Camelia, the Perl 6 bug

IRC log for #parrot, 2009-12-23

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:02 GeJ Good morning everyone.
00:10 cotto_work hi GeJ
00:12 GeJ heya cotto.
00:18 GeJ I have an error in the nqp tests, namely : src/gc/api.c:245: failed assertion 'PObj_is_PMC_TEST(obj)'
00:19 GeJ in t/20-list_access.t
00:19 GeJ is this a known issue or should I TT it?
00:20 dalek parrot-plumage: f8b2fdc | japhb++ | t/02-load-all.t:
00:20 dalek parrot-plumage: [TESTS] Add 02-load-all.t to test loading required modules
00:20 dalek parrot-plumage: review: http://gitorious.org/parrot-plumage/parrot-plumag​e/commit/f8b2fdc14bf57be931b85cf5181c3326d1d758fe
00:20 dalek parrot-plumage: ce84479 | japhb++ | t/0 (2 files):
00:20 dalek parrot-plumage: [TESTS] Stop using PGE-based like() in tests
00:20 dalek parrot-plumage: review: http://gitorious.org/parrot-plumage/parrot-plumag​e/commit/ce8447934616e9889f23ad822fe9fc3b2730894e
00:22 dduncan left #parrot
00:25 cotto_work GeJ, which tests?  nqp-rx from git passes all tests using whatever Parrot is build with --gen-parrot.
00:32 bacek joined #parrot
00:42 GeJ cotto_work: make test in compilers/nqp
00:42 GeJ which seems to be run everytime when I do `make test` in parrot's root directory.
00:46 cotto_work looks fine to me
00:46 cotto_work but you do have a special computer
00:47 dalek parrot-plumage: af730f6 | japhb++ |  (6 files):
00:47 dalek parrot-plumage: [LIB] Move last bits of Plumage-specific functionality from Util to Plumage::Util
00:47 dalek parrot-plumage: review: http://gitorious.org/parrot-plumage/parrot-plumag​e/commit/af730f6293caef2040a2ce9d324438478c85877b
01:02 GeJ cotto_work: that, I have :)
01:24 tetragon joined #parrot
01:43 bacek_at_work joined #parrot
02:07 dalek TT #1389 created by japhb++: [LHF] Add tests for Plumage's Util.nqp
02:09 dalek parrot-plumage: bdd74d1 | japhb++ | docs/hacking/add-tests.pod:
02:09 dalek parrot-plumage: [DOCS] Add new add-tests.pod hacking doc
02:09 dalek parrot-plumage: review: http://gitorious.org/parrot-plumage/parrot-plumag​e/commit/bdd74d126753ef9367c41f1ad1435674c4ecc72a
02:25 dalek parrot-plumage: 2ba369e | japhb++ | TASKS:
02:25 dalek parrot-plumage: [META] Starting to use Trac now; add section at top of TASKS accordingly, with links to canned ticket searches
02:25 dalek parrot-plumage: review: http://gitorious.org/parrot-plumage/parrot-plumag​e/commit/2ba369eb2346a47c55068c139df33155eb6c5f4c
02:26 Casan joined #parrot
02:47 JimmyZ joined #parrot
02:50 Coke ugh. if we're doing explicit deps, then every .o needs to explicitly depend on it's .c, yes?
02:50 Coke *its
02:54 cotto yup
02:54 cotto it'll be fun
02:54 purl Join the army he said, it'll be fun he said
02:55 cotto At least the ugliness can be generated.
02:56 dalek TT #1390 created by carlin++: Configure doesn't check for a 'curses' lib in the readline test
02:58 Coke cotto; at the moment, I'm just trying to verify. generation comes later.
02:59 Coke also, .h files can #include .h files, yes?
03:00 cotto it's valid to do so
03:00 cotto and it happens in parrot.h, most likely among others
03:02 dalek parrot: r43217 | coke++ | trunk (2 files):
03:02 dalek parrot: Cleanup makefile template
03:02 dalek parrot: * eliminate IMCC_H_FILES; ALL_H_FILES;
03:02 dalek parrot: * be as specific as possible (eschew GENERAL_H_FILES)
03:02 dalek parrot: * make checkdepend skip most generated files for now.
