Parrot 3.10.0 "Apple Pi" | parrot.org | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC
Set by moderator on 2 December 2011.
00:06 whiteknight joined 00:15 kid51 joined
dalek kudo/nom: bcf2a4e | jnthn++ | src/Perl6/Actions.pm:
Fix a couple of $/ handling issues. Of note, $/ was not available to code blocks inside of regexes, and the code that shoulda been setting it was binding, whereas we assign to $/.
 00:24
kudo/nom: dba6048 | jnthn++ | src/core/ (2 files):
Fix using make inside a regex (it only worked when used in the action method before now).
kudo/nom: d3471b9 | jnthn++ | t/spectest.data:
Run S05-match/make.t.
00:45 offby1 joined
whiteknight good evening, #parrot 00:52
cotto hio whiteknight 00:53
whiteknight ...are you kidding me? NQP uses :instanceof?
I didn't think that code was usable, much less used
cotto glad you saw that
whiteknight says a curse word
did somebody revert it? 00:54
cotto me too, now that I look at it
I don't think anyone has 00:55
whiteknight okay
shit. NQP uses it all over the place
cotto yup
whiteknight I've reverted it, I'm testing now 00:56
I'll test build NQP too 00:57
cotto I can imagine instanceof working reasonably well the way nqp uses it. 01:06
kid51 Darwin/PPC: make test PASS at c383a63 01:13
Linux/i386: make test PASS at HEAD: 4 ways: gcc g++ w/wo optimize 01:14
whiteknight hates 01:44
dalek rrot: bb02f3f | Whiteknight++ | compilers/imcc/ (7 files):
Revert "Remove the :instanceof flag from PIR."

This reverts commit 395ff548295f019d9a67ab122f1be32cf1a32731.
rrot: eb2b219 | jkeenan++ | MANIFEST.SKIP:
Rerun mk_manifest_and_skip.pl so that make manifest_tests passes.
 01:52
rrot: 03d92a8 | jkeenan++ | compilers/imcc/ (7 files):
Merge branch 'master' of git@github.com:parrot/parrot
03:30 jsut_ joined 03:36 jsut joined 03:38 Psyche^ joined 05:59 GeJ_ joined
dalek kudo/nom: 02d6448 | moritz++ | src/core/Duration.pm:
more conversion methods in Duration
 07:23
08:50 muixirt joined
muixirt good morning 08:50
08:51 mj41 joined
muixirt seen fperrad 08:52
aloha fperrad was last seen in #parrot 14 hours 34 mins ago joining the channel.
muixirt msg fperrad many parrot-lua tests don't run anymore, they choke on require 'Test.More' 09:25
aloha OK. I'll deliver the message.
moritz hi 09:26
can anybody please summarize the state of lorito for me? 09:27
muixirt moritz: would that help rakudo? 09:28
moritz muixirt: no, just satisfy my curiosity
muixirt moritz: the problem is the overall architecture, not the core byte interpreter 09:30
coding a bytecode interpreter is quite simple, isn't it? 09:31
*core bytecode interpreter*
moritz I was hoping for was answering questions like "what has been done, what remains to do, who does it?"
muixirt moritz: that's easy ;-) github.com/parrot/parrot/tree/m0-prototype 09:35
moritz muixirt: I see a source tree and a README that doesn't even mention m0. Not a summary. 09:36
09:38 lucian joined
muixirt moritz: are nqp-ops form a high-level interface to a vm? and how would m0/lorito fit into that? 09:40
moritz muixirt: there are two things, dynops in the nqp repo, and the nqp:: operator syntax in nqp and rakudo 09:41
muixirt: which one are you talking about?
muixirt nqp:: 09:42
moritz muixirt: ok. Yes, they form a kind of (shallow) abstraction of a VM. Currently they are just compiled to calls of ops or dynops
for example nqp::substr compiles to the 'substr' PIR op 09:43
and nqp::x to the 'repeat__PI' opcode
dalek kudo/nom: 8975299 | moritz++ | t/spectest.data:
track test file renaming
 11:10
muixirt moritz: what else has a vm (implementing all 125 nqp::ops) to supply to run rakudo? 11:33
moritz muixirt: well, first it needs some prerequisits like having closures 11:34
muixirt: then somebody needs to implement the codegen for that VM
muixirt: and then there's lots of parrot-specific code regarding exception handlers, module loading etc.
muixirt so rakudo relies on a lot of pir/parrot stuff that is not expressed by nqp::ops? 11:39
in the near term will there be a complete set of nqp::ops that provide all that and make it a real bytecode/intermediate language? 11:41
the boundaries seem to be fuzzy 11:42
moritz: the possibility switching to another platform/vm like clr/jvm (or a completely remodeled parrot) is far away 11:46
12:12 bluescreen joined 12:48 bluescreen joined
dalek kudo/nom: 0ac25d0 | jnthn++ | src/core/ (2 files):
First cut implementation of <prior>; seems to work, but sadly not for the m/.../ form.
