Camelia, the Perl 6 bug

IRC log for #parrot, 2009-01-25

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:09 AndyA joined #parrot
00:29 contingencyplan joined #parrot
00:30 skv joined #parrot
00:42 dalek r35979 | Whiteknight++ | branches:
00:42 dalek : branch hasn't been used yet, and is getting stale. We'll create a different branch for this later when we're ready
00:42 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35979
00:59 Whiteknight joined #parrot
01:03 Coke You can't ping an http url.
01:05 Coke Whiteknight: do you have a sample using the interpreter .hll_map method similar to my original posting using .hll_map ?
01:05 Whiteknight oh yeah, let me post it to the ticket
01:07 Whiteknight ETRACTOOSLOW
01:07 Coke no joke.
01:07 Coke back now.
01:08 Whiteknight ping trac.parrot.org
01:08 purl 10 packets transmitted, 10 received, 0% packet loss, time 8998ms, rtt min/avg/max/mdev = 78.817/80.015/88.549/2.861 ms
01:08 Whiteknight ...goood
01:08 NotFound Coke: I nopasted it some hours ago
01:09 NotFound Coke:  http://nopaste.snit.ch/15412
01:09 Coke NotFound: thanks.
01:09 NotFound Uh, no...
01:09 NotFound Ah, yes.
01:09 Coke =-)
01:09 Coke Whiteknight: please still add it to the ticket. =-)
01:10 NotFound Is late night in CET
01:11 Whiteknight I just did add it to the ticket
01:11 Whiteknight NotFound++
01:11 Whiteknight karma whiteknight
01:11 purl whiteknight has karma of 614
01:11 Whiteknight karma NotFound
01:11 purl notfound has karma of 64
01:12 NotFound karma numbers are also always even? ;)
01:12 purl numbers are also always even? ;) has neutral karma
01:12 Whiteknight (my wife wanted me to explain what karma was, so I showed her)
01:12 Whiteknight purl, you're retarded
01:12 purl Whiteknight: i'm not following you...
01:12 Whiteknight purl, whiteknight?
01:12 purl whiteknight is really looking forward to redoiing all the calling conventions nonsense or not bot savvy
01:12 NotFound purl: do you use twitter?
01:12 purl notfound: i don't know
01:12 Whiteknight purl forget whiteknight
01:12 purl Whiteknight: I forgot whiteknight
01:13 Whiteknight purl whiteknight is mailto:wknight8111@gmail.com
01:13 purl OK, Whiteknight.
01:13 Whiteknight purl, whiteknight is also the grand master funk
01:13 purl okay, Whiteknight.
01:13 Whiteknight :)
01:13 NotFound karma julianalbo
01:13 purl julianalbo has karma of 314
01:13 NotFound I'm jealous of myself X-)
01:14 Whiteknight karma wknight8111
01:14 purl wknight8111 has karma of 13
01:14 Whiteknight whiteknight?
01:14 purl whiteknight is mailto:wknight8111@gmail.com or the grand master funk
01:14 NotFound julianalbo?
01:14 Whiteknight NotFound?
01:14 purl well, NotFound is julian albo
01:15 Whiteknight very nice purl, very helpful
01:15 NotFound purl: julianalbo is NotFound
01:15 purl OK, NotFound.
01:15 NotFound purl: forget NotFound
01:15 purl NotFound: I forgot notfound
01:15 NotFound purl: NotFound is Julián Albo <julian.notfound@gmail.com>
01:15 purl OK, NotFound.
01:16 NotFound Good bot, take a cookie
01:16 purl :)
01:16 Whiteknight botsnack
01:16 purl :)
01:28 * Coke is failing to make the small sample from notfound work for tcl.
01:32 NotFound Coke: it fails, or just not work as expected?
01:34 Coke it seems to have no effect.
01:34 Coke of course, I've got load_bytecode, and .includes before getting to this bit.
01:35 * Coke is trying to complicate the example to figure out where it's failing.
01:35 GeJ Good morning everyone
01:37 Coke NotFound: found it.
01:38 Coke ... no i didn't. still checking
01:39 Coke NotFound: probably something stupid on my end. will try again in a bit.
01:42 NotFound How can I call the parent vtable from an vtable override?
01:43 Coke NotFound: moment.
01:44 Coke NotFound: http://code.google.com/p/partcl/source/​browse/trunk/src/class/tclconst.pir#250
01:44 shorten Coke's url is at http://xrl.us/bea6et
01:44 Coke but it doesn't work for all vtable entries.
01:47 NotFound Thanks. The example wroks with a more complex RPA class,
01:49 NotFound An overrided get_pmc_keyed_int is called from joe
02:03 kid51 joined #parrot
02:11 NotFound Coke: that's funny, the keyed getattribute does the same as not keyed. Both class and pmcproxy use default vtable get_attr_keyed
02:14 dalek r35980 | allison++ | trunk/docs/resources:
02:14 dalek : [doc] Clean up line spacing in generated HTML docs.
