Vision for 2.0: Production Users | Don't forget to post closed tickets in your report. | Note: This channel is only for our Tuesday status meetings; you probably want #parrot instead. | irclog: irc.pugscode.org/
Set by moderator on 2 February 2010.
00:38 cotto_work joined 07:02 bacek joined 07:44 bacek joined 10:09 cotto joined 12:23 bluescreen joined 13:22 plobsing joined
plobsing ~ What I Did: 13:24
* merged tt362 and opengl_dynamic_nci branches
* worked on nci_thunk_gen (fperrad++ for making it work on windows)
* made installable
* added extra_nci_thunks dynext library
What I Plan:
* will miss #ps today
* move some TTs along
* continue work on nci_thunk_gen
* add --without-extra-nci Configure.pl option to build without non-essential nci thunks
* move signatures no longer needed in core from src/nci/core_thunks.nci to src/nci/extra_thunks.nci
EOR
14:48 mikehh joined 17:05 bluescreen joined 17:09 cotto_work joined
mikehh What I did since my last report: 17:33
* building and testing parrot on amd64/i386, with gcc/g++, with and without --optimize
* a lot of branch testing 17:34
* then Coke++ included .t files in codetest
* spent the last couple of days getting codetest to PASS, all except Perl::Critic
* fixed hundreds of codetest failures, plus some where trailing spaces were required
* in the heredoc or equivalent CODE/OUT
* All test now PASS except t/codingstd/perlcritic.t (up to fulltest) which fails
* 33 subtests out of 659 - it used to run 306 subtests before the .t inclusion
What I intend to do in the next week:
* getting P::C to pass the .t tests
* other testing and fixing if I have any time
* looking at cleaning up some tests
* documentation
.eor
forgot to mention that there are now no missing C function docs - some still require documenting (boilerplate only) 17:38
and the I also updated the info in the wiki 17:39
18:14 bluescreen joined 18:30 allison joined
allison May be a bit late to #parrotsketch today. 18:31
Last week:
- Attended Python language and VM summits, spoke at PyCon.
- Two people from my talk submitted failing tests, which I committed with fixes for 2 out of 3 of the failures.
- Did a full review of the state of Pynie for my talk, and was impressed to see how close we are to the full pure Python feature set. It's given me a boost of motivation for the last mile.
- Started a branch for the PCC 'get_results' reorder.
- Some delay on getting Debian packages in, waiting for sponsors to upload them.
- Got a feature freeze exception for the Ubuntu packages to get them synced in Lucid from Debian even though we're a few days past the usual deadline.
Next week:
- Continue working on Pynie.
- Start changes to PCC branch.
EOR
18:41 bacek joined 18:53 pmichaud joined 19:50 chromatic joined
chromatic I fixed a few bugs (including a GC bug which led to 2.1.1). 19:51
I made some performance improvements, which directly show up in Rakudo's spectest run.
I reviewed a couple of patches.
I will continue to work on the MMD/HLL subclassing bug.
I'm blocking on time... and some help on the TT #389 branch.
19:55 davidfetter joined 20:02 NotFound joined
cotto_work EOR 20:05
20:09 darbelo joined
darbelo PAST: * Adopted the gdbmhash PMC: moved it out of the repo, and deprecated it. * Did the same for the digest dynpmcs, but then gave them back to fperrad. * Stumbled across TT #903. Me *WANT*. * Did some testing of rm_cflags on a non-gcc compiler. * Tweaked the OpenBSD hints file. * Not much else. 20:09
FUTURE: * No fixed plan.
END
BLOCKERS: * Tuits turned out to be square. Have to wait for round ones.
XTERM PASTE: * fail. 20:10
Let's try that again.
PAST
* Adopted the gdbmhash PMC, moved it out of the repo, and deprecated it.
* Did the same for the digest dynpmcs, but then gave them back to fperrad.
* Stumbled across TT #903. Me *WANT*. * Did some testing of rm_cflags on a non-gcc compiler. * Tweaked the OpenBSD hints file.
* Not much else. 20:11
FUTURE
* No fixed plan.
BLOCKERS
* Tuits turned out to be square. Have to wait for round ones.
END
Util # Done:
* Helped with Win32 0xC0000005 bug on perl6-compiler list.
* Played with Coke's `$conf->print_if_verbose` idea.
# Plan to do:
* Plans delayed 1 week.
# Blockers:
* Traveling this week for anniversary, and for Atlanta.pm.
= Might miss part of today's meeting, too.
.end
20:13 dukeleto joined
dukeleto Did: 20:13
* Worked on PL/Parrot 20:14
* Worked on porting Parrot to RTEMS and BUG platforms
* Wrote some tests to detect GC memory leaks, but still more are needed
* Answering questions on parrot-users
Will do:
* Write a blog post summarizing recent developments in the Parrot core + ecosystem
* More of the same
Blocks:
* Real life
bacek Done: 20:15
- mem_sys_foo cleanups
Todo:
- more mem_sys_foo cleanups.
- More Boehm GC work.
EOR
davidfetter Done:
* Inter-project communication between PL/Parrot & Pg 20:16
* Fixed immediate problem that caused the aforementioned
TODO:
* Moar PL/Parrot hacking
EOR
NotFound What I did: 20:18
- parrot: minor fixes.
- Winxed: minor improvements,
- parrot and Winxed:
* Working towards PIR/PASM compilers written in HLL, using the
packfile PMCs ande the new Op PMCs.
Got a disassembler and a very limited assmbler working.
What I will do:
* More on the same.
EOR
japhb Done: Resurrected one of two dead computers; got back to being able to compile parrot & plumage; minor help to kthakore with SDL/NCI stuff. Future: Hopefully produce some Plumage features for fperrad, who has gotten well ahead of me since I was out. EOR 20:23
