Camelia, the Perl 6 bug

IRC log for #parrot, 2008-12-15

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:14 gmansi joined #parrot
00:40 TiMBuS joined #parrot
01:05 Whiteknight joined #parrot
01:15 tetragon joined #parrot
01:42 dalek r33906 | pmichaud++ | assoc:
01:42 dalek : [rakudo]:  Update postcircumfix, remove unused get_pmc_keyed from Protoobject.
01:42 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33906
01:45 MariachiElf joined #parrot
02:00 Andy We need a Compiler Squad.
02:00 Andy http://leonardo-m.livejournal.com/73732.html
02:00 Andy Just to exercise everything.
02:00 Andy I AM GONNA DO SO MUCH PARROT SHIT WHEN THIS BOOK IS DONE.
02:01 Andy I miss writing tests and fixing compiler warnings.
02:01 Andy An honest man's work.
02:04 dalek r33907 | pmichaud++ | trunk:
02:04 dalek : [rakudo]:  Merge associative branch into trunk.
02:04 dalek : Rakudo now supports slicing on hashes.
02:04 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33907
02:04 dalek r33908 | pmichaud++ | assoc:
02:04 dalek : Delete obsolete branch (now merged to trunk).
02:04 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33908
02:06 Whiteknight Andy, what book are you working on?
02:07 Andy http://oreilly.com/catalog/9781934356265/
02:10 ChrisDavaz joined #parrot
02:11 Tene So, how would I catch just warnings, but not errors, in Perl 6?
02:11 * Tene checks S04, maybe
02:11 Tene HTTP is pretty broken for me right now.
02:12 Whiteknight how is warn() implemented? exception?
02:12 Tene It will be.
02:12 Tene Right now it just prints to stderr
02:12 Tene That's what I'm going to work on next, but I want to know what it looks like in Perl 6 before I implement in Parrot.
02:13 Whiteknight once it's exception-based, it should be easy enough to trap it.
02:13 Tene Right, I know how to do it on the Parrot level just fine.
02:17 pmichaud CONTROL is used to trap warnings, I think.
02:17 Tene Also, what exception type should Perl 6 warnings be?
02:17 Tene Parrot exception type?
02:17 pmichaud might to make one up, if one doesn't already exist.
02:18 Tene If it's CONTROL, is it CONTROL_WARN?
02:18 jimmy joined #parrot
02:19 pmichaud it'd be okay with me to create a CONTROL_WARN
02:20 Tene Or is it just: no 'warnings'; or something in a block?
02:20 pmichaud to catch the warning, it'd be    CONTROL { when <something> { ... } }
02:21 pmichaud where <something> is the way we distinguish warning exceptions from other types
02:22 pmichaud I don't know if that's specced.
02:23 Tene Hmm.
02:23 pmichaud (the <something> part, that is.)
02:23 pmichaud but yes, it'd be in S04.
02:23 pmichaud or perhaps in the test suite.
02:26 Tene So, if Parrot automatically resumes on WARNING and NORMAL exceptions, the default exception severity should probably be ERROR, not NORMAL like it is now.
03:08 jimmy seen bernhad
03:08 purl I haven't seen 'bernhad', jimmy
03:09 jimmy seen berney
03:09 purl I haven't seen 'berney', jimmy
03:09 cotto you mean barney?
03:09 jimmy seen barney
03:09 purl barney was last seen on #parrot 15 days, 12 hours, 8 minutes and 25 seconds ago, saying: will be back tomorrow evening, if there is internet access in Berlin  [Nov 29 15:00:47 2008]
03:09 jimmy yes
03:09 jimmy he was online,but not here.
03:20 cotto I'm certain he was online more recently that 15 days ago.
03:20 cotto s/that/than/
03:37 jimmy who updated pipp with bernhard account?
03:42 jimmy cotto: languages\pipp\t\php\string.t has double copyright and $id$, i don't think it is need to create a patch .
03:43 jimmy s/is//g
03:44 * jimmy is having lunch
04:01 cotto thanks, jimmy
04:02 elmex_ joined #parrot
04:04 dalek r33909 | cotto++ | trunk:
04:04 dalek : [pipp] remove redundant unnecessary copyright and svn id lines
04:04 dalek : jimmy++ for noticing
04:04 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33909
04:51 gaurav joined #parrot
04:53 gaurav joined #parrot
05:10 dalek r33910 | cotto++ | trunk:
05:10 dalek : [embed] fix a memory leak (CID #181)
05:10 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33910
05:10 gaurav joined #parrot
05:12 Tene Okay, I have auto-resume working.
05:12 Tene Let's see if tests pass.
05:39 vaidyagi joined #parrot
05:40 vaidyagi joined #parrot
05:40 petdance joined #parrot
05:47 dalek r33911 | pmichaud++ | trunk:
05:47 dalek : Forgot to commit MANIFEST earlier.
