Parrot 4.1.0 "Black-headed Parrot" | parrot.org | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC
Set by moderator on 28 February 2012.
dalek rrot: dcc70b5 | petdance++ | config/auto/warnings.pm:
removed -Wzero-as-null-pointer-constant because of too many false positives
 00:09
00:25 travis-ci joined
travis-ci [travis-ci] parrot/parrot#106 (master - dcc70b5 : Andy Lester): The build is still failing. 00:25
[travis-ci] Change view : github.com/parrot/parrot/compare/6......dcc70b5
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/795340
00:25 travis-ci left
benabik Is anyone building parrot with clang? Any idea why it's not working on travis? 00:25
00:32 plobsing joined
whiteknight I build parrot with clang every time 00:32
I think it's not building there because Ubuntu puts the header files in the wrong places. 00:33
benabik Half the travis CI failures are "Determine what C compiler and linker to use...Compilation failed with 'clang'" 00:34
The other half are timeouts. 00:35
whiteknight maybe it shouldn't build with clang, if it's always breaking for unknown reasons 00:36
in fact... 00:39
dalek rrot: 5863a87 | Whiteknight++ | .travis.yml:
[ci skip] Travis CI shouldn't attempt a build with clang. It's always failing config and we don't know why yet.
00:48 not_gerd joined
not_gerd whiteknight: dukeleto probably broke clang CI with github.com/parrot/parrot/commit/c6...334943b4f8 00:48
whiteknight: the sudo apt-get line shouldn't be commented out 00:49
whiteknight not_gerd: Okay, if we can fix it we can re-enable travis on that platform
I feel like we're running on a lot of configurations, we can probably do fewer
Has anybody else set up a GSOC profile and would want to be organization backup-admin? 00:54
Dukeleto has done it in years past. I don't know if he's set up a profile yet
benabik whiteknight: Some of the point of travis is to have it test a wide variety of platform/config. 00:58
whiteknight Yeah, I get that. There's a difference between getting good test coverage and just wasting server cycles
benabik Should probably just do a normal smolder instead of fulltest... No need to do code test with every config. 00:59
In fact, codetest could (should?) just be a separate task. 01:00
dalek p/new_getprop: 242f6e5 | bacek++ | / (6 files):
Update to new version of getprop op
 01:10
p/new_getprop: aef7cd8 | bacek++ | src/stage0/ (7 files):
Reboostrap to use new getprop op
01:15 not_gerd left
whiteknight msg dukeleto I've submitted an application for GSOC2012. I think we can still edit it. I listed you as backup admin, unless you don't want the job or somebody else volunteers. 01:42
aloha OK. I'll deliver the message.
whiteknight msg dukeleto I also put up a page on github wiki for project ideas, and one for proposal templates. The project ideas one is empty now but I'll be filling it briskly 01:43
aloha OK. I'll deliver the message.
02:14 kid51 joined
aloha (parrot/parrot) Issues opened : 724 (Bundle modern nqp with Parrot) by bacek : github.com/parrot/parrot/issues/724 02:49
dalek rrot/m0: b20eea4 | dukeleto++ | lib/Parrot/Harness/TestSets.pm:
Move m0 to the near core test group
 03:55
04:01 kshannon joined 04:10 kshannon joined
benabik msg whiteknight I copy-pasta'd the templates from the GSoC 2011 ideas page to the one for 2012. I also wrote up three projects of varying size for PACT. Feel free to expand/edit/remove/etc. 04:22
aloha OK. I'll deliver the message.