02:14 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35980
02:17 NotFound Ummm... no, the object pmc is the implied, but the result is the same :?
02:17 NotFound Too late, going to sleep
02:51 ask_ joined #parrot
03:17 skv joined #parrot
03:32 skv joined #parrot
03:40 dalek r35981 | allison++ | trunk/lib/Parrot/Docs (2 files):
03:40 dalek : [doc] Remove links to page top from generated HTML docs.
03:40 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35981
03:59 ask_ joined #parrot
04:04 dalek r35982 | cotto++ | trunk/t/tools/pmc2cutils:
04:04 dalek : [t] force some pmc2c tests generate .dump files (TT #219)
04:04 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35982
04:27 dalek r35983 | allison++ | trunk/docs:
04:27 dalek : [doc] Cleaning up Parrot release history.
04:27 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35983
05:04 Coke NotFound: found the problem; it's getting the tcllist, but it's not got any of the vtable overrides I've defined.
05:05 Coke almost like it took a snapshot of the class when I registered it.
05:09 Coke ... which is probably because chromatic just fixed parrot to distinguish between classes created in different HLLs.
05:12 Coke which makes doing the mapping much harder now.
05:12 Coke I want TclList to be in the HLL_map; I don't want to expose the TclList namespace to the user.
05:15 Coke (with it exposed, though, I'm finally passing tests again. NotFound++)
05:17 Coke (not all the tests, though. ah well.)
05:31 Coke whee, segfaults.
05:31 purl No whammies!
05:32 skv joined #parrot
05:35 Coke message chromatic regarding converting a tcl PMC to a PIR subclass; the primary test file for procesing this PMC runs in 3.3s before the conversion. 3.4s after; so, no speedup (again). I can't believe I let myself try to optimize for speed again without a profiler.
05:35 purl Message for chromatic stored.
06:12 ask_ joined #parrot
06:13 coke_reboot
06:17 DietCoke joined #parrot
06:35 HG` joined #parrot
07:22 ChrisDavaz joined #parrot
07:31 dalek r35984 | cotto++ | trunk/t/tools/pmc2cutils:
07:31 dalek : [t] fix the remaining failing tests caused by implementing ATTR inheritance
07:31 dalek : This closes TT #219.
07:31 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35984
07:34 cotto alright.  Now to get back to work changing important code.
07:40 dalek r35985 | cotto++ | trunk/t/tools/pmc2cutils:
07:40 dalek : [t] undo some accidentally committed code
07:40 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35985
07:41 cotto or maybe now
07:41 purl maybe now is a good time to tap into their phone line.
07:41 cotto purl++ #leaving much to the imagination
08:08 cotto What's be a good name for the place where an array stores its elements?
08:09 cotto s be/d be/
08:12 alvar joined #parrot
08:25 cotto items?
08:25 purl rumour has it items is around, regardless of what the class was named
08:25 cotto meh
09:08 clunker3 joined #parrot
09:13 ChrisDavaz joined #parrot
09:16 skv joined #parrot
09:20 cotto (VTABLE functions)++
09:37 Zaba joined #parrot
09:55 barney joined #parrot
10:08 iblechbot joined #parrot
10:11 bacek joined #parrot
10:19 Gerd joined #parrot
10:23 Gerd left #parrot
10:36 dalek r35986 | bernhard++ | trunk/src:
10:36 dalek : [codingstd] c_parens.t and linelength.t
10:36 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35986
10:38 dalek r35987 | bernhard++ | trunk (3 files):
10:38 dalek : [lazy-k] Add a Configure.pl
10:38 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35987
10:52 dalek r35988 | bernhard++ | trunk/languages/unlambda/config/makefiles:
10:52 dalek : [unlambda] update 'help' target
10:52 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35988
10:52 skv joined #parrot
10:53 dalek r35989 | bernhard++ | trunk (5 files):
10:53 dalek : [lazy-k] Generate lazy.pbc and use it for testing
10:53 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35989
10:59 cotto I love the smell of segfaults in the morning.
11:34 dalek r35990 | bernhard++ | trunk/languages (4 files):
11:34 dalek : [lazy-k] give up on unified languages testing for lazy-k, run tests in hll dir
11:34 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35990
11:34 alvar joined #parrot
11:37 dalek r35991 | cotto++ | trunk/src/pmc (2 files):
11:37 dalek : [pmc] get rid of UnionVal in *FloatArray
11:37 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35991
11:38 dalek r35992 | bernhard++ | trunk (10 files):
11:38 dalek : [lazy-k] Put the example scripts into the new dir 'examples'
11:38 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35992
11:40 dalek r35993 | bernhard++ | trunk (2 files):
11:40 dalek : [lazy-k] test.sh is not needed
11:40 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35993
11:51 masak joined #parrot
11:56 skv joined #parrot
12:00 Zaba joined #parrot
12:01 dalek r35994 | bernhard++ | trunk (5 files):
12:01 dalek : [lazy-k] Merge t/calc.t and t/cat.t into t/examples.t.