05:47 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33911
05:51 tewk joined #parrot
05:53 adam_ joined #parrot
05:56 bacek joined #parrot
06:10 Tene So... count_eh is broken and untested
06:11 Tene I'm done for the night, but someone needs to please file a ticket about this.
06:11 dalek r33912 | tene++ | trunk:
06:11 dalek : [exceptions]
06:11 dalek : * Nonfatal exceptions are automatically resumed, after printing the message
06:11 dalek : * Default exception severity is EXCEPT_ERROR
06:11 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33912
06:14 dalek r33913 | tene++ | trunk:
06:14 dalek : [rakudo]: Update warn() to use an exception.
06:14 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33913
06:14 Tene Should the default handler there for nonfatals also print a backtrace?
06:21 nopaste "tene" at 166.70.38.237 pasted "trivial count_eh demo; segfault" (3 lines) at http://nopaste.snit.ch/14951
06:21 Tene someone please make ticket of that, if there isn't one already
06:21 dalek r33914 | tene++ | trunk:
06:21 dalek : [NEWS]: updates
06:21 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33914
06:28 dalek r33915 | pmichaud++ | trunk:
06:28 dalek : [rakudo]: spectest-progress.csv update: 249 files, 5101 passing, 0 failing
06:28 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33915
06:30 Tene 100 new passing tests today?
06:30 Tene Nice.
06:31 pmichaud yeah.  Hard to know who to credit, so I credit everyone on this one.  Moritz++ did a lot of test (un)fudging that probably helped.
06:32 pmichaud we may see a similar gain in the next day or so, as I get the hash methods corrected.
06:32 gaurav joined #parrot
06:34 Tene Would it be a problem if i worked on loops refactor on my own tomorrow?  Or too close to release?
06:34 pmichaud do it in a branch.
06:34 Tene Clever.
06:34 pmichaud never hurts to work in a branch.  :-)
06:38 gaurav joined #parrot
06:50 Alias OK, building a Vanilla 6 Alpha I think is maybe good enough to let you guys play with, prior to real release tomorrow
06:51 Theory joined #parrot
06:51 Tene Alias: I hope to get control exceptions working tomorrow.  It would be great to get that into whatever release you make public.
06:53 pmichaud if I get hashes cleaned up I can help on the exceptions.
06:55 Tene pmichaud: thoughts on .HLL_map Exception to Perl6Exception, so I can add methods like .resume() and such, to start exploring things for working out the spec?
06:55 Tene Wait, I think we're already doing that.
06:55 Tene Nevermind.
06:55 pmichaud yes, we are.
06:55 Tene In that case, adding some unspecced potentially-useful methods to Perl6Exception?
06:55 pmichaud that's fine.
06:56 pmichaud watch the mailing list and #perl6 closely, though, because control exception handling (especially warn and return) is a current "hot topic"
06:56 Tene Good.
06:59 Theory joined #parrot
07:00 Alias Tene: I only build official tarballs, no exceptions
07:00 Alias So if something is in the tarball release, it goes in
07:00 Alias Attempting to patch of other modifying anything is just asking for maintenance nightmare
07:01 Tene Alias: I was confused by "prior to real release"
07:01 Alias Oh
07:01 Alias I'm building Vanilla Perl 6 0.8.1 Alpha 2
07:01 Tene "Tomorrow" for me meant "Monday", not "Tuesday" when the release happens.
07:01 pmichaud Alias comes from the future.  :-)
07:01 Alias Here in the future, it's already after business hours Monday evening
07:01 Tene Yes, I understand now, no problems.
07:02 Tene If I can think of a good example for the autoresume and warn-uses-exception changes i just made, I'll blog about it.
07:02 pmichaud blog it anyway, if only to say "we now follow the spec"
07:03 Alias One thing I'd very very much like is to be able to run make test
07:03 Alias I know chromatic calls then "stable" releases, but having to force tests really does say otherwise
07:03 pmichaud on the spectests?
07:03 Alias No, at all
07:03 Alias make test fails on Win32 for 0.8.1
07:03 pmichaud Parrot make test fails?
07:03 Alias yes
07:03 pmichaud ah.
07:04 pmichaud I don't know that we can do anything about it for the Dec release, but we can make that a critical milestone for Jan release (which I think it should be)
07:04 Alias The entire point of the test suite and skips is to decide which tests are so critical you would rather end users NOT use the package at all
07:05 Alias If test fails on Win32, it says you'd rather people not use it on Win32
07:07 * Alias ponders websites and what not
07:08 Alias I'd very much like to do something similar to strawberryperl.com
07:08 Alias Helps people to think of things as "real"
07:08 Alias BTw, have you SEEN perl6.org, it's utterly horrid
07:08 davidfetter joined #parrot
07:09 pmichaud I don't know who owns that domain.
07:09 pmichaud I have rakudoperl.com/.org/.net if we want to do something with those, though.