03:02 dalek parrot: * don't hide behind GENERAL_H_FILES for tested deps, be specific.
03:02 dalek parrot: review: http://trac.parrot.org/parrot/changeset/43217/
03:17 JimmyZ good morning
03:17 cotto hi JimmyZ
03:21 JimmyZ hello
03:27 Coke cotto: I think we can get away without the explicit .o : .c, as we're using a suffix rule to build all the .o files. whee.
03:28 Coke there. .h files covered.
03:32 cotto I'd like to know how the imcc.y.flag rule works when that file never appears to actually exist
03:34 Coke I don't think that rule is invoked.
03:34 Coke nothing seems to depend onit.
03:34 dalek parrot: r43218 | coke++ | trunk (2 files):
03:34 dalek parrot: .h can also other .h's, and so on ad infinitum.
03:34 dalek parrot: (Check for this; fix all instances found by the new improved check depend.)
03:34 dalek parrot: review: http://trac.parrot.org/parrot/changeset/43218/
03:36 nopaste joined #parrot
03:37 cotto Got it.  That file is only created if the .y or .l gets touched
03:46 TonyC joined #parrot
03:57 TiMBuS joined #parrot
04:06 nopaste joined #parrot
04:08 TonyC joined #parrot
04:12 japhb pmichaud, ping
04:14 szabgabx joined #parrot
04:18 dalek TT #1391 created by japhb++: [TODO] Generate per-parrot-build GUID/UUID, stored in e.g. interpinfo
04:24 * cotto wonders if yak shaving isn't a misspelling of yacc shaving.
04:27 japhb No, but determining the full answer to that question might be construed as such ....
04:39 GeJ !seen kid51
04:39 GeJ seen kid51
04:39 purl kid51 was last seen on #parrot 1 days, 38 minutes and 38 seconds ago, saying: must sleep  [Dec 22 03:59:14 2009]
04:40 dalek parrot: r43219 | cotto++ | trunk/config/gen/makefiles/pirc.in:
04:40 dalek parrot: [pirc] initial stab at making it easy to regenerate pirc's lexer and parser
04:40 dalek parrot: This is possibly wrong and should be used only with care for now.
04:40 dalek parrot: review: http://trac.parrot.org/parrot/changeset/43219/
04:42 GeJ message kid51 Sorry for the late answer. I just posted a comment on TT #1277. Coke++ patch seems to solve it in a better way than I suggested. Ticket should probably be closed. Thanks.
04:42 purl Message for kid51 stored.
04:45 cotto That patch could use review from someone who knows what he's doing.
04:59 cotto Would it be a good idea to move pirc's makefile template into root.in?
05:21 Coke cotto: my goal is to have a single top level makefile eventually. if we're modifying pirc's now, yes, move it over. you'll find most of the deps already in it, anyway.
05:22 Coke GeJ: ticket closed, thanks.
05:23 cotto It's moving time.
05:23 Coke cotto: (of course, that top level makefile would prolly be broken down in to smaller chunks, but they'd be includes, not standalone mfiles.
05:24 dalek TT #1277 closed by coke++: [PATCH] cleanup nqp-rx leftovers on `make realclean`
05:29 dukeleto 'ello
05:29 cotto hi
05:29 purl niihau, cotto.
05:29 cotto hi, not purl
05:35 cotto Hmm.  It'd be nice if kj were around.
