Parrotsketch Every Tuesday @ 18:30 UTC | IRC Log at irclog.perlgeek.de/parrotsketch/today
Set by moderator on 18 January 2009.
00:07 davidfetter joined 02:52 Tene_ joined 06:12 TimToady joined 12:34 rdice joined 14:25 PerlJam joined 14:38 rdice joined 16:21 Tene joined 16:25 Whiteknight joined
Whiteknight Leaving my report here now since I wont be at the meeting: 16:26
* Didn't accomplish anything real
* Dabbled with some tickets, got some ideas for work next week
* +1 on GeJ getting a commit bit, iff his CLA has arrived yet 16:27
EOR
17:06 pmichaud joined 18:09 moritz joined 18:13 jhorwitz joined, kj joined 18:17 diakopter joined, allison joined 18:31 chromatic joined
chromatic Hello, all. 18:32
pmichaud Hello. 18:33
Tene Hi, chromatic!
allison hi
18:33 Coke joined
kj hello all 18:33
moritz hello
chromatic allison?
18:34 NotFound joined
allison - Changed PIR/PASM/PBC library loading so an extension is required, and enabled checking for alternate extensions if the requested one isn't found. 18:34
- Generated Debian packages for the 0.9.0 release, and sent them to our Debian sponsors for review.
- Fixed the 'make html' target to generate HTML docs that look reasonably modern.
- Started cleaning up Parrot docs for 1.0.
jhorwitz arrives late
allison - Added a Pod parser written in PGE, and a set of PCT-based tree nodes for Pod document files. (Needs an actions file to transform the parse tree to the doc tree, if anyone wants a quick NQP task.)
- Merged string refactor branch into trunk. 18:35
EOR
18:35 Infinoid joined
chromatic I did the release last week; had to roll 0.9.0.1 for the CPAN to fix Rakudo. 18:35
I removed some deprecated features.
I challenge everyone here to close one ticket a day for the next week. I'll try to close two per day.
Coke?
Coke Parrot - ripped out several deprecated PMCs 18:36
partcl - see partcl.blogspot.com/2009/01/partcl-...-2009.html
Now running 62 test files, 4882 tests. 2879 passing.
cleaned up some extraneous code in PIR generation for a very slight speed boost.
Still have several parrot-related issues that are blocking various efforts:
code.google.com/p/partcl/wiki/ParrotIssues 18:37
EOR
chromatic cotto?
Coke (chromatic++ for pointing out obvious speedups, btw.)
chromatic Infinoid? 18:38
Infinoid Nothing to report.
chromatic japhb? 18:39
jhorwitz? 18:40
jhorwitz foundation:
* getting foundation financials in order
* hoping to get PDS expenses taken care of over the next two weeks
mod_parrot:
* TT #185 is still outstanding and blocking mod_perl6 -- my patch fixes the problem, but before committing i need a review to know if it's the right approach. if it is, i can close a ticket! :)
misc: fyi, OSCON proposals due in 1 week. 18:41
EOR
chromatic kj?
kj Slightly busy with $work recently;
== PIRC
+ Done pretty much no development except for:
+ staring for hours at Parrot segfaulting on a PIRC-generated PBC file.
The MSVC debugger indicates an access violation exception, but no idea how to solve that. Not sure if I'm finding bugs in parrot or in pirc. probably the latter.
chromatic moritz? 18:43
moritz nothing to report, just the usual Rakudo testing 18:44
chromatic NotFound?
NotFound Some bug haunting and cage cleaning
Applying some patches, notably the tcl/tk library
Added several things to pirric to be able to load languages and run intermixed pieces of code
1 question
EOR
chromatic particle?
particle just committed portable Inf/NaN support for parrot, and fixed rakudo to take advantage of it
+58 tests that previously passed only on linux
++ to all who helped
.end 18:45
chromatic PerlJam?
NotFound (The question is TT #210)
18:47 masak joined
chromatic pmichaud? 18:47
pmichaud * Didn't do much this week other than planning and observing
* Preparing to move rakudo out of parrot repository tomorrow or Thursday
* Taking a poll on "git" versus "svn" -- see xrl.us/bedxsq
* Will undoubtedly be dealing with build/install/migration issues once out of the repo
EOR
chromatic Tene?
masak is late but doesn't have much to report
Tene Nothing of note. I started trying to debug that load_bytecode issue I had recently, but lost the patch that Jonathan gave me. Hoping to do more this coming week. Maybe work on docs? .EOR
chromatic Did I miss anyone (or cut off kj)? 18:48
moritz masak came in late... 18:49
chromatic masak?
kj chromatic: no I was done, thanks.
masak * submitted many rakudobugs
.eor
chromatic particle, do you want to review our milestone progress? 18:50
particle yes. let's begin at the top: 18:51
pdd14-numbers
i've implemented nan/inf, but that's not yet specced properly.
not sure what else remains, but it's definitely not 'landed'. 18:52
c: what's up with the hll class registry?
chromatic I reviewed a ticket about that the other day, but the tests worked.
