Parrot 5.0.0 "Johnny Five Alive" | parrot.org/ | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC
Set by moderator on 23 January 2013.
cotto Sorry. I briefly forgot that sarcasm doesn't propagate effectively on irc. 00:00
sorear could maybe do something with the niecza IR
allison heh :)
Coke ah, whew.
pmichaud Sorry. My sarcasm detector is also on the fritz.
sorear it should be marginally easier to parse than perl6 ;)
allison "Perl 6: The Good Parts"
cotto allison: that seems like a whole lot of wheel to reinvent. 00:01
allison yup
cotto closer to a Dyson sphere
benabik Can someone let me know when they stop talking about Parrot on #perl6? ~.~
sorear benabik: we're not talking about parrot, we're talking about diakopter 00:02
pmichaud sorear++ :-)
allison sorear: are you a PaFo member? because that's where the post is 00:03
sorear allison: no
unless I'm automatically one because of my history as a smalltime parrot dev
I broke pdd21, I think that's most of what I did 00:04
00:04 swarley joined
allison sorear: all committers have the required qualifications for membership, but it's not automatic 00:04
sorear does that extend to patch-submitters?
allison sorear: Okay, then quick summary: PaFo is considering "rehosting" with the Software Freedom Conservancy, The Perl Foundation, or possibly Apache (though that option hasn't so far been popular) 00:05
sorear: it's just the same set of requirements as a commit bit, having a commit bit itself isn't required 00:06
sorear: so, possibly, depends on the level of patch-submission
sorear well I haven't filed a CLA so I guess I'm disqualified :D
allison sorear: not that the benefits of membership are huge, members just elect the board of directors
sorear: and the board of directors really just files the necessary paperwork to keep the foundation open 00:07
benabik Which hasn't happened in the two-ish years I've been participating.
allison benabik: the essentials have been taken care of
benabik: not a lot of fuss or fanfare, just submitted regularly 00:08
benabik allison: I was referring to the election.
allison benabik: yes, that hasn't happened
benabik: mentioned in backscroll (not sure when you were on)
and, Coke made the valid point that with limited resources, it might make more sense to spend them on closing the foundation than on running an election now 00:09
benabik Makes sense. 00:10
pmichaud I'm afk again for a while. If anyone needs an "official Rakudo opinion" on something, let me know and I'll come up with one. 00:12
00:23 kid51 joined
cotto hi kid51 00:30
00:33 not_gerd left 01:54 kid51 joined 02:59 diakopter left 04:43 Reini joined 06:08 Reini joined 07:09 Mike-PerlRecruiter_ joined
moritz good morning 09:03
after sleeping about yesterday's discussion a bit, I'm now able to express my concerns with a rakduo<->parrot re-merge better 09:04
first, independently of parrot's effort, rakudo will continue to target multiple VMs, so we'll want a loose coupling to parrot
(the reason is simply that we don't want to leave out the rather big market of java devices, and java library interop) 09:05
so, re-merging rakudo and parrot and maintaining loose coupling seems like contrary goals, though maybe something could be done about it 09:06
second point: in the early days of rakudo/parrot decoupling, it worked pretty well that parrot introduced a new feature, and rakudo used it pretty quickly, even though they were in separate repos 09:07
one of the reasons it doesn't work very well anymore is that there were very few parrot changes from which rakudo profited directly, so the motivation to update quickly (and deal with fallout related to the update) was low 09:08
if that is about to change, IMHO the change must come from the parrot side: FIRST improve stuff (potentially by focus on rakudo), THEN re-think about a merger 09:09
(phrased differently, show the commitment before we invest much work in its anticipation) 09:10
oh, and I think if there would be a merger, hypothetically, it would be more of a nqp <-> parrot merge, not a rakudo <-> parrot merge 09:11
so the summary is that it must be the parrot developers that develop parrot, be it in a merged or in a separate world. The last ~12 month didn't boost my confidence that the parrot developers have the time and/or motivation to do so (given the small exception of rurban++) 09:13
and before any potential merge, we must have faith in parrot's continued development, because we certainly won't merge anything with a dysfunctional or dead project 09:14
EOR
09:44 Psyche^ joined 10:07 muixirt joined
muixirt good morning 10:07
tadzik hello
muixirt (read what moritz has written and the discussion y'day) I still wonder if it's feasible to transfer the rakudo C parts to parrot 10:09
sorear good morning muixirt 10:11
muixirt hi sorear
another (lesser) concern is the future of winxed 10:13
dalek p: afaf87d | jnthn++ | src/ops/nqp.ops:
Corret nested SC write barrier enable/disable.

Unbusts the Rakudo build on latest NQP.		 10:19
11:00 bouncy joined 11:03 masak joined
masak I had thought about giving my two cents to the discussion, but now I see I can just say "what moritz++ said". 11:04
11:04 masak left
dalek p: 978dc43 | jnthn++ | src/ (3 files):
Prepare for eliminating clone callback prop use.
 11:56
p: 1b41b41 | jnthn++ | src/pmc/sixmodelobject.pmc:
Add a check to avoid a possible segfault.
p: f7eb701 | jnthn++ | src/NQP/World.pm:
Eliminate use of properties for clone callback.

