Camelia, the Perl 6 bug

IRC log for #parrot, 2009-07-24

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:00 tetragon joined #parrot
00:05 jrtayloriv joined #parrot
00:08 dalek parrot: r40243 | tene++ | trunk/src/pmc/class.pmc:
00:08 dalek parrot: Minor POD fix for class.pmc.
00:08 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40243/
00:16 dalek parrot: r40244 | whiteknight++ | branches/bsr_jsr_ret (3 files):
00:16 dalek parrot: [bsr_jsr_ret] fix some stuff that I broke because I'm not doing good work like I should be
00:16 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40244/
00:40 Whiteknight I am going to be so happy when bsr/jsr/ret are all completely removed
00:42 dalek parrot: r40245 | whiteknight++ | branches/bsr_jsr_ret/t/com​pilers/pge/p5regex/p5rx.t:
00:42 dalek parrot: [bsr_jsr_ret] another test to fix because I broke it last time I 'fixed' it
00:42 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40245/
00:51 wayland76 kid51++ for ticket sorting :)
00:58 dalek decnum-dynpmcs: r118 | darbelo++ | trunk/aux/decTest/src/ (3 files):
00:58 dalek decnum-dynpmcs: [decTest] Add support for the multiply_add operation (fma in the tests).
00:58 dalek decnum-dynpmcs: [decTest] report the 'testname' in the 'ok' line.
00:58 dalek decnum-dynpmcs: review: http://code.google.com/p/decnu​m-dynpmcs/source/detail?r=118
00:58 japhb joined #parrot
01:20 wayland76 joined #parrot
01:27 cotto darbelo, ping
01:27 darbelo cotto: pong
01:28 cotto darbelo, do you know why the decTest parser takes so long on the decTest files?
01:28 cotto It seems to be taking longer than it should for such a (relatively) simple format
01:29 pmichaud_ joined #parrot
01:31 darbelo I don't know really. I have no knowledge of what PCT does internally in order to tokeinze/parse/etc.
01:31 cotto It may just be that that pct-generated parsers are slow.
01:32 cotto The grammar looks sane and growth seems to be linearly proportional to the size of the input file, as one would expect.
01:32 chromatic The lack of LTM hurts.
01:33 wayland76 Who's working on LTM?  pm?
01:33 cotto wayland76, yes
01:33 pmichaud can I see the grammar?
01:33 Andy joined #parrot
01:33 pmichaud or whatever you all are referring to?  ;-)
01:34 darbelo cotto: FWIW I still plan to stick the generated pir in the repo, but right now at least one file dies from "out of memory"-ness, so that still needs some work.
01:35 darbelo pmichaud: http://code.google.com/p/decnum-dynpmcs/sourc​e/browse/trunk/aux/decTest/src/pct/grammar.pg
01:37 darbelo and a not-very-pathological example of what we are parsing is at http://code.google.com/p/decnum-dynpmcs/so​urce/browse/trunk/t/data/multiply.decTest
01:41 pmichaud you probably want to tell the compiler to assume ascii input instead of utf-8
01:41 pmichaud i suspect you get a significant speedup from that.
01:45 darbelo joined #parrot
01:45 darbelo pmichaud: How do I specify ascii-only input?
01:46 pmichaud I'm checking to verify that's actually the likely problem
01:47 pmichaud okay, I'm guessing it may already be coming in as fixed_8/ascii
01:47 pmichaud so that might not be it
01:47 pmichaud how long is "longer than it should" here?
01:49 cotto about 12 seconds for 730 lines of test file
01:49 pmichaud that does seem longish
01:49 darbelo cotto: Just parse or parse and run?
01:50 cotto like like about 620 lines of actual code
01:50 cotto that's just parsing
01:51 darbelo 'cause the "run" is pretty damm slow, and that's not PCT's fault ;)
01:54 cotto one thing at a time
01:55 * kid51 signs off, the better to contemplate the horrors of "Higher Order Perl"
01:56 darbelo cotto: Just making sure you weren't dumping all the blame in one place. :)
01:56 cotto btw, I don't think most of src/inc/procs.pir is necessary.
01:57 pmichaud maybe I'll try it here... (while watching Larry's SOTO)
01:58 cotto You can use a PAST::Op with :pirop('mul')
01:59 cotto although the way you're doing it does make the compiler simpler
02:00 darbelo cotto: I wanted to get it 'done' before getting it 'right', so to speak. Generating a dumb 'call' means I don't have to care if I made the operation a VTABLE or a METHOD.
02:00 cotto That's a valid strategy.
02:00 cotto I'm afk for a couple hours.  Bye
02:01 darbelo I'm going out for food in a few minutes, so msg me if you need to tell me something.
02:03 dalek decnum-dynpmcs: r119 | darbelo++ | trunk/src/pmc/decnum.pmc:
02:03 dalek decnum-dynpmcs: [DecNum] Add error-checking to a few METHODS. There are now a few less
02:03 dalek decnum-dynpmcs: unexplained NaNs floating about now.
02:03 dalek decnum-dynpmcs: review: http://code.google.com/p/decnu​m-dynpmcs/source/detail?r=119
02:23 pmichaud I'm not sure why it's taking so long to parse.  It could just be the length of the file, but it feels like there must be something else slowing it down.
02:25 pmichaud It doesn't appear to be a file encoding issue.
02:25 pmichaud It *could* be the number of case-insensitive matches required.
02:26 pmichaud (case-insensitive matching being a bit slow and requiring creation of strings)
02:29 tetragon joined #parrot
02:35 janus joined #parrot
02:40 dukeleto how do I run a single parrot test file?