Tene might know better. 18:53
Tene What's the question? What about the class registry?
particle TT #71
trac.parrot.org/parrot/wiki/ParrotRoadmap
allison particle: could you write up what you've implemented for nan/inf? 18:55
Tene particle: not fixed.
particle allison: pdd or list? yes.
allison particle: either
particle ok, then. thanks, tene. hopefully we get it sorted before this release :)
pmichaud both. :-) 18:56
Tene particle: pmichaud had a pretty good plan for fixing it, but last I heard PM and Allison didn't agree, or if they did I missed it.
particle ok, i'll update the pdd
pmichaud the parts that allison and I disagreed with weren't directly issues of the registry.
allison Tene: what was the disagreement?
pmichaud (and we now agree on the particulars, I think)
allison pmichaud: that's what I recall
Tene allison: Apparently nothing anymore. That was a vague memory, probably confused.
particle ok, so it seems there's a patch, will applying that be enough to mark that item 'landed'? 18:57
pmichaud is the registry used for anything now?
Tene I think there's just a plan, not a patch.
allison pmichaud: yes, it's used to assign a unique ID to every class 18:58
pmichaud every class, or every named class?
allison every class, named or not
particle sees class_registry.diff.
pmichaud how does the registry get used for unnamed classes?
allison (at least, last I checked)
pmichaud: it doesn't
pmichaud okay, I can't parse what you've written. 18:59
the registry gets used to assign a unique ID to every class, named or not?
but it doesn't get used for unnamed classes?
allison pmichaud: what the registry really should do is just store a pointer to the class object instead of storing a string name 19:00
allison looking at src/oo.c 19:01
chromatic I don't mean to interrupt, but if the discussion has turned into implementation questions, it may be more appropriate for #parrot than milestone review.
pmichaud anyway, I can confirm what chromatic reports: the test program in the ticket no longer produces an error
so we need to try a few things in Rakudo and other places to see if this is in fact a fix for the original problem. 19:02
particle ok, let's move on.
pmichaud (moving on)++
particle pdd26-ast, user docs. can we expect that this month, or will that be pre-1.0?
pmichaud I'd put it as being on track. It'll happen pre-1.0 19:03
particle ok. same question for pdd19.
pmichaud it may happen this month but there are other more important issues first.
particle agreed.
allison (aside pmichaud: anonymous classes *don't* get an entry in the registry)
Tene If I can get a more detailed description of what's wanted, I can work on it. The same goes for all user docs.
particle thanks for volunteering, tene. please add your name to the resources for those tasks. 19:04
btw i still haven't gotten any volunteers for creating trac tickets for each of these items.
Tene Oh, right, that list is actually used for something.
particle++
particle however, i may have the time later today. 19:05
pmichaud Tene: even better: create trac tickets and add yourself as a resource/owner :-)
Tene :)
particle ok, critical task of creating dynoplibs/dynpmclibs... any takers? there has been no progress.
chromatic The trick there is installing enough of the pmc2c and ops2c Perl code with an installed Parrot. 19:06
particle it shouldn't be too difficult to set up a test env for this one.
allison this is a build tools question, so Perl 5 code
particle yep
pmichaud we will undoubtedly be checking these issues as rakudo moves out of the parrot repo today+tomorrow
sorry, tomorrow+thu
(I keep thinking today is wednesday. My brane is messedup) 19:07
allison pmichaud: okay, let's do it as a Rakudo solution, with an eye for making it generally available
particle ok. i'll hack at it too, if my schedule allows.
allison pmichaud: and, you have as much of my time as you need to help make it work
pmichaud oh, I don't know what Rakudo will be able to solve it, but we'll be able to point where the issues are :-)
particle allison, parrot user docs? i assume you're tackling strings first 19:08
allison pmichaud: right, I mean "We'll fix it for Rakudo, and that'll fix it for others"
pmichaud: so, let me know what breaks
pmichaud allison: will do.
if rakudo moves to git, we'll be doing it tomorrow, if rakudo sticks with svn, it'll be thu+fri
I'm willing to live with broken repos for a day or two while we clean everything up. 19:09
(but want it all fixed by the weekend, and definitely in time for frozen perl)
allison particle: yes, I'm working on strings before extensively working on docs
particle ok. progress on strings? 19:10
allison pmichaud: sounds good, we'll make it happen
particle: first refactor checked in, about half-way through the total task
particle: that is, merged into trunk
particle: so "on track"
particle fab. on to pdd planning & milestone edits. i'm not quite sure what that means 19:11
allison me neither
milestone edits I'm pretty sure means "review our milestone planning and decide what's not going into 1.0"
pmichaud are they related to nlnet grant, perhaps? 19:12
chromatic That sounds familiar.