Still a reference in the setting, but that can go away with a bootstrap update.
12:13 PacoAir joined 13:21 kid51 joined 14:16 zby_home joined
dalek p: 7133fa1 | jnthn++ | src/ (4 files):
Preparations for positional/associative delegate.

This work will eliminate some of the v-table use that is in the way of certain bits of porting work, but should also provide some other wins.		 14:18
p: 7ec2668 | jnthn++ | src/6model/knowhow_bootstrapper.c:
Get KnowHOW bootstrap using REPR compose protocol.
p: 9c09d92 | jnthn++ | t/nqp/60-bigint.t:
Fix a test that used KnowHOW incorrectly.
p: 9ad9e37 | jnthn++ | t/serialization/02-types.t:
Fix another bad use of KnowHOW.
p: e93ea46 | jnthn++ | src/ops/nqp.ops:
Fix an op that failed to compose a package.

The op overall needs to go away as it's something of a hack, but easier to just fix it up for now.
p: e387b52 | jnthn++ | src/6model/reprs/P6opaque.c:
Finish removal of pre-REPR-compose code.
kudo/nom: b39b6f2 | jnthn++ | src/Perl6/Metamodel/ (2 files):
Make sure we call compose_repr early enough.

Should always call it before the first intantiation.
15:05 pjcj joined 15:07 awwaiid joined
dalek p: 723aae4 | jnthn++ | src/6model/ (3 files):
P6opaque adding list/hash delegation to P6opaque.
 15:24
p: 030ba9b | jnthn++ | src/ (4 files):
s/Indexing/Positional/ in REPR funcs.

Makes it more consistent with Perl 6 terminology and the naming on the JVM.
p: 5795fdf | jnthn++ | src/6model/ (3 files):
Toss a bunch of (thankfully unused) mis-design.
 15:34
p: 90e12d2 | jnthn++ | src/ (3 files):
Toss some unused v-table handler usages.
 15:57
p: 7676c23 | jnthn++ | src/6model/ (3 files):
Fill out positional REPR funcs more reasonably.
16:12 benabik joined 16:18 Reini joined 17:41 contingencyplan joined
dalek p: c4a988a | jnthn++ | src/ (2 files):
First pass at getting positional_delegate working.
 17:52
p: 6afcc54 | jnthn++ | src/NQP/Actions.pm:
Fix thinkos.
p: 6ad0c64 | jnthn++ | src/stage0/ (9 files):
Update bootstrap.
p: 51c43bf | jnthn++ | src/ (3 files):
Switch some things over to positional_delegate.
p: 27c96a9 | jnthn++ | src/QAST/Compiler.nqp:
Make VarWithFallback handle type-obj indexing.
p: d3431b6 | jnthn++ | src/stage0/ (9 files):
Update bootstrap.
 17:57
p: c123e22 | jnthn++ | src/core/NQPMu.pm:
Eliminate another Parrot v-table use.
18:11 benabik joined
cotto ~~ 18:12
kid51 Yo, cotto, how's life? 18:16
cotto hi kid51 18:20
I'm done with the startup and wfh now, so that's nice. 18:21
kid51 Well, if you have any energy for Parrot, we can certainly use it :-) 18:22
cotto I'm thinking about that quite a bit.
see also yesterday's discussion
kid51 I'll have to back-scroll that 18:23
cotto Grab a cup of tea.
kid51 That's the most sustained discussion we've had on #parrot in a long time 18:30
18:36 pjcj joined 19:08 Mike-PerlRecruiter_ joined 19:40 not_gerd joined
not_gerd you've got mail: lists.parrot.org/pipermail/parrot-d...07333.html 19:41
dalek kudo/nom: 7ad6e63 | jnthn++ | src/Perl6/Actions.pm:
Fix anonymous state variable regression.
 19:56
20:14 bluescreen joined 21:00 not_gerd left
dalek kudo/nom: d76f9b2 | (Solomon Foster)++ | src/core/Cool.pm:
Add Cool.set and Cool.bag.
 21:10
21:21 Reini joined
dalek p: bf54542 | jnthn++ | src/6model/ (2 files):
Add associative REPR API.
 22:02
p: 08684cd | jnthn++ | src/ (5 files):
Refactor elems.
p: 4436142 | jnthn++ | src/ (2 files):
Implement associative_delegate.
p: 02e8737 | jnthn++ | src/core/NQPCapture.pm:
Switch NQPCapture to use associative_delegate.
p: 8654841 | jnthn++ | src/QAST/ (2 files):
Switch QAST over to associative_delegate.
kudo/nom: a3869a0 | jnthn++ | / (2 files):
Update to build on latest NQP.

REPR API additions force a fresh sixmodelobject.h, so also needed to do an NQP_REVISION bump.		 22:09
22:22 PacoAir joined 22:51 Reini 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!