Camelia, the Perl 6 bug

IRC log for #parrot, 2008-06-28

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:03 dalek r28765 | jkeenan++ | autoicu:
00:03 dalek : [configure] To facilitate refactoring, explicitly code all 'else' conditions
00:03 dalek : and use whitespace extremely liberally.  Temporarily add a call to runstep()
00:03 dalek : to test file.
00:03 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28765
00:10 AndyA joined #parrot
00:31 ruoso joined #parrot
00:36 particle1 joined #parrot
01:06 bacek Morning everyone
01:06 bacek Congratulations! First 1000 passing tests in rakudo! Ho!
01:07 zostay joined #parrot
01:27 DietCoke woot
01:30 dalek r28766 | cotto++ | trunk:
01:30 dalek : [php] added an accidentally omitted target to make clean
01:30 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28766
01:30 cotto_home me++
01:38 * DietCoke wonders what he got credit for in pmichaud's blog entry.
01:41 Coleoid_ joined #parrot
01:41 dalek r28767 | cotto++ | trunk:
01:41 dalek : [dynpmc] This closes RTs 46387, 46389 and 46491.  All gdbm tests still pass and
01:41 cotto_home want karma
01:41 dalek : this seems to be the Right Thing based on what other PMCs do.
01:41 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28767
01:41 cotto_home yay
01:46 bacek perl6: say 'cotto++ # he want it :)'
01:46 polyglotbot OUTPUT[cotto++ # he want it :)␤]
01:46 bacek karma cotto
01:46 bacek pugs left again...
01:49 teknomunk joined #parrot
01:49 kid51 joined #parrot
02:14 ank joined #parrot
02:40 Coleoid joined #parrot
03:16 dalek r28768 | jkeenan++ | autoicu:
03:16 dalek : [configure] Begin refactoring of auto::icu::runstep.  First test file tests
03:16 dalek : the --without-icu command-line option.
03:16 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28768
04:10 apeiron_ joined #parrot
04:12 cotto_home purl--
04:33 purl joined #parrot
05:02 Psyche^ joined #parrot
05:10 Infinoid cotto_home: purl deftly dodged your decrement :P
05:19 Theory joined #parrot
05:20 apeiron joined #parrot
05:30 cotto_home cotto++ cotto++ purl--
05:30 cotto_home no he didn't
05:30 cotto_home s/he/she/
05:59 Psyche^ joined #parrot
06:16 apeiron_ joined #parrot
07:25 nnunley_ joined #parrot
07:34 Ademan joined #parrot
07:38 iblechbot joined #parrot
07:39 Ademan_ joined #parrot
08:15 barney joined #parrot
08:16 dalek r28769 | fperrad++ | libs4php:
08:16 dalek : [php] fix md5_file
08:16 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28769
08:17 dalek r28770 | fperrad++ | libs4php:
08:17 dalek : [php] missing operators
08:17 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28770
09:10 dalek r28771 | bernhard++ | trunk:
09:10 dalek : [Plumhead (Pipp)]
09:10 dalek : svn merge -r 28724:28770 https://svn.perl.org/parrot/bran​ches/libs4php/languages/plumhead
09:10 dalek : Mark all tests that are failing in the PCT variant as TODO.
09:10 dalek : Untodo all unexpected successes in the the PCT variant.
09:10 shorten dalek's url is at http://xrl.us/goi5o
09:10 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28771
09:21 dalek r28772 | bernhard++ | trunk:
09:21 dalek : [Plumhead PCT]
09:21 dalek : Use PhpArray instead of Hash.
09:21 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28772
09:25 dalek r28773 | bernhard++ | trunk:
09:25 dalek : [Plumhead PCT]
09:25 dalek : Use PhpString instead of String.
09:25 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28773
09:29 dalek r28774 | bernhard++ | trunk:
09:29 dalek : [Plumhead PCT]
09:29 dalek : Use PhpInteger instead of Integer
09:29 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28774
09:30 mire joined #parrot
09:31 masak joined #parrot
09:33 * barney reads http://www.artima.com/weblog​s/viewpost.jsp?thread=233654
09:34 dalek r28775 | bernhard++ | trunk:
09:34 dalek : [Plumhead PCT]
09:34 dalek : Use PhpFloat instead on Float.
09:34 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28775
09:40 schmalbe joined #parrot
09:55 dalek r28776 | bernhard++ | trunk:
09:55 dalek : [Plumhead PCT]
09:55 dalek : A function call is an expression.
