Parrotsketch Every Tuesday @ 18:30 UTC | IRC Log at irclog.perlgeek.de/parrotsketch/today
Set by moderator on 8 December 2008.
02:06 Whiteknight joined 06:45 particle joined 09:59 samlh joined 12:13 jimmy joined 12:14 jimmy left 13:09 Wknight8111 joined 14:39 PacoLinux joined 14:48 PacoLinux left 17:02 masak joined
masak I might not make it to the meeting today. but I have nothing much to report this week. 17:03
17:36 moritz joined 17:43 pmichaud joined
Tene PREREPORT 17:46
Got the beginning of Exception Handlers into PCT.
Hopefully finishing that tonight.
They won't work until PCT starts generating a prelude for .includes, though. 17:47
Started reviewing pdd24
I should have that done tomorrow.
KTHXBAI
17:59 Coke joined 18:10 PacoLinux joined 18:15 cognominal joined, jhorwitz joined 18:24 allison joined
moritz hello 18:30
18:30 chromatic joined
particle hi-o 18:30
Wknight8111 hello
allison hi
jhorwitz yo
Wknight8111 Tene and masak left messages/reports earlier
pmichaud hello. 18:31
chromatic Shall we begin? allison? 18:32
allison - Completed and merged in the I/O branch. 18:33
- Added a 'readline_interactive' method to the FileHandle PMC.
- At the Ubuntu Developer Summit.
EOR
chromatic I added a little bit of optimization. 18:34
I fixed all of the open Coverity Rung 1 problems. Now we're at Rung 2, with a lot more problems to fix. I can get more people accounts to help (cotto) and see about resetting passwords (tewk).
I didn't review NaN/Inf, but I will.
I have a couple of bugs to chase down. 18:35
Coke?
Coke - closed some tickets
- asked for help on next release
- tested io branch on tcl pre-merge
- have some volunteers to sort out for upcoming releases.
18:35 rurban joined
Coke - think tcl is still hitting GC issues in trunk, but will re-check post-io merge. Be nice if we could get that fixed with 0.8.2 18:35
EOR 18:36
chromatic japhb?
jhorwitz? 18:37
jhorwitz completed mod_parrot APR pool implementation. this was the last milestone for 0.5, so i will be prepping for a release this week, the first in almost a year.
recent postcircumfix[] changes in rakudo broke mod_perl6 horribly. helped pmichaud come up with a fix.
submitted a patch to get november working with mod_perl6.
EOR
chromatic moritz? 18:38
moritz * usual testing + bug reporting + rt cleaning
* as a user and tester I'm very pleased with Rakudo's progress
* thinking about writing a draft or RFC for (Perl 6) S24-Testing.
If somebody is interested in sharing thoughts or contributing, please /msg
me or ask on #perl6
.end
chromatic particle?
particle ~ tested pdd22io branch with allison, assign branch with pmichaud, calling_conventions branch with whiteknight 18:39
~ updated documentation in repo to point to parrot-dev rather than perl6-internals--which we should retire soon
~ looking at a funky bug in t/compilers/pct/complete_workflow.t
.end
chromatic rurban?
rurban still loaded with work. no time to look at install2
eor
chromatic tewk? 18:40
pmichaud? 18:41
pmichaud == Parrot stuff
: worked with tewk++ on :subid improvements
== PGE stuff
: Converted PGE::Match to use Capture PMC
** results in a ~13% improvement to rakudo building
== PCT stuff
18:41 kj joined
pmichaud : Converted PAST::Node to use Capture PMC 18:41
** results in a ~7% improvement to rakudo building
kj arrives late.
pmichaud : worked with Tene++ on exception handlers in PCT
: very good progress there
== Rakudo stuff
: list assignment now works
: array slicing now works
: ... so Rakudo now looks a lot more like Perl
: reduction operators ([+], [*], [/], ...) now work
: Perl 6 subs can now contain inlined PIR code
: temporarily added $*WARNINGS to disable "uninitialized value" warnings
.eor
18:41 jonathan joined
pmichaud queue 1 question. 18:41
jonathan (sorry I'm late!)
