eb | (Bart Wiegmans)++ | module/mod_parrot_pool. (2 files): Delete redundant file |
|||
| d_parrot: ed80765 | (Bart Wiegmans)++ | / (10 files): Updated loaders. Tests now pass |
|||
| d_parrot: d998114 | (Bart Wiegmans)++ | / (16 files): Merge sub-eval back |
|||
| d_parrot: a002179 | (Bart Wiegmans)++ | loader/common.winxed: Merge global storage in common |
|||
|
07:21
JimmyZ_ joined
10:26
JimmyZ joined
11:16
brrt joined
11:18
lucian joined
|
|||
| dalek | d_parrot: 95bbaeb | (Bart Wiegmans)++ | pudding/Server.pm: Verbose testing |
11:27 | |
| p: 23731de | jnthn++ | src/QAST/Compiler.nqp: Register type inference fix. |
12:08 | ||
| p: 83e2a14 | jnthn++ | src/QAST/Operations.nqp: Add some missing op variants. |
12:09 | ||
| kudo/use_unbox: cb8904d | jnthn++ | src/Perl6/Ops.pm: Switch set for unbox_i. |
|||
| kudo/use_unbox: dbeaeaa | jnthn++ | src/core/ (2 files): Various tweaks to avoid trying to unbox things that aren't 6model objects. |
|||
|
12:14
whiteknight joined
|
|||
| whiteknight | good morning, #parrot | 12:15 | |
| brrt | \\o whiteknight | ||
|
12:16
JimmyZ joined
|
|||
| brrt | you can help me out i think | 12:17 | |
| i need a title for my new blog post :-) | |||
| whiteknight | sure, what's it about? | 12:21 | |
| brrt | ehm, how i refactored the codebase to make it much more logical | 12:23 | |
| how i used an awesome hack to convert import route data | |||
| s/convert// | |||
| and how much more options i just gained | 12:24 | ||
| whiteknight | awesome hacks are awesome | ||
| brrt | slightly evil hack as well | ||
| i have a structure called mod_parrot_route | |||
| which is (sensibly) composed of all string fields | |||
| whiteknight | nice | ||
| brrt | containing, compiler, file, classname, routine | 12:25 | |
| whiteknight | is all that committed? | ||
| brrt | yes, in fact :-) the last two are currently optional | ||
| i'm thinking of adding logic to deduce the compiler | |||
| whiteknight | okay, I'll update soon and take a look at it | ||
| brrt | from the file | ||
| whiteknight | oh, like from the extension? | ||
| brrt | but the Great Hack is | ||
| from either the extension or the shebang | 12:26 | ||
| as in, if the shebang ends in winxed and we know winxed, thats probably our language | |||
| anyway, the great hack is i import that data as a string array | |||
| whiteknight | okay, that's a great idea | ||
| brrt | as if it were argc and argv | 12:27 | |
| which is ehm, sensible in a way | |||
| whiteknight | There is an API routine to import an argv/argc pair into a ResizableStringArray | 12:28 | |
| brrt | exactly :-) | 12:30 | |
| whiteknight | okay, it's a tidbit that isn't well-publicized | ||
| brrt | well i knew and i made use of it :-) | 12:33 | |
| whiteknight | great | ||
| brrt | but yeah, tests pass again | 12:34 | |
| whiteknight | mind if I make a small commit? | 12:38 | |
| dalek | kudo/use_unbox: 2219b21 | jnthn++ | src/Perl6/Ops.pm: Make p6bool a bit smarter. |
12:41 | |
| kudo/use_unbox: e84c9a8 | jnthn++ | src/core/MapIter.pm: Fix something we got away with but probably should not have. |
|||
| whiteknight | actually, I don't know if I have write permissions on that repo | ||
| whiteknight | maybe I'll make a fork | ||
| brrt | ... yeah please o | 12:43 | |
| do | |||
| i can give you write access | |||
| although, a fork and a pull request is probably easier | 12:44 | ||
| i'm off taking a nap | 12:47 | ||
| dalek | kudo/use_unbox: 3695086 | jnthn++ | src/core/Mu.pm: Fix mixins. |
||
| kudo/use_unbox: ad03484 | jnthn++ | src/core/List.pm: Add missing coercions. |
13:10 | ||
| kudo/use_unbox: effbe85 | jnthn++ | src/core/Temporal.pm: Fix for new use of unbox_i. |
|||
|
13:11
tuxit joined
|
|||
