Camelia, the Perl 6 bug

IRC log for #parrot, 2008-12-28

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:00 dalek r34458 | jkeenan++ | trunk/src/jit (10 files):
00:00 dalek : Add copyright and SVN Id tag.
00:00 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34458
00:02 Andy joined #parrot
00:09 AndyA joined #parrot
00:39 rdice joined #parrot
00:52 cotto joined #parrot
01:10 dalek r34459 | pmichaud++ | branches:
01:10 dalek : Create new branch for rakudo variable and parameter refactoring.
01:10 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34459
01:13 tewk pmichaud: ping
01:13 pmichaud tewk: pong
01:13 tewk runtime/parrot/library/PGE/Perl6Grammar.pir:185 pushes and exception handler.
01:14 tewk s/and/an/
01:15 pmichaud yes...
01:15 tewk one seoncd
01:17 tewk I took out throwing exceptions for duplicate classes.
01:18 tewk Each duplicate class name gets its own unique typeid, but only the first one gets put into the classname->typeid hash.
01:18 pmichaud ouch.
01:18 pmichaud it shouldn't recreate a class that already exists, though.
01:19 pmichaud i.e., newclass ['Foo']   should still throw an exception if a class already exists with that identifier
01:19 pmichaud it's just we shouldn't use strings as the way to decide if one already exists
01:19 pmichaud in this case, the ['Foo'] namespace will already have a class attached to it, so an exception should be thrown.
01:20 pmichaud (and we shouldn't (re-)create the class.)
01:20 tewk ok, I'll go back to the drawing board.
01:20 pmichaud basically I'm saying that namespaces become our "global" class registry, instead of the existing class registry.
01:21 pmichaud The existing class registry is just a shortcut for mapping strings to classes.
01:21 tewk yeah I get that now.
01:22 dalek r34460 | pmichaud++ | branches/rvar/languages/perl6/src/parser (2 files):
01:22 dalek : [rakudo]:  Refactor the grammar to match STD.pm, and rip out
01:22 dalek : lots of variable code so we can start rebuilding it from scratch.
01:22 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34460
01:28 TiMBuS joined #parrot
01:34 dalek r34461 | pmichaud++ | branches/rvar/languages/perl6/src/parser:
01:34 dalek : [rakudo]:  Add namespace qualified vars and globals.
01:34 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34461
01:39 dalek r34462 | pmichaud++ | branches/rvar/languages/perl6/src/parser:
01:39 dalek : [rakudo]:  Restore special variables ($_, $/, $_).
01:39 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34462
02:03 jimmy joined #parrot
03:06 dalek r34463 | jkeenan++ | trunk/tools/dev:
03:06 dalek : Add copyright notice and SVN Id tag.
03:06 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34463
03:35 Andy joined #parrot
03:45 kid51 Does anyone know what is the plan to get the links on this page working properly:  http://www.parrot.org/docs/ ?
03:59 jimmy_ joined #parrot
04:00 jimmy__ joined #parrot
04:03 * kid51 must sleep
04:03 purl $kid51->sleep(8 * 3600);
04:03 elmex_ joined #parrot
04:56 jimmy joined #parrot
04:56 jimmy Synopses?
04:56 purl well, Synopses is http://perlcabal.org/syn/ or http://dev.perl.org/perl6/doc/synopsis.html
04:57 apeiron_ joined #parrot
05:00 Andy joined #parrot
05:00 TimToady no, Synopses is http://perlcabal.org/syn/
05:00 purl okay, TimToady.
05:50 dalek r34464 | pmichaud++ | branches/rvar/compilers/pct/src/PAST:
05:50 dalek : [pct]:  Allow piropsig to specify an argument as a return value.
05:50 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34464
05:51 dalek r34465 | pmichaud++ | branches/rvar/languages/perl6/src/parser (2 files):
05:51 dalek : [rakudo]:  More variable refactoring -- "my TypeName $x".
05:51 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34465
06:04 galf joined #parrot
06:04 pedr joined #parrot
06:04 pedr left #parrot
06:28 dalek r34466 | allison++ | branches/pdd22io_part3/src/io:
06:28 dalek : [pdd22io] Change 'write' function to call existing function that already makes
06:28 dalek : a method call on the filehandle object.
