Camelia, the Perl 6 bug

IRC log for #parrot, 2010-04-07

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 cotto_work +meh for that as a gsoc project
00:04 cotto_work It could benefit Parrot some, but I don't see it as being particularly important right now.
00:04 chromatic I'd rather see NQP documented, bugs filed, and warts removed.
00:05 AndyA joined #parrot
00:07 payload1 joined #parrot
00:07 payload1 left #parrot
00:24 sorear kthakore: What's this you want from me?
00:24 bubaflub joined #parrot
00:29 tcurtis chromatic: warts such as?
00:29 chromatic I don't know; I haven't used NQP.  I assume they exist though.
00:30 Austin tcurtis: operator of death? vtable hackarounds? 50% of the features in kakapo?
00:34 sorear what's the operator of death?
00:34 sorear My favorite unfeature of NQP is the lack of anonymous hash constructors
00:35 chromatic
00:38 payload joined #parrot
00:44 d4l3k_ joined #parrot
00:44 particle joined #parrot
00:44 Maddingu1 joined #parrot
00:44 mj41_ joined #parrot
00:57 tcurtis Austin, sorear, anyone else familiar with NQP, what are the biggest problems you experience with NQP right now?
00:58 abqar joined #parrot
00:59 particle tcurtis: send a mail requesting nqp wishes to p6c, and look at the TODO file, ROADMAP, and whatever else is available in the nqp repo
00:59 particle workout &
01:00 Austin tcurtis: "problems" http://wknight8111.blogspot.co​m/2010/03/austins-tickets.html
01:01 Andy_ joined #parrot
01:02 Austin But "problems" as "I'm open for a Parrot Year-of-Drudgery project": I'd like to see PMCs become classes, and I'd particularly like to see someone produce some hard numbers about the "cost" of method calls versus vtable access. (And then I have some ideas about speeding that up.)
01:27 ash_ joined #parrot
01:41 ash_ I updated my GSoC
01:41 ash_ to include libffi work, anyone have any comments?
01:44 ash_ just wondering, does anyone know if parrot IO supports serial IO ?
01:46 ruoso ash_, do you need to configure the port with termio operations?
01:46 kid51 ash_:  Where is the updated proposal?
01:46 ruoso ascent, if not, and you're in Linux, you can just open /dev/ttyS0
01:47 ash_ kid51: http://docs.google.com/View?id=dfvcbmxs_51fd6qw7zj is a copy of it
01:49 ash_ ruoso: well, its more than just normal IO for opening a /dev/ttyS0 type port
01:49 ash_ its mostly normal IO
01:49 ash_ but you need to do things like set parity bits, stop bits, byte sizes, etc.
01:51 ash_ it would be nice to control baud rate too
01:52 ruoso ash_, that's termio
01:52 ash_ got ya
02:00 cotto seen chandon
02:00 purl chandon was last seen on #parrot 4 days, 5 hours, 57 minutes and 0 seconds ago, saying: It really depends on the application and the algorithm. For some things, nothing, for others, everything.  [Apr  2 20:03:14 2010]
02:22 integral joined #parrot
02:25 dalek rakudo: 8c434e8 | pmichaud++ | docs/spectest-progress.csv:
02:25 dalek rakudo: pectest-progress.csv update: 386 files, 30548 (79.2% of 38577) pass, 10 fail
02:25 dalek rakudo: Failure summary:
02:25 dalek rakudo: S02-literals/listquote-whitespace.rakudo aborted 8 test(s)
02:25 dalek rakudo: S06-multi/syntax.rakudo 21 - use multi method to override builtin lives
02:25 dalek rakudo: S06-multi/syntax.rakudo 22 - use multi method to override builtin
02:25 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/8​c434e8587abd38968d93c4258feeb844b682709
02:31 tcurtis_ joined #parrot
02:39 ash_ joined #parrot
02:48 janus joined #parrot
02:50 petdance joined #parrot
02:56 sorear tcurtis: It doesn't expose enough low-level functionalty without dropping to PIR.  No tailcalls, no :named :flat, no slurpy returns, no anonymous hash construction are the big ones
02:58 sorear lack of easy vtable overrides
03:03 ash_ joined #parrot
03:06 kurahaupo joined #parrot
03:10 Mokurai joined #parrot
03:14 payload joined #parrot
03:34 chromatic You know it's going to be worth reading when a message from Damian Conway has a title which reminds you of David Foster Wallace.
03:48 dukeleto this Emerging Languages conference seems interesting: http://emerginglangs.com/
03:49 petdance pig Coke
03:49 petdance ping Coke
03:49 petdance pig Coke sounds like Pork Soda.