07:09 Alias http://perl6.com/
07:09 Alias Even worse
07:10 Tene ... ew.
07:10 Alias No insult intended, but I'm trying to avoid using "Rakudo" anywhere
07:10 pmichaud none taken.
07:10 Tene perl6.com -- Jason Duke
07:11 Alias Although I shall of course say it's using Rakudo
07:11 Tene perl6.org -- Daniel Wright
07:11 Alias Both sites look like they mean well
07:11 Alias No actual squatting
07:15 Tene http://blogs.gurulabs.com/stephen​/2008/12/nonfatal-exceptions.html
07:17 Alias pmichaud: So could you do platform-specific skips for the failing Win32 tests you expect to fail, if indeed they aren't important
07:18 pmichaud Alias: at the moment I have no way of knowing which tests those are
07:18 Alias I can give you a list :)
07:18 pmichaud also, I don't muck around too much in Parrot's test suite -- I tend to focus on Rakudo and PGE
07:18 pmichaud but yes, istr there's a way to do platform-specific skips.
07:19 Tene Alias: if you could post a list to parrot-dev@lists.parrot.org that would be great
07:19 pmichaud or even as a ticket  (trac.parrot.org)
07:20 * Tene sleep, dream of exceptions
07:29 Alias OK, off to read my book while build finishes, back in an hour or so
07:32 UltraDM joined #parrot
07:37 adu joined #parrot
07:46 iblechbot joined #parrot
07:48 Alias Build seems to have worked
07:49 * Alias looks for somewhere to upload to
08:02 Alias OK, uploaded
08:03 ffwonko joined #parrot
08:03 Alias http://strawberryperl.com/downloa​d/vanilla-perl6-0.8.1-alpha-2.exe
08:04 Alias In about 15 minutes, the zip version should also be available
08:04 Alias http://strawberryperl.com/downloa​d/vanilla-perl6-0.8.1-alpha-2.zip
08:04 cotto but does it run on Linux?
08:04 cotto (sorry)
08:04 Alias That's ok
08:04 moritz Alias++ # packaging effort
08:04 Alias Next time, s/Linux/Gentoo/ (funnier)
08:05 cotto http://strawberryperl.com/downloa​d/vanilla-perl6-0.8.1-alpha-2.rpm
08:05 Alias Wouldn't be too hard
08:05 Alias People keep asking me about a Mac version
08:06 Alias I keep asking for a 17 inch powermac
08:06 Alias or mac book pro, or something shiny
08:06 Alias It's not really that hard, it's mostly just work...
08:06 Alias Anyways
08:06 Alias zip lands in 10 minutes ish
08:07 Alias Let me know how it goes, I'll be afk for a few hours or so readiung
08:12 adu joined #parrot
08:15 Hadi joined #parrot
08:30 alvar joined #parrot
08:33 ffwonko joined #parrot
08:45 Theory joined #parrot
08:56 ffwonko joined #parrot
08:59 ffwonko joined #parrot
09:06 alvar joined #parrot
09:32 allison joined #parrot
09:37 cotto joined #parrot
09:38 kj joined #parrot
09:51 Hadi left #parrot
09:52 donaldh joined #parrot
10:31 tomyan joined #parrot
10:31 jimmy vanilla-perl6-0.8.1-alpha-2.exe? ah, perl 6 alpha came out?
10:48 donaldh jimmy: more likely to be an alpha package of 0.8.1 for your platform
10:50 jimmy donaldh: thanks.
11:03 Theory joined #parrot
11:28 gaz joined #parrot
11:36 Alias Anyone got a chance to look at the Vanilla 6 alpha?
11:44 Lorn joined #parrot
12:10 gmansi joined #parrot
13:31 ffwonko joined #parrot
13:45 tetragon joined #parrot
14:02 iblechbot joined #parrot
14:09 gryphon joined #parrot
14:16 Wknight8111 joined #parrot
14:17 ruoso joined #parrot
14:28 PacoLinux joined #parrot
14:30 ffwonko joined #parrot
14:37 pmichaud Alias: after installing vanilla perl6, I get:
14:37 pmichaud (eek, can't copy/paste in windows)
14:38 kj pmichaud: can't copy/paste in command shell?
14:39 pmichaud having to learn how to do it :-)
14:39 nopaste "pmichaud" at 72.181.176.220 pasted "results of "perl6 -e 'say "hello world"' in vanilla perl6" (4 lines) at http://nopaste.snit.ch/14954
14:39 kj open properties window of the shell window
14:39 pmichaud I got it :-)
14:39 kj then in 'options', klik 'quickedit mode'
14:39 kj ok
14:39 ff-wonko joined #parrot
14:45 AndyA joined #parrot
15:07 dalek r33916 | pmichaud++ | trunk:
15:07 dalek : [rakudo]:  Refactor Array assignment.
