Camelia, the Perl 6 bug

IRC log for #parrot, 2008-10-18

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:10 AndyA joined #parrot
00:33 johbar joined #parrot
00:39 particle joined #parrot
01:14 Bzek joined #parrot
01:45 particle 25/35
01:45 purl 0.714285714285714
01:52 Andy joined #parrot
02:06 Whiteknight joined #parrot
02:07 dalek r32006 | Whiteknight++ | calling_conventions:
02:07 dalek : [calling_conventions] Fix problem with handling floatvals in the function signature. This fix is probably going to be mirrored in trunk soon.
02:07 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32006
02:16 silug joined #parrot
02:21 dalek r32007 | Whiteknight++ | calling_conventions:
02:21 dalek : [calling_conventions] codingstd trailing whitespace fix.
02:21 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32007
02:27 Whiteknight karma Whiteknight
02:27 purl whiteknight has karma of 391
02:28 Whiteknight yay!
02:28 dalek r32008 | Whiteknight++ | trunk:
02:28 dalek : [Core] Fix an omission in src/multidispatch.c:Parrot_bu​ild_sig_object_from_varargs, which might be negatively affecting the passing of FLOATVALS.
02:28 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32008
03:36 johbar_ joined #parrot
03:47 Bzek joined #parrot
05:42 nopaste joined #parrot
05:45 iblechbot joined #parrot
05:46 GeJ Is anyone experiencing a hang in t/stm/basic_mt_4.pir? There seems to be a ticket on RT about this. The ticket is from a cygwin setup but I experience the same symptoms on a FreeBSD setup.
05:47 GeJ RT#59790
05:47 GeJ it looks like I get stuck in the 'retry' loop on line 12.
06:28 omega joined #parrot
07:08 jq joined #parrot
07:35 rurban joined #parrot
07:58 iblechbot joined #parrot
08:01 clunker3 joined #parrot
08:43 rurban_ joined #parrot
10:36 masak joined #parrot
10:47 PacoLinux joined #parrot
10:58 Ademan joined #parrot
11:06 masak Rakudo says "Unable to open file" when an open fails.
11:06 masak There's a comment beside the die() call saying # XXX better message
11:06 masak I suppose there are two things that can be improved:
11:07 masak 1. the message could say why Parrot was unable to open the file
11:07 masak 2. the message could say which file it was
11:07 masak are both these a good idea, or is the second one bad from a security perspective?
11:10 masak rakudo: close $*OUT; say "OH HAI"
11:10 polyglotbot OUTPUT[Cannot put to a closed PIO handle␤current instr.: 'print' pc 14653 (src/gen_builtins.pir:9034)␤called from Sub 'say' pc 14672 (src/gen_builtins.pir:9043)␤called from Sub '_block11' pc 43 (EVAL_12:18)␤called from Sub 'parrot;PCT;HLLCompiler;eval' pc 851 (src/PCT/HLLCompiler.pir:494)␤called
11:10 polyglotbot ..from Sub 'parrot;PCT;HLLCompiler;evalfiles' pc 112...
11:21 cotto joined #parrot
11:32 masak could someone explain to me why line 73 of Junction.pir shouldn't be "goto false"?
11:36 masak rakudo: say 2..3 ~~ 1..4
11:36 polyglotbot OUTPUT[Parrot VM: Can't stat languages/perl6/perl6.pbc, code 2.␤main: Packfile loading failed␤]
11:52 masak rakudo: say 2..3 ~~ 1..4
11:52 polyglotbot OUTPUT[sh: ./parrot: No such file or directory␤]
11:53 masak different stages of parrot not existing...
11:53 masak couldn't a scheme be devised in which parrot compiles in the background?
12:13 tetragon joined #parrot
12:30 masak here goes, r32009. I'd be grateful if someone made sure I got the POD documentation right. I'm especially unsure about method signatures.
12:31 dalek r32009 | masak++ | trunk:
12:31 dalek : [rakudo]
12:31 dalek : * implemented ':e', for testing file existence, in any-str.pir
12:31 dalek : * implemented the 'ACCEPTS' method for pairs, which delegates to e.g. above
12:31 dalek : * renamed 'other' 'topic' in Failure.pir for consistency
12:31 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32009
12:56 Zaba_ joined #parrot
13:00 tetragon joined #parrot
13:02 Infinoid GeJ: I see an intermittant hang there on linux/x86_64.  looks like a pthread mutex race
13:03 Infinoid sometimes it goes through successfully; sometimes it gets stuck in pthread_mutex_lock (from Parrot_stm_wait_ic).
13:10 Infinoid masak: I think both things can be improved about the message.  I'm not really sure why saying the filename would be a security problem... after all, "cat" does the same thing.
13:10 Infinoid cat: /dev/foo: No such file or directory
13:12 Infinoid if the code doesn't want the user to see the message, they can catch the exception and handle it internally, right?