03:50 tetragon_ joined #parrot
04:07 dukeleto ss.r3nt
04:08 * dukeleto beckons the linenoise demons
04:10 bubaflub dukeleto: finally got my RTEMS basic setup to work, finishing my blog post about it now and then going to write my proposal
04:11 dukeleto bubaflub: sounds awesome
04:11 dukeleto bubaflub: the parrot+rtems project is very exciting
04:11 dukeleto bubaflub: i guess you should apply for it with TPF+Parrot
04:11 bubaflub dukeleto: the RTEMS guys said i should just apply for both
04:12 bubaflub and if i get accepted ya'll can coordinate
04:12 bubaflub makes it a little more flexible
04:12 dukeleto bubaflub: sounds fine, do that
04:14 bubaflub dukeleto: http://blog.bobkuo.com/2010/04/set​ting-up-rtems-part-2-hello-world/
04:14 bubaflub proposal will be coming up in shortly
04:25 dukeleto bubaflub: waiting on the edge of my seat :)
04:43 dalek parrot: r45409 | petdance++ | trunk/include/parrot/compiler.h:
04:43 dalek parrot: fixed the specifications of the attributeformat macro
04:44 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45409/
04:57 dukeleto bubaflub: i am reading your 2nd post
04:57 bubaflub dukeleto: cool, i'm putting up my proposal.  this is taking way longer than i thought it would.
04:58 dukeleto bubaflub: patience, my son :)
04:59 petdance ping chromatic
05:00 dalek parrot: r45410 | petdance++ | trunk (4 files):
05:00 dalek parrot: add sniffing of function attributes
05:00 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45410/
05:04 bubaflub dukeleto: http://blog.bobkuo.com/2010/04/google-summ​er-of-code-2010-proposal-parrot-on-rtems/
05:04 bubaflub needs some work, but at least it's done
05:14 dukeleto bubaflub: i still haven't sent you chris's patches+tarball stuff. will get on that
05:14 bubaflub dukeleto: chris sent me a link to the generated stuff, but the patches would be nice as well
05:16 dalek parrot: r45411 | petdance++ | trunk (3 files):
05:16 dalek parrot: headerizer now supports PARROT_HOT and PARROT_COLD
05:16 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45411/
05:24 dukeleto bubaflub: i just sent you a tarball
05:24 bubaflub dukeleto++
05:24 dukeleto bubaflub: can you send me the link chris johns sent you?
05:24 bubaflub sure thing dukeleto
05:25 bubaflub (it's in your email)
05:25 bubaflub dukeleto: i've got two more proposals for TPF
05:25 bubaflub one is to do the Math::GSL bindings for parrot
05:25 bubaflub one is to continue with Math::Primality and replace some Perl 5 code to use it
05:25 bubaflub for the first one, how should i scope the project?
05:26 bubaflub and for the second, how should i figure out which modules should be replaced?
05:27 dukeleto bubaflub: for the 1st, i would say add support for special functions and then a few other important subsystems
05:27 dukeleto bubaflub: SF has the easiest function signatures
05:29 dukeleto bubaflub: for Math::Primality -> we need Crypt::Primes and Crypt::Random replacements that use Math::Primality and Math::GMPz
05:30 bubaflub dukeleto: awesome.  i'll look into the GSL special functions and both of those modules.  i'll post some more proposals on my blog (hopefully tonight)
05:30 sorear bubaflub: I've got this eerie feeling we're duplicating work
05:31 bubaflub sorear: oh.  whatcha doing?
05:31 bubaflub (i don't want to step on any toes)
05:31 dukeleto bubaflub: you should read the original Math::GSL gsoc proposal from 2008 for ideas
05:31 sorear binding every CPAN module to Parrot
05:31 dukeleto sorear: not talking about binding every CPAN module to Parrot
05:31 sorear no, you're just binding one
05:32 dukeleto sorear: we are binding to the GSL C library
05:33 dalek parrot: r45412 | petdance++ | trunk (4 files):
05:33 dalek parrot: adding a few PARROT_COLD funcs for balance
05:33 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45412/
05:33 dukeleto sorear: it was very hard to bind Perl 5 to GSL
05:34 dukeleto sorear: my hope is that if Parrot has a GSL binding, every language on parrot gets that for free
05:37 dalek TT #1545 created by petdance++: Properly split the headerizer
05:37 dalek TT #1545: http://trac.parrot.org/parrot/ticket/1545
05:40 bubaflub sorear: either way, it's just a proposal.  if it's duplicate or unnecessary work it'll get shot down.