02:53 jdv perl t/harness [tests] ?
02:59 dukeleto jdv: thanks, that works
03:41 dukeleto joined #parrot
03:49 dalek cardinal: 7112008 | (Ted Reed)++ |  (3 files):
03:49 dalek cardinal: Remove the old Failure stuff and add Exception and its hierarchy.
03:49 dalek cardinal: review: http://github.com/cardinal/cardinal/commit​/7112008065f5e541eba155b82d31d750810644fa
04:08 cognominal joined #parrot
04:14 wayland76 joined #parrot
04:16 dukeleto jdv: of course, prove works as well. I just wasn't grasping that Parrot tests are plain ol' perl tests
04:22 bacek_at_work dukeleto: not all of them
04:22 purl not all of them are modifying my brain
04:22 dalek parrot: r40246 | dukeleto++ | trunk (2 files):
04:22 dalek parrot: Add some basic testing for pgegrep
04:22 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40246/
04:22 dukeleto bacek_at_work:  good to know
04:22 dukeleto i don't think pgegrep works
04:24 dalek TT #874 created by dukeleto++: pgegrep is broke
04:31 dukeleto fixed it :)
04:34 bacek_at_work ship it!
04:35 dukeleto is there any guidelines for add files which are needed by tests? i want to add tests for pgegrep on files. i could just test from stdin, but that seems dirty
04:36 bacek_at_work bacek@illusion:~/src/parrot$ find t -name testdata
04:36 bacek_at_work t/native_pbc/testdata
04:36 bacek_at_work dukeleto: something like this
04:50 dalek parrot: r40247 | dukeleto++ | trunk (4 files):
04:50 dalek parrot: Fix pgegrep by updating the mode argument to open() and add a test
04:50 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40247/
04:52 dalek TT #874 closed by dukeleto++: pgegrep is broke
04:54 dukeleto joined #parrot
04:59 mokurai joined #parrot
05:05 dukeleto joined #parrot
05:20 masak joined #parrot
05:25 mokurai left #parrot
05:30 davidfetter joined #parrot
05:31 cotto If we had profiling, it'd be pretty simple to figure out why the decTest parser is slow.  Maybe I should get back to working on that.
05:35 cotto msg darbelo It'd be nice if there were a make target to build/run the decTest tests from the tree's root.
05:35 purl Message for darbelo stored.
05:37 Zak joined #parrot
06:10 dalek parrot: r40248 | dukeleto++ | trunk/t/tools/pgegrep.t:
06:10 dalek parrot: Add tests for the basic command-line options of pgegrep
06:10 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40248/
06:15 cotto That sounds like it could be an interesting tool.
06:17 cotto Hmm.  I think that code wants to be rewritten in nqp.
06:21 bacek_at_work cotto: I'm going to finish OrderedHash branch over weekend and jump back to ops_pct branch.
06:23 cotto fortunately the ops_pct stuff helps me prepare for working with profiling with chromatic
06:25 maettu joined #parrot
06:26 iblechbot joined #parrot
06:27 maettu I have parrot checked out & installed. Now I want to try some languages. BASIC :-) e.g. It's web-link is: https://trac.parrot.org/parrot​/browser/trunk/languages/BASIC How do I check it out with svn?
06:30 moritz mkdir parrot/languages/; cd parrot/languages/; svn co https://trac.parrot.org/parrot​/browser/trunk/languages/BASIC or something like that
06:31 moritz but I don't know if BASIC works against an installed parrot, or if it needs the build tree available
06:32 maettu svn: »https://trac.parrot.org/parrot​/browser/trunk/languages/BASIC« path not found :-(
06:34 moritz ah wait
06:35 bacek_at_work https://trac.parrot.org/languages
06:35 moritz the link you've given about doesn't work either
06:35 Andy joined #parrot
06:35 NotFound I think BASIC doesn't work since some time ago.
06:36 bacek_at_work https://trac.parrot.org/parrot/brow​ser/trunk/languages/BASIC?rev=37396
06:37 bacek_at_work worksforme...
06:37 moritz maybe I should get some more sleep.
06:39 NotFound I think it was droppen when languages leave the nest because no one asked for it to stay or maintained.
06:39 bacek_at_work patch^W volunteers welcome! :)
06:40 NotFound I'd like better an implementation of the gambas Basic dialect.
06:41 maettu bacek_at_work: yes, I can see the code in my browser, but I can't download the thing using svn
06:43 moritz svn co -r $revsion $svn_url
06:44 moritz the $svn_url is probably https://svn.parrot.org/par​rot/trunk/languages/BASIC/
06:45 maettu no
06:47 moritz which revision did you chose?
06:52 maettu 37396
06:55 moritz no idea then
07:26 dukeleto joined #parrot
07:34 dukeleto joined #parrot
07:36 Andy joined #parrot
07:43 wayland76 does anyone know whether ncigen works?
07:46 chromatic I think it does.
07:53 Andy joined #parrot
07:55 cotto pbc?
07:55 purl pbc is not a good form for direct execution
07:55 cotto no, pbc is peanut butter cookie
07:55 purl okay, cotto.
07:55 cotto pbc is also parrot bytecode
07:55 purl okay, cotto.
08:02 cotto I guess it could also be a lead cookie.