09:55 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28776
09:57 barney joined #parrot
10:01 bacek joined #parrot
10:25 dalek r28777 | bernhard++ | trunk:
10:25 dalek : [Plumhead PCT]
10:25 dalek : Echo can take several arguments.
10:25 dalek : Arguments are separated by comma.
10:25 dalek : Untodo many test cases.
10:25 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28777
10:27 moritz is there a consensus already for the new name of plumhead?
10:27 moritz or a favourite?
10:27 tlb joined #parrot
10:29 bacek joined #parrot
10:34 tlb How do I increment a pack'ed pointer from Perl? eg. my $memptr = pack( 'P', $memory ); ?
10:35 tlb wrong channel, should have looked at the header, sorry :)
10:35 barney The favourite is Pipp. Alternatives are at http://www.perlfoundation.org/pa​rrot/index.cgi?plumhead_renaming
10:35 shorten barney's url is at http://xrl.us/d2gpk
10:40 dalek bernhard.schmalhofer@gmx.de | Plumhead:
10:40 dalek link: http://www.perlfoundation.or​g/parrot/index.cgi?plumhead
10:42 dalek bernhard.schmalhofer@gmx.de | plumhead_renaming:
10:42 dalek link: http://www.perlfoundation.org/pa​rrot/index.cgi?plumhead_renaming
10:42 shorten dalek's url is at http://xrl.us/d2gpk
10:56 cognominal joined #parrot
10:57 tlb joined #parrot
11:32 bacek joined #parrot
11:32 bacek g'localtime
11:33 moritz hi bacek ;)
11:36 rdice joined #parrot
11:39 bacek moritz: congratulations for your commit bit
11:40 bacek and first 1000 passing tests
11:40 moritz thanks
11:40 moritz and thanks to you for helping!
11:42 bacek welcome!
11:42 purl Heh, not likely around here you won't be.
11:42 bacek purl: how is going?
11:42 purl i haven't a clue, bacek
11:42 bacek stupid bot.
11:42 purl So I passed Turing's test?
11:43 bacek oh wait. Stupidbot is mine
11:43 * bacek very disappointed...
11:44 bacek My dog dropped laptop from table... Disk died... My rakudo's repo died... Other projects died...
11:46 rjbs but due to your restraint, dog didn't die
11:50 bacek no... But she was close to death.
12:15 paco joined #parrot
12:17 kid51 joined #parrot
12:20 bacek heh. I broke 'mutating $_' test in S29-list/grep.t in my implementation of lazy list.
12:25 bacek http://192.168.1.3/cgit/parrot/tree/languag​es/perl6/src/classes/List.pir?h=bacek&i​d=98825289fb25d8a19f31e0679b9bd9ac23ab7854
12:25 shorten bacek's url is at http://xrl.us/hi2j2
12:25 bacek message pmichaud http://xrl.us/hi2j2 my current version of lazy list (without ListIterator but with broken 'mutating $_')
12:26 bacek yak...
12:26 bacek She doesn't like me
12:27 AndyA joined #parrot
12:27 kid51 She's been very petulant lately -- running away at *anyone's* attempt to leave a private msg.
12:28 kid51 hachi was looking at that
12:29 Ivatar joined #parrot
12:31 bacek I can load Infobot plugin in stupidbot.
12:32 bacek message pmichaud http://xrl.us/hi2j2 my current version of lazy list (without ListIterator but with broken 'mutating $_')
12:32 bacek stupidbot: message pmichaud http://xrl.us/hi2j2 my current version of lazy list (without ListIterator but with broken 'mutating $_')
12:32 bacek oh...
12:59 bacek moritz: ping
12:59 moritz bacek: pong
13:00 moritz bacek: giving URLs with local IP addresses isn't a good idea ;-)
13:00 bacek yak..
13:01 bacek http://bacek.homeip.net/cgit/parrot/tree/lang​uages/perl6/src/classes/List.pir?h=bacek&​id=98825289fb25d8a19f31e0679b9bd9ac23ab7854
13:01 shorten bacek's url is at http://xrl.us/hjit3
13:01 bacek this is better :)
13:01 dalek r28778 | jkeenan++ | autoicu:
13:01 dalek : [configure] Refactor some code out of runstep() into internal sub
13:01 dalek : _handle_search_for_icu_config(), then test it.
13:01 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28778
13:01 bacek moritz: can you test 'test3.diff' from RT #55438?