12:01 dalek : Bring lazy-k back into unified languages testing.
12:01 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35994
12:09 namenlos joined #parrot
13:24 masak pmichaud: [perl #62704] and [perl #62730] are showstoppers for me right now.
13:32 jan joined #parrot
13:32 skv joined #parrot
14:16 Whiteknight joined #parrot
14:21 dalek r35995 | julianalbo++ | trunk/examples/pir:
14:21 dalek : pirric: change internal method name conventions, add NEW instruction
14:21 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35995
14:23 Zaba joined #parrot
14:54 kid51 joined #parrot
14:56 bkuhn joined #parrot
14:56 skv joined #parrot
15:06 Whiteknight rakudo: my Int $x = 5; $x.perl.say;
15:07 polyglotbot OUTPUT[5␤]
15:07 Whiteknight rakudo: my @x = (1,2,3); @x.perl.say;
15:07 polyglotbot OUTPUT[[1, 2, 3]␤]
15:07 Whiteknight rakudo: my @x = [1,2,3]; @x.perl.say;
15:07 polyglotbot OUTPUT[[[1, 2, 3]]␤]
15:07 Whiteknight rakudo: my @x = {1,2,3}; @x.perl.say;
15:07 polyglotbot OUTPUT[[{ ... }]␤]
15:07 Whiteknight rakudo: my %x = {:first(1),:second(2),:third(3)}; %x.perl.say;
15:07 polyglotbot OUTPUT[{"first" => 1, "second" => 2, "third" => 3}␤]
15:08 Whiteknight rakudo: my %x = (:first(1),:second(2),:third(3)); %x.perl.say;
15:08 polyglotbot OUTPUT[{"first" => 1, "second" => 2, "third" => 3}␤]
15:08 Whiteknight rakudo: my %x = [:first(1),:second(2),:third(3)]; %x.perl.say;
15:08 polyglotbot OUTPUT[Odd number of elements found where hash expected␤current instr.: 'parrot;Perl6Hash;!STORE' pc 7488 (src/classes/Hash.pir:154)␤called from Sub '_block14' pc 141 (EVAL_15:51)␤called from Sub '!UNIT_START' pc 18152 (src/builtins/guts.pir:321)␤called from Sub 'parrot;PCT;HLLCompiler;eval' pc 950
15:08 polyglotbot ..(src/PCT/HLLCompiler.pir:527)␤called from Sub '...
15:10 gravity joined #parrot
15:11 Whiteknight rakudo: my %x = ["first" => 1, "second" => 2, "third" => 3]; %x.perl.say;
15:11 masak Whiteknight: you might know this already, but @x = [1,2,3] and %x = { :first(1) } both create one layer of container too many.
15:11 polyglotbot OUTPUT[Odd number of elements found where hash expected␤current instr.: 'parrot;Perl6Hash;!STORE' pc 7488 (src/classes/Hash.pir:154)␤called from Sub '_block14' pc 159 (EVAL_15:57)␤called from Sub '!UNIT_START' pc 18152 (src/builtins/guts.pir:321)␤called from Sub 'parrot;PCT;HLLCompiler;eval' pc 950
15:11 polyglotbot ..(src/PCT/HLLCompiler.pir:527)␤called from Sub '...
15:11 Whiteknight rakudo: my %x = ("first" => 1, "second" => 2, "third" => 3); %x.perl.say;
15:11 polyglotbot OUTPUT[{"first" => 1, "second" => 2, "third" => 3}␤]
15:11 masak Whiteknight: even the parens are optional.
15:11 Whiteknight masak: yeah, I'm just experimenting with them all to see current behavior
15:12 masak oki.
15:12 Whiteknight [] creates an array ref object, and {} creates a hash ref, right?
15:12 Whiteknight Or is that me thinking about Perl 5?
15:13 Whiteknight my %x; %x<first> = 1; %x.perl.say;
15:13 Whiteknight rakudo: my %x; %x<first> = 1; %x.perl.say;
15:13 polyglotbot OUTPUT[{"first" => 1}␤]
15:14 masak Whiteknight: {} creates a hash ref under certain circumstances. and we don't usually use the bad word "ref" in the Perl 6 world.
15:14 masak we just call it a hash.
15:14 Whiteknight yeah, that's what I was thinking
15:15 Whiteknight rakudo: my $x = (1, 2, 3);
15:15 polyglotbot RESULT[[1, 2, 3]]
15:15 Whiteknight oh right, I should produce some kind of output with that
15:18 masak otherwise the Perl 6 bot will default to printing the value of the last statement.