06:28 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34466
06:30 dalek r34467 | allison++ | branches/pdd22io_part3/src/pmc:
06:30 dalek : [pdd22io] Change several direct references to the 'pmc' variable to the
06:30 dalek : standard 'SELF' macro.
06:30 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34467
06:32 Ademan_ joined #parrot
06:33 dalek r34468 | petdance++ | trunk/src/pmc:
06:33 dalek : remove an unused arg in a static function
06:33 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34468
06:41 Theory joined #parrot
06:46 pedr joined #parrot
06:48 pedr left #parrot
07:02 dalek r34469 | allison++ | branches/pdd22io_part3/src (2 files):
07:02 dalek : [pdd22io] Invert the test for 'eof' so public interface calls a method on the
07:02 dalek : filehandle object.
07:02 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34469
07:07 Andy The cut & paste code, it makes me cry
07:21 dalek r34470 | petdance++ | trunk/lib/Parrot/Pmc2c (2 files):
07:21 dalek : hoisted the export macros into a PMC method
07:21 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34470
07:25 mberends joined #parrot
07:30 lathos
08:29 galf1 joined #parrot
08:33 iblechbot joined #parrot
09:27 kj joined #parrot
09:44 jimmy joined #parrot
09:54 dalek r34471 | kjs++ | trunk/compilers/pirc/new (2 files):
09:54 dalek : [pirc] refactoring of conversion function for invocation object to instruction sequences.
09:54 dalek : + layout stuff.
09:54 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34471
10:13 dalek r34472 | kjs++ | trunk/compilers/pirc/new:
10:13 dalek : [pirc] add and update function documentation.
10:13 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34472
10:30 dalek r34473 | kjs++ | trunk (6 files):
10:30 dalek : [pirc] refactor a bit of #defines, adding a new header file. + update MANIFEST.
10:30 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34473
10:48 dalek r34474 | kjs++ | trunk/compilers/pirc (7 files):
10:48 dalek : [pirc] fix tests and add function docs.
10:48 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34474
10:55 Zaba joined #parrot
11:09 barney joined #parrot
11:11 dalek r34475 | kjs++ | trunk/compilers/pirc/new (2 files):
11:11 dalek : [pirc] store the number of arguments and results in the invocation object.
11:11 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34475
11:16 Theory joined #parrot
11:25 dalek r34476 | fperrad++ | trunk/languages/lua:
11:25 dalek : [Lua]
11:25 dalek : - more credits from branches
11:25 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34476
11:58 jimmy flw: ping ?
11:58 purl I can't find ? in the DNS.
12:30 Zaba joined #parrot
12:47 dalek r34477 | rurban++ | branches/pdd30install_stage3/config (4 files):
12:47 dalek : 39742-installed-conflict
12:47 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34477
12:48 dalek r34478 | rurban++ | branches/pdd30install_stage3​/lib/Parrot/Configure/Step:
12:48 dalek : trac94-pod2man
12:48 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34478
12:49 dalek r34479 | rurban++ | branches/pdd30install_stage3/tools/dev:
12:49 dalek : 56544-install_files
12:49 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34479
12:51 dalek r34480 | rurban++ | branches/pdd30install_stage3 (3 files):
12:51 dalek : 57006-opengl-cyg
12:51 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34480
12:51 dalek r34481 | rurban++ | branches/pdd30install_stage3 (2 files):
12:51 dalek : 58034-config_args
12:51 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34481
13:02 dalek r34482 | rurban++ | branches/pdd30install_stage3 (48 files):
13:02 dalek : 56554-make-install-lang
13:02 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34482
13:03 dalek r34483 | rurban++ | branches/pdd30install_stage3/config/init/hints:
13:03 dalek : 56998-cygdll_versioning
13:03 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34483
13:04 dalek r34484 | rurban++ | branches/pdd30install_stage3:
13:04 dalek : 51944-README_cygwin
13:04 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34484
13:06 dalek r34485 | rurban++ | branches/pdd30install_stage3 (4 files):
13:06 dalek : 56996-fhs-runtime
13:06 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34485
13:12 idemal joined #parrot
13:12 dalek r34486 | bernhard++ | trunk/languages/pipp (3 files):
13:12 dalek : [Pipp] Start with support for closures.