08:03 mikehh codetest, distro_tests FAIL , All others PASS (pre/post config, smolder, fulltest) at r40248 - Ubuntu 9.04 amd64
08:04 nopaste "mikehh" at 86.167.197.184 pasted "codetest, distro_test failures at r40248" (85 lines) at http://nopaste.snit.ch/17377
08:04 nopaste "mikehh" at 86.167.197.184 pasted "PATCH for codetest failure in r40248" (19 lines) at http://nopaste.snit.ch/17378
08:04 MoC joined #parrot
08:06 mikehh t/tools/pgegrep.t and t/tools/testdata need svn properties set
08:06 cotto yay!  easy karma
08:07 moritz mikehh++ cage cleaning
08:08 mikehh ok now I am going to try rakudo
08:08 moritz it won't build yet.
08:09 moritz jonathan wanted to give it a shot today (iirc), but doesn't seem awake yet ;-)
08:09 cotto moritz, does it need the ops patch?
08:09 moritz cotto: presumably
08:09 dalek parrot: r40249 | cotto++ | trunk/compilers/ncigen/NCIGENP6.pm:
08:09 dalek parrot: [codingstd] trailing space fix, mikehh++ for noticing
08:09 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40249/
08:09 cotto just a sec
08:10 mikehh oh dear - yes I see that
08:10 nopaste "cotto" at 74.61.2.46 pasted "ops update for rakudo" (25 lines) at http://nopaste.snit.ch/17379
08:10 cotto moritz, try that
08:11 * moritz tries
08:11 cotto (that's for post r40198)
08:11 cotto er, r40198 or after
08:12 moritz seems to re-enable build
08:12 moritz cotto++
08:12 flh joined #parrot
08:12 moritz I'll just bump PARROT_REVSISION to current HEAD
08:12 cotto I broke it in the first place. ;)
08:12 dalek parrot: r40250 | cotto++ | trunk/t/tools (2 files):
08:12 dalek parrot: [codingstd] set svn file properties, mikehh++ for noticing
08:12 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40250/
08:13 moritz it now builds the stage 1 compiler, and then dies with "Direct creation of Iterator"
08:13 moritz which seems to be a different problem
08:14 moritz something that bacek++ "broke", presumably
08:15 cotto Dude.  You get karma for breaking stuff without fixing it?
08:16 dryahetzeph joined #parrot
08:17 moritz I just hope that bacek_at_work will feel guilty for getting karma thusly, and also contributes a patch
08:18 dryahetz1ph joined #parrot
08:18 moritz ;-)
08:18 moritz cotto: I pushed your patch to a branch in Rakudo, so that somebody (maybe jonathan) can pick it up easily
08:19 cotto The Crazy Russian feels no remorse.
08:19 moritz (just for reference, the branch is called build-parrot-head)
08:21 moritz somehow I don't understand that
08:22 moritz ah, never mind
08:23 dukeleto joined #parrot
08:23 moritz seems the Iterator fix is also easy...
08:23 moritz testing
08:37 dalek rakudo: fc60e1d | (Matthew Walton)++ | Test.pm:
08:37 dalek rakudo: Updated Test.pm planless testing support.
08:37 dalek rakudo: Omit your call to plan, and call done_testing when all tests have run.
08:37 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/f​c60e1d0ffdeaadc190bba420d0ac81d09819c06
08:37 dalek rakudo: b198624 | moritz++ |  (3 files):
08:37 dalek rakudo: re-enable building against parrot HEAD
08:37 dalek rakudo: Patch courtesy (mostly) by cotto++
08:37 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/b​198624acd6cf85cf4b38a81b79cd682b00e3cea
08:42 payload joined #parrot
08:50 bacek joined #parrot
08:51 dalek parrot: r40251 | dukeleto++ | trunk/t/tools/pgegrep.t:
08:51 dalek parrot: Add svn Id to t/tools/pgegrep.t
08:51 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40251/
08:53 theory joined #parrot
08:54 bacek o hai
08:54 bacek Who called my name?
08:54 * moritz did
08:54 moritz but it's fixed now anway
08:55 bacek moritz: that's good :)
08:55 * bacek looking for brake something else.
08:57 bacek without remorse and regrets!
09:22 mikehh any idea wht nqp_test is included in make test and not smoke/smolder or fulkltest
09:23 mikehh s/fulkl/full/
09:23 mikehh s/ wht/why/
09:24 chromatic Possibly an oversight.
09:31 mikehh chromatic: been following your blog on Modern Perl - I don't understand all the objections
09:31 mikehh chromatic: to your comments I mean - It makes perfect sense to me
09:33 chromatic David Golden made some good points about vision recently: http://www.dagolden.com/index.php/​331/what-do-you-want-perl-5-to-be/
09:33 moritz that would make a good perlmonks poll ;-)
09:35 mikehh I read that - I still feel if you want to use an old(er) version you can get it from git
09:35 chromatic p5p is not a democracy though, and it's unlikely to be so.  The best we can do is be clear about our motivations and try to help shape development in the directions we want to go.
09:39 mikehh BTW - I just got a copy of Linux Magazine this morning (UK edition) - It has an article on Parrot by Martin Streicher (Birdsong)
09:42 mikehh that's the September 2009 edition
09:45 mikehh All tests PASS (pre/post config, smolder, nqp_test, fulltest) at r40251 - Ubuntu 9.04 amd64
09:47 TiMBuS joined #parrot
09:51 mikehh rakudo now builds on parrot r40251 - make test PASS running make spectest
09:58 donaldh joined #parrot
10:02 mikehh rakudo (b198624) builds on parrot r40251 - make test/make spectest (up to 27696) PASS
10:09 bobke joined #parrot
10:17 dalek rakudo: 9c44089 | jnthn++ |  (8 files):
10:17 dalek rakudo: Separate out the Rakudo metaclass (calling it ClassHOW) from the Parrot one, and make sure everything we create uses this metaclass. This means we poke less stuff into p6object's namespace, and paves the way for various other refactors and additions.