13:01 stupidbot Error calling said() for rt: Internal Server Error
13:02 stupidbot joined #parrot
13:02 bacek RT #55438
13:02 stupidbot Error calling said() for rt: Internal Server Error
13:03 bacek Looks like it from rt.perl.org...
13:04 kid51 RT has occasional micro-timeouts, but I was just able to reach it.
13:06 bacek stupidbot uses REST interface. Probably it broken.
13:09 moritz bacek: I'm of two minds regarding your patch
13:09 moritz bacek: the testing interface is very ugly, and it hides state information
13:10 moritz bacek: maybe expect_fail should take a closure, and only in that closure a test negates its meaning?
13:10 moritz don't know if that's really better, though
13:11 bacek moritz: I actually prefer version from test2.diff (with overriding proclaim), but pmichaud thinks that latest is better.
13:12 * moritz looks at test2.diff, just a second...
13:13 moritz both are "spooky action at a distance"
13:14 moritz which is why I'd like expect_fail { ok 0, 'OK works' }, which would hide the spooky action behind Test.pm's scenes
13:15 moritz it can then set $Test::expect_fail locally, execute the closure, and reset $Test::expect_fail
13:19 bacek hmm...
13:20 moritz but wait for pmichaud's opinion before you implement it ;)
13:22 bacek :)
13:22 bacek Can you check, that this patch works (it works for me)?
13:24 moritz tryiing...
13:26 moritz so far everything seems fine
13:26 moritz (spectest_regression still running)
13:28 bacek What about t/02-test-pm/6-full.t?
13:29 moritz passes
13:29 bacek it's good :)
13:36 bacek Did I miss something and Num is actually role now?
13:36 moritz don't think so (although I wish it were)
13:37 moritz did you read "rakudo test miscellanea" thread on p6l?
13:37 bacek in progress.
13:37 bacek "The following are all defined in the Num role:"
13:38 bacek this is from http://feather.perl6.nl/syn/S29.html
13:38 moritz perl6: say (1.234).WHAT
13:38 polyglotbot OUTPUT[Num␤]
13:39 moritz can there be a role and a class of the same name?
13:40 * bacek switch purl mode
13:40 bacek i haven't a clue, moritz
13:40 * bacek switch off purl mode
13:48 iblechbot joined #parrot
13:49 moritz t/spec/S03-operators/comparison.t
13:49 moritz # Looks like you failed 17 tests of 18
13:49 moritz ouch
13:51 bacek hmm... works for me... Can you run it separately?
13:52 moritz all but second test are "not ok"
13:53 pmichaud I think that when S29 was written it wasn't clear if Num would be a role or class.  (It's still not clear.)
13:53 Whiteknight joined #parrot
13:53 bacek moritz: yak... It's impossible
13:54 moritz it's still not clear to me how we should work with numeric types that don't fit into our usual type system
13:54 dalek r28779 | Whiteknight++ | gsoc_pdd09:
13:54 dalek : [gsoc_pdd09] miscellaneous changes from yesterday, including adding some useful macros and uses of the new object_lives function
13:54 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28779
13:54 pmichaud comparison.t fails because the numeric constants were switched to enums (that rakudo doesn't understand yet)
13:54 pmichaud I'll fix that shortly.
13:55 moritz is a hash a list of pairs? ie can I index it with postcircumfix:<[ ]>?
13:56 bacek Whiteknight: consider to use 'git-svn' and 'git add --interactive'. Then you can commit parts of file separately. And 'git svn dcommit' it to svn
13:56 pmichaud moritz:  I think it'd be %hash.pairs[...]
13:57 Whiteknight bacek, I'll look at it.
13:58 dalek r28780 | Whiteknight++ | gsoc_pdd09:
13:58 dalek : [gsoc_pdd09] update to trunk from r28740
13:58 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28780
14:01 Whiteknight bacek, git got a good tutorial?
14:02 bacek Whiteknight: http://www.kernel.org/pub/softwa​re/scm/git/docs/user-manual.html
14:02 shorten bacek's url is at http://xrl.us/hkcf9
14:03 Whiteknight thanks
14:03 moritz so far git resisted all my wishes to work as I intended
14:03 bacek http://www-cs-students.sta​nford.edu/~blynn/gitmagic/
14:03 bacek moritz: how so?