13:13 dalek : Passing an argument to a closure is still broken.
13:13 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34486
13:24 kid51 joined #parrot
13:27 jimmy good morining kid51 :)
13:29 jimmy does php 5.2 surport  closure?
13:29 kid51 jimmy good morning to you
13:29 * kid51 knows nothing about php 5.2
13:29 jimmy barney: does php 5.2 surport  closure?
13:31 jimmy kid51->hasSleep(9*3600);
13:31 barney jimmy: No, only 5.3
13:32 barney http://www.ibm.com/developerworks/op​ensource/library/os-php-5.3new2/inde​x.html?S_TACT=105AGX44&S_CMP=EDU
13:32 shorten barney's url is at http://xrl.us/beatxo
13:32 ffwonko joined #parrot
13:32 jimmy oops, pipp exceeds php5.2 now.
13:34 kid51 jimmy:  In your patch in TT#95, you have this:  if (n < NUM_REGISTERS & !ru[typ].reg_count[n]++)
13:34 kid51 Why '&' and not '&&'?
13:34 jimmy ah
13:37 * jimmy makes a mistake
13:40 dalek r34487 | kjs++ | trunk/compilers/pirc/new (3 files):
13:40 dalek : [pirc] work on parameter passing bytecode.
13:40 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34487
13:43 jimmy kid51: do I need to reupload it ?
13:44 kid51 Yes, you should, with a title indicating correction (assuming you can't delete a patch once it's uploaded).
13:45 * kid51 doesn't know whether Trac permits deletion of attachments and, if so, by whom
13:47 kid51 jimmy:  The changes you propose for src/pmc/role.pmc seem to affect horizontal whitespace only -- correct?  I ask because the alignment doesn't look any better (or worse) than before.
13:48 TiMBuS joined #parrot
13:48 jimmy kid51: yeah? do you look it with firefox?
13:49 dalek r34488 | bernhard++ | trunk/languages/pipp (2 files):
13:49 dalek : [Pipp] Fix passing of arguments to anonymous functions.
13:49 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34488
13:51 alvar joined #parrot
13:52 kid51 I've looked at it two ways:  (1) Clicked on "Unified", which displays in monospace font; (2) downloaded attachment and viewed in in text reader with monospace font.
13:52 jimmy kid51: I used vim
13:54 jimmy when I used firefox, it is really worse than before.
13:54 kid51 Well, now I too am looking at it with vim
13:56 kid51 Oh, I see, you're trying to align on the assignment operators.
13:56 jimmy yeah
14:06 contingencyplan joined #parrot
14:14 jimmy kid51: seems that can't create the patch with a title indicating correction, because the typo is not in svn.
14:16 dalek r34489 | kjs++ | trunk/compilers/pirc/new (5 files):
14:16 dalek : [pirc] fix parameter instruction generation.
14:16 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34489
14:18 kid51 jimmy:  I'm not sure what "typo" you're referring to.  We simply need to know that this is a different patch from your first submission.
14:18 kid51 "this" meaning a patch that corrects '&'
14:19 pmichaud jimmy: create a completely new patch with the correction in it, not a patch to your patch
14:19 jimmy i think i should create another patch to replace it.
14:19 pmichaud jimmy: yes.
14:19 pmichaud basically say "the old patch is incorrect, here's the correct one"
14:20 jimmy i can repalce it at all.
14:22 Whiteknight joined #parrot
14:24 jimmy kid51: uploaded
14:24 jimmy thanks pmichaud
14:27 * kid51 will look at it once he finishes his Xmas shopping ;-)
14:34 dalek r34490 | pmichaud++ | trunk/languages/perl6/docs:
14:34 dalek : [rakudo]: spectest-progress.csv update: 264 files, 5913 passing, 0 failing
14:34 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34490
14:38 kj jonathan: ping
14:40 PacoLinux joined #parrot
14:41 jimmy pmichaud++, Is there roadmap about perl6 ?
