Parrotsketch Every Tuesday @ 18:30 UTC | IRC Log at irclog.perlgeek.de/parrotsketch/today
Set by moderator on 8 December 2008.
00:11 Whiteknight joined 00:36 Whiteknight joined 01:10 davidfetter joined 01:21 particle joined 03:05 tewk joined 03:32 particle1 joined 05:44 particle joined 08:04 particle1 joined 10:22 cotto joined 10:24 particle joined 10:35 cotto joined 11:12 rdice joined 12:33 particle1 joined 13:58 Whiteknight joined 14:54 particle joined 15:26 pmichaud joined 16:51 davidfetter joined 17:14 particle1 joined 17:25 wknight8111 joined 17:36 andrew_ joined 17:49 barney joined 17:55 kj joined 17:59 allison joined
kj Pre-reporting here, as I'll have to leave now. 18:00
== PIRC
+ solved the bytecode generation for keys problem, which turned out to be not too difficult
+ all bytecode stuff should now be working, kinda (needs A LOT OF testing)
+ now working on tying up all sorts of loose ends, such as namespaces, :outer, etc.
+ converted tests so that they'll actually be run by Parrot
+ finishing PIRC first version should not take very long; expect few weeks.
== Question
+ I'd like to have a string_unescape_cstring() function that returns a C-string,
not a Parrot STRING *. Can that be fixed? (Otherwise I'll be implementing it
myself in PIRC).
EOR
18:02 masak joined
allison I have a meeting 30min before #parrotsketch, so may be late. Putting my report in channel: 18:03
- Spent the week mostly on branch/ticket review.
18:03 chromatic joined
allison - Merged in the calling conventions branch. 18:03
- Merged in the GC branch.
- Caught up on parrot email, whittled my parrot-dev box down to 30 messages.
- Started talking with with Simon on strings refactor.
EOR
18:26 jonathan joined 18:29 coke joined 18:30 jhorwitz joined
chromatic hello 18:30
jonathan ahoj
jhorwitz hi
barney hi 18:31
coke ~~
Whiteknight hello
masak sa'
chromatic allison pasted earlier. Let's go in alphabetical order.
barney?
barney Added a Configure.pl to befunge, pipp and hq9plus.
Whiteknight allison and kj left their reports in the channel
barney Started playing with git.
HQ9+ has left the nest and is now on github.com/bschmalhofer/hq9plus/tree/master.
Fiddled with Parrot::Distribution.
.eor
chromatic Applied a few patches.
Worked on the GC refactoring branch, which meant moving around functions and files.
Will work with Whiteknight on specifying a GC API behind which to hide GC specifics. 18:32
Poked at Parrot::Embed.
Will work on ops review and PMC review, looking for things to deprecate.
coke?
coke lists.parrot.org/pipermail/parrot-d...00920.html
(and q1q)
chromatic cotto? 18:33
cotto * various minor test improvements (args assert, pirification, string->key conversion)
* resolve meth conflicts (much easier than IRL)
* fix a Gash misbehavior wrt freeze/thaw and clone
* planning on attacking PMC_foo_val deprecation
* various patches from GeJ
chromatic japhb?
jhorwitz? 18:34
jhorwitz got mod_parrot working on solaris and freebsd.
spent time fixing mod_perl6 for recent parrot releases -- still have issues that i'll need some help with. will post on #parrot or the list. 18:35
EOR
chromatic jonathan?
jonathan [Parrot]
* Got more bytecode annotations stuff done; added support to ParrotInterpreter after feedback from the list
* As particle++ pointed out, still need a bit more testing on it to check the persistence accross bytecode; did check it worked while doing it, but needs tests
[Rakudo]
* Helped pmichaud a little with the rvar branch, and it got merged in
* Finished (hopefully) type registration; the class/sub distinction on starting with a first uppercase letter is gone; no regressions that I know of
* Need to do a bit more on that stuff, so we get some better error messages; studying STD.pm to see more of what it does
* Currently working on initial support for parametric roles
* Once I've got initial parametric role stuff in place and essentially working, will look at dispatchy stuff. Hope that by this time next week we'll have junction auto-threading working. 18:36
.end
chromatic masak?
masak * Took up November again. Made all tests pass in the master branch.
* Next day, rvar branch in Rakudo merges. Spent weekend collecting the debris.
* Turns out it wasn't so bad, just a lot of small things.