15:33 jimk joined #parrot
15:50 Hunger joined #parrot
15:50 skv joined #parrot
15:54 dalek r35996 | julianalbo++ | trunk/examples/pir:
15:54 dalek : [pirric] added GETPARROTINTERP function and -p command line option
15:54 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35996
15:58 tetragon joined #parrot
16:08 gryphon joined #parrot
16:17 * Coke_z ponders writing a summary of trying to convert the tcl list PMC into a PIR subclass, and suspect it would just come off all ranty. :|
16:17 Coke *suspects
16:19 NotFound Coke: Did you already solved all problems?
16:20 Coke NotFound: no.
16:20 Coke IMO, both a pir class and a PMC should work the same; they don't, in many respects. and as it turns out, it was a complete waste of time for me to do the switch, as the resulting class is slower than the PMC.
16:21 Coke (which is one of 2 main reasons I attempted the switch in the first place.)
16:21 Coke (the other being to simplify the code; for these PMCs, the resulting PIR is often much easier to read.)
16:22 Coke once I saw the speed comparison, I lost much of my incentive to fix the remaining PIR subclass issues.
16:22 jimk Coke:  Am working on https://trac.parrot.org/parrot/ticket/223
16:23 * jimk is really kid51; need to change identity
16:23 * jimk nods at #parrot and gives him a small kipper
16:23 Coke jimk: danke.
16:24 kid51 joined #parrot
16:24 * kid51 is back
16:24 kid51 Coke:  Can I pose you some questions that may be relevant to TT 223?
16:25 NotFound Coke: if you write the summary, we can take a look at the things that doesn't work the same and see if they can be fixed.
16:32 kid51 Coke (or anyone):  Do we still have any source files known as const*.pmc?  Such as:  src/pmc/constexception.pmc ?
16:33 NotFound kid51: I don't see any
16:33 kid51 At the point in time when I began to write tests for the configuration system, we did have such pmcs.
16:34 kid51 And there is one statement in config/auto/pmc.pm which depends on whether a particular pmc is a ^const pmc or not.
16:34 kid51 const pmcs seem to be something Leo worked on back in the mists of time.
16:35 kid51 But if don't have them any more -- and aren't going to have them -- then that statement in auto::pmc can be excised, which would in turn eliminate some of the code in t/steps/auto_pmc-01.t to which Coke objects.
16:36 kid51 That's question #1.
16:46 Tene joined #parrot
16:46 Coke kid51: all core pmcs are in src/pmc/*.pmc
16:47 Coke if there's no const stuff, then when it was ripped out, no one knew enough to rip out the tests for it.
16:47 kid51 There is one oblique reference to constant PMCs in docs/pdds/pdd17_pmc.pod
16:48 Coke NotFound: there are tickets opened for 2 of the big ones already.
16:48 kid51 =item singleton
16:48 kid51 The PMC is a singleton, created in the constant PMC pool.
16:48 Coke constant there refers to the pool.
16:49 Coke not a particular type of PMC.
16:49 Coke (at least, that's my understanding)
16:50 kid51 okay, then that will enable us to eliminate one of the two long lists in t/steps/auto_pmc-01.t
16:50 kid51 Now, the second question.  The other list of dummy PMCs in t/steps/auto_pmc-01.t exists so as as to test the --pmc option to Configure.pl.
16:51 dalek r35997 | rurban++ | trunk/config/gen/makefiles:
16:51 dalek : Fix TT #225, make perl6 installable
16:51 dalek : - add make installable_perl6 to make installable target
16:51 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35997
16:51 kid51 i.e., if an option to Configure.pl exists, it ought to be tested.  Where test it?  In the configuration step that uses that option.
16:52 NotFound Kill the option
16:53 kid51 So we need to demonstrate, at least in a formal sense, that config/auto/pmc.pm DTRT when the user provides her own list of PMCs to Configure.pl.
16:53 kid51 NotFound:  Well, that's one possible approach.
16:53 NotFound Testing and maintaining options that no one use nor plan to use is a waste of tim
16:54 NotFound time
16:54 kid51 My philosophy when writing those tests was:  Assume that our Parrot ancestors knew what they were doing re configuration functionality.  Hence, preserve such functionality but make sure all nooks and crannies are exercised during testing.
16:54 gryphon joined #parrot
16:54 kid51 But I don't know/understand under what circumstances someone would use the --pmc option.
16:55 NotFound And I don't see why one might want to build without some core pmc. They are core for some reason, isn't it?
16:56 kid51 True, but that option would not have been put there if at some point in time, someone somewhere felt there was a good reason to build with a user-selected list of PMCs.
16:57 kid51 So eliminating the --pmc option would, at bare minimum, require an RFC-type TT and a deprecation cycle if accepted.
16:58 NotFound kid51: if that people does not remember or is no longer involved in parrot development, maybe his reasons were not good enough ;)
16:58 kid51 All of which is beyond the scope of TT 223.
17:13 rurban joined #parrot
17:13 dalek r35998 | jkeenan++ | trunk/config/auto:
17:13 dalek : Population of $pmc_list is better handled as a branch (using ternary op) rather than as a condition (using || op).