10:17 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/9​c4408925727778c32b08f9b75210ee995ac07b3
10:23 dalek rakudo: e989e65 | jnthn++ | build/PARROT_REVISION:
10:23 dalek rakudo: Oops, mis-merged PARROT_REVISION (Matt-W++ for noticing).
10:23 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/e​989e65d065d764a8d79ac7f5402532528f118c2
10:34 dalek parrot: r40252 | bacek++ | branches/orderedhash_revam​p/src/pmc/orderedhash.pmc:
10:34 dalek parrot: [pmc] Fix thawing OrderedHash.
10:34 dalek parrot: Hooray! Build is unbroken!
10:34 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40252/
10:34 dalek parrot: r40253 | bacek++ | branches/orderedhash_revam​p/src/pmc/orderedhash.pmc:
10:34 dalek parrot: [pmc] Reimplement OrderedHash.defined_keyed_* and exists_keyed_*
10:34 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40253/
10:34 dalek parrot: r40254 | bacek++ | branches/orderedhash_revam​p/src/pmc/orderedhash.pmc:
10:34 dalek parrot: [pmc] Implement OrderedHash.delete_keyed
10:34 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40254/
10:36 maettu left #parrot
10:49 masak I'd like to take a moment to point out that https://trac.parrot.org/parrot/ticket/536 is slightly annoying for Rakudo application developers.
10:52 skv joined #parrot
11:03 wayland76 joined #parrot
11:12 jonathan masak: It sure is. :-|
11:13 jonathan And it's only going to get a bigger problem in the near future.
11:13 * masak sulks
11:16 jonathan Well, since the first proposed solution involves removing code.. ;-)
11:17 bacek jonathan: patched welcome! :)
11:17 bacek s/d /s /
11:17 jonathan bacek: Aye, well, I think allison++ wants the difference benchmarked before we rip it out.
11:17 jonathan I've no clue what difference it will make tbh.
11:18 bacek it's better to be slow but correct...
11:19 bacek And any optimisations shouldn't be handled by some low-level assembly compiler.
11:20 jonathan correct++
11:21 donaldh joined #parrot
11:21 bacek *sigh*
11:21 * bacek just realised that he forgot to put OrderedHash into DEPRECATION.pod...
11:24 rob joined #parrot
11:27 wayland76 joined #parrot
11:57 wayland76 masak: You're incredible
11:59 masak wayland76: thank you. for what do I deserve the honour?
11:59 wayland76 Well, you've heard of the incredible sulk, yes?  :)
12:00 masak not until right now, no. :P
12:00 wayland76 (sorry :) )
12:00 wayland76 (I can never resist a joke)
12:00 masak you're forgiven.
12:01 * pmichaud now understands what that breeze passing directly over his head was.
12:33 iblechbot joined #parrot
12:40 payload joined #parrot
12:41 tetragon joined #parrot
13:37 wayland76 purl: help
13:37 purl #perl is not a help channel, and I'm not a help bot.  If you want Perl help, try #perl-help or #metallica. or (see the 'help channel' factoid as well) or
13:38 masak purl: you have some attitude, you know that?
13:38 purl bugger all, i dunno, masak
13:38 wayland76 purl: @karma
13:38 purl wayland76: huh?
13:39 moritz karma wayland76
13:39 purl wayland76 has karma of 1
14:06 ruoso joined #parrot
14:28 mj41 joined #parrot
15:20 dalek TT #823 reopened by coke++: packfile pmc tests failing
15:20 donaldh joined #parrot
15:27 integral joined #parrot
15:46 dalek pynie: r78 | isop44++ | trunk/ (4 files):
15:46 dalek pynie: Use iter pirop instead of instantiating Iterator's. Parrot doesn't allow the
15:46 dalek pynie: latter anymore.
15:46 dalek pynie: review: http://code.google.com/p/pynie/source/detail?r=78
15:48 dalek parrot: r40255 | jonathan++ | trunk/runtime/parrot/library/P6object.pir:
15:48 dalek parrot: [p6object] If a class already does P6object, don't try to add it again. This enables us to register derived metaclasses.
15:48 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40255/
15:53 dalek rakudo: ca2e4f1 | moritz++ | t/spectest.data:
15:54 dalek rakudo: [t/spectest.data] 5 more passing test files
15:54 dalek rakudo: How did we miss those? there was no (new) fudging required to make them pass.
15:54 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/c​a2e4f1f66a1f8ee85125fc7062d4940df1ea9f8
16:06 dalek rakudo: f78cf0b | jnthn++ | src/ (3 files):
16:06 dalek rakudo: This gets us creating the metaclass right at the start of class creation and operating on it. We dispatch traits on it also, which many of the fixes in here are helping towards. We also make mentions of ClassHOW work more properly, by having it registered, so now it's a bit less special. For now, there's more mess, but this is mostly an enabling refactor for more cleanups.
16:06 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/f​78cf0be1d194dd81b72b97fc50c21c824abf385
16:06 dalek rakudo: 546c253 | jnthn++ | :
16:06 dalek rakudo: Merge branch 'master' of git@github.com:rakudo/rakudo
16:06 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/5​46c25380e9c7a3a44265a0f71280527ca1e7b36
16:08 darbelo joined #parrot
16:12 dalek rakudo: d4d2319 | moritz++ | t/spectest.data:
16:12 dalek rakudo: [t/spectest.data] fix test name, moritz--
16:12 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/d​4d2319eb75cadb1fadb10a34494351b1035b4ff
16:26 mokurai joined #parrot
16:31 cotto bacek_at_work, how exciting.  You get the have the first deprecation eligible after 2.0.