chromatic Wknight8111? 18:42
Wknight8111 * Worked on docs/book more
* Removed .param type "name" => value syntax from IMCC
* Worked on and talked about the calling_conventions branch problems
* Might not be able to merge this branch until after the release next week
* Helped register with a few questions, watching his progres
* Add one question to the queue
chromatic jonathan?
jonathan * Was flying home last Tuesday from long weekend in UK, so wan't at PS
* Started branch for bytecode annotations stuff
* Various bits of Rakudo hacking
* Nearly finished (1 hour to go) using up the DeepText MMD grant
chromatic Let's go to questions then. pmichaud has the first one. 18:43
jonathan Think that's about it...
.end
pmichaud Can we eliminate or otherwise disable the perl6-internals mailing list "soon"?
I expect we'll get a lot more rakudo users/hackers in the next two weeks, and it would be very nice for all concerned if rakudo messages weren't spilling into the perl6-internals mailing list. 18:44
(it's okay with me if parrot-porters continues to work, I'd just like to get rid of the name "perl6-internals")
particle pmichaud: i'm working on it now, or anyway the repo/website related changes 18:45
is everyone from p6i on parrot-dev?
pmichaud I'd like the email to bounce, or return an intelligent "if you meant to write for Rakudo, see p6c; if you meant to write for Parrot development, see parrot-dev" response
but let's get perl6-internals out of folks' address books.
particle i suppose we need to get ask/robrt involved in that 18:46
pmichaud I'll craft a suitable response message.
allison pmichaud: yes, it could be possible to change quickly
particle send it to me, i'll take charge
pmichaud we've already had a couple of instances of Rakudo messages going to p6i, and it will likely get worse very soon. 18:47
allison particle: and, yes, I subscribed the whole p6i list to parrot-dev
pmichaud okay, thanks. I think that's all I had for that question.
chromatic kj?
kj + work on PIRC bytecode generation
+ local branches work (gotos etc.)
+ non-branching instructions work pretty much as well (except PMC constant stuff)
+ PCC ops need to be generated, but I need to do a lot of thinking how to exactly.
.end
Coke remembers one other thing he did.
(migrate some wiki) 18:48
chromatic jonathan also had a question.
jonathan Did he?
Wknight8111 i did
chromatic "Did he?" counts.
Wknight8111?
Wknight8111 Who knows the JIT system? I got a lot of questions about it from register, and there are some documentation holes that I want to fill in there 18:49
he's keen on doing the LLVM backend
chromatic tewk probably knows the most right now.
For what it's worth, I went to a talk last night about SpiderMonkey's tracing JIT. We could easily do something like that. 18:50
For various values of "easily".
Wknight8111 and various values of "we"
okay, thanks
chromatic Other questions? 18:51
rurban but we don't trace yet
chromatic Tracing is almost trivial.
particle please update the roadmap. 18:52
it seems, with one week to go, that we're behind on our goals for the release
it would be nice to have a more accurate picture, though--by having more data
kj quickly goes to check out the roadmap... 18:53
particle trac.parrot.org/parrot/wiki/ParrotRoadmap
allison for the 1.0 vision, I'm thinking a post on parrot.org makes the most sense, does anyone want more?
or less?
or different? 18:54
pmichaud works for me.
chromatic Works for me.
particle +1
jonathan +
1
Coke +i
allison ok, I'll work on that next
pmichaud we may need to push pct honors hll to after release.