13:16 tetragon joined #parrot
13:20 masak Infinoid: one gets wary about security after being burned once by SQL injection attacks. showing the filename would be just the kind of thing that would give an attacker too much information in a server environment. but maybe it's more a question of not showing $! in a 500 message
13:24 Infinoid yeah, that should probably go to apache.err
13:29 tetragon joined #parrot
13:31 Infinoid I dunno, its debatable.  as long as it goes to stderr and not stdout, I think the apache case is okay.  otherwise the behavior for this kind of thing seems to vary a lot from one implementation to the next
13:32 Infinoid perl's native open (and the one in IO::File) fails silently.  IO::All emits an error message to stderr containing the filename
14:11 tetragon_ joined #parrot
14:20 Whiteknight joined #parrot
14:22 dalek r32010 | Whiteknight++ | trunk:
14:22 dalek : [PDDs] Codingstd fix on PDD20, fixing very long lines.
14:22 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32010
14:26 dalek r32011 | Whiteknight++ | trunk:
14:26 dalek : [IMCC] Activate ".tailcall" syntax in IMCC, as perl RT#58974.
14:26 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32011
14:30 tetragon joined #parrot
14:34 dalek r32012 | Whiteknight++ | trunk:
14:34 dalek : [t] Update t/compilers/imcc/syn/tail.t tests to use ".tailcall" in tailcall context instead of ".return". As per RT#58974.
14:34 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32012
15:00 PacoLinux Generic VMs Key To Future of Coding  -->  http://tech.slashdot.org/t​ech/08/10/18/0352235.shtml
15:29 jan joined #parrot
15:33 tetragon joined #parrot
15:36 iblechbot joined #parrot
15:41 particle whiteknight: please update news with your imcc & other changes
15:45 Whiteknight okay, will do. I'm still working on more IMCC changes, so I'll update it in a bit
15:45 Whiteknight you're the release manager this month?
15:52 particle i am
15:53 particle i'll be doing builds this weekend on various OSes
15:53 particle but now, breakfast &
16:28 Zaba joined #parrot
16:31 tewk_ joined #parrot
16:36 Whiteknight t/pmc/ro.t is failing for me. It was passing for me last night
16:36 Whiteknight There's some kind of problem finding Perl6Scalar, something I wouldn't have touched
16:39 Whiteknight actually, something I definitely haven't touched all morning
16:42 wknight8111 joined #parrot
16:42 dalek r32013 | Whiteknight++ | trunk:
16:42 dalek : [IMCC] Fix a memory leak in IMCC as per RT#42383. t/pmc/ro.t is failing, but it was doing so before this change.
16:42 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32013
16:48 rdice joined #parrot
16:49 Theory joined #parrot
16:54 dalek r32014 | jkeenan++ | trunk:
16:54 dalek : Added an IMPLEMENTATION head1 to quiet a coding standards test.
16:54 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32014
16:54 kid51 joined #parrot
17:02 cotto Whiteknight, does t/pmc/ro.t work when perl6 is built?
17:02 kj joined #parrot
17:03 cotto That test originally depended on Perl6Integer (which no longer exists) and was only passing because the test didn't know which kind of exception to catch.
17:03 Whiteknight i dont know, I'll have to test it
17:03 Whiteknight we really shouldn't have a parrot test that depends on perl 6 being built
17:04 cotto I wasn't sure why Perl6Integers were special, so I translated the test.
17:04 cotto It's probably safe to rip that one out.
17:06 kid51 Failure in t/pmc/ro.t appears to have occurred between r31992 and r32000 (from Smolder reports).
17:07 cotto (instant bisection)++
17:07 kid51 The smolder report at r31992 was from me, and I know that I did not have Perl 6 build at the time that smolder report was run ('cause kid51 doesn't work like that!)
17:08 cotto that sounds like the failures happened around the time I updated t/pmc/ro.t
17:08 Whiteknight cotto, that fixes it
17:08 Whiteknight cotto, you want to rip it out, or should I?
17:09 Whiteknight or, instead of ripping it out entirely, we could move it over into the rakudo test suite
17:09 cotto go ahead.
17:09 cotto I have to take off.
17:09 Whiteknight okay, I'll TODO it or something
17:11 Whiteknight no, I take that back, I don't know how to make a test TODO.
17:12 cotto I'll move it when I get back then.
17:12 cotto afk &
17:15 dalek r32015 | Whiteknight++ | trunk:
17:15 dalek : [t] Ripped out a test in t/pmc/ro.t that relies on Perl6Scalar. This requires that perl6 be built. In the long term, Perl6 won't be in the Parrot repo, so it's not a good idea for our test suite to be dependent on it.
17:15 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32015
17:16 Whiteknight okay, i'm done with all my haxxoring for now.