pmichaud as in, making sure all of the nlnet grant milestones are completed?
allison nlnet grant is nearly completed, one finished with strings, and last one (for me) finished with an initial implementation of sandboxing 19:13
we've also got one doc milestone remaining, related to PCT
pmichaud okay, that's mine.
particle yep, ok then. i'll put allison, patrick and myself on that task, and mark on track.
pmichaud and yes, I'll make sure to have that for feb release.
particle pdd23-exceptions. 19:14
i think tene could apply some tuits here
pmichaud raises Tene's hand. :-)
chromatic MOAR TESTS PLZ
Tene Yeah, I can exceptions exceptions.
more exceptions tests?
pmichaud I'm pretty sure Tene++ can handle exceptions. :-)
chromatic Especially C handlers and exceptions throwing exceptions.
Tene Oh, right! 19:15
I remember what I was going to do for that.
That part is going to be fun.
pmichaud I think the main thing is making sure that pdd23 describes reality, which it presently doesn't do.
chromatic I'm sure we have some bugs, but if we can reproduce them in a couple of dozen lines of code, we can fix them.
pmichaud (or didn't do last time I checked)
particle on to the dreaded pdd20. pmichaud, you're volunteering. 19:16
pmichaud yes, I'll do that one.
particle fab, thanks!
Tene One more item about exceptions... the PDD wanted us to have exception classes instead of just integer types. Is this something we want to try to do this release, or just punt that to sometime after 1.0? 19:17
pmichaud after 1.0, please.
particle prefers classes, but doesn't want the disruption 19:18
pmichaud there's been discussion that they really should be roles instead of classes, iirc.
particle ah, yes, there was that discussion, indeed.
Moses SO SHALL IT BE
Tene Okay.
pmichaud I'm not sure that I want to deal with mapping a big hierarchy of parrot exception class types into my hll's idea of exception class hierarchy.
particle next, who wants to attack the support policy? someone surly, i hope? 19:19
pmichaud (for whatever "my hll" might happen to be)
chromatic I already edited the document to take support policy. I'll write one tomorrow
particle joy.
windows porting... anyone? ;) i'll take a stab, but i'm not going to do it alone.
pmichaud q1q
particle time just won't allow. 19:20
for me, not the q
allison particle: what needs to be done for windows porting?
diakopter particle: ditto ?
allison particle: works fine with Strawberry Perl or Cygwin
particle i think that will come mainly out of what's currently skip/todo on windows 19:21
allison particle: I've got a bug with ActivePerl, that would be nice to fix
particle ok then, we need tickets!
hopefully this one will take care of itself in short order.
mmd & subclasses?
allison particle: not sure what that is 19:22
particle me neither. subclasses of Sub, perhaps?
VTABLE_invoke?
pmichaud mmd doesn't work for subclasses of PMC 19:23
I think that's all of the tickets marked "mmd:" in RT
(maybe they've been fixed already, but I think that's what the item is referring to)
allison okay, put my name on it, and add "resolve mmd tickets"
pmichaud for example, subclassing Complex caused issues with mmd
allison pmichaud: that sounds familiar 19:24
pmichaud we worked around them in rakudo, but they're definitely (ugly) workarounds
particle finally, who's doing the february release? 19:25
it's just two wonderful weeks away
pmichaud three, yes?
today is still january.
particle yes, three.
but only two of those weeks are wonderful. 19:26
:P
pmichaud fair point.
particle ok, we'll throw that release manager question to the list. 19:27
thanks, all. question time.
chromatic NotFound?
NotFound TT #216 : drop the destination parameter in string_repeat 19:28
And start implementing it with the new name Parrot_string_repeat 19:29
allison NotFound: that ticket seems to be about HLL_map
NotFound Oooops, TT #210
allison NotFound: approved, with name "Parrot_str_repeat" 19:30
NotFound Ok, I'll start with it right now 19:31
allison NotFound: (The new standard in the current refactors is "Parrot_str_*")
NotFound EOQ 19:33
chromatic Was there another question? 19:35
pmichaud yes, I had one.
the parrot repository is scheduled to move at 19h00 utc on Thursday. Is there a plan to update parrot websites and documentation shortly thereafter? (Does there need to be such a plan?) 19:36
allison pmichaud: yes, though not a ticket yet, and should be 19:37
pmichaud I'll probably do some ticket work tonight -- will add a ticket if nobody else beats me to it. 19:38
chromatic Anything else? 19:39
Let's call it a week then. Good work everyone. 19:41
19:42 moritz left 19:43 Infinoid left, masak left 19:52 chromatic left 19:53 diakopter left, allison left 21:05 NotFound left 23:33 japhb 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!