04:25 travis-ci joined
travis-ci [travis-ci] parrot/parrot#107 (m0 - b20eea4 : Jonathan "Duke" Leto): The build is still failing. 04:25
[travis-ci] Change view : github.com/parrot/parrot/compare/0......b20eea4
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/796277
04:25 travis-ci left
dalek rrot: 48317e5 | petdance++ | src/runcore/subprof.c:
Updating function parameters
 04:32
04:46 travis-ci joined
travis-ci [travis-ci] parrot/parrot#108 (master - 48317e5 : Andy Lester): The build is still failing. 04:46
[travis-ci] Change view : github.com/parrot/parrot/compare/5......48317e5
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/796359
04:46 travis-ci left 05:41 jsut joined
dalek rrot: 220feb9 | petdance++ | config/auto/warnings.pm:
adding some C++-specific warnings and a cage check for redundant declarations
 06:18
rrot: 2c7436f | petdance++ | include/parrot/packfile.h:
removed old function declarations
rrot: 608aa0a | petdance++ | include/parrot/interpreter.h:
remove redundant declaration
rrot: 2c959c2 | petdance++ | config/auto/warnings.pm:
add -Wstrict-overflow to see what that turns up
06:33 travis-ci joined
travis-ci [travis-ci] parrot/parrot#109 (master - 2c959c2 : Andy Lester): The build is still failing. 06:33
[travis-ci] Change view : github.com/parrot/parrot/compare/4......2c959c2
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/796667
06:33 travis-ci left
aloha (parrot/parrot) Issues opened : 725 (Integrate m0 makefile into main parrot makefile) by leto : github.com/parrot/parrot/issues/725, 726 (Add makefile target to run m0 tests under valgrind) by leto : github.com/parrot/parrot/issues/726 07:01
dalek rrot: 3741c0d | dukeleto++ | config/auto/ (2 files):
Merge pull request #715 from gerdr/gerdr/709-clang

fix issue #709 on clang		 07:04
aloha (parrot/parrot) Issues closed : 715 (fix issue #709 on clang) by gerdr : github.com/parrot/parrot/issues/715 07:06
(parrot/parrot) Issues closed : 568 (examples/nci/ls.pir broken) by bubaflub : github.com/parrot/parrot/issues/568 07:11
dalek rrot: 075a574 | petdance++ | config/auto/warnings.pm:
adding some more warnings
 07:15
07:18 travis-ci joined
travis-ci [travis-ci] parrot/parrot#110 (master - 3741c0d : Jonathan "Duke" Leto): The build is still failing. 07:18
[travis-ci] Change view : github.com/parrot/parrot/compare/2......3741c0d
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/796840
07:18 travis-ci left 07:31 travis-ci joined
travis-ci [travis-ci] parrot/parrot#111 (master - 075a574 : Andy Lester): The build is still failing. 07:31
[travis-ci] Change view : github.com/parrot/parrot/compare/3......075a574
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/796923
07:31 travis-ci left 07:52 johbar_work joined 08:23 fperrad joined
bacek ~~ 08:46
dalek rrot: 10ae7c6 | bacek++ | docs/book/ (2 files):
Update pir book to new getprop syntax.
 08:50
rrot: cebe8d7 | bacek++ | ext/nqp-rx/src/stage0/ (2 files):
Update nqp-rx in-place to new getprop syntax
rrot: a7790bf | bacek++ | compilers/pct/src/P (2 files):
Update PCT to new getprop version
rrot: 5c3c2fe | bacek++ | t/ (4 files):
Update tests to use new getprop version
rrot: 29f72df | bacek++ | examples/sdl/tetris/b (2 files):
Update tetris to use new getprop version
rrot: 4b23603 | bacek++ | runtime/parrot/library/ (4 files):
Update runtime to use new getprop version
08:55 mj41 joined
dalek rrot: 179831d | bacek++ | t/pmc/prop.t:
Update test to new getprop version
 08:59
09:06 travis-ci joined
travis-ci [travis-ci] parrot/parrot#112 (master - 4b23603 : Vasily Chekalkin): The build is still failing. 09:06
[travis-ci] Change view : github.com/parrot/parrot/compare/0......4b23603
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/797339
09:06 travis-ci left 09:18 travis-ci joined
travis-ci [travis-ci] parrot/parrot#113 (master - 179831d : Vasily Chekalkin): The build is still failing. 09:18
[travis-ci] Change view : github.com/parrot/parrot/compare/4......179831d
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/797374
09:18 travis-ci left 09:36 alester_ joined 09:41 alester joined
aloha (parrot/parrot) Issues opened : 727 (Memory related (non-GC version) macros are deprecated and will be removed.) by bacek : github.com/parrot/parrot/issues/727 10:13
10:29 schmooster joined
dalek rrot: 67b0e5e | bacek++ | include/parrot/memory.h:
Add mem_internal_realloc_n_typed for consistency
 11:21
rrot: 6444a67 | bacek++ | / (2 files):
Update PointerArray to use mem_internal_* functions.