05:41 japhb Tene++ again for => syntax in NQP-rx
05:42 dalek parrot-plumage: 73be904 | japhb++ |  (2 files):
05:42 dalek parrot-plumage: [plumage,DOCS] Use hash() and => to initialize global hashes instead of going through JSON -- easier to read, removes need for various hacks, and reduces startup time considerably; remove fixup_commands() and fix add-command doc to match
05:42 dalek parrot-plumage: review: http://gitorious.org/parrot-plumage/parrot-plumag​e/commit/73be904c13c6b489e0fb9a35828281780f222702
05:51 japhb pmichaud, ping
06:00 bacek joined #parrot
06:09 dalek parrot-plumage: 472e7fe | japhb++ | TODO:
06:09 dalek parrot-plumage: [META] Fix URI in TODO
06:09 dalek parrot-plumage: review: http://gitorious.org/parrot-plumage/parrot-plumag​e/commit/472e7fe3cef01c3644e154f926d32133a9396f61
06:09 dalek parrot-plumage: df3b137 | japhb++ | src/plumage.nqp:
06:09 dalek parrot-plumage: [plumage] Convert straggling root_new call from Q:PIR to pir:: syntax
06:09 dalek parrot-plumage: review: http://gitorious.org/parrot-plumage/parrot-plumag​e/commit/df3b137f40884a6cfe6113d61e3aad161c6054aa
06:10 * Coke wonders why timtowtdi.org seems so slow.
06:18 dalek parrot: r43220 | coke++ | trunk (2 files):
06:18 dalek parrot: when checking deps, check all _DIR vars. Fixup deps on newly caught files.
06:18 dalek parrot: review: http://trac.parrot.org/parrot/changeset/43220/
06:30 dalek partcl-nqp: 0cf7a57 | duff++ | src/Partcl/commands/main.pm:
06:30 dalek partcl-nqp: Add pwd command
06:30 dalek partcl-nqp: review: http://github.com/partcl/partcl-nqp/commit​/0cf7a57a217ed31815cd08073076ec29b4aa3546
06:32 PerlJam Coke: ping
06:33 pdcawley joined #parrot
06:34 dalek parrot: r43221 | coke++ | trunk/tools/dev/checkdepend.pl:
06:34 dalek parrot: Complain if the rule for a given file depends on things that are NOT included.
06:34 dalek parrot: This test now fails. Some care should be taken to insure that the dependencies are not implicit, and if they are, this file should be skipped for now.
06:34 dalek parrot: review: http://trac.parrot.org/parrot/changeset/43221/
06:36 Tene joined #parrot
06:39 PerlJam Coke: I notice from the docs that the globbing syntax for "string match" is slightly different from that used for "glob".  ("string match" doesn't do tilde-expansion nor does it support alternation, while "glob" does).   Would it be fine with you if I broke up Glob into StringGlob and FileGlob?
06:42 Coke PerlJam: +1
06:42 purl 1
06:43 Coke -> abed
06:43 PerlJam g'night
06:50 dalek parrot: r43222 | mikehh++ | trunk/tools/dev/checkdepend.pl:
06:50 dalek parrot: fix codetest failure - trailing whitespace
06:50 dalek parrot: review: http://trac.parrot.org/parrot/changeset/43222/
07:06 dalek parrot: r43223 | mikehh++ | trunk/tools/dev/checkdepend.pl:
07:06 dalek parrot: fix codetest failure - Cuddled else
07:06 dalek parrot: review: http://trac.parrot.org/parrot/changeset/43223/
07:06 dalek parrot: r43224 | cotto++ | trunk/docs/pdds/pdd13_bytecode.pod:
07:07 dalek parrot: [docs] more pdd13 changes
07:07 dalek parrot: review: http://trac.parrot.org/parrot/changeset/43224/
07:07 dalek parrot: r43225 | mikehh++ | trunk/docs/pdds/pdd13_bytecode.pod:
07:07 dalek parrot: fix codetest failure - max pod linelength 78
07:07 dalek parrot: review: http://trac.parrot.org/parrot/changeset/43225/
07:11 japhb msg pmichaud Util.nqp has entirely replaced Glue.pir, and the last bits of Plumage-specific functionality have been stripped from Util.nqp; TT #1389 now tracks improved tests for it.  Can you take a look and let me know what other changes you would like to see before some or all of it can move into the NQP-rx repo?  http://gitorious.org/parrot-plumage/parr​ot-plumage/blobs/master/src/lib/Util.nqp
07:11 purl Message for pmichaud stored.