17:23 mberends joined #parrot
17:57 contingencyplan joined #parrot
18:17 chromatic joined #parrot
18:23 contingencyplan joined #parrot
18:42 mj41_ joined #parrot
19:03 dalek r32016 | chromatic++ | trunk:
19:03 dalek : [lib] Prevented Parrot::Ops2c::Utils from attempting to create lib/Parrot/OpLib
19:03 dalek : directory in its constructor.  A race condition can cause the program to die
19:03 dalek : and the build to fail -- and Parrot::Ops2pm creates this directory as needed
19:03 dalek : only when it's actually necessary anyway (RT #59680, reported by Moritz Lenz).
19:03 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32016
19:05 chromatic Whee, SVN timeouts again.
19:09 chromatic (Yes, I did report them to the Perl.org admins.)
19:59 dalek r32017 | chromatic++ | trunk:
19:59 dalek : [t] Cleaned up file created during testing (RT #46829).  This will need
19:59 dalek : revising during ParrotIO changes, but you can't yet delete files within Parrot,
19:59 dalek : so this will do for now.
19:59 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32017
20:17 dalek r32018 | allison++ | trunk:
20:17 dalek : [pct] Better error reporting of null nodes in the fallback PAST method for
20:17 dalek : unrecognized node types.
20:17 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32018
20:19 dalek r32019 | chromatic++ | trunk:
20:19 dalek : [charset] Tidied file and fixed some documentation typos; no functional changes.
20:19 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32019
20:24 dalek r32020 | chromatic++ | trunk:
20:24 dalek : [src] Removed logic to strip leading __ from old multidispatch names.
20:24 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32020
20:24 dalek r32021 | allison++ | trunk:
20:24 dalek : [pct] Update 'astgrammar' to convert a string class name into an array of the
20:24 dalek : class name, splitting on '::'.
20:24 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32021
20:33 bacek joined #parrot
20:33 bacek morning
20:36 dalek r32022 | chromatic++ | trunk:
20:36 dalek : [IMCC] Removed parser support for pseudo-ops that start with n_.  Actual ops
20:36 dalek : which start with n_ remain unaffected.
20:36 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32022
21:26 Andy joined #parrot
21:30 dalek r32023 | chromatic++ | trunk:
21:30 dalek : [MMD] Removed deprecated Parrot_MMD_method_idx().  This also resolves RT #45973.
21:30 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32023
21:45 dalek r32024 | chromatic++ | trunk:
21:45 dalek : [t] Ensured that temporary files created by File::Temp get unlinked at the end
21:45 dalek : of the test run.
21:45 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32024
22:05 dalek r32025 | chromatic++ | trunk:
22:05 dalek : [config] Added S_ISREG support to generic POSIX and Win32 stat files.  You need
22:05 dalek : to reconfigure Parrot to get it.  This allows the Parrot stat system to test
22:05 dalek : that a file is a regular file (not a directory or special file).  See RT
22:05 dalek : #46153.
22:06 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32025
22:15 dalek r32026 | chromatic++ | trunk:
22:15 dalek : [src] Added check to Parrot_readbc() that the file to read is a regular file.
22:15 dalek : See RT #46153.
22:15 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32026
22:17 chromatic 665 total tickets/623 open.
22:19 chromatic Okay, now 664/622.
22:32 TiMBuS joined #parrot
22:42 japhb Anybody here have rights on Planet Perl 6?
22:42 Tene I think you want Coke
22:42 japhb It looks like the entire O'Reilly blog list just got ascribed to *me*.
22:43 japhb .oO( Gee, I don't *remember* having posted about my "Photoshop Lightroom 2 Adventure", especially since I've never owned Photoshop .... )
22:43 chromatic Which feed?
22:43 purl it has been said that Which feed is that?
22:44 japhb It says my link is: http://www.oreillynet.com/pub/au/2333
22:44 japhb Oh -- the feed URL looks really weird: http://feeds.oreilly.com/orei​lly/digitalmedia/mac?au=2333
22:44 japhb Someone mess with it recently?
22:45 chromatic Yeah, we switched to a tag-based feed system.
22:45 japhb ah ... so can you just nuke me from Planet Perl 6?  I don't ever blog on oreillynet anymore.
22:53 dalek r32027 | allison++ | trunk:
22:53 dalek : [tge] Change TGE to use keyed classnames both internally and in generated code.
22:53 dalek : See RT #59912.
22:53 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32027
23:24 dalek r32028 | allison++ | trunk:
23:24 dalek : [lua] Update Lua to use keyed classnames for TGE.
23:24 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32028
23:33 johbar joined #parrot
23:37 dalek r32029 | kjs++ | trunk:
23:37 dalek : [pirc/new] move some #defines around so it's a bit cleaner + disable some warnings on windows for parser and lexer (generated code, cannot be changed)
23:37 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=32029
23:50 Andy joined #parrot

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

Parrot | source cross referenced