14:41 pmichaud there's an outdated one in languages/perl6/ROADMAP
14:42 pmichaud which I need to update.
14:42 pmichaud when I get tired of writing code today I'll probably do that.
14:46 jimmy I have tired ;)
14:49 * jimmy is going to sleep now.
14:52 tewk pmichaud: class registry is coming along. fixing hll_maps right now.
14:52 tewk its fair to say that you have to use [ ] syntax in hll maps as well, right?
14:53 tewk or, if you don't use [] I could default to checking the current namespace first before checking the global registry.
15:01 dalek r34491 | bernhard++ | trunk/languages/pipp/t/php:
15:01 dalek : [Pipp] add two test cases, where a function returns an anonymous sub
15:01 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34491
15:04 dalek r34492 | kjs++ | trunk/compilers/pirc/new (4 files):
15:04 dalek : [pirc] fix emission of bytecode for params. no :flags yet, only types and reg numbers.
15:04 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34492
15:05 kid51 nopaste?
15:05 clunker3 http://pasta.test-smoke.org/ or http://paste.husk.org/ or http://nopaste.snit.ch:8001/ or http://rafb.net/paste or http://poundperl.pastebin.com/ or http://paste.scsys.co.uk/
15:05 purl nopaste is at http://nopaste.snit.ch/ (ask TonyC for new channels) or http://rafb.net/paste or http://poundperl.pastebin.com/ or http://paste.scsys.co.uk/ or App::Nopaste or tools/dev/nopaste.pl or at http://www.extpaste.com/ or http://paste.scsys.co.uk (for #catalyst, #dbix-class, #moose  and others)
15:14 nopaste "kid51" at 68.237.11.157 pasted "Big increase in warnings during 'make', mainly from Parrot::Pmc2c::PCCMETHOD" (570 lines) at http://nopaste.snit.ch/15102
15:15 pmichaud tewk: (hll maps)  do you mean the .HLL_map statement?
15:15 pmichaud because Rakudo probably won't be using it.  :-)
15:16 pmichaud but yes, I would think that .HLL_map would keep the same syntax
15:16 pmichaud (as the other opcodes)
15:17 dalek r34493 | jkeenan++ | trunk/src (3 files):
15:17 dalek : Applying cage-cleaning patch submitted in Trac #95.  jimmy++.
15:17 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34493
15:20 apeiron joined #parrot
15:26 kid51 In the paste I just submitted, the line in Parrot::Pmc2c::PCCMETHOD most frequently cited in the warnings is line 436:      if (PObj_get_FLAGS(_ccont) & SUB_FLAG_TAILCALL) {
15:27 kid51 Why '&' here rather than '&&'?
15:27 pmichaud we're testing a flag.
15:27 pmichaud not two boolean conditions.
15:27 kid51 okay.  so why all the warnings?
15:27 pmichaud That said, this is probably not the actual error line.
15:28 pmichaud the actual error line would have Parrot_push_context on it.
15:29 pmichaud I'm guessing that "make headerizer" changed the signature of Parrot_push_context somewhere and these warnings started showing up.
15:29 kj ISTR that there was some tweaking with warnings flags recently
15:29 tewk Andy said something about more warnings
15:29 nopaste "kid51" at 68.237.11.157 pasted "Recent changes in lib/Parrot/Pmc2c/PCCMETHOD.pm" (24 lines) at http://nopaste.snit.ch/15103
15:30 kid51 What is purpose/function of 'make headerizer'?
15:30 pmichaud it adds warning flags to functions
15:31 pmichaud to catch brainos such as sending NULL as an argument to something that can't possibly take NULL as an argument
15:31 pmichaud aha
15:31 pmichaud _n_regs_used was changed to have a const modifer
15:31 pmichaud *modifier
15:31 pmichaud that's why Parrot_push_context is complaining.
15:31 nopaste "kid51" at 68.237.11.157 pasted "Part of PCCMETHOD.pm cited in 'make' warnings" (7 lines) at http://nopaste.snit.ch/15104
15:32 pmichaud sometimes the compiler gets confused about the actual location of errors in generated code
15:33 kid51 Was the change in _n_regs_used necessary?   Excessive?