15:07 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33916
15:12 ruoso joined #parrot
15:24 jhorwitz joined #parrot
15:35 ruoso joined #parrot
15:53 dalek r33917 | pmichaud++ | trunk:
15:53 dalek : [rakudo]:  Refactor Scalar coercions.
15:53 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33917
16:07 gaurav joined #parrot
16:08 gaurav joined #parrot
16:11 dalek r33918 | Whiteknight++ | trunk:
16:11 dalek : [Book] Add more information about some of the pmclass flags "abstract", "has_ro", "is_ro", and "is_shared". Many of these are almost completely undocumented elsewhere, so my descriptions here might not be completely accurate.
16:11 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33918
16:12 DietCoke joined #parrot
16:12 * Coke now has power at home, work, and everywhere in between.
16:15 pmichaud yay
16:15 pmichaud wb, Coke
16:24 hercynium joined #parrot
16:27 gmansi joined #parrot
16:28 * Coke sees tcl is not on https://trac.parrot.org/parrot/wiki/Languages
16:29 * Coke suggests we kill languages/LANGUAGES.STATUS and just use the wiki.
16:29 PerlJam well, it's a wiki ...
16:29 Coke PerlJam: someone went to great trouble to copy an existing set of documentation and leave a note to keep them in sync.
16:29 Coke PerlJam: and tcl was already in the other file.
16:29 * Coke is fixing it.
16:29 PerlJam weird
16:34 particle Alias: your claim of windows tests failing is baseless without a bug report
16:34 * Coke hates wiki table editing.
16:34 gaurav PerlJam, Coke: oh, hey, that was me.
16:35 PerlJam gaurav: Do you have some aversion to tcl?  Did Coke tick you off?  :-)
16:35 gaurav heh, i *like* wiki table editing, which is partially why I did it. I'll add tcl to the list if you like.
16:36 gaurav PerlJam: No, I *like* Coke. I hate Pepsi. :)
16:36 gaurav I must have missed it while running all the tests
16:36 PerlJam gaurav: probably because tcl is no longer in the parrot repo?
16:37 gaurav joined #parrot
16:37 Coke gaurav: I just added it.
16:38 Coke Thanks.
16:41 * Coke opens a ticket on my suggestion. Feel free to vote on it if you care.
16:41 Coke gaurav++ since that won't show up here.
16:42 Coke guarav, have you been converting the old wiki, was that you?
16:42 ruoso joined #parrot
16:42 gaurav No, just the Languages thing (and a couple of edits to NewParrotDevelopers...) so far.
16:45 dalek r33919 | infinoid++ | trunk:
16:45 dalek : [src] Fix the following warnings:
16:45 dalek : src/events.c:808: warning: null argument where non-null required (argument 1)
16:45 dalek : src/events.c:912: warning: null argument where non-null required (argument 1)
16:45 dalek : src/exceptions.c:215: warning: ignoring return value of 'die_from_exception', declared with attribute warn_unused_result
16:45 dalek : src/exceptions.c:295: warning: ignoring return value of 'die_from_exception', declared with attribute warn_unused_result
16:45 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33919
16:45 dalek r33920 | infinoid++ | trunk:
16:45 dalek : [cage] Fix the following codetest failures:
16:45 dalek : * t/codingstd/c_cppcomments.t
16:45 dalek : * t/codingstd/c_operator.t
16:45 dalek : * t/codingstd/trailing_space.t
16:45 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33920
16:49 Coke is fperrad in here?
16:50 dalek r33921 | infinoid++ | trunk:
16:50 dalek : [cage] t/codingstd/pdd_format.t fails on pdd23 with some line length
16:50 dalek : errors, split the lines.
16:50 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33921
16:52 dalek r33922 | pmichaud++ | trunk:
16:52 dalek : [rakudo]: Refactor hash construction and assignment.
16:52 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33922
17:03 chromatic joined #parrot
17:03 barney joined #parrot
17:05 tomyan left #parrot
17:34 baest_ joined #parrot
17:53 khisanth_ joined #parrot
18:03 PerlJam joined #parrot
18:16 rdice joined #parrot
18:18 alvar joined #parrot
18:21 ffwonko joined #parrot
18:31 Debolaz joined #parrot
18:37 Hadi joined #parrot
18:44 Theory joined #parrot
18:48 jan joined #parrot
18:50 Tene pmichaud: ideas about including except_*.pasm yet?
18:50 Tene I'm wanting to fix continue() in when{} blocks
18:54 dalek r33923 | coke++ | trunk:
18:54 dalek : Allow this function to be used in class.pmc and avoid linkage issues.
18:54 dalek : Patch Courtesy Andrew Dougherty <doughera@lafayette.edu>
18:54 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33923
18:56 chromatic Ah, good.
18:57 Wknight8111 :( my first book proposal just got rejected
18:57 Tene pmichaud: possibly in PAST::Compiler.to_post?