rrot: 8213cff | bacek++ | src/ (19 files):
Don't use mem_sys_* macros. They are deprecated
11:35 travis-ci joined
travis-ci [travis-ci] parrot/parrot#114 (master - 8213cff : Vasily Chekalkin): The build is still failing. 11:35
[travis-ci] Change view : github.com/parrot/parrot/compare/1......8213cff
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/798036
11:35 travis-ci left 11:46 Tene joined
dalek kudo/nom: bc677b4 | jonathan++ | src/ (3 files):
Add storage space in Block for phasers, plus some methods for adding/grabbing them.
 11:58
kudo/nom: 753fbc5 | jonathan++ | src/Perl6/World.pm:
Start to break some parts of create_code_object out in to separate Perl6::World methods.
kudo/nom: 428f954 | jonathan++ | src/Perl6/World.pm:
Toss now-unused $slot variables.
kudo/nom: a9209d1 | jonathan++ | src/Perl6/World.pm:
Replace some pir:: with nqp::.
kudo/nom: 05464d4 | jonathan++ | src/Perl6/World.pm:
Split the final setup of a code object out; create_code_object becomes a helper method that just delegates to the methods implemetning the various steps.
kudo/nom: c5e9349 | jonathan++ | src/Perl6/ (2 files):
Swtich routines over to creating code object and putting it in $*DECLARAND before we go parsing the body.
kudo/nom: 060b8ea | jonathan++ | src/Perl6/ (2 files):
Similar refactor for methods and regexes.
moritz oh hai 12:04
my latest parrot/master build segfaults
tadzik moritz: did you try 'make clean' and make again? 12:06
That helped me yesterday 12:07
moritz I'll try 12:09
now some other error with -j4 12:10
12:32 bacek_at_work joined 12:34 aloha joined 12:47 bluescreen joined
dalek p: 3332c0a | moritz++ | src/HLL/Compiler.pm:
Revert "Merge pull request #30 from sergot/master"

This reverts commit 990347106f770bec366eb27a27cbabfceb85589a, reversing changes made to d44ab2eba183738e8d58079901479e5393bed7db.
This actually didn't produce nicer error messages, just looped infinitely.		 13:36
kudo/nom: 192c8f5 | moritz++ | tools/build/NQP_REVISION:
bump NQP to improve error messages
 13:37
kudo/macros3: fc713e3 | masak++ | src/ (2 files):
check whether an AST comes from a quasi or not
 13:56
14:55 contingencyplan joined, contingencyplan_ joined 14:57 PacoAir joined
dalek kudo/macros3: c781128 | masak++ | src/Perl6/Actions.pm:
allow macros to return Nil

Nil will be inserted into the mainline. This isn't spec, but it makes for shorter macro definitions. :-)		 15:00
15:36 schmoo joined
aloha (parrot/parrot) Issues closed : 718 (update auto-sizes to latest master) by gerdr : github.com/parrot/parrot/issues/718 15:45
15:45 Psyche^ joined
alester Are any of y'all running under g++, rather than gcc? 15:54
dalek kudo/nom: eebcd27 | jonathan++ | src/Perl6/ (2 files):
Refactor pblock and block to make the code object before parsing the block.
 15:59
kudo/nom: 3479bec | jonathan++ | src/Perl6/Grammar.pm:
Add stub parsing for missing phasers and re-organize them to follow the ordering in STD.
kudo/nom: b771b34 | jonathan++ | src/Perl6/ (2 files):
Small refactor to phaser handling so we'll be able to attach phasers to blocks.
kudo/nom: a6d75bc | jonathan++ | src/Perl6/ (3 files):
First crack at implementing ENTER phasers.