05:42 dukeleto bubaflub: feel free to start a parrot-gsl github repo
05:42 dukeleto bubaflub: i want to hack on that, regardless if it is part of gsoc
05:42 bubaflub dukeleto: roger that.  it'll be done through the NCI interface, right?
05:42 dukeleto bubaflub: back back in a bit
05:42 dukeleto bubaflub: probably
05:42 bubaflub dukeleto: k
05:44 sorear does this depend on the new framebuilder?
05:45 sorear I just thought of a way to make the POSIX generic framebuilder idea almost efficient
05:46 sorear separate the framebuilder internal API into "request" and "demand" phases
05:46 sorear when a NCI PMC is created, add the signature to the "request" list
05:46 sorear when any unresolved NCI PMC is called, create thunks for *all* outstanding requested signatures
05:47 sorear so if parrot-gsl is implemented correctly, the portable backend only needs to make one round-trip to the compiler
05:48 bubaflub interesting sorear; i'm not too familiar with the framebuilder
05:49 jsut joined #parrot
05:49 bubaflub actually dukeleto there is a proposal to rework the BigNums and BigInts into a more full GMP binding; i think i'll do that
05:50 sorear bubaflub: the framebuilder is the part of Parrot which takes a description like "tfi" and converts it into a PCC/C-ABI thunk function
05:50 bubaflub though having GSL bindings isn't a bad idea - the more we have at the low level the more attractive it is to develop on parrot
05:50 sorear currently, there is 1 frame"builder"; it works by having a fixed list of thunks (created by a preprocessor), and using hash lookup
05:51 sorear the "portable framebuilder", my invention and not endorsed by any Parrot committer at this time, generates C code at runtime and passes it to /usr/bin/c99 and dlopen(3), both of which are specced by POSIX
05:51 dduncan joined #parrot
05:52 sorear there are also GSoC proposals to implement framebuilders based on the LLVM and libjit libraries
05:54 bubaflub sorear: interesting.  POSIX always sounds nice
05:56 sorear it's nice in theory, but remember that 90% of the computer market doesn't follow it and the remaining 10% is comprised of FreeBSD and Linux
05:57 bubaflub yeah, *actual* portability is always a problem
06:10 dduncan is there a distinction between being POSIX and qualifying as a UNIX?
06:11 Austin Heh
06:11 Austin Yeah.
06:11 Austin People actually use Unix.
06:11 dduncan I don't know the exact significance, but supposedly starting with version 10.5 the Mac OS qualifies as something UNIX related where it didn't before ... I had thought this related to POSIX, but maybe not
06:12 chromatic They fixed some of the dynamic loading nonsense inherited from NeXT with 10.3 or 10.4, I know.
06:13 uniejo joined #parrot
06:14 chromatic ... not that you're going to convince me that relying on the presence and sanity and portability of /usr/bin/c99 will work for an NCI thunk generator.
06:14 Austin WFM. Ship it.
06:41 sorear chromatic: obviously this would be part of a pluggable fallback system
06:41 sorear with core.nci and the static generator (and maybe even extra.nci) retained
06:41 sorear and libffi/llvm backends added later for performance/lorito
07:03 payload joined #parrot
07:05 bubaflub dukeleto: i've just submitted a few more proposals to TPF; they are posted on blog.bobkuo.com.  i welcome any feedback or comments.
07:14 sorear bubaflub: you seem to have typoed the Crypt::Primes/Crypt::Random one
07:14 sorear it says Artistic 2.0 as Parrot
07:14 bubaflub doh! copy and paste --
07:14 sorear but Crypt::Primes and Crypt::Random are under the *Perl* license
07:14 sorear which is not compatible with A2
07:15 sorear except by conversion clauses which allow both to be moved to GPL
07:15 bubaflub good eye sorear++
07:15 bubaflub would it be compatible to say "same as perl" ?
07:15 sorear that's the existing license of Crypt::both, so yesd
07:16 bubaflub excellent; made the change.
07:55 aukjan joined #parrot
07:57 sorear in Perl 5, it is possible to overload an object such that it behaves differently when used as an array versus used as a hash
07:57 sorear how should this be handled in the get_pmc_keyed VTABLE?
07:57 sorear in particular, with PMC keys (which can be stringified and numified)
07:59 moritz sorear: use overload '%{ }'
08:00 sorear Should I special case INTVAL keys?
08:00 sorear or, "how to weigh two violations of the PoLS"
08:01 Austin Do whatever perl does.