16:31 cotto although if you're not changing the API, you're clear.
16:33 pmichaud joined #parrot
16:35 flh joined #parrot
16:39 Zak joined #parrot
16:42 dalek partcl: r543 | coke++ | trunk/ (2 files):
16:42 dalek partcl: Add --skip option to spec testing, to specifically test skipped tests (unfudge)
16:42 dalek partcl: Minor code cleanup.
16:42 dalek partcl: Add a potentially useless SIGCHLD handler
16:42 dalek partcl: review: http://code.google.com/p/p​artcl/source/detail?r=543
16:47 dalek rakudo: 5167835 | pmichaud++ | docs/spectest-progress.csv:
16:47 dalek rakudo: spectest-progress.csv update: 418 files, 11913 passing, 0 failing
16:47 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/5​167835aac57c0c6e20b976998f402bd359d5300
16:49 Psyche^ joined #parrot
16:58 dalek cardinal: ba66eaf | (Danius Michaelides)++ |  (2 files):
16:58 dalek cardinal: Implement Array.each_index with test; also adds a test for .each_with_index.
16:58 dalek cardinal: Signed-off-by: Ted Reed <ted.reed@gmail.com>
16:58 dalek cardinal: review: http://github.com/cardinal/cardinal/commit​/ba66eafe01d1a018567e00689f1c40919fee7a7e
17:01 particle great to see cardinal progressing with new contributors
17:01 rob joined #parrot
17:04 treed Yeah.
17:04 treed It'd be great to grow a community around it.
17:05 particle every commit helps :)
17:05 cotto I'm happy.  More people developing means more pressure to make Parrot awesome.
17:06 treed I got a pull request yesterday, and that patch today.
17:06 treed Two different people.
17:07 jonathan Wow! Nice! :-)
17:08 treed The patch was froom a guy who says he's a newbie with the whole git thing. In my reply thanking him for the patch I offered to help teach him about git.
17:10 PerlJam treed++ building community
17:10 treed Thanks.
17:11 particle feel free to set up #cardinal if it'll help
17:11 treed I think he's in a drastically different timezone than me, though.
17:11 treed I was thinking about it.
17:11 particle i'll join, to be a parrot resource
17:11 treed His name is Greek and the patch came in at 4AM my time.
17:12 treed I just gotta figure out how to have that many IRC channels.
17:12 treed (irssi)
17:13 treed Well, rather, how to navigate to them.
17:13 PerlJam treed: what client do you use?
17:14 particle sounds like irssi
17:14 treed Yeah.
17:14 treed I got rid of a query window so now #cardinal fits in at #10.
17:15 PerlJam I use Alt-1, Alt-2, etc. to switch between my windows
17:15 PerlJam for numbers > 10, it's Alt-Q, Alt-W, etc.
17:15 particle maybe Infinoid can get dalek reporting cardinal commits in #cardinal
17:15 treed Ah.
17:15 PerlJam (qwerty keyboard)
17:15 treed That'd be better if I still used QWERTY>
17:15 particle dvorak++
17:15 particle 12 years and counting
17:15 treed 7 years here, I think.
17:16 treed I remember it was December and I was living in an apartment in Ontario.
17:16 treed Which would have to have been either 2002 or 2003.
17:16 treed Pretty sure it was 2002.
17:16 * Coke idly wonders if he can qwerty faster than particle can dvorak.
17:17 treed Because I was unemployed at the time. (Actually the last time I was unemployed before the last couple weeks.)
17:17 treed I have a hard time accurately measuring my typing speed. Largely because so many tests have that stupid "accuracy" thing.
17:17 particle acuraccy--
17:17 treed Look, let me type things to their completion. If it's not 100% accurate, I'm not done.
17:17 * treed has a bad habit of typing arary instead of array.
17:17 treed Also acconut instead of account.
17:18 particle i can usually tell dvorak users by their typos
17:18 treed Heh. I remember a quote from a dvorak user. "Actually, don't listen tto $band. I want you to kill yourself. ... KILL yourself. Dammit."
17:18 treed er
17:18 treed the first kill was kiss
17:19 treed LOL, I typoed it.
17:20 particle :)
17:20 jdv79 are there any plans for (or does it already exist) memory metrics for mem profiling and such?
17:21 treed (Also, lag (typeahead( ssucks)
17:21 jdv79 we're running into this at $work with p5 and its inadequate facilities for mem profiling.
17:24 mikehh_ joined #parrot
17:27 dalek rakudo: dd5767c | (Kyle Hasselbacher)++ | src/ (2 files):
17:27 dalek rakudo: Simple implementation of [//] and [||] operators
17:27 dalek rakudo: (Improvements by Moritz: use slurpy arg, and register ops in grammer-oper.pg)
17:27 dalek rakudo: Signed-Off-By: Moritz Lenz <moritz@faui2k3.org>
17:27 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/d​d5767c0cbfb4f1d70641e8499e21091be03729c
17:31 dalek cardinal: 4a4c0b6 | (Ted Reed)++ | README:
17:31 dalek cardinal: Add a note about #cardinal to the README.
17:31 dalek cardinal: review: http://github.com/cardinal/cardinal/commit​/4a4c0b6bce2ec5abebcf9d59e67b4ef8c115e368
17:36 cotto jdv79, the only profiling work I know of is what chromatic and I will be working on, which is focusing on execution time.