19:02 rurban joined #parrot
19:05 Infinoid perlcritic doesn't like it when files start with "#!/home/tewk/srcs/parrot/perl6".  for a rakudo .pm file, is the hashbang even necessary?
19:07 Tene No.
19:07 Infinoid that said, perlcritic should probably just whitelist perl 6 files, as it can't understand them
19:07 particle yep
19:08 particle and within two months, we won't have to worry about most languages/ anyway
19:09 gmansi joined #parrot
19:10 dalek r33924 | chromatic++ | trunk:
19:10 dalek : [t] Fixed an uninitialized value warning in test for Parrot::Test.
19:10 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33924
19:17 cognominal joined #parrot
19:18 particle some of these tests are maddening
19:18 Andy joined #parrot
19:19 dalek r33925 | particle++ | trunk:
19:19 dalek : [t] account for different name for null device on windows
19:19 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33925
19:22 dalek r33926 | Whiteknight++ | trunk:
19:22 dalek : [Core] remove unreachable code from src/multidispatch.c:Parrot_m​md_find_multi_from_sig_obj()
19:22 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33926
19:25 Infinoid should all perl 6 libraries contain a "use v6;"?  if so, I can make perlcritic skip them based on that
19:25 particle no
19:25 particle there was discussion on that topic, but i can't remember the latest trend
19:25 Infinoid sadly, ext/SQLite3/DBDI.pm doesn't have much I can use to filter on
19:26 particle put it in an exclusions list, we already have one
19:26 chromatic Coke, ping.
19:26 particle might be nice to move it to ext/perl6/... or something
19:27 particle because we could easily also put python, perl 5, ruby modules in ext
19:27 * Infinoid doesn't see an exclusions list in perlcritic.t
19:27 Infinoid (excluding languages, yes, but not ext files)
19:28 Infinoid its too bad, really.  the "grep { $_->read !~ m/use v6;/ }" I added was working great for a while
19:29 particle ok, actually, v6 is still specced as preferred
19:29 dalek r33927 | particle++ | trunk:
19:29 dalek : [t] use path-independent shell quotes from parrot's config
19:29 particle so change the files to include "use v6;"
19:29 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33927
19:29 dalek r33928 | chromatic++ | trunk:
19:29 dalek : [src] Delayed a string copy until absolutely necessary during an inplace chop.
19:29 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33928
19:29 Infinoid so I can go around adding it to untagged perl6 files and make perlcritic ignore them?
19:30 particle yes
19:30 chromatic particle, r33925 could use File::Spec->devnull()
19:30 Coke chromatic: pong.
19:30 particle chromatic: you need to read the test...
19:31 chromatic Okay, that test is crack.
19:31 particle i don't have the energy to fight, just want to get it passing
19:32 nopaste "chromatic" at 69.64.234.10 pasted "Coke: Tcl Sub NameSpace PMC Patch" (30 lines) at http://nopaste.snit.ch/14957
19:33 tewk chromatic: t/pmc/key.t has a todo test that should pass w/ fix
19:33 cout left #parrot
19:34 chromatic tewk, it's not passing for me here.  Hmm.
19:34 chromatic particle, I see what you mean.
19:35 tewk chromatic: my todo job was ugly, you have to uncomment the test
19:36 tewk How should I todo a segfault style failure, skip?
19:36 Coke chromatic: with that, #10 is resolved. (if I run it against the parrot_io branch in partcl.)
19:36 dalek r33929 | particle++ | trunk:
19:36 dalek : [t] un-todo bonus MSWin32 dynoplibs test
19:36 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33929
19:37 chromatic Excellent.  Let me see how Perl 6 and Pheme fare.
19:37 Coke chromatic++
19:37 Coke tewk++
19:40 Coke chromatic: perl6 seems ok
19:40 Coke (OS X/intel_)
19:40 alvar joined #parrot
19:41 Coke chromatic: pheme test failure, but it looks like an odd syntax failure.
19:41 chromatic It should be a warning, not a failure.
19:41 chromatic I need to add operator precedence; I just haven't had time.
19:41 Coke ah, yes. just verbose.
19:41 Coke otherwise ok.
19:41 Tene chromatic: did you see my warn() commit?
19:42 dalek r33930 | infinoid++ | trunk:
19:42 dalek : [cage] More codingstd fixes.
19:42 dalek : * t/codingstd/perlcritic.t (ignore perl6 files based on containing "use v6;")
19:42 dalek : * t/codingstd/trailing_space.t (again)
19:42 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33930
19:42 dalek r33931 | chromatic++ | trunk:
19:42 dalek : [PMC] Worked around TT #10 temporarily by skipping deep Sub-like introspection
19:42 dalek : when using a PIR-level subclass of Sub.  The right solution is making Sub's
19:42 dalek : attributes into PMC ATTRs, but that's too big to do before tomorrow's release.
19:42 dalek : Reclaimed a related test for the Key PMC.
19:42 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33931
19:42 chromatic Tene, very nice.