08:02 sorear Austin: Perl has different syntax for accessing hash references and array references
08:02 sorear Parrot has a single op for both
08:02 kurahaupo joined #parrot
08:02 Austin Hmm..
08:02 Austin Parrot has get_...keyed_int and keyed_str
08:02 sorear that's what I meant by "special case INTVAL keys"
08:03 sorear or a special case of it :)
08:03 sorear the ambiguity comes when a PMC variable is passed
08:03 moritz in perl $thing[3.3] will automatically round
08:03 moritz (iirc)
08:04 sorear I think there's a 'use warnings' for that
08:05 Austin Check if one or the other is meaningless - that is, if the object "does array" versus "does hash." If it does both, default to array.
08:05 Austin (50% chance of being correct behavior.)
08:06 sorear sounds good, *TODO*
08:09 sorear my biggest source of confusion on this is determining whether any given occurance of "PMC*" should have value or container semantics
08:16 * sorear wonders whether to steal S_refgen
08:24 dalek parrot: r45413 | mikehh++ | trunk/include/parrot/compiler.h:
08:24 dalek parrot: fix codetest failure - unwrapped macro arguments
08:24 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45413/
08:24 dalek parrot: r45414 | mikehh++ | trunk/include/parrot/compiler.h:
08:24 dalek parrot: fix codetest failure - there should be one space or a newline after a comma (should have added the spaces in the previous commit)
08:24 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45414/
08:41 payload joined #parrot
08:45 clinton joined #parrot
08:50 mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#33035), fulltest) at r45414 - Ubuntu 10.04 beta amd64 (gcc with --optimize)
09:08 barney joined #parrot
09:23 dalek blizkost: 9b2f488 | jonathan++ | src/pmc/bkmarshal.c:
09:23 dalek blizkost: Need a my_perl declared on some platforms in order to use SvREFCNT_inc; this partly unbreaks the Win32 build.
09:23 dalek blizkost: review: http://github.com/jnthn/blizkost/commit/​9b2f4887290d270e7f5d404a192412d335931557
09:33 AndyA joined #parrot
09:34 dalek blizkost: 90ed467 | jonathan++ | build/src/pmc/Makefile.in:
09:34 dalek blizkost: nmake seems unable to expand $(O) inside a substitution rule, so just use @o@ there instead. This fixes the build on Win32.
09:34 dalek blizkost: review: http://github.com/jnthn/blizkost/commit/​90ed467437e57bf514cd944520202e38b22123cd
09:38 dduncan left #parrot
09:50 mikehh rakudo: builds on parrot r45414 - tests spectest_smolder #33038 - 2 failures
09:51 payload joined #parrot
10:10 fperrad joined #parrot
10:25 bacek o hai
10:28 karim joined #parrot
10:28 karim hi
10:28 purl bonjour, karim.
10:31 lucian joined #parrot
11:25 aukjan joined #parrot
11:45 payload joined #parrot
11:51 bluescreen joined #parrot
11:54 payload joined #parrot
11:57 bluescreen joined #parrot
11:58 dalek parrot: r45415 | bacek++ | branches/immutable_strings_part1:
11:58 dalek parrot: Create branch immutable_strings_part1
11:58 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45415/
11:59 jan joined #parrot
12:15 tcurtis joined #parrot
12:24 Mokurai1 joined #parrot
12:27 whiteknight joined #parrot
12:28 Coke particle: pong.
12:28 particle pong
12:29 particle i had some osx questions for you yesterday, trying to help a friend unzip a bz2 file.  it's taken care of, thanks.
12:29 particle yesterday? actually, monday. whatever.
12:54 Coke joined #parrot
12:54 * Coke tests.
13:03 whiteknight good morning everybody
13:06 bacek hello Mr. Whitworth!
13:13 ttbot Parrot trunk/ r45416 i386-linux-thread-multi make error http://tt.taptinder.org/file/cmdout/254016.txt ( http://tt.taptinder.org//bui​ldstatus/pr-Parrot/rp-trunk/ )
13:14 bacek SHIT
13:15 ttbot Parrot trunk/ r45423 i386-linux-thread-multi make error http://tt.taptinder.org/file/cmdout/254029.txt ( http://tt.taptinder.org//bui​ldstatus/pr-Parrot/rp-trunk/ )
13:16 ttbot Parrot trunk/ r45424 i386-linux-thread-multi make error http://tt.taptinder.org/file/cmdout/254037.txt ( http://tt.taptinder.org//bui​ldstatus/pr-Parrot/rp-trunk/ )
13:19 Coke bacek - why are you breaking trunk?