17:36 estrabd joined #parrot
17:39 treed I wonder, too, if it'd be possible to set up dalek to get information about the issue tracking system on cardinal.
17:42 Coke where is cardinal hosted?
17:42 treed github
17:43 Infinoid does github have issue tracking?
17:43 treed Yeah.
17:43 treed http://github.com/cardinal/cardinal/issues
17:43 treed It's fairly rudimentary, but it works.
17:44 Infinoid cool.  That means I can add support for issue tracking once and it'll automatically work for all the github projects (which are many)
17:44 mikehh__ joined #parrot
17:45 treed There's an RSS feed on github, but it's for the whole project, commits and all.
17:45 Infinoid yeah.  http://github.com/Infinoid/dalek-plugins/​blob/master/modules/local/githubparser.pm parses that
17:45 treed Actualyl, doesn't seem to do issues or wiki page changes.
17:45 Infinoid if it's some other feed URL, shouldn't be hard to poll that too
17:46 treed There doesn't seem to be one for issues.
17:46 treed http://wiki.github.com/car​dinal/cardinal/wikis.atom
17:46 treed That's the one for the wiki.
17:49 particle opbots: trust treed
17:49 slavorg Ok
17:49 slavorgn Ok
17:49 treed Huh.
17:50 treed That's a ... simple security model.
17:50 particle you seem vaguely human.
17:50 particle that's our criterion for op privs here
17:50 particle congrats! :)
17:50 treed Oh, so I need to work on my mask, then? Where is it showing through?
17:50 treed Haha, thanks.
17:50 Andy joined #parrot
17:52 treed Apparently in order to get slavorg to join #cardinal, it needs to trust me on that channel.
17:53 particle slavorg: owner?
17:53 purl owner is hachi, see also #purl
17:53 particle hachi can get that set up for you
17:53 treed k
17:54 treed message hachi Hi! Would it be possible to set up slavorg in #cardinal? Thanks!
17:54 purl Message for hachi stored.
17:54 particle probably better to /msg hachi, he's not here that often
17:54 treed k
17:55 hachi isn't slavorg a person?
17:56 treed Huh?
17:57 treed If they are, they type very fast and very convincingly botlike.
17:57 hachi well, in any case... I am not their keeper... and I don't even know who is
17:58 particle oh, silly purl
17:58 purl La la la. I can't hear you
17:58 particle i asked slavorg for owner, and purl responded
17:58 treed Oh, heh.
17:59 treed Sorry to bug you, hachi.
18:08 chromatic joined #parrot
18:17 darbelo treed: I think slavorg/slavorgn is run by Infinoid
18:17 Coke note that purl was answering for herself.
18:18 treed Yeah, we noted that after the fact.
18:18 treed darbelo: He was just here. :-P
18:19 Coke rant: coredumps are overriding each other!
18:19 Coke (so if I have 20 coredumps in the partcl spec test suite, I have only one core file. whah.)
18:24 hercynium joined #parrot
18:24 Tene treed: github has IRC reporting on freenode.  You could move #cardinal to freenode.
18:25 treed Hm.
18:25 treed How does it handle that?
18:25 treed Is there a configuration for that somewhere?
18:25 Tene Yes.
18:25 Tene in the admin page
18:25 Tene you'll need to log in as 'cardinal'
18:25 * treed nods.
18:26 treed If only I didn't have an entirely different nick on freenode. :-P
18:27 treed Shit.
18:27 treed #cardinal on freenode is already taken.
18:27 Tene orly?
18:27 purl YA RLY.
18:27 Tene thx purl
18:27 treed Only three people in there.
18:28 Tene we could just use it anyway and /ignore them.
18:28 Tene ;)
18:28 treed Haha.
18:28 treed Whoever's in there isn't the registered owner.
18:29 treed It hasn't been used by the registered owner in 5 years.
18:33 Tene treed: actually, github has irc notifications for any irc server.
18:33 treed Oh.
18:33 Tene just checked
18:33 Tene https://github.com/cardinal/cardinal/edit/hooks
18:34 treed Aha.
18:34 treed Hm.
18:34 treed Best to keep it here, then maybe.
18:35 Tene Could be.  The reason to move to freenode is to reinforce the separation from Perl 6.  I know that's an issue for Pynie.
18:35 Tene The python folks don't much like perl, or something, so allison has had to be very explicit in not letting anything with 'perl' in the name be a part of Pynie.
18:36 Coke Tene: given that perl6 is on freenode, I'm confused. =-)
18:36 theory joined #parrot
18:36 treed Ah.
18:36 treed That's a good point.
18:36 Tene Coke: "on the same network as #perl6" is less of an association with perl than "on irc.perl.org"
18:36 Coke I certainly hope the non-perl gyrations are worth the effort. So far, any of those migrations have been a PITA.
18:36 Coke Tene: so you mean perl, not perl6?
18:37 Tene Coke: from what I've heard, seeing the word "perl" mentioned anywhere associated with a Parrot-hosted language is enough to drive some people away from contributing.
18:37 Tene I've heard.
18:38 darbelo Tene: I think irc.parrot.org works as an alias to irc.perl.org
18:38 Coke darbelo: yes, that's all it is.
18:39 Coke the switch away on mailing lists and ticketing systems was a bit more problematic.
18:39 Coke (and still not resolved, after... what, a year?)
18:39 treed Yeah, I found a rant from someone about how they gave up on Parrot  because it's just a Perl6 clique blah b lah blah.
18:39 Tene Coke: I know, for example, that pynie absolutely can't use P6object while it still has that name, but potentially could if it was renamed.