| dalek | kudo/nom: 39b18fa | jnthn++ | src/Perl6/Grammar.pm: Pull in C<repeat> parsing fix from STD. |
13:30 | |
| kudo/nom: cb8904d | jnthn++ | src/Perl6/Ops.pm: Switch set for unbox_i. |
13:32 | ||
| kudo/nom: dbeaeaa | jnthn++ | src/core/ (2 files): Various tweaks to avoid trying to unbox things that aren't 6model objects. |
|||
| kudo/nom: 2219b21 | jnthn++ | src/Perl6/Ops.pm: Make p6bool a bit smarter. |
|||
| kudo/nom: e84c9a8 | jnthn++ | src/core/MapIter.pm: Fix something we got away with but probably should not have. |
|||
| kudo/nom: 3695086 | jnthn++ | src/core/Mu.pm: Fix mixins. |
|||
|
13:32
kid51 joined
|
|||
| kudo/nom: ad03484 | jnthn++ | src/core/List.pm: Add missing coercions. |
|||
| kudo/nom: effbe85 | jnthn++ | src/core/Temporal.pm: Fix for new use of unbox_i. |
|||
| kudo/nom: a548a46 | jnthn++ | src/ (6 files): Merge branch 'use_unbox' into nom |
|||
| kudo/nom: a07e65b | jnthn++ | tools/build/NQP_REVISION: Bump to latest NQP for QAST improvements. |
|||
| whiteknight | brrt: pull request sent. | 13:47 | |
| dalek | rrot/whiteknight/io_cleanup1: 6ea7fc1 | Whiteknight++ | src/platform/win32/io.c: Fix macro name for windows build |
13:50 | |
| rrot/whiteknight/io_cleanup1: 494ae0d | Whiteknight++ | src/platform/win32/io.c: On second thought, remove this dang macro all together. |
13:51 | ||
| kudo/nom: 3a24210 | jnthn++ | src/Perl6/ (2 files): Fix double-quoted strings used in regexes; they now interpolate their contents properly (and are parsed with the right language). |
13:52 | ||
| kudo/nom: ea5f78c | jnthn++ | docs/ROADMAP: Remvoe a completed ROADMAP item. |
13:57 | ||
| rrot/whiteknight/io_cleanup1: eabb359 | Whiteknight++ | src/pmc/socket.pmc: Don't mix declarations and code to fix c++ build |
|||
| rrot/threads: 39d88fc | Whiteknight++ | src/scheduler.c: Several small cleanups, comments and a few TODO notes in scheduler.c. No functional changes |
14:06 | ||
| rrot/threads: 0a11e25 | nine++ | / (2 files): Unshare runloop_id_counter to fix races runloop_id_counter should be local to the interp instead of global. There's no reason to share it between threads. Leaving it global was an oversight. |
|||
| rrot/threads: 3451668 | nine++ | src/scheduler.c: Merge branch 'threads' of github.com:parrot/parrot into threads |
|||
| rrot/threads: 78cb477 | Whiteknight++ | src/pmc/alarm.pmc: Cleanup Alarm PMC. Add some comments/docs. Cleanup some code. Add the ability to get the alarm time as a PMC from get_pmc_keyed_int, in case anybody wants to try that. |
|||
| rrot/threads: 33a45e3 | Whiteknight++ | src/pmc/scheduler (2 files): A few small cleanups to scheduler and schedulermessage PMCs. No functional changes |
|||
| rrot/threads: 9596588 | Whiteknight++ | t/pmc/nci.t: Merge branch 'threads' of github.com:parrot/parrot into threads |
|||
| whiteknight | hmmm... I wonder why those commits got pushed again? | ||
| kid51 | Yes, same messages but different commit numbers! | 14:10 | |
| whiteknight | and I didn't munge any history, I don't think | ||
| JimmyZ | sometimes git removes changes of the old commit without a new commit | 14:11 | |
| moritz | that happens after a merge commit | 14:12 | |
| whiteknight | I checked the diffs. Nothing nefarious popped in | 14:15 | |
| also, I figured out why my windows VM is randomly failing so many tests: Out of disk space | |||
| kid51 | To think that in 2012 somebody, somewhere, would ever run out of disk space! | 14:16 | |
| moritz | the sha1 of a commit is a checksum over the diff, commit message and parent commit(s) | ||
| so if you do a rebase anywhere, the parent changes, and thus the sha1 | 14:17 | ||
| whiteknight | kid51: I created this VM with far too little disk space | ||