19:42 Wknight8111 I'm getting test failures in t/compilers/pct/complete_workflow.t in trunk after an update and a realclean
19:42 Wknight8111 can anybody else duplicate that?
19:43 tewk chromatic++ test reclamation
19:43 particle Wknight8111: win32? yes.
19:43 Tene Did anyone get alias's report of win32 test fialures?
19:44 Wknight8111 I didn't
19:45 particle nope
19:45 Coke Are they the normal test failures?
19:45 particle Wknight8111: i've fixed 4 win32 test failures today already, working on pct now
19:45 Coke *cringe*
19:46 * Coke can try strawberry perl on his laptop tonight if there's time.
19:46 particle Wknight8111: if you have tuits for t/pmc/eval, go for it
19:46 Wknight8111 i probably don't till later tonight
19:51 particle r32058 | particle | 2008-10-20 16:17:09 -0700 (Mon, 20 Oct 2008) | 3 lines
19:51 particle [t] create utility module for parrot tests
19:51 particle ~ add function for createing cross-platform tempfiles
19:51 particle ~ convert tests to use it
19:51 particle i have absolutely no memory of this commit.
19:51 Tene complete_workflow.t passes fine for me
19:51 chromatic October Surprise
19:51 Wknight8111 this test has it in for me
19:52 Wknight8111 this is almost the same test errors that were appearing in the calling_convention branch
19:54 chromatic Is it possible that your computer is haunted?
19:54 Tene Have you tried submerging it in holy water?
19:54 Infinoid hauntedness is a little too meta for my tastes
19:55 Wknight8111 it's two separate computers though, and three operating systems
19:55 chromatic Okay, then you're haunted.
19:56 Wknight8111 the test *knows* where I am and what I am working on, then then fails right there
19:56 dalek r33932 | Whiteknight++ | trunk:
19:56 dalek : [Core] mark an undocumented function as being subject to RT#48260
19:56 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33932
20:03 Coke I dislike RT#48260 markings.
20:03 * Infinoid will test -rHEAD on strawberry perl, if "svn st" ever gets around to returning
20:03 Coke we -have- a test. Let's just use it.
20:04 mberends joined #parrot
20:06 elmex joined #parrot
20:07 chromatic I kind of like them, because I really dislike the 'RT#' all scrunched together, so whenever I tidy a file, I'm always tempted to write documentation to get rid of them altogether.
20:08 dalek r33933 | coke++ | trunk:
20:08 dalek : Apply remainder of patch from Andy Dougherty that I thought was already applied.
20:08 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33933
20:14 riffraff joined #parrot
20:15 Wknight8111 I like the RT#48260 markers too. They're easy to spot when you're working on a file, and easy to find all of them with an ack
20:17 Coke Wknight8111: (easy to find) so would be the codingstd test.
20:17 Coke chromatic: I can just add the failing codingstd test back into 'make test'. =-)
20:18 chromatic Or we could review diffs and patches and reject the latter without docs.
20:18 Coke ah, crap. 'make test' with latest tcl & latest parrot is failing some tests, and hanging in t/cmd_expr.t
20:19 riffraff hi
20:19 Wknight8111 Coke there are, by my count, about 300 documented functions in the repo
20:19 Wknight8111 adding a test for it would basically guarantee that test would fail forever
20:19 Wknight8111 300 "un"documented functions
20:20 PerlJam Wknight8111: your patch is rejected because you didn't provide documentation  :-)
20:20 Coke Wknight8111: there IS a test for it.
20:20 Wknight8111 I know there is a test, and that test will always fail
20:20 Coke the RT notes make the TEST PASS.
20:20 Coke which is insanity.
20:20 chromatic Welcome, riffraff.
20:20 Wknight8111 or, almost always fail, in the foreseeable future
20:20 Tene riffraff: welcome
20:28 * Coke sees tools/docs/func_boil* and wonders a) if that really helps, and b) why it doesn't work as adverted on the tin.
20:29 Coke ah. it does work as adverted on the tin. It doesn't modify the files in question, though, which makes it <useful.
20:30 peters joined #parrot
20:30 peters Hi folks =)
20:31 nopaste "peters" at 83.183.129.106 pasted "URI (RFC 3986) grammar with tests" (312 lines) at http://nopaste.snit.ch/14958
20:31 dalek r33934 | particle++ | trunk:
20:31 dalek : [t] slight refactoring for readability
20:31 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33934
20:31 dalek r33935 | particle++ | trunk:
20:31 dalek : [t] whitespace cleanup
20:31 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33935
20:31 peters could use a few more tests, but it's a good start =)
20:32 chromatic Very nice, peters.
20:32 Coke peters++
20:32 peters =)
20:33 particle peters: you should see if rakudo supports MAIN() well enough for you to use it for the tests
20:34 particle that way, if you run the file directly, it'll run the tests, but if you 'use' the file, it won't.
20:34 * pmichaud has refactoring MAIN on his todo list.