18:39 Coke that guy on list, yah. who refused to contribute even after everyone nicely followed up with him.
18:39 Coke I'm not entirely convinced those are the people we want to be attracting anyway.
18:40 Coke Tene: that's insane.
18:40 Coke I'm not saying you're wrong, to be clear.
18:40 Tene Coke: As I understand it, allison considers it a completely reasonable fear.
18:40 darbelo Tene: to Python6Object? :)
18:40 chromatic I'll go further.  Those are people I want nothing to do with.
18:41 Coke chromatic: I was trying to be diplomatic.
18:41 treed Tene: looks like the IRC hook on github only does commits?
18:41 Tene treed: what else is there?
18:41 Coke perhaps that'll change if I get voted out. =-)
18:41 treed Tene: Issue tracker, wiki.
18:41 Tene ah
18:41 Tene Um, dunno.
18:42 chromatic I don't see the point of diplomacy versus someone willing to post angry screeds.  I'm all for discouraging angry screeds.  We're all volunteers and we should treat each other with kindness and respect.
18:42 treed The wiki has an atom feed, but there's nothing for the issue tracker.
18:45 PerlJam chromatic: http://www.randsinrepose.com/archiv​es/2009/06/21/a_toxic_paradox.html  (he says we need these people but doesn't make a good enough argument for me)
18:49 chromatic "I'm not touching your project because the domain name of your bug tracker has the word 'perl' in it!" is far, far beyond "I feel upset because I believe your backwards compatibility policy will lead to rewriting 500k sloc in the next year instead of doing new development."
18:49 chromatic One of them is a disagreement over vision, implied or expressed.
18:51 chromatic "I will only contribute if you change your bug tracker, migrate to other domain names, and set up a separate 501 c 3 corporation to hold your copyrights and trademarks" is a big request.
18:59 Coke chromatic: and yet we did it.
18:59 Coke I haven't seen a huge influx of contributors, though. =-)
19:00 chromatic There's no *good* time to make any of those changes.  I won't deny that they're useful changes though.
19:03 Coke I don't think "changing our bug tracker" is useful. It was theoretically useful. in practice it has been a huge time sink and is less featureful now.
19:04 PerlJam sunk cost though at this point.
19:06 Coke sure, NOW it is.
19:06 Coke I know. why don't we switch back to RT, then? then it'll be another sunk cost. =-)
19:09 Coke PerlJam: to be less snarky, it was an avoidable cost, and it gave us very little bang for our buck. We're not even done paying for it yet.
19:10 Coke so the next time we make a big non-technical change like that, we should probably examine our goals a little more closely.
19:14 iblechbot joined #parrot
19:17 hachi treed: no worries, I'm hoping to try and find a better answer for you
19:19 Infinoid treed: I run slavorgn, but not slavorg
19:19 Infinoid so I can give you 1 out of 2
19:20 donaldh joined #parrot
19:33 theory joined #parrot
19:35 payload joined #parrot
19:38 flh joined #parrot
19:39 dalek parrot: r40256 | tene++ | trunk/runtime/parrot/library/Curses.pir:
19:39 dalek parrot: Add a pair of missing fucntions to Curses.pir
19:39 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40256/
19:42 particle how do i convince perltidy to align vertically on =>
19:42 dalek parrot: r40257 | cotto++ | branches/ops_pct (391 files):
19:42 dalek parrot: bring branch up-to-date with trunk, remove some references to deleted files
19:42 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40257/
19:47 MoC joined #parrot
19:47 jevin joined #parrot
20:03 jevin joined #parrot
20:22 bacek joined #parrot
20:29 cotto howdy bacek
20:36 eternaleye joined #parrot
20:37 bacek morning cotto
20:38 * bacek trying to understand why he woke up...
20:44 eternaleye joined #parrot
20:46 Andy joined #parrot
20:48 cotto clock?
20:48 purl cotto: LAX: Fri 1:48pm PDT / CHI: Fri 3:48pm CDT / NYC: Fri 4:48pm EDT / LON: Fri 9:48pm BST / BER: Fri 10:48pm CEST / IND: Sat 2:18am IST / TOK: Sat 5:48am JST / SYD: Sat 6:48am EST /
20:48 cotto kangaroo?
20:48 purl kangaroo is an aboriginal word meaning "I don't know." Applied to a bouncy marsupial due to a conversation that went something like. "What is that thing?" "Kangaroo." "Oh."
20:48 treed Infinoid: Thanks.
20:49 jonathan Is that really ture?!
20:50 cotto http://en.wikipedia.org/wiki/Kangaroo#Terminology
20:50 cotto because everything on Wikipedia is true
20:52 payload FAILGOAL is defined very similar in PGE/Regex.pir and in PCT/Grammar.pir
20:52 Coke bacek: ping
20:53 bacek cotto: poooong
20:53 Coke bacek: packfile tests failing on i686. I reopened the ticket.
20:53 Coke (culled from smolder reports)
20:54 not_coke That should clear up a few things.
20:54 not_cotto sure will!
20:54 bacek Coke: I saw it :-/
20:54 * zombie_bacek brainz... brainz...
20:55 not_coke zombie_bacek, can zombies still code?
20:55 chromatic They can't steer.
20:55 zombie_bacek In brainless more
20:55 Andy joined #parrot
20:55 not_coke Hmmm.
20:55 not_cotto and that's no bull?
20:55 not_coke I guess that works.
20:56 not_coke Sometimes I wonder if I've been coding like that.