17:13 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35998
17:14 skv joined #parrot
17:14 dalek r35999 | jkeenan++ | trunk/config/auto:
17:14 dalek : For consistency with other config step classes, move runstep() to sit immediately after _init().
17:14 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=35999
17:15 rurban stm_start FATAL: not in ops.num nor ops.skip
17:17 rurban sorry, make realclean fixed it.
17:21 NotFound installed parrot_config is linked with build tree libparrot, not with installed libparrot
17:22 rurban which platform?
17:22 purl I'm running on OS/2 on an Atari, can you help?
17:22 NotFound Linux i386
17:22 rurban is it the install_parrot_config binary or the installed parrot_config binary?
17:23 NotFound The installed one
17:23 rurban because the install perl script often has problems with wrong MANIFEST files.
17:23 rurban it could be that the ./parrot_config was wrongly copied
17:24 rurban the logic is first copy all binaries, and then if found the installables
17:24 rurban if the installable is missing the wrong one is installed
17:24 rurban without warning or error
17:27 rurban Are you sure you did make installable before make install?
17:28 NotFound Oh, nice, yet ahother important thing no one cares to document?
17:28 NotFound Looks like the Makefile take care of that
17:29 rurban I would really fix install_files.pl not copying any [main]bin without installable prefix
17:30 rurban The makefile has that dependency, right. But you might forget it calling it standalone.
17:31 rurban perl6 is another candidate friend. I just fixed that in trunk
17:42 rurban rpath is also not yet right for the installables, see tt#208
17:42 purl okay, rurban.
17:42 Vetinari joined #parrot
17:42 NotFound Looks like libparrot_ldflags from config is used to build it, and it has the value for the build tree
17:43 barney joined #parrot
17:43 rurban yes
17:44 rurban I just wanted to fix that but got distracted
17:45 NotFound Maybe will be easier and more consistent to compile and link the installable from the Makefile, letting pbc_to_exe just to generate the .c file
17:45 rurban hmm
17:46 rurban why not
17:50 Theory joined #parrot
17:50 skv joined #parrot
17:54 skv joined #parrot
18:05 Whiteknight Would it make sense for every non-method subroutine to have "self" reference the subroutines Sub PMC?
18:07 Whiteknight actually, I take that back. That would get confusing I think
18:11 dalek r36000 | jkeenan++ | trunk (2 files):
18:11 dalek : Per discussion in https://trac.parrot.org/parrot/ticket/223, in
18:11 dalek : config/auto/pmc.pm eliminate test
18:11 dalek : for PMC files with names beginning 'const' -- we don't have them any more --
18:11 dalek : and eliminate tests in auto_pmc-01.t testing for such files.
18:11 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=36000
18:14 skv joined #parrot
18:16 dalek r36001 | jkeenan++ | trunk/t/steps:
18:16 dalek : Provide explanatory inline comment.
18:16 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=36001
18:22 kid51 Whiteknight:  Agreed that that would be confusing.
18:36 Hunger joined #parrot
18:40 moritz I have a passing TODO test, t/pmc/os.t test 15
18:40 moritz erm, 17
18:40 moritz am I the only one?
18:42 NotFound moritz: same here
18:43 moritz let's ask the other way round - are there any platforms on which it doesn't pass?
18:50 NotFound moritz: that thing makes no sense. How can it unlink a link that it has been unable to create?
18:50 moritz with magic? ;-)
18:50 skv joined #parrot
18:50 barney Looks like smolder, http://smolder.plusthree.com/app​/public_projects/smoke_reports/8, doesn't answer that easily
18:50 shorten barney's url is at http://xrl.us/beaqvk
18:51 NotFound moritz: by lack of logic. The 'tools' directory actually has 8 links
18:51 NotFound No wonder that check for more than 3 gets true
18:52 skv joined #parrot
18:53 NotFound Uh, no, lack of perl knowledge from me %-)
18:54 NotFound Even better, greater than 1
18:55 NotFound But, what is the point of testing a thing that will never work?
18:55 moritz it seems pretty pointless to have a test that usually don't work on most platforms
18:55 moritz right
18:56 moritz on my machine even root can't create directory hard links
18:56 NotFound moritz: i think that there is no unix filesystem that allow that.