* Fixed Druid too, by applying pmichaud++ to the problem.
* 15 rakudobugs since last week! Each of them earnest and meticulate.
* S29 and C<unpack> haven't gotten more care and feeding since last week.
* Just about to pull together a documentation effort for Perl 6.0.0.
EoR
chromatic particle?
particle ~ coordinated some roadmap tasks, answered questions 18:37
~ encouraged hlls to leave the nest
~ largely finished with Perl 6 Synopsis 19, command-line syntax
~ started preliminary work on a parrot lib to parse perl 6 command line last night
~ looking for work is a full-time job, not enough time for parrot and perl6
~ hopefully soon i find work and have more time to hack
~ currently ~84 failing rakudo tests due to Inf/NaN support, working on a parrot patch for that
.end
chromatic PerlJam?
PerlJam lurking. 18:38
eor
chromatic pmichaud?
pmichaud * mostly worked on the rvar/rvar2 branches this week, merged to trunk
* jonathan++ for his help with that
* since then I've just been cleaning up tickets and other items that were waiting on the refactor 18:39
* also fixed up Rakudo's handling of Scalar variables
* plan for this week is more cleanups, do a Cursor refactor for PGE so that we get line numbers in error messages
EOR
chromatic Tene? 18:41
tewk? 18:42
Whiteknight? 18:43
Whiteknight BOOK
- Wrote and updated sections about Exceptions and Exception Handlers
- Wrote sections about objects and classes
- Received Perl 6 Microgrant to do more book-work
JIT
- talked with chromatic++ about the new GC API
- abandoning pdd09gc_part1 branch for now to do some more cleanup work like this
CALLING CONVENTIONS
- started laying some groundwork for :lookahead parameters
- created pcc_invoke_convert branch to continue 18:44
- Working to convert calls from Parrot_PCCINVOKE to Parrot_pcc_invoke_method_from_c_args
STM
- ripped out the STM subsystem in the removing_stm branch
- hoping to merge that branch back in today
MISC
- Fixed 'new , [Classname]' syntax in IMCC (TT#137)
- Dug into ExceptionHandler subclassing problems (TT#154), no resolution yet
- followed around donaldh++ and dealt with some tickets/patches he submitted
- Removed pmc->real_self, patch courtesy donaldh++
- dealing with a bunch of tickets, especialy in preparation for next weeks's release
- practicing install on various platforms and raising issues related to that
- Been playing with Padre and Padre::Plugin::Parrot
QUESTIONS
- queue one question about TT#102
EOR 18:45
chromatic Anyone I missed?
coke kid51's report. 18:47
>>
A. Same infrastructural concerns as last week: 1. parrot.org has many dead links. See: trac.parrot.org/parrot/ticket/45.
2. Trac system zaps contributor accounts when you set Preferences. See: rt.perl.org/rt3/Ticket/[.]lay.html?id=61870.
B. Code review on trac.parrot.org/parrot/ticket/160.
C. GeJ++ for commit bit.
<<
cotto C addresses my question. 18:48
chromatic Coke?
cotto C in kid51's report
coke regarding my email about the DEPRECATIONs, please consider grabbing one of those remaining tickets if you can.
this is a high priority for the upcoming release. 18:49
er, my question being, I suppose "can anyone help me close these tickets?"
.
chromatic Any other comments about GeJ's commit bit?
Whiteknight +1
coke +1
chromatic +1 18:50
Who would like to mentor?
Whiteknight I will if nobody else has dibs
chromatic Alright, walk him through sending in the CLA.
Whiteknight okey-dokey
chromatic Whiteknight, you had a question. 18:51
particle oh, i have a comment.
pmichaud I have a comment.
Whiteknight yes, we have a few opcodes such as "cpu_ret" that say "don't use", and if you use them anyway it causes bad thigns to happen
(segfaults, etc)
obviously we want to avoid this with more force then a simple "don't use" message, something graceful 18:52
I thought about hacking IMCC to catch these few in the lexer and throwing an exception, but that doesn't seem like a great solution
any other ideas for this? (TT#102)
chromatic Fix CGP so it doesn't need that? 18:53
particle cgp is a gcc-only runcore
Whiteknight that's a better idea certainly, but requires a few tuits
particle no other compiler supports that construct.