20:35 peters particle: cool, I'll try that later
20:35 Tene pmichaud: objections to refactoring the controltypes hash initialization in pastcompiler to just store the ints instead of strings of the macros, if we're not going to get prelude with includes this release?
20:36 Coke chromatic: ../../parrot tcl.pbc t/cmd_break.t segfaults with latest of everything now.
20:36 pmichaud Tene: no object.  If I manage to decide how I want to do prelude before the release we can (easily) switch back.
20:37 pmichaud *objection
20:37 Coke (not saying it's your patch.)
20:37 particle pmichaud: any particular reason pct generates 8 extra spaces before every sub line? it's just making generated pir files larger and imo harder to read
20:37 pmichaud ...pct does this?
20:37 * particle makes sure.
20:38 particle it happens in complete_workflow.t
20:38 nopaste "particle" at 76.121.106.245 pasted "a temp pir file generated from complete_workflow.t" (187 lines) at http://nopaste.snit.ch/14959
20:38 pmichaud oh.  That's PGE.
20:39 particle yep, so it is
20:39 pmichaud it's because of the way codegen works in PGE.  We can remove the spaces but it makes the code *much* harder to read.
20:39 Tene chromatic: if you're looking for an easy-to-reproduce segfault, try using count_eh
20:40 pmichaud because then one can't easily see what is PGE code and what is PGE generating.
20:40 particle ok, maybe a filter that removes X spaces when writing to a file or something
20:40 pmichaud at this point, readability of PGE source is more important than readability of what PGE generates.
20:40 chromatic Tene, is there a test file somewhere?
20:40 particle true enough, it's a minor nit
20:41 nopaste "tene" at 97.117.74.5 pasted "count_eh test for chromatic" (3 lines) at http://nopaste.snit.ch/14960
20:41 Tene chromatic: coutn_eh is completely untested
20:41 pmichaud ultimately I hope that PGE will generate POST, and then it'll be a non-issue :-)
20:41 particle :)
20:42 pmichaud I don't know that count_eh is going to be all that useful, fwiw.
20:43 pmichaud at least, not as it exists now.
20:50 Coke tt #48 is also (ime) very reproducable.
20:52 Coke what is the difference (on trac) between 'mac' and 'darwin' ?
20:54 hercynium joined #parrot
20:54 Tene peters: do all of those tests pass in your URI grammar?
20:54 mugwump maybe mac means MacOS 9-
20:55 Coke mugwump: which we don't support or plan to any time soon. I suspect it was just a copy over from RT. Deleting.
20:55 Coke ... or it's built into trac and I can't.
20:55 peters Tene: all 71 tests pass using r33858
20:56 mugwump mmm.  and how's the Z80 port going?
20:56 Tene peters: very good work. :)
20:56 Tene mugwump: hasn't been touched in ages
20:56 mugwump blast, was really looking forward to getting it running on my OS-X box
20:57 peters Tene: thanks, when grammar inheritance has been fixed I'll start on more URI schemes
20:57 nopaste "Infinoid" at 96.238.213.50 pasted "currentish failures on strawberry perl (takes a while on this box)" (29 lines) at http://nopaste.snit.ch/14961
20:58 Infinoid all tests pass on linux/x86-64
20:58 Wknight8111 w00t
20:59 particle Infinoid: there's a TT for ubuntu 64 failing configure. can you have a look?
21:01 Infinoid I can try.
21:01 particle Infinoid++
21:02 Infinoid segfault at VTABLE_get_string_keyed(INTERP, agg, key) looks awfully familiar though
21:02 Coke GAH
21:02 Coke our new ticket handling sucks.
21:03 Coke Can we get someone to cleanup parrot-tickets so we can't keep mis-posting there?
21:03 particle oh, i can't post there? uggh.
21:03 * Infinoid drives home through the snow storm
21:05 particle coke: what version of trac are we running?
21:06 cotto 0.10.4, if the footer is telling the truth
21:07 particle that's pretty old
21:07 particle i believe it supports email2trac plugin, though...
21:08 particle https://subtrac.sara.nl/oss/email2trac/wiki/
21:11 Coke particle: TT#49
21:13 Coke ->
21:17 cotto Anyone have a clue what compilers/imcc/parser_util.c:multi_keyed does?  It's only used in one place, which has been #if'd out for most of Parrot's life.
21:18 cotto It looks like it could be removed without any pain.
21:22 chromatic No idea.
21:24 cotto should I remove it?
21:24 particle wednesday
21:24 chromatic No rush to remove it before the release.
21:25 cotto will do
21:35 davidfetter joined #parrot
21:41 riffraff in PGE how do I make a rule so that it matches a dot-terminated string?
21:42 riffraff I was thinking rule foo { <[^.]*> {*} } would have worked but it seems it doesn0t
21:42 pmichaud you mean a string ending in dot?   / '.' $ /   will work.