PerlJam allison: and links on perlbuzz.com, perlmonks.org, etc. if you can
kj particle: what tasks do we seem to be behind on? 18:55
particle those without status or champions
allison PerlJam: good idea, will do
chromatic Are there other questions? 18:56
allison particle: okay, everything has a champion now
PacoLinux I have a question 18:57
for 1.0 we want documents translations (the importants ones) ?
i can translate some documentation to spanish
allison PacoLinux: any translations we get are good, but we're not making translations a requirement until a later release (I think it was 2.0, Jan 2010) 18:58
PacoLinux ok
chromatic That means we won't hold a release for translations, but we'll gladly accept them!
allison PacoLinux: and any help you need to get started, we'll gladly offer
a translation can take a while, so start anytime 18:59
PacoLinux ok, I have some documentation translated, but I want to know what are the most -prefered- docs
Coke we don't know yet, I don't think. 19:00
No one has gone through the /english/ docs yet.
allison makes sense, if you post a message to parrot-dev, we can at least pick the top 2 (requires some thought)
PacoLinux I have the faq, and some more ..
allison your review is a good one, because you can imagine what a new user will need most 19:01
pmichaud in many cases, the "most preferred" docs would be the user-level documentation (language implementors, in this case), and several of those don't exist yet (planned Jan 2009)
PacoLinux ok
thanks
kj should we go through the tasks for 0.8.2 and see where extra effort/help is required? 19:03
particle we could do that quickly. 19:04
pmichaud: opcode set... on track?
kj w.r.t. the opcode review: there's a bunch that's not in 'ops.num'. Maybe now is the time to fix that? 19:05
compilation of parrot starts with 20 lines '... is not in ops.num' messages.
pmichaud slightly behind -- I could do it on the plane tomorrow though.
I don't think it's "at risk". Nor is it critical if it bumps to Jan (other than we don't want too many things to bump)
chromatic That one's pretty easy to do though. 19:06
particle agreed. list it as on track
chromatic I'll pencil that in for Thursday or Friday.
particle vision we've addressed
calling coventions?
seems at risk to me, given wk's earlier comment. Wknight8111? allison? 19:07
Wknight8111 there's still a strange test error that prevents merging
pmichaud does calling conventions work with rakudo spectest?
also, keep in mind that I'm unavailable tomorrow+Thu to do any troubleshooting with pct or rakudo
Wknight8111 not sure
particle i'll run spectest now in that branch and let you know 19:08
Wknight8111 if the branch doesnt merge today or tomorrow I'll probably push it off till after the release
pmichaud so if a merge occurs and "all hell breaks loose", I can't say that we'd be able to clean it up on the rakudo side before release.
particle ok, at risk it is. next: pct honors hll. 19:09
Wknight8111 exactly, I dont want to release hell before my first release :)
pmichaud we're behind on that one, only because there's a lot of dependencies.
pct itself can honor hll no problem, but I was wanting to see at least one HLL actually using it before calling it "done"
kj w.r.t. the HLL stuff: ISTR that HLL_map wasn't working well. Maybe I was dreaming ,but just checking: should it work?
pmichaud kj: there have been some .HLL improvements since then.
kj ok. missed them on the svn logs then. 19:10
19:10 allison joined
pmichaud I really think it'd be better to simply say we'll have hll in Jan release. 19:10
kj it would allow for a lot more testing 19:11
if that can be fixed early jan.
particle i'll push it there now
chromatic: nan/inf?
pmichaud yes, there aren't a lot of blockers, there are just a lot of things that have to be coordinated to make it happen.
chromatic particle, I believe that's just documenting what we need and making it happen and harassing platform people to make it so.
particle yes. will it happen by tuesday? 19:12
chromatic Yes.
kj maybe that can be split into a 'document it==pdd14' and a 'make it so', if necessary
allison chromatic: at the moment, the task is just to document it
particle great. tene: sounds like events is on track. exceptions, too?
allison make it so is later 19:13
pmichaud ENOTENE, but it's on track.
particle ok, right.
pct loops?
pmichaud also on track.
kj particle: is the 'make it so' part for numbers on the roadmap?
particle ok, that's everything, except... Wknight8111, are you ready for the release? 19:14
Tene particle: Exceptions is on track.