13:19 Coke I think you meant to do that in branch.
13:21 Coke ah, there's the "SHIT" I missed.
13:22 dalek parrot: r45416 | bacek++ | trunk (2 files):
13:23 dalek parrot: Brake the build - remove COW functions
13:23 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45416/
13:23 dalek parrot: r45417 | bacek++ | trunk (2 files):
13:23 dalek parrot: Remove str_set function. We can't set content of immutable strings
13:23 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45417/
13:23 dalek parrot: r45418 | bacek++ | trunk/src/ops/set.ops:
13:23 dalek parrot: Replace str_set in assign op with just assign.
13:23 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45418/
13:23 dalek parrot: r45419 | bacek++ | trunk/src/call/args.c:
13:23 dalek parrot: Remove str_new_COW in args passing.
13:23 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45419/
13:23 dalek parrot: r45420 | bacek++ | trunk (2 files):
13:23 dalek parrot: Remove Parrot_str_resize function. "Resizing of rock-stable rocks? What
13:23 dalek parrot: you are talking about?"
13:23 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45420/
13:23 dalek parrot: r45421 | bacek++ | trunk (2 files):
13:23 dalek parrot: Remove Parrot_str_copy. "Rock is always a rock. Why copy it?"
13:23 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45421/
13:23 dalek parrot: r45422 | bacek++ | trunk/src (3 files):
13:23 dalek parrot: Remove few str_copy calls
13:23 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45422/
13:23 dalek parrot: r45423 | bacek++ | trunk (23 files):
13:23 dalek parrot: Remove a bunch of str_copy usage
13:23 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45423/
13:23 dalek parrot: r45424 | bacek++ | trunk/src/pmc (2 files):
13:23 dalek parrot: Fix usage of long-standing bug with wrong returns from *Handle.encoding functions
13:23 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45424/
13:43 bacek sorry to everyone... My fault...
13:44 bacek I didn't ensure that git-svn switched to new branch.
13:47 dalek joined #parrot
13:47 dalek parrot: r45425 | bacek++ | trunk (27 files):
13:47 dalek parrot: Fix usage of long-standing bug with wrong returns from *Handle.encoding functions
13:47 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45425/
13:51 Coke bacek: ok, you're still committing to trunk.
13:52 bacek Coke, It's one of the old commits. I reverted them all...
13:53 Coke bacek;which commit reverted them?
13:53 Coke I see 45426 which mentions the /last/ of the reverting, but nothing before that indicating a revert.
13:54 bacek Coke, erm... Let me check
13:55 Coke does 45425 have an incorrect log message?
13:56 bacek Coke, it is not...
13:56 bacek Looks like git-svn lost revert commit somehow
13:57 Coke WHEE.
13:57 ash_ joined #parrot
13:59 Coke bacek: are you sure 45425 isn't the revert commit? it seems to add back in Parrot_str_copy...
14:01 bacek Coke, you right.
14:01 bacek 45425 is actual revert commit...
14:03 patspam joined #parrot
14:03 * bacek bang his head against desk and swear never dcommit anything close to midnight...
14:04 lucian joined #parrot
14:05 dalek parrot: r45426 | bacek++ | trunk/src/call/args.c:
14:05 dalek parrot: Revert last bits of wrong commits. bacek--
14:05 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45426/
14:14 bubaflub joined #parrot
14:20 Andy Coke++
14:20 Andy attribute sniffing FTW
14:29 payload joined #parrot
14:31 kthakore .j #perl-help
14:34 Coke Andy: i hope I deserve that karma, but I'll take it. =-)
14:40 theory joined #parrot
14:40 ruoso joined #parrot
14:43 Andy I wonder if the PARROT_HOT is going to do any good, though.
14:57 Coke Andy: is there a non-gcc analogue?
14:57 Andy In other compilers?  Probably
14:58 ruoso joined #parrot
15:36 frzntoz joined #parrot
15:55 payload joined #parrot
15:55 cotto_work joined #parrot
15:56 cotto_work joined #parrot
16:00 frzntoz left #parrot
16:13 aukjan1 joined #parrot
16:37 ruoso joined #parrot
17:20 payload joined #parrot
17:32 PerlJam socghop should have some sort of switch to deal with the scoring.  Currently the score is completely cumulative which allows the same person to inflat the total by sumbitting multiple scores.  A saner approach IMHO would be to have a switch to allow for "last score wins" for each user.
17:34 Austin Or maybe they just DQ the entire project if someone tries to pump the score.