18:58 Rahly or linux
18:58 NotFound And don't even talk about FAT systems X-)
18:59 moritz they don't even support soft links ;)
19:03 Rahly they don't even have support for proper filenames
19:04 NotFound We must write a new filesystem, in order to be able to test that feature X-)
19:05 NotFound Someone wants to open a ticket for that X-)
19:06 dalek r36002 | bernhard++ | trunk/src/pmc (2 files):
19:06 dalek : [codingstd] trailing_space.t
19:06 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=36002
19:07 NotFound http://www.usenix.org/events/hotdep06/tech/​prelim_papers/henson/henson_html/node8.html --> According to this, there must be some Unix that allows it
19:07 shorten NotFound's url is at http://xrl.us/bedq6a
19:09 NotFound http://lists.apple.com/archives/d​arwin-dev/2007/Dec/msg00029.html --> And this says that apple OS X has it
19:09 shorten NotFound's url is at http://xrl.us/bedq6i
19:14 dalek r36003 | bernhard++ | trunk (8 files):
19:14 dalek : [unlambda] Moved to http://github.com/bschmalhofer/unlambda/
19:14 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=36003
19:21 dalek r36004 | julianalbo++ | trunk (5 files):
19:21 dalek : fix count_eh opcode and add test for it, TT #212 rg++
19:21 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=36004
19:23 geof joined #parrot
19:25 Whiteknight joined #parrot
19:32 dalek r36005 | bernhard++ | trunk (6 files):
19:32 dalek : [lazy-k] Moved to git://github.com/bschmalhofer/lazy-k.git
19:32 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=36005
19:36 dalek r36006 | bernhard++ | trunk (2 files):
19:36 dalek : [codingstd] regenerate MANIFEST and MANIFEST.SKIP
19:36 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=36006
19:36 dalek r36007 | julianalbo++ | trunk/t/pmc:
19:36 dalek : fix fake positive in t/pmc/os.t
19:36 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=36007
19:38 ask_ joined #parrot
19:59 Whiteknight joined #parrot
20:12 skv joined #parrot
20:20 dalek r36008 | julianalbo++ | trunk/languages/ecmascript/src/builtin:
20:20 dalek : [ecmasript] make the print builtin behave as in SpiderMonkey js shell
20:20 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=36008
20:21 Zaba joined #parrot
20:24 kj joined #parrot
20:42 Theory joined #parrot
20:48 alvar joined #parrot
20:49 Zaba_ joined #parrot
20:49 moritz if anybody is interested, I have a complete read-only git mirror of parrot here: git://timtowtdi.org/parrot
20:50 moritz the mirror is updated every 20 minutes
20:51 moritz (but I don't know if you can easily git-svn dcommit because it uses a mirror of the parrot svn repo, not the original)
20:57 moritz ... and it seems the update isn't working. Ignore me...
21:00 Vetinari left #parrot
21:01 Zaba joined #parrot
21:08 dalek r36009 | julianalbo++ | trunk/languages/ecmascript/src/builtin:
21:08 dalek : [ecmascript] add 'quit' builtin like SpiderMonkey js shell
21:08 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=36009
21:17 Zaba joined #parrot
21:18 chromatic joined #parrot
21:24 dalek r36010 | julianalbo++ | trunk/languages/ecmascript/src/builtin:
21:24 dalek : [ecmascript] add 'readline' builtin like SpiderMonkey js shell
21:24 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=36010
21:24 GeJ Good morning everyone
21:34 Whiteknight Good morning GeJ
21:34 skv joined #parrot
21:37 GeJ hello Whiteknight.
21:37 Whiteknight I haven't heard anything about your CLA, have you?
21:44 Whiteknight Any of the P6 people around?
21:45 moritz I know Perl 6, but I don't really hack rakudo
21:45 moritz do I qualify? ;-)
21:46 Whiteknight I'm trying to figure out what's going on in src/classes/Object.pir
21:46 Whiteknight Like what's the difference between .hash and .Hash, for instance (or is there not any difference)?
21:48 moritz I can't comment on that
21:49 kj it seems that there's a method tailcall going on over there. I thought they were broken..
21:49 kj hi , btw :-)
21:49 GeJ Whiteknight: not yet either. But living on a small island in the middle of nowhere, I wouldn't be surprised if the CLA didn't make it to WA yet.
21:55 Whiteknight GeJ, where do you live?
22:00 Whiteknight I'm also trying to figure out where .HOW is defined
22:03 Whiteknight rakudo: my $x = 5; $x.HOW.say
22:03 polyglotbot OUTPUT[Could not locate a method 'say' to invoke on class 'P6metaclass'.␤current instr.: 'die' pc 16767 (src/builtins/control.pir:204)␤called from Sub 'parrot;P6metaclass;dispatch' pc 311 (src/classes/ClassHOW.pir:149)␤called from Sub '_block14' pc 117 (EVAL_16:49)␤called from Sub '!UNIT_START' pc
22:03 polyglotbot ..18152 (src/builtins/guts.pir:321)␤called from Su...
22:03 Whiteknight rakudo: my $x = 5; $x.HOW.perl.say;
22:03 polyglotbot OUTPUT[Could not locate a method 'perl' to invoke on class 'P6metaclass'.␤current instr.: 'die' pc 16767 (src/builtins/control.pir:204)␤called from Sub 'parrot;P6metaclass;dispatch' pc 311 (src/classes/ClassHOW.pir:149)␤called from Sub '_block14' pc 117 (EVAL_15:49)␤called from Sub '!UNIT_START' pc
22:03 polyglotbot ..18152 (src/builtins/guts.pir:321)␤called from S...