Wknight8111 I'm ready 19:15
Tene AFAIK, unless someone can bring up eny other issues with parrot exceptions
Wknight8111 at least, I think I am
pmichaud and, another report:
** Rakudo spectests (r33717): 235 files, 4842 passing, 0 failing
that's +153 from last week.
chromatic 158 to go...
pmichaud we might make 5k by tuesday.
allison pmichaud: excellent! 19:16
pmichaud There are a lot of tests and tickets that need reviewing.
Coke (parrot exceptions) there's a bunch of tic.... right.
jonathan I've got maybe another 15 coming from hypers. :-)
particle kj: i don't see it immediately
pmichaud and a lot of tests that weren't possible yesterday are now possible today.
kj particle: neither can I. Should it be added?
particle yes. find a good place 19:17
chromatic kj, particle, put that in for January.
I'll work on it.
kj ok. will do.
particle updates: trac.parrot.org/parrot/wiki/ParrotRoadmap
pmichaud oh, another item I forgot to report: www.pmichaud.com/perl6/commits-by-month.gif
shows our monthly commits and the current trend
particle please add any tickets to the roadmap if they're related. 19:18
Tene Coke: can you point me to the exceptions tickets, or assign them to me or something? 19:19
pmichaud (assign tickets)++
particle i'm looking forward to see more 'landed' status this week. great work, all. 19:20
chromatic Other questions?
Tene I'm unclear on the ful list of what "exceptions, debugging, exception types, resumable exceptions" means.
If that could be expanded somewhere so I can be confident about it, I'd appreciate it. 19:21
allison Tene: it was a direct type in of several sticky notes we grouped together at the summit
pmichaud resumable exceptions is done, yes?
allison Tene: the heart was just debugging resumable exceptions
Tene Yes, unless that implies auto-resuming non-fatal exceptions or something.
allison Tene: no, that's not spec'd 19:22
pmichaud "exception types" was initially meant as a review of our existing exception types
Tene exceptions thrown from PIR are resumable just fine.
allison pmichaud: ah, let's make that a separate task for next month
pmichaud that doesn't have to be done for dec.... right
it should probably be part of pdd23 review 19:23
particle (keeping the roadmap honest)++
Tene and the "exceptions" task?
pmichaud I think that was a generic label, not a specific task
particle yep
Tene Okay.
allison pmichaud: ok, then for Feb release
pmichaud the big thing I was looking for was that resumable exceptions worked
we also need to make sure we're clear on throw versus rethrow semantics
allison I'll move exception types to pdd23 review 19:24
Tene pmichaud: can you list the deps for pct+hll somewhere?
pmichaud Tene: yes, I'll do that. File a ticket for me to trac?
I can respond to the ticket.
(a little hectic here at the moment)
Tene nods.
pmichaud specific tasks for each of the roadmap items probably ought to become tickets. 19:25
Tene Should I make a ticked about making thos etickets?
;)
Coke no. 19:26
kj seems we've gone through this month's list then? 19:31
pmichaud yes.
do we have any release blockers? 19:32
reminder to everyone to update NEWS, PLATFORMS, etc.
Wknight8111 thank you pmichaud, I was going to make that reminder too 19:33
chromatic It's good to remind the list too.
Wknight8111 I will. I have a lot of release-related tasks in my hiveminder queue 19:34
kj seems we're done then ? :-) 19:37
particle aye. until next week. 19:38
Wknight8111 till next week
19:40 chromatic left, Coke left 19:42 moritz left, allison left 19:50 jonathan left, kj left 20:01 Wknight8111 left 20:16 rurban left 20:17 PacoLinux left 20:32 davidfetter joined 20:39 davidfetter joined 20:44 contingencyplan joined 21:27 particle joined 22:09 Whiteknight joined 22:47 register 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!