17:35 jan joined #parrot
17:57 dukeleto PerlJam: yeah, that is wonky
18:01 Coke scoring is internal to an org, though.
18:02 Coke so it can be a social solution instead of a global technical solution that might not fit every team.
18:03 darbelo It also means students only need to bribe one person to inflate their scores.
18:06 darbelo It'd get prohibitively expensive if we had to pay to all of the mentors...
18:07 Austin I thought that was the whole reason for mentoring...
18:07 Austin Can't afford Google stock? Become a mentor...
18:10 darbelo I doubt it's going to be enough. There's only so much you can sqeeze out of a student, after all.
18:15 dukeleto i can see each vote, so if I see funny business, I will get out my Big Stick. Problem Solved.
18:23 darbelo dukeleto++ # speak softly...
18:49 chromatic joined #parrot
18:49 whiteknight I'm happy to see so many proposals in
18:50 whiteknight 4 so far, 5 if you count the two separate proposals from Robert
18:51 dukeleto whiteknight: i need to poke the interwebs and remind people that the deadline is friday
18:52 whiteknight dukeleto: I'll blog about it. Anything else you want me to do?
18:53 dukeleto whiteknight: remind people that the deadline is Friday and include a link to the timeline for exact time + others dates/etc
18:53 whiteknight do you have that link?
18:53 dukeleto whiteknight: encourage students to send their apps to some kind of mailing list/irc channel first
18:54 dukeleto whiteknight: http://socghop.appspot.com/document/sho​w/gsoc_program/google/gsoc2010/timeline
18:54 whiteknight nice. Thanks
18:54 dukeleto whiteknight: April 9:
18:54 dukeleto 19:00 UTC is the app deadline
18:54 Coke (in addition to blogging, post a news story on parrot.org)
18:54 Coke (and perhaps email the parrot-dev list.)
18:54 dukeleto Coke++
18:54 dukeleto i need to email p5p, p6a and friends again
18:55 * dukeleto has been enjoying vacation for the last few days
18:57 bubaflub dukeleto: nice!
18:58 dukeleto bubaflub: indeed
19:01 dukeleto bubaflub: nice blog posts
19:01 dukeleto bubaflub: could you make your timeline breakdown weekly
19:02 dukeleto bubaflub: just estimate stuff. things will inevitably change
19:02 bubaflub dukeleto: sure thing.
19:06 darbelo whiteknight: http://www.timeanddate.com/counters/customc​ounter.html?month=04&day=09&year=20​10&hour=19&min=0&sec=0&p0=0
19:06 darbelo That'll help resolve timezone issues ;)
19:07 darbelo ZOMG! Only 172360 seconds to go!
19:07 clinton joined #parrot
19:17 chromatic Oh, git-svn!  You let me cherry-pick bacek's accidental commits to trunk to the branch where he meant to commit them.
19:18 dukeleto chromatic: is that good or bad? i think cherry-pick is pure git-fu
19:19 chromatic It's very good.
19:20 ash_ joined #parrot
19:22 dukeleto this is pretty cool: http://news.perlfoundation.org/2010/0​4/grant-report-changing-the-perl.html
19:22 dukeleto except for all of the failing tests
19:27 bubaflub that comment was pretty harsh - has anyone else actually used this code?  does it work?
19:30 Coke I must admit that failing your own tests doesn't look good.
19:31 * Coke pings the GC on that one.
19:31 chromatic I understand this to be a work in progress, not something at a milestone of completion yet.
19:37 dukeleto looks like we have about 12 decent gsoc proposals already. cool.
19:42 darbelo Nice. How many of those are for parrot?
19:43 joeri joined #parrot
19:46 davidfetter joined #parrot
19:46 dukeleto bubaflub: add your possible mentors to your app as well
19:46 bubaflub dukeleto: roger that.
19:47 dukeleto darbelo: 5
19:47 dukeleto whiteknight: ping me when you make your blog post, and i will link to it in my upcoming post
19:48 whiteknight dukeleto: going to have to wait till after work
19:48 dukeleto whiteknight: no big deal, i will link to your ideas list
19:48 whiteknight ok, awesome
19:53 Andy joined #parrot
19:58 ash_ whiteknight: i updated my GSoC proposal, to be libffi and llvm stack frame builder
19:58 whiteknight ash_: nice
19:58 whiteknight benchmarks? comparisons?
19:59 ash_ should I put those into the proposal? talk about comparisions between the current one and my new work i'd do?
20:00 TimToady phone
20:02 whiteknight ash_: if you plan to do them, put hem in the proposal
20:04 hercynium joined #parrot
20:20 particle anybody know a way to make a trac wiki page viewable only by a permission group or set of users?