22:04 GeJ Whiteknight: New Caledonia, a tiny set of islands lost somewhere on the north-east of Australia.
22:04 GeJ And I would be suprised if mail to the US would be carried by sea turtles.
22:04 Whiteknight GeJ: Wow! That's quite out of the way.
22:06 Whiteknight Have you ever been to the isle of pines?
22:07 Whiteknight I've always wanted to go myself, seems to be the source of some amazing photography
22:10 GeJ Went there the first time I came here. "amazing" is correct. if only I hadn't left my bag (with my camera in it) on the beach while I went snorkling I may have had pictures of the trip :/
22:10 Whiteknight So what brought you all the way out to New Caledonia? Job? Family?
22:12 GeJ A fair trade with my girlfriend. I brought her with me for 2 years in Germany. For some reason, she didn't quite enjoy winters with 10 inches of snow. So I decided to follow her back to her place.
22:14 moritz wow. Fair trade indeed ;)
22:17 GeJ we've been here 2 years and a half already. And for sonme reason I don't miss winters with 10 inches of snow so much. :)
22:18 moritz how did your first Christmas in the Summer feel?
22:18 Whiteknight Urg, tell me about it. We're having a terrible winter here in PA. Freezing damn cold
22:18 GeJ but I confess that I miss a Schneeballschlacht from now and then with my former boss.
22:23 bacek good morning
22:24 GeJ heya bacek
22:24 bacek hi GeJ
22:28 moritz bacek: if you commit tests to pugs repo that die on rakudo, please #?rakudo skip them, not just todo them
22:28 bacek moritz: oh. Sorry. It actually passed on my Rakudo :)
22:41 bacek moritz: around?
22:41 moritz bacek: yes
22:41 bacek moritz: will you mind to apply my patch from #61610? One element List.redure
22:41 bacek reduce
22:41 * bacek needs more coffee
22:42 kid51 joined #parrot
22:43 bacek rakudo: say "foo" ~~ ["foo", "bar"]
22:43 polyglotbot OUTPUT[Non-dwimmy hyperoperator cannot be used on arrays of different sizes or dimensions.␤current instr.: 'die' pc 16767 (src/builtins/control.pir:204)␤called from Sub '!HYPEROP' pc 15914 (src/builtins/assign.pir:373)␤called from Sub 'parrot;List;ACCEPTS' pc 5539 (src/classes/List.pir:103)␤called
22:43 polyglotbot ..from Sub '_block14' pc 102 (EVAL_16:46)␤called f...
22:43 moritz bacek: looking at it right now...
22:43 bacek moritz: ok
22:44 bacek there is another patch coming to #62742. I even added tests in spectest :)
22:45 kid51 GeJ:  Do you have any patches currently in need of review?
22:45 moritz bacek: so with the patch one-item lists warn with .reduce, and without it they don't?
22:46 bacek moritz: not actually. Without it doesn't work correctly
22:46 GeJ kid51: nope. not currently. I have been forced to put my Parrot activity on the side for now. $life got in the way.
22:46 bacek but yes, it produces warn about uninitialised vars
22:46 bacek rakudo: say (1).reduce: { $^a * $^b }
22:46 kid51 Note apropos earlier discussion of t/pmc/os.t:  I, took, was getting the passing TODOed test in t/pmc/ost.
22:46 polyglotbot OUTPUT[1␤]
22:47 kid51 ... err, t/pmc/os.t.  But it's not yet passing on Darwin.
22:47 GeJ kid51: and the fact that I answer on IRC is just because it happens that I have 3 vim opened in the same desktop as my irssi.
22:47 GeJ but otherwise, I'm busy... for real. :)
22:47 * GeJ &
22:48 chromatic Anyone online right now want a Coverity account who doesn't have one?
22:49 kid51 Coverity?
22:49 purl somebody said Coverity was a commercial tool for Automated Error Prevention and Source Code analysis, See,  http://www.coverity.com/main.html or it has been used to measure the quality of the LAMP stack and other major source projects
22:50 kj chromatic: would it be useful if you're on windows?
22:51 moritz bacek: now I see the difference... but is the new behaviour actually specced? it seems counter-intuitive, and perl5's List::Util::reduce does it the "old" way
22:51 * bacek thinking...
22:51 purl Oooh he is soooo fine!!!
22:52 chromatic Yes, they run the scans on their computers; we just use a web application to read the reports.
22:52 kj chromatic: ok, cool. Can you help me with that?
22:52 chromatic Sure thing.
22:53 kj great, thanks
22:53 bacek rakudo: say 1*undef
22:53 polyglotbot OUTPUT[Use of uninitialized value␤0␤]
22:53 bacek rakudo: my @a=(1); say @a.shift * @a.shift
22:53 polyglotbot OUTPUT[Undefined value shifted from empty array␤0␤]
22:53 * kid51 looks at Coverity site
22:54 kid51 chromatic:  Which aspects of Coverity do we use, or might we use?