14:04 moritz bacek: I followed tutorials, and got errors that weren't covered in the tutorials
14:04 moritz bacek: I tried to read the manuals without understanding the internal data structures, and failed
14:04 moritz I modified something, and couldn't get it to discard that change. Subsequent git-pull's failed
14:04 moritz etc
14:05 bacek moritz: to discard changes you should use git-checkout
14:05 bacek It will checkout fresh copy from repo
14:06 moritz but why does it fail when I deleted a file in the working copy?
14:07 moritz anyway, I'll read that magic article. And afk
14:07 bacek moritz: what do you mean?
14:08 apeiron joined #parrot
14:09 bacek pmichaud: (about lazy list) we need ListIterator. Or discard 'mutating' grep/map/etc
14:11 dolmen joined #parrot
14:14 dalek r28781 | jkeenan++ | autoicu:
14:14 dalek : [configure] Refactor code into internal sub _handle_autodetect(), then test
14:14 dalek : it.  Test command-line option --icu-config=none.
14:14 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28781
14:16 moritz bacek: dunno, it just failed when I tried it
14:17 bacek moritz: failed what? git pull?
14:17 moritz git-checkout IIRC
14:19 bacek no way! :)
14:41 bacek afk #sleep
14:42 dolmen Cursor5?
14:44 dolmen purl, Cursor?
14:47 bacek joined #parrot
14:49 ambs joined #parrot
14:55 * DietCoke yawns.
15:00 dalek r28782 | jkeenan++ | autoicu:
15:00 dalek : [configure] Refactor some code inside runstep() into _handle_icuconfig_opt()
15:00 dalek : and test it.
15:00 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28782
15:11 * kid51 nudges DietCoke about ops_renum.mak :-)
15:14 dalek r28783 | jkeenan++ | autoicu:
15:14 dalek : [configure] Have variables distinguish between what use set as command-line
15:14 dalek : options and the current state of our attempt to configure those options.  Move
15:14 dalek : hard-coded list of ICU headers to data point in auto::icu object.
15:14 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28783
15:25 iblechbot joined #parrot
15:26 Andy joined #parrot
15:33 dalek r28784 | jkeenan++ | autoicu:
15:33 dalek : Take hard-coded string deep inside runstep() and make it a data point in
15:33 dalek : auto::icu object.  This should eliminate one if/else block.
15:33 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28784
15:36 Andy joined #parrot
15:49 dalek r28785 | pmichaud++ | trunk:
15:49 dalek : [rakudo]:
15:49 dalek : * spectest-progress update for 6-28:  75 files, 1072 passing, 17 failing
15:49 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28785
16:14 tetragon joined #parrot
16:15 PacoLinux joined #parrot
16:17 dalek r28786 | pmichaud++ | trunk:
16:17 dalek : [rakudo]:
16:17 dalek : * Add temporary implementation of Order::Same, Order::Increase,
16:17 dalek :   Order::Decrease
16:17 dalek : * t/spec/S03-operators/comparison.t passes again
16:17 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28786
16:45 teknomunk joined #parrot
17:18 dalek r28787 | moritz++ | trunk:
17:18 dalek : [rakudo] delete.t in spectest_regression.data is no longer "pure",
17:18 dalek : dolmen++ added new tests.
17:18 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28787
17:40 cjfields joined #parrot
17:46 japhb joined #parrot
17:58 dalek r28788 | pmichaud++ | trunk:
17:58 dalek : [rakudo]:
17:58 dalek : * Fix exclusive from ranges (1^..5)
17:58 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28788
18:08 dalek r28789 | pmichaud++ | trunk:
18:08 dalek : [rakudo]:
18:08 dalek : * Make sure Array.delete flattens its arguments in list context.
18:08 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28789
18:14 tewk_ non-generational GC is killing me.
18:14 tewk_ I'
18:15 dalek r28790 | fperrad++ | libs4php:
18:15 dalek : [php] implement parse_parameters()
18:15 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28790
18:16 tewk_ m able to parse and generate asts for fairly complicated C files (including system header files).
18:16 pmichaud nice!
18:17 tewk_ PGE + PCT + perl6 could eventually turn in to a pretty cool static analysis / refactoring tool.
18:18 tewk_ Unfortunately runtimes are growing exponential in the number of source lines :(
18:19 pmichaud it's probably due to the parsing
18:19 pmichaud that should improve in a few weeks.
18:22 tewk_ I think its gc because the ast tends to hang on to objects it allocates.
18:22 tewk_ Why would parsing get slower the longer the input?
18:22 dalek r28791 | pmichaud++ | trunk:
18:22 dalek : [rakudo]:
18:22 dalek : * Adjust <=>, cmp, leg to return Order::* values.