20:57 estrabd_ joined #parrot
20:59 zombie_bacek not_cotto: (OrderedHash) My new implementation "reorded" integer keys after delete values.
20:59 zombie_bacek And it differ from "old" OrderedHash...
21:01 not_coke zombie_bacek, Hmm.  That does sound suspiciously like an API change.
21:01 not_not_cotto maybe that'll help
21:05 treed Infinoid: If you could add commit notifications or even just this feed https://github.com/cardinal.atom to #cardinal, that'd be great.
21:05 treed (I think that's the feed)
21:06 bacek Coke: it is API change :(
21:08 Coke bacek: what?
21:08 cotto You might mention it (by proxy?) at #ps.  I think deletion from OrderedHashes is mostly unused (because it's an ugly hack).
21:08 treed Yeah, probably with s/https/http/
21:08 Coke were you talking to cotto again?
21:08 treed That'll include the wiki page edits and issue tracker, as well.
21:08 bacek Bah!
21:08 * bacek need second coffee
21:08 cotto Yes, you do.
21:08 * Coke takes his coffee and goes home
21:16 bacek cotto: Deletion from "old" OrderedHash is hack. But it works (fsvo). We can't change this behaviour without deprecation notice...
21:21 cotto bacek, could you make the new code act like the old code in a way that'd be easy to change after 2.0?
21:22 bacek cotto: there is not point to do it...
21:23 cotto It sounds easier than dealing with 6 months of changes before applying it.
21:24 bacek not quite true...
21:25 bacek support "old" behaviour will increase code (and complexity of core) dramatically.
21:26 cotto Ok.  I was afraid that'd be the case.
21:26 bacek (at least with my current OH implementation)
21:26 cotto How many files does it touch?
21:26 cotto (the branch)
21:26 bacek May be I'll investigate different approach for implementing OrderedHash.
21:26 bacek 1
21:26 cotto no test changes?
21:27 bacek tests failing.
21:27 bacek (for deleted values from OH)
21:27 cotto ok.  For some reason I though that the branch was ready to merge (apart from the missing deprecation)
21:27 bacek (actually 3 files)
21:28 bacek cotto: it almost ready. Some final polishing required.
21:30 cotto Is the only externally-visible difference in how OH behaves when you delete elements?
21:31 bacek cotto: yes afaiu
21:31 cotto because if so, it doesn't sound like it'd be too much work to keep it in a branch, occasionally sync with trunk and merge it after 2.0.
21:31 cotto but adding a deprecation notice now wouldn't be a bad idea
21:32 bacek indeed
21:35 bacek msg Coke I can't reproduce packfile failure on my Linux/i686...
21:35 purl Message for coke stored.
21:37 theory joined #parrot
21:41 Whiteknight joined #parrot
21:43 bacek good morning Whiteknight
21:43 Whiteknight good morning bacek
22:03 theory joined #parrot
22:09 treed message Infinoid If you could add commit notifications or ideally just this feed http://github.com/cardinal.atom (includes wiki edits and issue tracker as well as commits) to #cardinal, that'd be great.
22:09 purl Message for infinoid stored.
22:09 Infinoid Yeah, working on that now
22:09 treed Oh, okay.
22:09 treed Thanks.
22:09 treed Wasn't user you saw that earlier.
22:09 dalek joined #parrot
22:10 Infinoid Huh?  Understand sentence I don't your.
22:10 Infinoid Oh, s/user/sure/?
22:19 Whiteknight joined #parrot
22:24 treed Yeah.
22:24 treed Sorry.
22:24 treed That's a funny typo.
22:28 dalek cardinal: aa7cbfb | cardinal++ | index.html:
22:28 dalek cardinal: github generated gh-pages branch
22:28 dalek cardinal: review: http://github.com/cardinal/cardinal/commit​/aa7cbfb404d650e43a4f853d139366af7fcaa9af
22:28 dalek cardinal: 03380ae | (Ted Reed)++ | :
22:28 dalek cardinal: Merge branch 'gh-pages' of git@github.com:cardinal/cardinal
22:28 dalek cardinal: review: http://github.com/cardinal/cardinal/commit​/03380aef1e20e996be2fd766171d7c7684577bb2
22:28 dalek cardinal: 2c093da | (Ted Reed)++ | Rakefile:
22:28 dalek cardinal: Add more items to the test suite report, and also to the CLEAN FOR COMMIT check.
22:28 dalek cardinal: review: http://github.com/cardinal/cardinal/commit​/2c093da59eaaf2d8cc88bc7a71bac95264cf041d
22:35 donaldh_ joined #parrot
22:37 rg joined #parrot
22:43 dalek parrot: r40258 | Infinoid++ | trunk/CREDITS:
22:43 dalek parrot: Add treed++ to CREDITS for karma aliasing.
22:43 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40258/
22:47 dalek cardinal: 604c940 | (Ted Reed)++ | index.html:
22:47 dalek cardinal: Remove index.html, which somehow fell into master from gh-pages.
22:47 dalek cardinal: review: http://github.com/cardinal/cardinal/commit​/604c940868b39662738d98d2e2a0d7a32cf0997b
22:47 bacek joined #parrot
22:51 tetragon joined #parrot
23:06 Limbic_Region joined #parrot
23:20 donaldh joined #parrot
23:22 kid51 joined #parrot
23:31 bacek joined #parrot
23:53 payload joined #parrot
23:56 dalek parrot: r40259 | whiteknight++ | branches/bsr_jsr_ret (3 files):
23:56 dalek parrot: [bsr_jsr_ret] update two more tests
23:56 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40259/

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

Parrot | source cross referenced