22:55 chromatic We read the report and figure out if they've indentified real errors.
22:56 hudnix joined #parrot
22:56 skv joined #parrot
22:57 kid51 which particular report?
23:00 bacek moritz: heh. Patch is actually wrong...
23:02 NotFound kid51: the passed todo was a failure on the test.
23:08 bacek rakudo: say (1..2).reduce: { "$^a-$^b-$^c" }
23:08 polyglotbot OUTPUT[Use of uninitialized value␤1-2-␤]
23:09 dalek r36011 | Whiteknight++ | trunk/docs/book:
23:09 dalek : [Book] Added some stuff about hll mapping
23:09 dalek review: http://www.parrotvm.org/svn​/parrot/revision?rev=36011
23:10 Whiteknight rakudo: say (1..9).reduce: { "$^a-$^b-$^c" }
23:10 polyglotbot OUTPUT[1-2-3-4-5-6-7-8-9␤]
23:10 Whiteknight rakudo: say (1..9).reduce: { "$^a-$^b }
23:10 polyglotbot OUTPUT[Unable to parse block; couldn't find final '}' at line 1, near "\"$^a-$^b }"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤called from Sub 'parrot;Perl6;Grammar;block' pc 29774 (src/gen_grammar.pir:2481)␤called from Sub 'parrot;Perl6;Grammar;pblock'
23:10 polyglotbot ..pc 29094 (src/gen_grammar.pir:2219)␤called from S...
23:11 Whiteknight rakudo: say (1..9).reduce: { "$^a-$^b" }
23:11 polyglotbot OUTPUT[1-2-3-4-5-6-7-8-9␤]
23:11 moritz rakudo: say (1..9).reduce: { "$^a-$^b|$^c" }
23:11 polyglotbot OUTPUT[1-2|3-4|5-6|7-8|9␤]
23:11 chromatic kid51, there's just one Coverity report, on their web site.
23:11 Whiteknight rakudo: say (1..9).reduce: { $^a + $^b }
23:11 polyglotbot OUTPUT[45␤]
23:12 Whiteknight so reduce just keeps applying the trasformation until there is a single scalar left?
23:12 moritz rakudo: say (1..9).reduce: { $^a + $^b, $a^*2 }
23:12 polyglotbot OUTPUT[Statement not terminated properly at line 1, near "2 }"␤␤current instr.: 'parrot;PGE;Util;die' pc 129 (runtime/parrot/library/PGE/Util.pir:83)␤called from Sub 'parrot;Perl6;Grammar;eat_terminator' pc 32087 (src/gen_grammar.pir:3388)␤called from Sub 'parrot;Perl6;Grammar;statementlist' pc
23:12 polyglotbot ..30735 (src/gen_grammar.pir:2855)␤called from Sub 'p...
23:12 bacek moritz: I can shut 'Use of uninitalised' warnings in List.reduce. But I'm not sure is it kosher way.
23:13 moritz rakudo: say (1..9).reduce: { $^a + $^b, $a*2 }
23:13 polyglotbot OUTPUT[114␤]
23:13 moritz bacek: I don't think it's a very important feature atm
23:13 bacek moritz: Just replace "elem = 'undef'()" with "elem = new 'String'" on line 506 in any-list.pir
23:13 bacek moritz: just cage cleanup :)
23:13 moritz but why a String?
23:14 bacek empty String can be casted to any type
23:14 bacek without warnings
23:14 bacek rakudo: my Str $a; say +$a
23:14 polyglotbot OUTPUT[Use of protoobject as value␤0␤]
23:14 bacek rakudo: my Str $a; say $a
23:14 polyglotbot OUTPUT[Str␤]
23:15 bacek rakudo: my Str $a = Str.new; say $a;
23:15 polyglotbot OUTPUT[␤]
23:15 bacek rakudo: my Str $a = Str.new; say +$a;
23:15 polyglotbot OUTPUT[0␤]
23:15 moritz rakudo: my @a = ""; say @a.perl
23:15 polyglotbot OUTPUT[[""]␤]
23:16 bacek rakudo: my Str $a = Str.new; say $a.perl;
23:16 polyglotbot OUTPUT[""␤]
23:20 bacek rakudo: say [1..4].elems
23:20 polyglotbot OUTPUT[4␤]
23:25 sjn joined #parrot
23:32 skv joined #parrot
23:41 kj chromatic: does Coverity also analyze sources of compilers/pirc?
23:43 chromatic kj, not currently, but if we included it in the build it might.
23:43 chromatic We wouldn't have to link it into libparrot.
23:43 kj right
23:43 kj it would be very useful for me, as I'm running into very nasty exceptions these days
23:44 Whiteknight exceptions are the worst
23:44 kj access exceptions :-(
23:51 Tene exceptions are fun
23:52 kj when running the code, the debugger will point to code in the ops file, in libparrot.
23:52 kj I'm not sure if that implies that the problem is in libparrot..

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

Parrot | source cross referenced