Camelia, the Perl 6 bug

IRC log for #parrot, 2008-07-14

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:01 ruoso joined #parrot
00:03 dalek r29412 | chromatic++ | trunk:
00:03 dalek : [PMC] Made NameSpace PMC use the hash function interface, rather than poking
00:03 dalek : around in its hash internals directlyi.
00:03 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29412
00:10 AndyA joined #parrot
00:16 dalek r29413 | chromatic++ | trunk:
00:16 dalek : [PMC] Cleaned up more hash usage in AddrRegistry and NameSpace PMCs I missed in
00:16 dalek : r29411 and r29412.
00:16 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29413
00:23 contingencyplan joined #parrot
00:41 dalek r29414 | jkeenan++ | parallel:
00:41 dalek : Consolidate multiple test files per configuration step into a single file.
00:41 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29414
00:50 dalek r29415 | jkeenan++ | parallel:
00:50 dalek : Consolidate multiple test files per configuration step into a single file.
00:50 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29415
00:56 dalek r29416 | Whiteknight++ | gsoc_pdd09:
00:56 dalek : [gsoc_pdd09] Stop sweeping const_PMC pools for now (causes weird error). Added a note about this.
00:56 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29416
00:57 dalek r29417 | fperrad++ | trunk:
00:57 dalek : [CORE] Make Emacs coda read-only in MANIFEST.SKIP
00:57 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29417
01:06 dalek r29418 | fperrad++ | trunk:
01:06 dalek : [emacs]
01:06 dalek : - add :lexid directive
01:06 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29418
01:34 Zaba joined #parrot
01:54 Zaba joined #parrot
02:07 Zaba_ joined #parrot
02:08 davidfetter joined #parrot
02:29 Andy joined #parrot
03:20 Zaba joined #parrot
03:20 clunker3 joined #parrot
04:35 Zaba_ joined #parrot
05:18 Psyche^ joined #parrot
05:19 Ademan joined #parrot
05:19 Ademan_ joined #parrot
05:44 TiMBuS joined #parrot
05:50 cotto_home anyone mind de-warnocking my reply to #47109?
06:14 uniejo joined #parrot
06:16 uniejo joined #parrot
06:16 TiMBuS|Away joined #parrot
06:18 uniejo joined #parrot
06:22 masak joined #parrot
06:24 Zaba joined #parrot
06:41 uniejo joined #parrot
06:45 Zaba_ joined #parrot
06:56 masak joined #parrot
07:07 Zaba joined #parrot
07:08 iblechbot joined #parrot
07:29 Theory joined #parrot
07:37 Tango_ joined #parrot
07:38 Tango_ are u all in the designing of the parrot virtual machine ??
07:38 Tango_ btw hello all
07:38 barney joined #parrot
07:39 moritz hello
07:39 purl hey, moritz.
07:39 masak hi Tango_
07:39 moritz Tango_: 99% of the time we're just implementing, not designing ;)
07:39 Tango_ :)
07:39 moritz and I just work on rakudo (perl6 on parrot), not parrot itself
07:39 Tango_ ok
07:40 moritz is there anything specific you want to know (or comment on)?
07:41 Tango_ i want to know is this project anyhow maintained by the jamaica Group also ?
07:42 moritz never heard of it
07:42 * Zaba neither
07:42 Tango_ http://intranet.cs.man.ac.uk/apt/p​rojects/jamaica/os.php#challenges
07:43 Tango_ here i seen your name first
07:43 Zaba I thought jamaica was a country
07:43 Tango_ how the "type safe making" for the JVMs(parrot) are going on ?
07:44 moritz Tango_: you can search https://svn.perl.org/parrot/trunk/CREDITS for members for the jamaica group
07:44 Tango_ ok..watching
07:44 moritz what do you mean by "type safe makiing"?
07:45 moritz it's designed to work well with dynamically typed languages
07:46 Tango_ http://en.wikipedia.org/wiki/Lambda_calculus
07:47 Tango_ is it a JIT or a simple bytecode interpreter ??
07:48 moritz parrot has multiple run cores, some of them do JIT
07:50 Tango_ ok nice
07:50 Zaba run cores?..
07:51 * Zaba is far from understanding parrot's architecture
07:52 moritz Zaba: a runcore is the part that actually executes the byte code
07:52 Zaba why have multiple of them?
07:53 moritz for example there is a GC debug runcore, a normal interpreter like, a computed goto one, some JIT runcores for different platforms
07:53 Zaba hmm.
07:53 moritz they have different advantages each
07:53 moritz for example the fastest ones are usually not very cross-platform
07:53 Zaba because they use platform-specific things?