21:42 moritz the * should be outside the <[...]>
21:42 pmichaud oh, you want to match everything but a dot?
21:42 moritz token foo { <-[.]>* \. $ }
21:42 riffraff everything but a dot yes
21:43 pmichaud or if you want to match everything up to a dot (but not necessarily require the dot to be at the end of the string):   / (<-[.]>*) '.' /
21:43 riffraff I nitially tried putting the * outside of the brackets but the parse three did not contain the data
21:43 pmichaud there aren't any captures present, so "no parse tree".
21:44 riffraff ah right :)
21:48 riffraff ah, it seems i got the negated character class wrong
21:48 pmichaud ...anyone besides allison have admin privileges to the parrot-tickets list?  If so, we can disable all postings except from trac.
21:49 riffraff I must reread the apocalypse I'm afraid
21:49 pmichaud the apocalypse may be out of date in key areas.  The synopsis is more up to date.
21:50 moritz aye, S05++
21:50 moritz pmichaud++ # slowly killing the $str ~~ Grammar::rule meme
21:51 pmichaud mostly I'm closing tickets.  :-)
21:52 moritz pmichaud: I'm a bit concerned about RT #61308...
21:52 pmichaud how so?
21:52 moritz pmichaud: the test that i provided in the ticket doesn't work
21:53 pmichaud your test had 'xab' instead of 'xay'
21:53 moritz ah
21:53 moritz I was too tired to find that yesterday night
21:53 moritz no concern then ;-)
21:53 pmichaud Martin found it :-)
21:53 pmichaud I didn't see it either.
21:54 moritz ah, right
22:00 Whiteknight joined #parrot
22:01 dalek r33936 | pmichaud++ | trunk:
22:01 dalek : [pct]:  Fix -e option to execute '' or 0 argument if given (RT #57018).
22:01 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33936
22:05 jonathan pmichaud++ # fixing/closing lots of tickets
22:11 peters pmichaud++ # RT #61338
22:14 Whiteknight particle still here?
22:14 ff-wonko joined #parrot
22:15 Limbic_Region joined #parrot
22:16 particle ayep
22:17 Whiteknight particle, what was that test you said I should look at tonight?
22:17 particle i get a win32 failure with t\pmc\eval.t:10
22:18 Whiteknight ...great, need to reboot into windows
22:18 Whiteknight failure have a trac ticket?
22:18 particle iunno
22:19 Whiteknight brb
22:39 andrew_ joined #parrot
22:44 MariachiElf joined #parrot
22:56 dalek r33937 | particle++ | trunk:
22:56 dalek : [t] increased use of parrot test library functions, portability, test description accuracy, documentation. also fixes six failing tests on MSWin32
22:56 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33937
22:57 chromatic Was that the culprit?
22:58 particle yeah
22:58 particle a few bugs in that code. specifically, needed to s/nqp/pm/ in one place
22:59 particle secondly, windows wasn't looking in the proper dir for the input file to the compiler, so needed to use absolute path
22:59 particle the rest makes ugly code a little less ugly
23:00 * particle makes test
23:04 Zaba_ joined #parrot
23:32 particle andrew_: as of r33938, all tests pass with msvc/win32
23:32 dalek r33938 | particle++ | trunk:
23:32 dalek : [t] cannot 'rm' an open file on windows
23:32 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33938
23:33 andrew_ thanks particle++. I was working on it then got sidetracked
23:33 andrew_ wife started baking cookies, and I'm a sucker for cookie dough
23:34 particle ok, maybe now i'll turn my attention to making gifts for my nieces and nephews
23:34 jonathan make gift
23:34 wknight8111 what programming language are you using to make their gifts? :)
23:35 particle looks like we'll get parrot.org updated to trac 0.11.2 this week
23:35 wknight8111 what version are we at now?
23:35 * jonathan looks forward to playing with it, to see if it's worth upgrading the installation he looks after
23:35 particle mostly i'm using acrobat, illustrator, and photoshop for the gifts :)
23:35 particle we're at 0.10.1 iirc
23:36 particle 0.10.4 i mean
23:41 ruoso joined #parrot
23:51 wknight8111 I like what I've seen of padre so far, but wxWidgets can be such a pain in the ass to install
23:51 wknight8111 of course, it's no harder then Gtk2, installing that on Windows was almost impossible
23:58 GeJ What are the "Tier 1" languages in Parrot? Rakudo, tcl, pipp? others?
23:58 jonathan LOLCODE
23:59 chromatic Pheme
23:59 chromatic (because it's so fast to test)
23:59 dalek r33939 | pmichaud++ | trunk:
23:59 dalek : [rakudo]:  Remove obsolete 'hash' method from List.
23:59 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=33939
23:59 dalek r33940 | pmichaud++ | trunk:
23:59 dalek : [rakudo]:  Make sure Protoobjects don't participate in list flattening.

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

Parrot | source cross referenced