15:33 pmichaud I don't know -- checking.
15:34 pmichaud but this is an excellent example of why make headerizer is helpful (more)
15:34 pmichaud src/gc/register.c:345 defines the Parrot_push_context function:
15:34 pmichaud Parrot_push_context(PARROT_INTERP, ARGMOD(INTVAL *n_regs_used))
15:34 pmichaud that "ARGMOD" says that that the second parameter might be modified
15:35 pmichaud and so the compiler is correctly complaining that we're sending a const value to a function that might try to modify it
15:35 pmichaud so, let's see if it really does try to modify it.  :-)
15:36 pmichaud and I think the answer is "no"
15:37 dalek r34494 | bernhard++ | trunk/languages/pipp (3 files):
15:37 dalek : [Pipp] try to support closures
15:37 dalek : Add one working and one failing test case.
15:37 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34494
15:39 pmichaud making test now
15:41 dalek r34495 | kjs++ | trunk/compilers/pirc/new (5 files):
15:41 dalek : [pirc] clean up documentation.
15:42 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34495
15:44 pmichaud oops, that wasn't it.
15:44 pmichaud trying again.
15:49 pmichaud r34496 should fix it.
15:49 dalek r34496 | pmichaud++ | trunk (2 files):
15:49 dalek : [core]: eliminate "Parrot_push_context discards qualifiers" warning (kid51++)
15:49 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34496
15:57 jhorwitz joined #parrot
15:59 kid51 Yes, that does silence the warnings.  Thanks, pmichaud.
16:00 Infinoid any IMCC experts around?
16:01 Zaba joined #parrot
16:01 pmichaud If there are, can we shoot them?  ;-)
16:03 nopaste "Infinoid" at 75.5.247.153 pasted "The call to IMCC_info() seems to be the only place output_file is used in this function." (22 lines) at http://nopaste.snit.ch/15105
16:04 jimmy joined #parrot
16:05 Infinoid the headerizer nonnull-assert hack has found several cases so far of NULLOK arguments not being marked as such, but I could use a little sanity checking on this one.
16:05 jimmy kid51: you ci the wrong patch
16:05 kid51 Can you paste error?
16:06 jimmy kid51: still was &, not &&
16:06 * kid51 checks
16:07 kid51 Correct
16:07 kid51 I will fix in a minute.
16:08 jimmy and list.c seems that you have modified it
16:11 kid51 I believe that the only change I made in list.c was to correct a spelling error in a comment:  'wee' -> 'we'
16:12 jimmy kid51++
16:12 kid51 ... which was not your error; it was already there.
16:12 jimmy yep
16:16 kid51 rebuilt successfully with && -- now running 'make test'
16:16 masak joined #parrot
16:17 iblechbot joined #parrot
16:20 kj Infinoid: I have some knowledge of IMCC
16:22 Infinoid kj: does http://nopaste.snit.ch/15105 look sane?
16:22 Infinoid I am wondering if the condition on obj_file is a typo.
16:23 kj I 'll have to look at the file itself, to check the context
16:23 kj 1 moment
16:23 kj no I think it's correct
16:24 Infinoid ok, thanks.  the logic is a little weird there
16:24 kj i'll check, though. I suspect it's correct, because obj_file is an integer, and somewhere (I think) a flag is set whether an output file is requested
16:24 Infinoid that would make sense, if obj_file is set whenever output_file is non-NULL
16:25 Infinoid I've been putting in asserts to validate nonnull assumptions about input arguments, and that function is being called with a NULL output_file.
16:26 Infinoid I'm sure you're right, I'll back out the condition part of my patch
16:26 Infinoid kj++
16:26 kj both obj_file and output_file are used only in the IMCC_info call
16:27 kj so both are just there for some verbosity
16:27 Infinoid makes sense, thanks
16:27 kj np
16:29 dalek r34497 | jkeenan++ | trunk/src:
16:29 dalek : Fix errors in r34490; jimmy++.