07:15 dalek partcl-nqp: a112f2a | duff++ |  (10 files):
07:15 dalek partcl-nqp: rename Glob to StringGlob
07:15 dalek partcl-nqp: review: http://github.com/partcl/partcl-nqp/commit​/a112f2aaeb8884aa843852553f3755fd823ebfe7
07:15 dalek partcl-nqp: 32cde5f | duff++ |  (5 files):
07:15 dalek partcl-nqp: create FileGlob as subclass of StringGlob, but adds alternation
07:15 dalek partcl-nqp: review: http://github.com/partcl/partcl-nqp/commit​/32cde5f719cc647d323e46fc9b3b56bf9c100b8a
07:29 mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#31264), fulltest) at r43225 - Ubuntu 9.10 amd64 (gcc)
07:34 integral joined #parrot
07:51 contingencyplan joined #parrot
07:51 Tene joined #parrot
07:55 dalek parrot: r43226 | cotto++ | trunk/src/pmc/parrotinterpreter.pmc:
07:55 dalek parrot: [pmc] avoid function-wide one-character variable name
07:55 dalek parrot: review: http://trac.parrot.org/parrot/changeset/43226/
08:19 iblechbot joined #parrot
08:36 bacek joined #parrot
08:42 TiMBuS joined #parrot
09:11 bacek joined #parrot
09:19 elgatito joined #parrot
09:28 elgatito left #parrot
09:33 baest joined #parrot
09:36 TiMBuS joined #parrot
09:41 fperrad joined #parrot
10:10 estrabd joined #parrot
10:24 fperrad_ joined #parrot
11:12 bacek_ joined #parrot
11:18 pdcawley joined #parrot
11:22 estrabd_ joined #parrot
11:27 estrabd joined #parrot
11:51 bacek_ joined #parrot
12:04 estrabd joined #parrot
12:11 mikehh joined #parrot
12:18 bluescreen joined #parrot
13:05 payload joined #parrot
13:42 iblechbot joined #parrot
14:10 iblechbot joined #parrot
14:13 patspam joined #parrot
14:14 Coke mikehh++ # cleaning up my codemess.
14:17 iblechbot joined #parrot
14:17 Coke PerlJam++ # that's shorter than I expected.
14:28 Coke looks like everything that tools/dev/checkdepend.pl is carping about is an easy fix if someone wants some karma.
14:35 bacek joined #parrot
14:42 dalek parrot: r43227 | coke++ | trunk/tools/dev/checkdepend.pl:
14:42 dalek parrot: Format diagnostic slightly better.
14:42 dalek parrot: review: http://trac.parrot.org/parrot/changeset/43227/
14:43 JimmyZ joined #parrot
14:44 JimmyZ good evening
14:44 purl Ah, evening. The tumultuous mind tarries and contemplates, reveling in the silence afforded by the diurnal proletariat. Good evening, indeed.
14:44 Coke greedy coke wonders what else he can convince perljam to hack on!
14:45 szabgab joined #parrot
14:55 fperrad joined #parrot
15:04 bubaflub joined #parrot
15:04 Psyche^ joined #parrot
15:05 * JimmyZ wonders when he would get the commit bit.
15:10 Coke there needs to be a CLA, two votes, and a mentor.
15:11 Coke I know we have the CLA.
15:20 JimmyZ last ps losts.
15:22 JimmyZ And I put a question there, but no answers
15:43 jsut_ joined #parrot
15:49 theory joined #parrot
16:03 mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#31285), fulltest) at r43227 - Ubuntu 9.10 amd64 (g++ with --optimize)
16:17 Coke mikehh: tools/dev/checkdepend.pl could use some love. (I added a new condition last night, it's showing some makefile brokenness.)
16:17 iblechbot joined #parrot
16:18 Coke mikehh: (I just figured I'd tell you since if it were a real distro test, you'd have fixed it already. =-)
16:20 mikehh Coke: ok I will have a look
16:20 Coke muahahaha
16:40 plobsing joined #parrot
17:14 dukeleto gut moroning
17:15 cotto_work hello?