07:54 moritz or CPU specific things
07:57 * barney read http://blogs.open.collab.net/sv​n/2008/07/subversion-merg.html
08:01 barney Zaba: Strange, I've never heard about http://parakeet.wiki.sourceforge.net/, the JVM executing Parrot bytecode
08:03 Zaba neither have I
08:11 dalek r29419 | bernhard++ | trunk:
08:11 dalek : [docs] Mention the other Parakeet in LANGUAGES_STATUS.pod.
08:11 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29419
08:12 * barney is glad that he didn't rename 'Plumhead' to 'Parakeet'
08:12 mire joined #parrot
08:57 Zaba_ joined #parrot
09:00 jan joined #parrot
09:03 dalek r29420 | bernhard++ | trunk:
09:03 dalek : [docs] Update CREDITS and fiddle with release_manager_guide.pod.
09:03 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29420
09:06 dalek r29421 | jkeenan++ | parallel:
09:06 dalek : Consolidate multiple test files per configuration step into a single file.
09:06 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29421
09:23 Whiteknight joined #parrot
09:41 GeJ joined #parrot
09:52 particle1 joined #parrot
09:57 * barney discovered      svn diff -c
09:57 tewk_ joined #parrot
09:57 moritz hey cool, I didn't know that either
09:58 moritz maybe I should start reading the docs of all these tools I use
10:00 barney moritz: I'm looking for sombody to add the Rakudo news to NEWS
10:00 moritz barney: I can draft a first draft, but in the end I think that pmichaud or jonathan should do it properly
10:00 moritz when is release? tomorrow?
10:01 barney Yes.
10:01 Infinoid joined #parrot
10:01 moritz ok, I'll put something together in case nobody of the others finds the tuits
10:01 barney moritz++
10:03 moritz when was last release? 2008-06-10 ?
10:10 moritz oh no, 2008-06-17
10:19 dalek r29422 | moritz++ | trunk:
10:19 dalek : [NEWS] first shot at news for Rakudo
10:19 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29422
10:24 dalek r29423 | bernhard++ | trunk:
10:24 dalek : [docs] Update NEWS with changes up to r29000.
10:24 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29423
10:34 dalek r29424 | moritz++ | trunk:
10:34 dalek : [NEWS] more rakudo news, including number of newly passing tests
10:34 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29424
10:49 dalek r29425 | bernhard++ | trunk:
10:49 dalek : [docs] more NEWS up to HEAD
10:49 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29425
10:49 barney moritz++: Danke
10:50 moritz Bitte ;-)
10:54 confound joined #parrot
10:55 confound joined #parrot
11:00 Zaba joined #parrot
11:12 Zaba_ joined #parrot
11:37 Whiteknight joined #parrot
11:42 dalek r29426 | bernhard++ | trunk:
11:42 dalek : [codingstd] Add the languages eclectus, hq9plus, m4 and pipp to Perl::Critic testing
11:42 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29426
11:55 Zaba joined #parrot
12:04 iblechbot joined #parrot
12:10 verve joined #parrot
12:49 bacek joined #parrot
12:49 bacek g'night
12:50 bacek moritz: around?
13:00 gryphon_ joined #parrot
13:08 moritz bacek: yes
13:08 moritz (but slightly distracted)
13:11 dalek r29427 | fperrad++ | trunk:
13:11 dalek : [RELEASE]
13:11 dalek : update LANGUAGES_STATUS.pod from languages-smoke results
13:11 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29427
13:15 bacek moritz: can you take a look at S02-builtin/type.t?
13:16 moritz yes
13:16 bacek There is fudged test 'try{} doesn't keep test from failing'
13:16 moritz aye
13:17 bacek actually it fail to parse.
13:18 bacek rakudo treats 'is(try{}, foo, bar)' as 'is((try{}, foo, bar))'
13:19 bacek so, removing 'try{}' will make this test passing
13:19 moritz sounds good
13:20 nopaste "bacek" at 122.110.92.165 pasted "Proposed patch for type.t" (23 lines) at http://nopaste.snit.ch/13576
13:21 moritz go right ahead and commit it
13:22 moritz or do you want me to steal your karma?
13:22 moritz it seems that try is parsed a function, not an unary op
13:24 bacek moritz: yes, it is.
13:26 pmichaud try is a statement_control
13:26 pmichaud (and is parsed that way)
13:30 bacek pmichaud: probably no. It's statement_prefix in src/parser/grammar.pg
13:30 pmichaud my apologies, you're correct -- statement_prefix
13:30 pmichaud but it's still parsed as a statement_prefix :-)
13:30 pmichaud from STD.pm:  token statement_prefix:try     { <sym> <.ws> <statement> {*} }
13:31 pmichaud and {$foo = 42},    42,    'Int is an integer'   is (so far) a valid statement.