 16:00
kudo/nom: 1a49d73 | jonathan++ | src/Perl6/ (3 files):
Implement START phaser.
kudo/nom: 84e7c42 | jonathan++ | t/spectest.data:
Run S04-phasers/start.t.
16:10 bluescreen joined
dalek kudo/nom: 4e02c27 | jonathan++ | src/Perl6/ (2 files):
Attach FIRST/NEXT/LAST phasers to blocks, though we don't do anything with them yet.
 16:24
kudo/nom: b4b9971 | jonathan++ | src/core/MapIter.pm:
Run the NEXT and LAST phasers for for loops.
cotto `~ 16:38
dalek sella: a14b556 | Whiteknight++ | t/date/Date.t:
[Date] fix test for get_string_formatter
 16:39
sella: 5c7c3a5 | Whiteknight++ | src/core/Globals.winxed:
[Core] Re-do the way the globals hash is stored. Store it in the Rosella.Globals namespace instead of using a coroutine.
rrot: b960fc1 | petdance++ | src/gc/mark_sweep.c:
removed unused label
 16:43
16:50 dmalcolm joined 17:02 travis-ci joined
travis-ci [travis-ci] parrot/parrot#115 (master - b960fc1 : Andy Lester): The build is still failing. 17:02
[travis-ci] Change view : github.com/parrot/parrot/compare/8......b960fc1
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/799894
17:02 travis-ci left
aloha (parrot/parrot) Issues opened : 728 (Remove all function definitions from header files) by petdance : github.com/parrot/parrot/issues/728 17:21
17:23 jsut_ joined 17:27 cogno joined 17:30 cogno_ joined
dalek kudo/nom: 5c1b4e6 | jonathan++ | src/ (4 files):
Implement FIRST.
 17:34
kudo/nom: 53aa9c6 | jonathan++ | t/spectest.data:
Run S04-phasers/first.t.
rrot: bdc7c8d | petdance++ | src/interp/api.c:
renamed a shadowed variable.
rrot: dbc3edb | petdance++ | src/events.c:
removed unused variable
rrot: 857acac | petdance++ | config/auto/warnings.pm:
adding more GCC/G++ warning flags for --cage
17:43 rurban joined 17:53 mj41 joined 18:06 lucian joined
dalek kudo/nom: 98f9b74 | jonathan++ | src/Perl6/ (2 files):
Fix use of phasers in the mainline.
 18:08
kudo/nom: 3c5838b | jonathan++ | t/spectest.data:
Run S04-phasers/ascending-order.t.
18:13 rurban left
dalek kudo/world-API: f1b49f2 | moritz++ | src/ (4 files):
start passing $/ to Perl6::World methods to make error reporting easier
 18:19
18:27 cogno joined
benabik alester: I occasionally run with g++, but IIRC nqp or rakudo doesn't build cleanly with it. 19:33
alester rakudo does for me.
and tests.
You have to link with g++ as well.
benabik Hm. Maybe the codebase is cleaner now.
alester Could be.
benabik I think parrot defaults to g++ for linking.
(It was pointer conversion issues, as I recall)
dukeleto we test parrot with gcc, g++ and clang on Travis. I think clang might be having some issues right now, but I might have fixed it last night. 19:35
alester I've added about 10 new warnings yestday and today. 19:38
Warnings make me happy.
well, warning flags do. Warnings make me sad. 19:39
benabik git builds with -Wall and -Werror. I like that.
alester My next project to beat on is libgit2.
benabik dukeleto: I looks like travis is running fulltest. Doesn't that include all the code tests and the like that won't change? I know some of the compile runs are timing out and that makes me wonder if we should cut down on some tests.