17:15 dukeleto i totally missed parrotsketch yesterday, didn't I? duh.
17:16 cotto_work There wasn't much of a #ps anyway.
17:16 cotto_work allison had to miss it and I guess chromatic is afk for the holidays
17:22 tetragon joined #parrot
17:36 davidfetter joined #parrot
17:38 tetragon joined #parrot
18:22 patspam1 joined #parrot
18:26 plobsing how should I put the contents of a Buffer* into a STRING*?
18:34 Coke I would have to dig through string/api.c to find out.
18:34 Coke mikehh: any luck?
18:35 cotto_work plobsing, there doesn't appear to be a function to do that.
18:37 plobsing should there be?
18:37 purl should there be is probably this something I should be wanting to do?
18:38 cotto_work I don't know.  If you want to do it, that indicates that there may also be a future need for it.
18:38 plobsing purl, forget should there be
18:38 purl plobsing: I forgot should there be
18:39 plobsing push @yaks_to_shave, $string_system
18:41 santsman joined #parrot
18:42 santsman hola
18:42 purl salut, santsman.
18:42 santsman bones purl!
18:42 santsman soc nou per a qui
18:42 santsman de que va aquest xat?
18:44 santsman hola???
18:44 purl well, hola is niihau
18:44 santsman hiihau
18:44 cotto_work hi
18:44 purl what's up, cotto_work.
18:44 santsman hi cotto
18:44 santsman where are you from?
18:45 cotto_work Washington
18:45 purl Washington is a real pit of a city , is this true ? or yes
18:45 cotto_work (the state)
18:45 mikehh coke: was away for a bit - looking now
18:45 cotto_work as opposed to DC
18:45 santsman there ara someone from Barcelona?
18:45 cotto_work istr that NotFound is from Spain
18:46 santsman ???
18:46 Coke we do have a map for people to record this sort of thing.
18:47 Coke ... except that frappr is now dead?
18:47 santsman where is the map?
18:47 purl the map is probably just a table
18:48 cotto_work So that's what Frappr is.  I was wondering why I got an email from when when I didn't recall signing up.
18:48 Coke http://www.frappr.com/parrotcoders/map
18:49 * Coke will download the KML sos we don't lose it.
18:50 Coke coke now has a .kmz file with that data.
18:50 Coke no one on the map there from spain, but frappr iz old.
18:55 PacoLinux NotFound is from Santiago de Compostela, Spain
19:12 japhb joined #parrot
19:43 dukeleto that map is pretty cool
19:43 dukeleto Coke: i am adding myself now, then your kmz will be out of date :)
19:43 dukeleto Coke: maybe we should tell people on list to add themselves, and then we can export at the last minute
19:46 Coke dukeleto: I'd rather get it migrated and then ask people to update it there.
19:46 cotto_work migrated to where though?
19:47 dukeleto Coke: sure. it is hanging on me adding myself, anyway
19:48 Coke cotto_work: facebook?
19:48 purl somebody said facebook was a  social networking site used at most large universities or about 30% less annoying than myspace or too high friction to be really useful or http://i174.photobucket.com/album​s/w89/pickledpizza/facebook2.jpg or a high school reunion where you can weed out the losers.  Venue still sucks, though or failbook or assbook or a broken, glorified telephone
19:50 Coke http://apps.facebook.com/mapping/ might look promising if it didn't have broken images everywhere.
19:53 Coke cotto_work: ah. or we could just host the KML file and point people at google earth.
19:57 Coke There you go. someone write a perl6 webapp for this.
20:11 dukeleto you can render KML with javascript
20:11 dukeleto i can hack on that
20:12 * dukeleto puts another thing on a stack of "stuff to do" that looks like the leaning tower of Pisa
20:14 dukeleto fuck facebook in the face
20:14 cotto_work I'm not sure if I feel quite as strongly, but Facebook doesn't strike me as the right tool for a developer map.