| kid51 | Well, my hat is off to you for being bold enough to try to build VMs! | 14:18 | |
| Wait, come to think of it, I'm not wearing a hat. | |||
| whiteknight | I've been using VMs for a long time. I can't run itunes on linux | ||
| so I copied my itunes vm, installed parrot, and have been using that | |||
| kid51 | Ah, yes, that is one draw back to Linux on desktop | 14:19 | |
| With new $job, I no longer have an ubuntu desktop at my disposal. | |||
| whiteknight | I used to use opensolaris on a vm, but after the oracle merger I haven't found a distro I liked. And I use 32-bit linux on a vm too, for testing | ||
| kid51 | Well, for 6 years I used Ubuntu as my $job desktop linux. Particularly in the last 2-3 years I had no complaints. | 14:21 | |
| (except that they laid off Allison ;-) ) | |||
| whiteknight | yeah, I'm a big ubuntu fan | 14:23 | |
| kid51 goes in search of breakfast | 14:24 | ||
|
14:26
brrt joined
|
|||
| dalek | kudo/nom: 7ec7a2b | jnthn++ | src/Perl6/ModuleLoader.pm: Explicitly disallow pre-compilation of a module unless its dependencies have been pre-compiled already. Instead, fail with an informative error message. Before, we went ahead with the pre-compilation, but the resulting output would not work. |
14:26 | |
| d_parrot: eae1a8d | Whiteknight++ | loader/ (4 files): Add several forward class declarations to silence winxed warnings during the build. No functional changes. |
14:32 | ||
| d_parrot: 4c4d1c6 | (Bart Wiegmans)++ | loader/ (4 files): Merge pull request #1 from Whiteknight/master Silence Winxed build warnings |
|||
| brrt | whiteknight++, tests also still pass :-) | 14:33 | |
| i'm actually so optimistic right now, i could imagine running parrot.org on mod_parrot | 14:39 | ||
| that is | |||
| if you are willing and happy to run apache with mpm_prefork | |||
|
14:50
kid51 joined
15:08
preflex_ joined
|
|||
| dalek | rrot/threads: 6c765e1 | jkeenan++ | compilers/imcc/pbc.c: [codingstd] Conform to line length standard. |
15:31 | |
| whiteknight | brrt: honestly, I have no idea what mpm_prefork is, or what the advantages/disadvantages would be | 16:00 | |
| arnsholt | IIRC, prefork is the "traditional" Unix approach where you fork off worker processes that handle incoming requests | 16:02 | |
| Infinoid | ...as a way to avoid using any threads at all | 16:03 | |
|
16:15
brrt joined
|
|||
| whiteknight | msg brrt I get this error when I test now: Cannot load /home/andrew/projects/mod_parrot/build/mod_parrot.so into server: libparrot.so.4.6.0: cannot open shared object file: No such file or directory | 16:26 | |
| aloha | OK. I'll deliver the message. | ||
| whiteknight | hmmm, wait. I feel like I remember this. I don't remember how I fixed it | 16:27 | |
| damnit, I forgot ldconfig again | 16:30 | ||
| nine | rurban: make test passed for me on both mips and powerpc. | 16:42 | |
|
16:58
brrt joined
|
|||
| brrt | ping whiteknight | 17:06 | |
| ldconfig problem again? ;-) | |||
| maybe there I can do something about that | |||
| statically link parrot during compilation, or something. | 17:07 | ||
| I don't k now | |||
|
17:23
brrt joined
|
|||
| brrt | also, prefork is indeed single-threaded | 17:24 | |
|
17:58
mtk joined
|
|||
| dalek | kudo/nom: 0dd82c7 | jnthn++ | src/Perl6/Actions.pm: If symbols are declared in a block, it certainly cannot be a hash constructor. Fixes segfault discovered by ChoHag++. |
18:01 | |
|
18:12
davidfetter joined
19:21
lucian joined
19:37
benabik joined
|
|||
| whiteknight | brrt: yeah, I've added ldconfig to my build script, like I should have done months ago | 19:48 | |
| I'm out for the evening. Later | |||
|
20:32
benabik joined
21:06
perlite_ joined
21:55
lucian joined
23:20
slavorg joined
|
|||