16:29 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34497
16:44 dalek r34498 | kjs++ | trunk/compilers/pirc/new (2 files):
16:44 dalek : [pirc] calculate flags for arguments.
16:44 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34498
16:47 jhorwitz joined #parrot
16:48 particle1 joined #parrot
16:48 particle1 left #parrot
16:49 tetragon joined #parrot
16:51 dalek r34499 | bernhard++ | trunk/languages/pipp/src (2 files):
16:51 dalek : [Pipp] small cleanup in grammar and in php_array.pir
16:51 dalek : Courtesy of Daniel Keane.
16:51 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34499
16:54 Zaba joined #parrot
16:59 dalek r34500 | kjs++ | trunk/compilers/pirc/new:
16:59 dalek : [pirc] calculate PCC flags for parameters/result-parameters.
16:59 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34500
17:01 dalek r34501 | pmichaud++ | branches/rvar/languages/perl6/src/parser:
17:01 dalek : [rakudo]:  Put a type property on a container only if there is one.
17:01 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34501
17:03 Zaba joined #parrot
17:10 dalek r34502 | rurban++ | branches/pdd30install_stage3 (58 files):
17:10 dalek :
17:10 dalek : - [cage] VERSION 0.8.2
17:10 dalek : - [install] 57548-CONDITIONED_LINE_enh
17:10 dalek : - [install] finish leftovers from 56554-make-install-lang
17:10 dalek : - [Zode] prepare ZCode to be installable (single file Zcode.pbc without z3main)
17:10 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34502
17:22 Zaba joined #parrot
17:25 rurban joined #parrot
17:27 rurban branch pdd30install_stage3 is now ready. I've splitted the svn commits into the existing patches. What is not working yet: befunge(syntax error), lua (mmd error), ZCode (single-file merge missing)
17:27 rurban I've added a new feature: make -C languages reallyall (which is for d in *; do test -e $d/Makefile && make -C $d all; done)
17:29 dngor joined #parrot
17:35 workbench joined #parrot
17:35 Andy joined #parrot
17:39 TimToady joined #parrot
17:44 nopaste "kid51" at 68.237.11.157 pasted "diff of pdd30install_stage3 branch vs trunk" (10470 lines) at http://nopaste.snit.ch/15106
17:44 kid51 rurban++ for work on branch
17:44 kid51 Now for some questions ;-)
17:45 kj is anyone aware of some documentation about the why and how of using INTVAL/FLOATVAL instead of just "int" or "double"?
17:46 kid51 You've added 'auto::pod2man' to list of configuration steps.  But I don't see any config/auto/pod2man.pm.  Nor do I see any tests for such a class, i.e., t/steps/auto_pod2man-01.t.  Have you run 'perl Configure.pl --test' with this branch?
17:46 rurban sorry; I've forgot to add. will do.
17:48 dalek r34503 | rurban++ | branches/pdd30install_stage3 (3 files):
17:48 dalek : [cage] add missing pod2man configure files
17:48 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34503
17:49 rurban kid51: The overall diff is huge. It's simplier to review the single commits I did. They match the tickets
17:51 kid51 Yes, it is huge.  But I was trying to get a sense of what this branch was all about and whether it affected the configuration system (which it did).
17:51 dalek r34504 | pmichaud++ | branches/rvar/compilers/pct/src/PAST:
17:51 dalek : [pct]:  Add :pasttype('stmts').
17:51 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34504
17:52 * kid51 appreciates t/steps/auto_pod2man-01.t is provided immediately along with config/auto/pod2man.pm.
17:55 dalek r34505 | rurban++ | branches/pdd30install_stage​3/languages/dotnet/config:
17:55 dalek : [install] dotnet leftovers for 57548-CONDITIONED_LINE_enh.patch
17:55 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34505
17:56 rurban Esp.  57548-CONDITIONED_LINE_enh changes the configure system totally
17:56 rurban ascii2man is also needed but only the languages for now. So I didn't add it.