20:17 dukeleto cotto_work: Facebook is a ghetto. and I agree with you.
20:23 dalek parrot: r43228 | whiteknight++ | trunk/src/ops/core.ops:
20:23 dalek parrot: fix some paths on include files after the pmc headers were moved to include/pmc/*. Not a functional change, but I have found some compilers that can get a little bit confused on that.
20:23 dalek parrot: review: http://trac.parrot.org/parrot/changeset/43228/
20:30 lucian joined #parrot
20:30 Coke dukeleto: please keep that stuff off channel.
20:31 Coke (the over the top cursing)
20:32 dukeleto Coke: 10-4, boss. there is #perl for that
20:32 pdcawley joined #parrot
20:32 Coke exactly. =-)
20:32 Coke pdcawley: PIERS!
20:32 Coke come, write some perl6 for me.
20:32 Tene What's up, Coke?
20:33 dukeleto Tene: coke wants something in perl6 to render kml
20:33 Coke dukeleto, tene: well, that was just a throwaway suggestion. I really want people to hack on partcl-nqp for me.
20:33 Coke PerlJam++
20:38 Zak joined #parrot
20:38 bluescreen joined #parrot
20:44 patspam joined #parrot
20:48 joeri joined #parrot
20:55 dukeleto Coke: so we need a partcl-nqp kml renderer?
20:56 Coke dukeleto: those two projects are orthogonal. =-)
21:00 dukeleto Coke: no, you just need to write a web framework for TCL :)
21:04 Tene dukeleto: no, I just need to go through and update all of the languages to the new HLL interop protocol (again), and then TCL can use Rakudo's Web.pm
21:05 Coke dukeleto: tcl has an httpd in the stdlib.
21:05 Coke also: http://tcl.apache.org/
21:06 dukeleto Coke: cool
21:11 dalek parrot: r43229 | plobsing++ | branches/pmc_freeze_cleanup (3 files):
21:11 dalek parrot: make visit_info.pos an array index and stop using Parrot_str_new_init on 'Buffer *'s
21:11 dalek parrot: This fixes segfaults in examples/benchmarks/freeze.pasm and when DISABLE_GC_DEBUG isn't defined. It
21:11 dalek parrot: is dirty, clumsy, and breaks encapsulation on strings, but it works.
21:11 dalek parrot: visit_info.pos:
21:11 dalek parrot: Pointers into buffers will fail after a gc string compaction, only Buffer.bufstart is guaranteed to
21:11 dalek parrot: work.
21:11 dalek parrot: Parrot_str_new_init vs. Buffer*:
21:11 dalek parrot: Passing Buffer.bufstart to a routine that might allocate memory (and therefor might compact strings)
21:11 dalek parrot: might fail for a similar reason to that mentioned above. Separate routines taking 'Buffer *'s in
21:11 dalek parrot: stead should fix this, but these don't exist yet.
21:11 dalek parrot: review: http://trac.parrot.org/parrot/changeset/43229/
21:14 jjore joined #parrot
21:25 bacek joined #parrot
21:28 lucian joined #parrot
21:36 GeJ Good morning everyone!
21:44 lucian joined #parrot
21:48 Andy joined #parrot
21:52 xenoterracide joined #parrot
22:02 bacek joined #parrot
22:12 mikehh joined #parrot
22:46 Patterner joined #parrot
23:03 bacek joined #parrot
23:09 abqar joined #parrot
23:24 lucian joined #parrot
23:34 theory joined #parrot
23:45 tewk_ ping plobsing
23:46 mikehh joined #parrot
23:49 Coke mikehh: I'm back and off work, looking at fixing up those deps now.
23:55 dalek winxed: r332 | julian.notfound++ | trunk/winxedst1.winxed:
23:55 dalek winxed: fix operator + for int-string in stage 1
23:55 dalek winxed: review: http://code.google.com/p/w​inxed/source/detail?r=332

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

Parrot | source cross referenced