20:22 darbelo Trac has permission groups?
20:22 darbelo For the wiki, I mean.
20:27 cotto_work You'll probably have to find an extension.
20:31 particle PrivateWikiPlugin would do it, it seems
20:31 particle thanks for the pointer
20:40 chromatic http://calculist.blogspot.com/2010/0​4/design-space-of-continuations.html
20:53 mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#33050), fulltest) at r45426 - Ubuntu 10.04 beta amd64 (g++ with --optimize)
20:54 mikehh nothing seems to be broken at the moment
20:57 chromatic You sound surprised.
20:58 mikehh thought there were some mis-commits to trunk rather than a branch
20:59 darbelo They got undone.
20:59 mikehh ah
21:04 Coke (not that you'd know it looking at the commit logs messages. =-)
21:06 darbelo 11:04 <+dalek> parrot: Revert last bits of wrong commits. bacek--
21:07 darbelo Not his most verbose, but gets the ide across.
21:07 darbelo s/ide/idea/
21:10 kid51 joined #parrot
21:11 GeJ Good morning everyone.
21:12 kid51 good morning Geraud
21:15 cotto_work chromatic: PARROT_HOT and PARROT_COLD strike me as attributes that would be best applied in a somewhat automated fashion, similar to a pgo build.  Do you think it'd be a worthwhile optimization?
21:16 chromatic I think they vary too much between runs.
21:16 chromatic I can tell you the biggest hotspots: string comparisons, GC runs, and hashing.
21:16 chromatic Beyond that, it's too subjective.
21:17 chromatic Incoming cherry picks to the right branch in 3... 2... 1...
21:17 moritz 0... -1... -2 ...
21:18 dalek parrot: r45427 | chromatic++ | branches/immutable_strings_part1 (2 files):
21:18 chromatic Dalek is still fighting the time war.
21:18 dalek parrot: Broke the build; removed COW functions.
21:18 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45427/
21:18 dalek parrot: r45428 | chromatic++ | branches/immutable_strings_part1 (2 files):
21:18 dalek parrot: Remove str_set function. We can't set content of immutable strings
21:18 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45428/
21:18 dalek parrot: r45429 | chromatic++ | branches/immutable_strings_part1/src/ops/set.ops:
21:18 dalek parrot: Replace str_set in assign op with just assign.
21:18 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45429/
21:18 dalek parrot: r45430 | chromatic++ | branches/immutable_strings_part1/src/call/args.c:
21:18 dalek parrot: Remove str_new_COW in args passing.
21:18 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45430/
21:18 dalek parrot: r45431 | chromatic++ | branches/immutable_strings_part1 (2 files):
21:18 dalek parrot: Remove Parrot_str_resize function. "Resizing of rock-stable rocks? What
21:18 dalek parrot: you are talking about?"
21:18 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45431/
21:18 dalek parrot: r45432 | chromatic++ | branches/immutable_strings_part1 (2 files):
21:18 dalek parrot: Remove Parrot_str_copy. "Rock is always a rock. Why copy it?"
21:18 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45432/
21:18 purl i already had it that way, dalek.
21:18 dalek parrot: r45433 | chromatic++ | branches/immutable_strings_part1/src (3 files):
21:18 dalek parrot: Remove few str_copy calls
21:18 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45433/
21:18 dalek parrot: r45434 | chromatic++ | branches/immutable_strings_part1 (23 files):
21:18 dalek parrot: Remove a bunch of str_copy usage
21:18 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45434/
21:18 dalek parrot: r45435 | chromatic++ | branches/immutable_strings_part1/src/string/api.c:
21:18 dalek parrot: [string] Removed some string copy operations, as their functions have
21:18 dalek parrot: disappeared.
21:18 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45435/
21:18 ash_ do i need to send a CLA (the one on the perl foundation) for GSoC?
21:19 ash_ http://www.perlfoundation.org/​contributor_license_agreement is what I am referring to
21:19 moritz ash_: only if you want to contribute to Rakudo or Perl 5 core
21:19 darbelo ash_: Only if you intend on working on something that would require it.
21:19 moritz for parrot, that's not necesary
21:19 ash_ okay, just checking, wanted to make sure i had everything in place
21:19 moritz there you need the PaFo CLA
21:20 kid51 Parrot has its own CLA, which you need to get a commit bit (ISTR)
21:20 ash_ k, i'll look for that then
21:20 darbelo Last year I completed my summer of code project without having to send in *any* CLA.