17:56 rurban only FOR some of the languages
17:59 dalek r34506 | bernhard++ | trunk/languages/pipp/src/pct (2 files):
17:59 dalek : [Pipp] Rename rule arguments to argument_list
17:59 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34506
18:02 dalek r34507 | kjs++ | trunk/compilers/pirc/new (2 files):
18:02 dalek : [pirc] Allow .call "foo" in a .begin/.end_call sequence. imcc does not allow this, but I think it adds to consistency to allow this.
18:02 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34507
18:06 dalek r34508 | rurban++ | branches/pdd30install_stage3/config/gen/makefiles:
18:06 dalek : [cage]  trailing whitespace in lines 49,76
18:06 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34508
18:10 kid51 rurban:  When do you expect to be merging this branch into trunk?
18:12 rurban I could merge it immediately, but allison or some other reviewer should give the okay.
18:12 rurban The longer we wait the more work it is.
18:13 rurban And the WMLScript, forth and Zcode maintainers should speak up also.
18:17 kj I don't think zcode is still being maintained
18:19 rurban Looks like so, yes.
18:19 rurban befunge being the worst.
18:22 dalek r34509 | rurban++ | branches/pdd30install_stage3/t/steps:
18:22 dalek : [cage] fix test number
18:22 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34509
18:22 dalek r34510 | rurban++ | branches/pdd30install_stage3/langu​ages/ecmascript/config/makefiles:
18:22 dalek : [install] workaround HLLNAME ne dirname
18:23 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34510
18:24 rurban this pdd30install_stage3 fixes the problem building with already installed (different) shared libparrot. But testing not yet.
18:37 dalek r34511 | bernhard++ | trunk/languages/perl6/src/builtins:
18:37 dalek : [doc] Change the filename in PDO to the real filename
18:37 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34511
18:45 dalek r34512 | kjs++ | trunk/compilers/pirc/new (5 files):
18:45 dalek : [pirc] fix -o commandline option, so it works with -H and -E.
18:45 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34512
18:51 dalek r34513 | pmichaud++ | branches/rvar/languages/perl6/src/parser (2 files):
18:51 dalek : [rakudo]:  Rebuild parts of parameter and signature handling.
18:51 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34513
18:52 alvar joined #parrot
18:54 apeiron_ joined #parrot
19:03 kid51_afk rurban:  I did coverage analysis on your branch.  It suggests that in config/auto/pod2man.pm, sub _handle_no_pod2man() is never called and is therefore deletable:  http://thenceforward.net/parrot/coverage/c​onfigure-build/config-auto-pod2man-pm.html
19:03 shorten kid51_afk's url is at http://xrl.us/beauoj
19:04 * kid51 prepares smolder report on rurban's branch
19:05 rurban kid51: there's still a logic error with the prefix logic which is not yet tested.
19:06 rurban I'll write a test for the installables now
19:07 kid51 k
19:14 dalek r34514 | bernhard++ | trunk/src:
19:14 dalek : [codingstd] satisfy c_parens.t
19:14 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34514
19:16 dalek r34515 | bernhard++ | trunk (3 files):
19:16 dalek : [Pipp] Add support for $x = ( $y = 2 ) * 3;
19:16 dalek : as a special case.
19:17 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34515
19:27 dalek r34516 | kjs++ | trunk/compilers/pirc:
19:27 dalek : [pirc] update the todo.
19:27 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34516
19:43 kid51 Here's a Smolder report on rurban's branch:  http://smolder.plusthree.com/app/pu​blic_projects/report_details/10133
19:44 shorten kid51's url is at http://xrl.us/beauqw
19:50 rurban_ joined #parrot
20:15 davidfetter joined #parrot
20:16 Zaba joined #parrot
20:35 ruoso_ joined #parrot
20:42 kj joined #parrot
20:52 Zaba joined #parrot
21:02 Zaba joined #parrot
21:06 dalek r34517 | pmichaud++ | branches/rvar/compilers/pct/src/PAST:
21:06 dalek : [pct]:  If a PAST::Var register node doesn't provide a name, generate one.
21:06 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34517
21:07 dalek r34518 | pmichaud++ | branches/rvar/languages/perl6/src (3 files):
21:07 dalek : [pct]: More updates to parameters -- array parameters now work.
21:08 dalek : * Build and attach Signature objects to blocks.