18:22 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28791
18:22 pmichaud because of things like     token stmt {   <a> | <b> | <c> | <d> | <e> }
18:22 pmichaud every additional statement adds 6 more options
18:22 pmichaud er, 5 more options
18:23 pmichaud although I guess that would be linear.
18:23 tewk_ Yeah that what I was about to say.
18:24 pmichaud I dunno, I just know that in the other things I've done with parsing, larger programs take much longer.  It may be due to increased memory allocation.
18:24 pmichaud (where "much longer" seems like "more than linear").
18:24 tewk_ Thats my bet.  Parsers and compilers tend to allocate and hold until the very end of the program.
18:25 pmichaud well, they don't have to -- the action methods could process and throw away parts of the subtree
18:25 pmichaud but I've been leaving that for a future optimization :-)
18:25 tewk_ The hanging on to objects behavior is not good for n^2 mark and sweep gc algorithms
18:26 tewk_ Yeah, but If your trying to do whole file analysis or optimization, you tend to want to hang on to the entire tree.
18:27 pmichaud yes, true.
18:53 cognominal I think there is a cut rule in S05, to let the parser forget information
19:04 rch joined #parrot
19:11 Theory joined #parrot
19:17 dalek r28792 | fperrad++ | libs4php:
19:17 dalek : [php] refactor parameters
19:17 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28792
19:29 Whiteknight joined #parrot
19:30 rch left #parrot
19:53 davidfetter purl, seen jhorwitz
19:53 davidfetter hrm
19:53 davidfetter anybody?
19:56 dalek r28793 | fperrad++ | libs4php:
19:56 dalek : [php] refactor optional parameters
19:56 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28793
20:01 Andy joined #parrot
20:04 Tene davidfetter: haven't seen him online since the 25th
20:04 davidfetter k
20:04 davidfetter thanks for the heads-up, Tene :)
20:04 Tene np
20:05 Tene I owe him a lolcode var-handling rewrite.
20:12 Andy I would like a Rakudo task.
20:12 Andy Is there something I can get assigned?
20:12 Andy Because I really want to start actually doing Perl 6 things.
20:12 Andy and don't know where to start.
20:12 Tene I usually start by looking at spectest failures.
20:13 Tene I think there are still a few "builtin not implemented"s left.
20:14 dalek r28794 | fperrad++ | libs4php:
20:14 dalek : [php] add caller in wrong_param_count
20:14 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28794
20:14 Tene You could also check RT.
20:14 Tene Lemme think what else I've seen people say needs to be done recently...
20:16 Andy The Perl 6, it is beyond me.
20:16 Andy I am afraid.
20:16 Andy Someone hold me.
20:21 Andy So we fully expect the spectest to fail under rakudo.
20:47 peepsalot joined #parrot
21:20 cotto_home seen barney
21:21 Patterner if you see purple dinosaurs, you should stop smoking those mushrooms...
21:24 cotto_home If I were smoking mushrooms, that'd probably be the effect I'd be hoping for.
21:24 Patterner you don't get much out, do you?
21:32 cognominal I needed google to figure out this one
21:34 Patterner amazing what you can learn on the webtubes.
21:36 dalek Christoph Otto | plumhead_renaming:
21:36 dalek link: http://www.perlfoundation.org/pa​rrot/index.cgi?plumhead_renaming
21:36 shorten dalek's url is at http://xrl.us/d2gpk
22:19 dalek r28795 | chromatic++ | trunk:
22:19 dalek : [IMCC] Fixed a double-free error introduced with :lexid handling in r28723.
22:19 dalek : Now all tests pass for me, including Rakudo and Pheme.
22:19 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28795
22:22 dalek r28796 | chromatic++ | trunk:
22:22 dalek : [Rakudo] Fixed SVN metadata.
22:22 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28796
22:29 dalek r28797 | chromatic++ | trunk:
22:29 dalek : [GC] Fixed #define indentation to match coding standards.
22:29 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28797
22:31 bacek joined #parrot
22:34 cjfields joined #parrot
22:43 kid51 joined #parrot
23:20 TiMBuS joined #parrot
23:50 dalek r28798 | jkeenan++ | autoicu:
23:50 dalek : Refactor additional code from runstep() to internal subroutines, then test
23:50 dalek : them.
23:50 dalek diff: http://www.parrotvm.org/svn​/parrot/revision?rev=28798

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

Parrot | source cross referenced