13:32 moritz so the tests are wrong?
13:33 pmichaud I remarked about this at some point in the past -- but yes, as written it seems to me that is() only gets one argument (the result of the try)
13:33 moritz bacek: one more reason to fix the tests
13:34 dalek r29428 | fperrad++ | trunk:
13:34 dalek : [RELEASE]
13:34 dalek : update LANGUAGES_STATUS.pod from manual tests
13:34 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29428
13:35 pmichaud if using try in an argument list, it probably needs to be   (try {...})
13:35 bacek moritz: there is probably more places where such bogus syntax used...
13:35 moritz bacek: aye, I just ack'ed for them
13:35 moritz moritz@trudi:~/pugs/t/spec>ack 'is.*try'|wc -l
13:35 moritz 85
13:35 moritz so many potential errors
13:35 bacek pmichaud: yes. It works with (). But try is redundant in this cases.
13:36 pmichaud perhaps the test should be using lives_ok
13:36 pmichaud with a separate test to make sure the variable changed to 42
13:36 moritz aye, that's even better
13:37 pmichaud S29-trig/trig.t is using try{} and probably doesn't need to be.
13:38 moritz I'm removing that while you're writing ;)
13:38 dalek r29429 | coke++ | trunk:
13:38 dalek : [codingstd] exec counts as an explicit way to leave a sub.
13:38 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29429
13:38 dalek r29430 | bernhard++ | trunk:
13:38 dalek : [m4] backticks need to be quoted in Perl 6 regexes.
13:39 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29430
13:55 ron joined #parrot
13:55 ron left #parrot
13:56 donaldh joined #parrot
14:07 dalek r29431 | coke++ | trunk:
14:07 dalek : [codingstd] - Allow user to specify a directory to test. Useful for language
14:07 dalek : maintainers, or just as a way to focus cleanup efforts.
14:07 dalek : - Move some variables into the scope in which they are used.
14:07 dalek : - Rename a confusing variable.
14:07 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29431
14:10 Andy joined #parrot
14:20 jhorwitz joined #parrot
14:37 Zaba joined #parrot
14:39 dalek r29432 | coke++ | trunk:
14:39 dalek : [codingstd] [tcl] perltidy our perl code. This allows us to pass all the
14:39 dalek : perlcritic tests, even the --theme=extra ones.
14:39 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29432
14:54 Zaba_ joined #parrot
15:12 skv_ joined #parrot
15:13 cotto_work particle1, ping
15:29 dalek r29433 | coke++ | trunk:
15:29 dalek : [codingstd] Unbreak regexp for stripping out files not to test.
15:29 dalek : Pretty would be nice, but working wins.
15:29 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29433
15:32 Zaba joined #parrot
15:38 Zaba_ joined #parrot
15:43 dalek r29434 | coke++ | trunk:
15:43 dalek : [codingstd] Add another perl critic policy we should aspire to.
15:43 dalek : While in here, update [tcl] to respect this policy.
15:43 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29434
15:50 dalek r29435 | coke++ | trunk:
15:50 dalek : [distro] 'OpLib' existed and was empty in the repo. Now that it's
15:50 dalek : not in the repo, we need to ignore it explicitly.
15:50 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29435
16:02 rurban joined #parrot
16:10 dalek r29436 | coke++ | trunk:
16:10 dalek : [codingstd] require version in a way that avoids a complaint from perlcritic
16:10 dalek : and is theoretically more readable anyway. (Reini Urban++)
16:10 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29436
16:22 rurban hell. I open irc and just read my name :)
16:29 Infinoid rurban++
16:47 dalek r29437 | coke++ | type_ids:
16:47 dalek : svn merge -r28032:28100 from trunk
16:47 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29437
17:05 Infinoid joined #parrot
17:16 dalek r29438 | coke++ | type_ids:
17:16 dalek : merge Skipped this file for some reason
17:16 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29438
17:19 Limbic_Region joined #parrot
17:21 Theory joined #parrot
17:23 purl joined #parrot
17:23 NotFound svn diff
17:23 purl svn diff is probably clean
17:24 NotFound Wrong window
17:25 Zaba joined #parrot
17:27 dalek r29439 | coke++ | type_ids:
17:27 dalek : svn merge -r28101:28200 from trunk
17:27 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29439
17:46 dalek r29440 | bernhard++ | trunk:
17:46 dalek : [ook] fix 'cd languages/ook && make test'
17:46 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29440
18:00 * Limbic_Region yawns
18:04 Zaba joined #parrot
18:18 cjfields joined #parrot
18:35 dalek r29441 | coke++ | type_ids:
18:35 dalek : merge -r28201:29436 from trunk
18:35 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29441
18:35 Zaba joined #parrot
18:37 dalek r29442 | coke++ | trunk:
18:37 dalek : [deprecated] avoid type ids
18:37 dalek : [codingstd]  avoid useless interpolation
18:37 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29442
18:40 sjansen joined #parrot
18:51 gmansi joined #parrot
18:53 cotto_work particle1, ping
18:55 vhold joined #parrot
18:59 cotto-work joined #parrot
19:08 dalek r29443 | pmichaud++ | trunk:
19:08 dalek : [rakudo]: spectest-progress.csv update:  95 test files, 1691 passing tests
19:08 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29443
19:08 Zaba joined #parrot
19:17 dalek r29444 | coke++ | type_ids:
19:17 dalek : merge -r 29437:29442 from trunk
19:17 dalek : (that particular deprecation in trunk is fatal here.)