21:08 dalek : * Add Array() and Scalar() coercion functions.
21:08 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34518
21:08 ruoso joined #parrot
21:20 dalek r34519 | kjs++ | trunk/compilers/pirc/new (3 files):
21:20 dalek : [pirc] fix a small bug; the output options is a bit messy.
21:20 dalek : + implement add_const_key.
21:20 dalek : + start of build_key, but imcc's version very complex to copy from.
21:20 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34519
21:24 alvar joined #parrot
21:25 tewk .HLL_map is close to useless, the runtime version makes a lot more sense.
21:28 pmichaud agreed.
21:28 pmichaud .HLL_map does help if your things to be mapped are all PMCs, though.
21:31 kj tewk: what's the runtime version?
21:33 tewk it looks like the parrotinterpreter.pmc has a hll_map method
21:34 pmichaud it does
21:34 donaldh joined #parrot
21:34 pmichaud we added that in because .HLL_map doesn't help much for classes created at runtime.
21:34 PacoLinux joined #parrot
21:35 tewk Is the pir directive useful? Can you make a class at compile time? other that :immediate?
21:35 pmichaud I wouldn't eliminate the pir directive.
21:35 pmichaud It's useful if we know that a class already exists.
21:35 pmichaud (at the time the code is loaded)
21:36 pmichaud but yes, the .hll_map method seems much more generally useful.
21:53 dalek r34520 | kjs++ | trunk/compilers/pirc/new:
21:53 dalek : [pirc] refactoring + updates for emitting bytecode for keys. Not complete yet, and generated bc doesn't work yet, although it seems correct.
21:53 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34520
21:59 chromatic joined #parrot
22:03 Zaba joined #parrot
22:07 kid51 joined #parrot
22:11 Zaba_ joined #parrot
22:13 dalek r34521 | allison++ | branches/pdd22io_part3/src/pmc:
22:13 dalek : [pdd22io] Don't calculate the integer value of flags twice.
22:13 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34521
22:14 dalek r34522 | allison++ | branches/pdd22io_part3 (2 files):
22:14 dalek : [pdd22io] Add StringHandle PMC to test polymorphism in I/O subsystem, and for
22:14 dalek : use in I/O mocking.
22:14 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34522
22:21 dalek r34523 | pmichaud++ | branches/rvar/languages/perl6/src (3 files):
22:21 dalek : [rakudo]:  Add type checking to scalar parameters.
22:21 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34523
22:23 TiMBuS joined #parrot
22:27 dalek r34524 | kjs++ | trunk/compilers/pirc/new (3 files):
22:27 dalek : [pirc] start with generate_multi_signature function.
22:27 dalek : + some brackets.
22:27 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34524
22:35 dalek r34525 | infinoid++ | trunk/compilers/imcc:
22:35 dalek : [IMCC] Add NULLOK to a couple of function arguments.
22:35 dalek : These arguments can reasonably be NULL (and are sometimes NULL during the
22:35 dalek : course of normal operation), yet didn't have the NULLOK flag set.
22:35 dalek : (Found by the headerizer NONNULL assert() patch.)
22:35 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34525
22:37 dalek r34526 | kjs++ | trunk/compilers/pirc/new:
22:37 dalek : [pirc] add bits from imcc/pbc.c to pirc/bcgen.c for lex_info creation.
22:37 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34526
22:38 kid51 joined #parrot
22:45 kid51 joined #parrot
22:48 dalek r34527 | kjs++ | trunk/compilers/pirc/new:
22:48 dalek : [pirc] add comments, and optimize string creation slightly
22:48 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=34527
22:51 kid51 joined #parrot
23:14 Zaba joined #parrot
23:35 * Coke kicks off a build on his old and underpowered *bsd box that hasn't been turned on in ... a long time.
23:54 Coke slooooow.
23:55 Coke did the proposed changes I saw to 'newclass' in scrollback go in?
23:55 pmichaud I think they're still proposed, or else they're in a branch.
23:57 Coke k
23:57 Coke just wanted to retest some hanging tcl spec tests once they go in.

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

Parrot | source cross referenced