Whiteknight I'm pretty sure cpu_ret is used by JIT as well
particle it gives us speed, but not portability
ok, then that's a different story. 18:54
chromatic Are you suggesting that we don't need CGP?
particle we don't *need* cgp.
chromatic I think a context-threaded runcore might be more portable, but I don't have time to write it and I don't exactly know how to do it yet.
particle i'm happy if parrot 1.0 has two cores, slow and jit.
coke that's one core for me: slow, and slow. =-)
chromatic We need a third core: veryslowtcl. 18:55
Whiteknight if portability is a major concern, JIT doesn't hardly fit the bill anyway
particle i suppose we'd still have trace and gcdebug cores too
coke ... I've already got one! </python>
Whiteknight okay, so best solution is avoid the dependence on these opcodes in the first place, and then we can deprecate/remove them entirely?
barney Bytecode needs to be portable, not runcores 18:56
chromatic Improving JIT support is a SMOP, while making CGP work on other compilers means patching those compilers, or slowly buying a place on Microsoft's board so that Visual Studio will eventually support the C language.
particle barney, true enough. 18:57
Whiteknight EOQ 18:59
chromatic Queued comments?
particle gsoc2009 is gearing up. we need some folks do volunteer for the hype machine. make yourselves known. 19:00
pmichaud We seem to have a lot of cruft left over from website transitions, mailing list transitions, etc. Who are the ones we contact for updates/fixes to these? 19:01
cotto Could we get a gsoc2009 tasklist wiki page?
pmichaud "could we get a" --> "I'll go make a" 19:02
coke pmichaud: trac tickets are probably the best way to go there.
cotto I ask because I don't know what needs to be done.
pmichaud coke: I already filed trac tickets, so did kid51. I'm wondering who/what takes them from there.
coke since some of them require svn access to perl.org, some require OSU, some require ask & robrt...
cotto but a stub is easy 19:03
coke pmichaud: complaining here is probably a good first step.
pmichaud coke: right, but who is empowered to start working on them?
so far I'd just be forwarding them to particle, but I'm not sure that's correct.
coke empowered - anyone with the appropriate commit bits? if there are volunteers, we can work on getting them access.
pmichaud I'd be glad to help with them, but I don't think I have appropriate bits. 19:04
and I don't know who the current owners-with-commit-bits are.
coke I'll try to go through the open tickets and see what I can do.
(in terms of giving them away. =-)
pmichaud can I assign tickets like that to you for the time being, then? ;-)
particle allison and i have an open relationship with osuosl and perl.org. i'll take ownership of any infrastructure tickets.
pmichaud you can re-assign as appropriate.
19:05 allison joined
particle anyone can reassign them, right? 19:05
(like rt)
pmichaud sure, I'm just wanting to see that they get taken care of.
particle er, anyone with bugadmin, which is anyone we care about ;)
coke there's lots of stuff on my plate. I'll see what I can do.
allison is also a good person to go through the website/mailing list-related trac tickets. 19:06
(looking for people to give permissions to.)
pmichaud over the next two weeks I expect to be cleaning rakudo documentation issues, it would be convenient to do (or arrange to have done) the parrot items as well.
chromatic Can't think of a better time to do it!
pmichaud another comment: next tuesday is release day, get NEWS, PLATFORMS, etc. updates in 19:07
chromatic ESPECIALLY DEPRECATION NOTICES
allison who's taking this release?
(not listed in release manager guide)
pmichaud chromatic gets the 0.9.0 release, iirc :-) 19:08
allison oh, right, it's still january. then it is listed
Whiteknight it's chromatic++
chromatic Mostly because I've had a codename for this release for over a year.
Other questions? 19:09
coke I have a just added a report by component to trac:
trac.parrot.org/parrot/report/12
barney trac.parrot.org/parrot/ticket/170 might be a GC issue
coke (should help find all the, say, website tickets.)
pmichaud coke++
coke (and now it even works) 19:11
chromatic Other comments?
cotto There's a stub at trac.parrot.org/parrot/wiki/GSOC2009Tasklist now.
coke cotto++
cotto it's stubby
chromatic Other comments or questions? 19:14
pmichaud Parrot is -Ofun. 19:16
19:36 particle1 joined 19:44 chromatic left 20:41 coke left 20:50 jhorwitz left 20:51 allison joined 21:47 Tene joined 21:56 particle joined 22:05 Whiteknight joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!