chromatic Hello. 20:30
mikehh hi there
darbelo Hola. 20:31
cotto_work hi
Util Hello 20:32
japhb o/
NotFound Hola
davidfetter nin3 hao3
chromatic Let's review last week.
How are PCC deprecations going?
darbelo ENOALLISON 20:33
chromatic HLL subclassing and MMD: delayed thanks to the GC bug.
I'm working on it now; I think I know the problem and have a fix, but getting it past C is... well.
Other tasks from last week? 20:34
bacek aloha 20:35
chromatic Should we continue to make these priorities, or do we have others?
bacek I think PCC is our priority. 20:36
pmichaud hello, all
chromatic Okay. We need to get some tasks out of allison and start doling them out. 20:37
Suggestions or volunteers? 20:38
mikehh got most of codetest working again, still got to fix perlcritic stuff 20:39
20:39 allison joined
mikehh and here is alison 20:39
missing l 20:40
allison hi all (bit late)
Util Reference: trac.parrot.org/parrot/wiki/AllisonTasklist
dukeleto 'ello 20:41
chromatic allison, we're talking about the PCC task list. 20:42
dukeleto i am going to work on PDD18 (security) so I will take that off of allisons task list 20:43
do i just delete it or say that I am taking care of it? 20:44
mikehh I would think the latter 20:46
dukeleto mikehh: just did it 20:47
allison: what does "security sandbox" mean on your task list? 20:48
davidfetter dukeleto, it's stuff analogous to Safe.pm
i.e. being able to restrict operations the VM can do 20:49
NotFound The skipped part of embed.t is shit. I skipped it instead of delete to remember that the functionality must be tested.
chromatic Let's set priorities for the coming week then. 20:50
Besides PCC, is there anything else?
dukeleto what does a priority of PCC actually mean?
NotFound My personal priority is the packfile and oplib PMCs.
bacek I'm going to finish GC cleanups/Boehm encapsulation.
mikehh for me getting perlcritiic working
so all tests pass again 20:51
chromatic PCC refactoring means reordering the get/set params/results ops.
bacek NotFound, we can expose SUB_PF_foo flags into PIR. I think it's much better than providing methods on Packfiles.
dukeleto chromatic: who is actually going to work on that? Can that be a single person or a small groups priority? Because most people (in my opinion) do not feel comfortable working on that 20:52
NotFound bacek: methods has the advantange of being easily usable from HLLs.
dukeleto it seems more beneficial to have a priority that many people can work on
bacek NotFound, "int flags; flags = .SUB_PF_MAIN;" is hard? 20:53
chromatic I'd like to have multiple people working on it.
NotFound bacek: if you want to talk about that, better in #parrot
bacek NotFound, ok.
chromatic ... but we need a list of concrete tasks (and maybe tests). 20:55
bacek ... and IMCC expert.
allison dukeleto: yeah, davidfetter has the right explanation of "security sandbox" 20:56
chromatic I can do the IMCC parts.
dukeleto allison: ok, i will do the "security sandbox" and PDD18 stuff.
allison: it goes well with my work on PL/Parrot 20:57
chromatic Any questions we need to discuss? 21:02
Tene My report: Busy all week, had no time available. 21:06
I plan to work on exceptions this week, if I have time. I'd love to work on PCC too. 21:07
chromatic allison, what can we do to make the PCC tasks easier and self-contained? 21:08
allison back 21:18
the tasks for this branch are pretty self-contained 21:19
I'm limiting this branch to the smallest set possible to reorder the two opcodes 21:20
NotFound allison: There is some transition plan, or will be a full incompatibility step?
allison I can work with anyone who wants to take pieces of it
Only PASM code should experience any incompatibility 21:21
for PIR code, it'll work the same (just do something different behind the syntactic sugar)
I'm not aware of any large PASM codebases outside the Parrot repository
darbelo So, IMCC hacking required?
dukeleto allison: let me know if any PASM tests need to be fixed in yor PCC work. i can help with that
the largest PASM codebase is our test suite, AFAICT 21:22
allison but if there are any, we can work with them
NotFound allison: yes, but I ask because of my current experiments on pbc generation.
I'll delay the PCC part if is about to change.
allison dukeleto: excellent, yes, that's the largest collection of PASM code
NotFound: it won't change substantially enough to get in the way of the PBC generation
will just be a small change 21:23
NotFound allison: ok
allison so, please don't delay that
and, hopefully I can merge the branch in within 2 weeks
NotFound Great
chromatic Anything else we need to discuss? 21:25
dukeleto i have one question 21:29
is it possile to get TPF grants to work on Parrot? How does that work?
if that is not appropriate for #ps, we can continue in #parrot 21:30
NotFound That remembers me: if someone wants to contact Nokia for sponsoring, tell them we have parrot working in the N900 21:32
dukeleto hears crickets and thinks the meeting may be over 21:38
chromatic Yep, looks that way.
dukeleto ENODISHES 21:39
dukeleto moves back into #parrot
21:39 dukeleto left 21:41 NotFound left 21:45 davidfetter left 21:57 PacoLinux left 22:32 chromatic left 22:37 Whiteknight joined 22:57 darbelo left 23:04 plobsing left

Powered by Cro and the Raku® Programming Language.

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

Thank you!