21:20 darbelo But it was an exception, rather than the rule.
21:21 cotto_work That's because your code lives outside Parrot's svn repo.
21:21 ash_ so, is that something you have to sign if your code is going to live inside parrot's svn?
21:22 darbelo Right. I had to send in a parrot CLA to get my parrot commit bit.
21:22 darbelo purl: CLA?
21:22 purl rumour has it CLA is Contributor License Agreement or http://www.perlfoundation.org/​contributor_license_agreement or http://www.parrot.org/foundation/legal or http://www.parrot.org/files/parrot_cla.pdf
21:22 darbelo http://www.parrot.org/files/parrot_cla.pdf
21:23 particle ash_: yes, if you want patches accepted to a pafo repo, we need a signed cla
21:24 ash_ k, well, my work (i think) might end up in a parrot repo. so i guess i should go on and sign one just in case? it can't hurt i don't think
21:24 particle yes, please.
21:24 ash_ ah boy, i get to print something! ... is it sad that i don't print stuff, like ever...
21:27 particle chromatic: is the tpf cla now headed your way, instead of kurt's?
21:29 Whiteknight joined #parrot
21:32 GeJ kid51: Heya James. How are you?
21:32 chromatic I don't think so, particle.
21:32 particle (2:18:05 PM) bkeeler: Hmm, there appear to be two CLAs on the perlfoundation website, with different mailing addresses
21:32 particle (2:18:15 PM) bkeeler: http://www.perlfoundation.org/​contributor_license_agreement and http://www.perlfoundation.o​rg/attachment/legal/cla.txt
21:32 particle (2:18:56 PM) bkeeler: I sent mine to the Michigan address, not the Oregon one
21:33 particle ok, just curious.
21:33 purl i think just curious is all
21:34 dukeleto ash_: GSoC requires no CLA
21:35 dukeleto ash_: but feel free to sign the Parrot CLA, then you can commit to core, fix docs, add tests, etc...
21:35 Whiteknight yeah, I had to fill out a CLA for my GSoC project
21:35 Whiteknight but I was working in the core repo
21:36 chromatic I know Onyx Neon doesn't have that PO Box anymore.
21:38 TiMBuS joined #parrot
21:39 particle ok, i'll mail directors@tpf about it
21:40 kurahaupo joined #parrot
21:48 tcurtis joined #parrot
21:50 Austin_Hastings joined #parrot
22:14 bacek ~~
22:14 bacek morning
22:17 Coke darbelo: yes, but the commit before that one is the one that does 99% of the reverting. =-)
22:17 cotto_work hi bacek
22:18 darbelo Coke: Oh. Hadn't noticed that.
22:19 Coke it's all good now. =-)
22:20 darbelo I figured the commit that said 'revert' was the one reverting ;)
22:20 darbelo I should know better by now, but I'm a slow learner.
22:22 bacek aloha cotto_work
22:27 GeJ G'Day bacek.
22:40 kurahaupo joined #parrot
22:46 payload joined #parrot
22:56 dalek parrot: r45436 | bacek++ | branches/immutable_strings_part1/src/pmc (2 files):
22:56 dalek parrot: Fix usage of long-standing bug with wrong returns from *Handle.encoding functions
22:56 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45436/
22:56 dalek parrot: r45437 | bacek++ | branches/immutable_strings_part1/src/pmc (2 files):
22:56 dalek parrot: Fix *Handle.mode functions to return correct result
22:56 dalek parrot: review: http://trac.parrot.org/parrot/changeset/45437/
22:57 darbelo bacek: didn't chromatic already re-commit your changes?
22:58 darbelo http://irclog.perlgeek.de/p​arrot/2010-04-07#i_2206683
22:58 chromatic I wasn't sure if those last two belonged on the branch.
22:59 darbelo Oh, my bad then.
23:05 bacek They are belong to branch.
23:05 bacek I double checked it...
23:07 bacek Anyway, $dayjob time.
23:07 bacek See you!
23:07 ruoso joined #parrot
23:13 tetragon joined #parrot
23:13 Mokurai joined #parrot
23:24 Whiteknight allison++
23:28 Whiteknight I like the dev meetings
23:30 darbelo Ouch. Looks like I'll miss it.
23:34 Whiteknight noooooooooooooo!!!!!!!!!!!!!!!!
23:35 darbelo It's not *that* dramatic.
23:35 darbelo You make is sound like I'm valuable or something ;)
23:36 cotto_work Me too.  it'd be nice for me if it were later in the afternoon Sunday.

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

Parrot | source cross referenced