19:17 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29444
19:18 dalek r29445 | chromatic++ | trunk:
19:18 dalek : [docs] Updated NEWS.
19:18 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29445
19:26 gryphon__ joined #parrot
19:29 dalek r29446 | pmichaud++ | trunk:
19:29 dalek : [docs]:  More NEWS updates
19:29 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29446
19:46 Zaba_ joined #parrot
20:39 rurban_ joined #parrot
20:47 wknight8111 joined #parrot
20:54 bacek joined #parrot
21:01 Zaba joined #parrot
21:26 Zaba_ joined #parrot
21:35 Limbic_Region joined #parrot
21:59 bacek joined #parrot
22:06 dalek r29447 | Whiteknight++ | gsoc_pdd09:
22:06 dalek : [gsoc_pdd09] update to trunk r29446
22:06 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=29447
22:08 teknomunk joined #parrot
22:10 NotFound parrot_config gives the revision number from .parrot_current_rev, which is not updated by configure or make, and is not deleted by make clean.
22:24 Whiteknight .parrot_current_rev--
22:33 rurban NotFound: sure. that's a feature
22:33 rurban svn up will update it
22:33 NotFound rurban: negative
22:34 NotFound $ cat .parrot_current_rev
22:34 NotFound 29431
22:34 NotFound $ svn up
22:34 NotFound En la revisión 29447.
22:34 shamu joined #parrot
22:34 rurban hmm...
22:35 NotFound The file is created by lib/Parrot/Revision.pm , but only if not exists.
22:35 rurban it might be some svn server hook which updates it
22:36 NotFound It's supposed to be a cache.
22:36 Limbic_Region pmichaud ping
22:37 rurban I see. It's the live svn --xml info
22:37 rurban Indeed a bug.
22:37 * Limbic_Region tosses rurban a cockroach
22:38 rurban There's method missing to force update the cache, which should be added to the makefile
22:39 NotFound Was reported yesterday for an smoke testers, the recommendation to him was to do a make realclean in his smoke script, but the problem has other implications.
22:40 NotFound Not very importante, but having wrong revision info in bug reports hardly helps anything.
22:40 rurban NotFound: write it to the bugtracker
22:41 rurban tommorrows release should not go with wrong one.
22:41 NotFound I'll do it tomorrow, now I'm going to bed :O
22:41 rurban me too :)
22:49 jonathan pmichaud: About?
22:50 kid51 joined #parrot
22:52 shamu can parrot be made reliably with make -j n>1 ?
22:53 Infinoid yes, I regularly use -j8
22:53 shamu platform?
22:53 Infinoid linux/x86-64
22:53 shamu can it reliably be tested in parallel?
22:54 Infinoid I don't think that's been implemented yet, but it is a much desired feature
22:54 cotto_work iirc, kid51 is working on parallel testing
22:54 shamu ah ok
22:54 shamu thanks a bunch
22:57 bacek joined #parrot
23:00 kid51 cotto_work:  No that branch's name is something of a misnomer.  It refers to a (since abandoned) idea to create an object whose growth parallels that of the Parrot::Configure object.
23:00 kid51 It's not the sort of "testing in parallel" which, e.g., they're talking about on perl.qa.
23:01 kid51 The branch is now simply the place where I'm working on http://rt.perl.org/rt3/Tic​ket/Display.html?id=56928
23:09 cotto_work my mistake
23:28 bacek joined #parrot
23:36 kid51 joined #parrot
23:53 particle .parrot_current_rev should always reflect the last time you ran 'perl Configure.pl' successfully
23:55 sandra_f joined #parrot
23:58 Whiteknight joined #parrot

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

Parrot | source cross referenced