19:40 benabik left, benabik joined
alester I love splint. I just ran it for the first time in forever and hey look, an error. 19:40
benabik hasn't tried splint
alester I'll tell you all you want to know. 19:41
benabik "It's awesome"? ;-)
alester yes, but not 100% so
19:44 not_gerd joined
not_gerd dukeleto: clang on travis is broken, and whiteknight disabled it (but actually forgot one configuration) 19:45
dalek rrot: cd67cf8 | petdance++ | src/debug.c:
const a bunch of variables. Make some ints into UINTVALs. Simplify a big for() by turning it into a while().
not_gerd dukeleto: see irclog.perlgeek.de/parrot/2012-03-05#i_5246198 for discussion and possible fix
not_gerd dukeleto: also, the build sometimes fails due to timeout, possibly depending on travis load 19:47
dukeleto not_gerd: travis is seeing some timeouts due to flaky ubuntu mirrors 19:52
not_gerd: you are also right that I broke the clang build 19:53
dukeleto--
alester Who knows what a hop_bucket is? 20:00
I'll settle for knowing what a hop is, for that matter. 20:01
dukeleto alester: a pre-cursor to beer?
alester I assume it's some sort of "H op"
ack -w hop_buckets
dukeleto alester: hash opcode bucket? Just guessing. 20:02
alester 'cause I think we have a memory leak with them.
20:02 travis-ci joined
travis-ci [travis-ci] parrot/parrot#117 (master - cd67cf8 : Andy Lester): The build is still failing. 20:02
[travis-ci] Change view : github.com/parrot/parrot/compare/8......cd67cf8
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/800998
20:02 travis-ci left
benabik static HOP *hop_buckets. 20:02
hop = struct { op_info_t *info, struct hop *next } 20:03
Hashed Ops ? 20:04
alester Well, this is definitely a bug 20:12
github.com/parrot/parrot/issues/729
aloha (parrot/parrot) Issues opened : 729 (Memory leak with hop_buckets in ops) by petdance : github.com/parrot/parrot/issues/729 20:13
dalek rrot: 795a444 | dukeleto++ | .travis.yml:
[ci] Re-enable clang, which I accidentally broke by commenting out the before_script command which installs it
 20:18
dukeleto blarg 20:30
dukeleto is talking to the travis people about stuff 20:33
Coke is reminded about the travis kerfluffle that Coke should get back to muddle. 20:36
(so we can kill smolder.)
20:44 not_gerd left
dukeleto Coke: travis submits to smolder 20:45
Coke: they overlap, but they both excel at different things
dukeleto will have to look into travis stuff later
Coke knows that, yes. 20:46
Coke is looking to replace smolder only.
20:50 perlite joined
moritz when I try to build NQP on newest parrot, I get some errors and warnings 20:51
cc1: warning: ../install/include/parrot/4.1.0-devel: No such file or directory
cc1: warning: ../install/include/parrot/4.1.0-devel/pmc: No such file or directo
ry
dispatchersub.c:17:27: error: parrot/parrot.h: No such file or directory
benabik moritz: --gen-parrot?
moritz benabik: no
oh, might be my fault 20:52
benabik Looks like the install path is goofy
moritz seems that nqp can't handle relative prefixes in parrot installation
dalek rrot: 774e4e5 | dukeleto++ | .travis.yml:
[ci] Fix improper spacing in our .travis.yml, thanks lint.travis-ci.org/parrot/parrot
 21:24
21:47 travis-ci joined
travis-ci [travis-ci] parrot/parrot#119 (master - 774e4e5 : Jonathan "Duke" Leto): The build is still failing. 21:47
[travis-ci] Change view : github.com/parrot/parrot/compare/7......774e4e5
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/801783
21:47 travis-ci left 23:14 PacoAir joined 23:25 whiteknight joined
whiteknight good evening, #parrot 23:25
benabik++
dukeleto julialang.org/ 23:39
they have an LLVM JIT. Let's steal it.
whiteknight it's not a matter of us not having a JIT, it's a matter of us not having a JITable opset 23:42
M0 is the hard part. Once we have that, JIT won't be nearly so bad 23:44
well, basic JIT. Complicated ones will be harder
23:55 kid51 joined
dukeleto kid51: howdy 23:58

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!