 13:08
14:52 benabik joined, PacoAir joined
benabik ~~ 14:53
(backlogging) Blah, I should have just reverted that instanceof thing myself instead of carping. benabik-- 14:55
PerlJam Who is handling the release today? 15:00
And is trac being abandoned completely? 15:03
benabik PerlJam: I think we're trying to leave trac, yes. 15:04
moritz - Jan 17, 2012 - 4.0.0 - whiteknight 15:05
from project/release_manager_guide.pod 15:06
benabik Hm. We should probably s/trac/github/ in our docs before the release. 15:08
PerlJam I was just noticing that 16 files in the docs direcotry mention trac.parrot.org
several of them talk about the roadmap and issues and such.
benabik I don't think we've been using the roadmap stuff for a while already. :-/ 15:09
moritz $ moritz@jacq:~/p6/rakudo/parrot>ack -aQ 'trac.parrot.org'|wc -l 15:10
169
benabik There's going to be a lot of references to trac in api.yaml 15:11
... And we should add links to github issues from the trac tickets. :-/
I wonder if we have a list of TT# to gh# 15:12
PerlJam Whatever did the bulk move should
benabik It was run by someone at github, I think. 15:13
PerlJam oh, each github issue has a link to the original trac ticket. 15:15
(I think)
benabik So the data isn't lost, at least.
moritz hrmpf 15:16
docs/submissions.pod seems quite out of date
benabik docs/** needs love
dalek rrot/github-links: dc3d65b | moritz++ | / (4 files):
replace some references to trac.parrot.org with the appropriate links to github issues
 15:19
moritz four mentions down, 165 to go :-)
benabik hah.
aloha (parrot/parrot) Issues opened : 694 (Docs still refer to trac.parrot.org) by Benabik : github.com/parrot/parrot/issues/694 15:20
PerlJam benabik: the docs in the github wiki also point to trac.parrot.org :( 15:23
benabik Well, the transition was done only a few days ago, so it's not surprising. 15:24
PerlJam rather than muck with the docs right away (i.e., before the release), perhaps it would be easier (for now) to make trac.parrot.org redirect to github for a little while. 15:25
If someone really needs to access trac, it could be another URL 15:26
s/be another/be via another/
moritz PerlJam: because of the immanent release I've started doing it in a branch
dalek rrot/github-links: 4b038fb | moritz++ | docs/ (2 files):
more doc updates
 15:27
PerlJam makes his comment on issue 694 15:28
benabik When is the release supposed to be? Today?
PerlJam - Jan 17, 2012 - 4.0.0 - whiteknight 15:29
benabik doesn't pay much attention to releases, obviously.
Hm.
PerlJam er, someone give me commit on the parrot repo and I'll help update the links with moritz 15:35
moritz PerlJam: you don't have one?
PerlJam apparently not.
(I haven't rally touched parrot in a long time.) 15:36
s/rally/really/ even
apparently the last time I committed anything to the parrot repo, it was still in svn 15:38
NotFound PerlJam: What's yout github account? PerlJam?
PerlJam NotFound: perlpiolot
er, perlpilot
Is there a roadmap-a-like in github? 15:40
NotFound This github.com/perlpilot ? 15:41
PerlJam NotFound: aye
NotFound Added to Developers team. If I remember well, that gives you the write permission on the repos. 15:42
dalek rrot/github-links: 00e4c7d | duff++ | docs/translations/README. (6 files):
Mention github issues in translations
PerlJam ding! :)
NotFound Yes, Push & Pull grant.
moritz just to be clear... are we abandoning all of trac? or just the tickets?
benabik I think we're trying to leave trac. 15:43
It's been problematic to keep running and keep clear of spam.
dalek rrot/github-links: 66823eb | moritz++ | examples/ (2 files):
fix links to abc and befunge
 15:47
rrot/github-links: 386b6b0 | benabik++ | docs/book/draft/appe_source_code.pod:
Mention github in "Use the Source"
 15:53
rrot/github-links: aa2df56 | benabik++ | docs/book/draft/ch01_introduction.pod:
book: Mention github in Chapter 1
moritz down to 147 mentions 15:56
alvis moritz++ 15:57
PerlJam Some of these will require an extensive overhaul of the text 15:58
docs/submissions.pod for instnace 15:59
moritz aye
alvis in submissions.pod: does anyone know, is there an github email address to use? something to tickets@parrot.org? ...
moritz works on ticket_triaging.pod
alvis or is it just parrot-dev?
PerlJam moritz: that one is highly trac-oriented too
moritz alvis: I don't there is an email-to-github-issues gateway
alvis moritz: so ... parrot-dev for now, do you think? (and we come up with something later?) 16:00
moritz alvis: parrot-dev for discussion, gh issues for submission
dalek rrot/github-links: 00d9dc0 | benabik++ | docs/book/p (2 files):
book: Remove last trac mentions

A lot of this is very redundant, I'm noticing...
rrot/github-links: 80199fd | moritz++ | docs/project/ticket_triaging.pod:
fix URLs in ticket_triaging.pod; the rest still needs updating I fear
 16:02
alvis moritz: ok then. i'll pull it out. 16:03
dalek rrot/github-links: 1408b09 | duff++ | docs/project/cage_cleaners_guide.pod:
new github cage url
 16:04
16:05 dmalcolm joined
dalek rrot/github-links: b60e8a0 | benabik++ | docs/ (3 files):
docs: misc link updates to github
 16:11
rrot/github-links: 9c18519 | duff++ | / (2 files):
update Languages url
 16:12
rrot/github-links: f3d6241 | benabik++ | src/ops/experimental.ops:
src/ops/experimental: Use github issues, not trac
 16:14
rrot/github-links: 1400ed7 | benabik++ | config/gen/makefiles/root.in:
Makefile: Ask people to submit patches to github
 16:17
rrot/github-links: e97ba05 | benabik++ | examples/languages/squaak/doc/tutorial_episode_1.pod:
Squaak: Point to github languages list
 16:18
PerlJam What maintains api.yaml? Is it generated or hand tooled or what? 16:20
benabik I believe it's by hand for theoretical automated consumption. 16:21
dalek rrot/github-links: 748f522 | benabik++ | t/tools/install/testlib/README:
Update a README to github from trac
 16:23
benabik I think I got a lot of the LHF... I don't have the tuits to search for updated ticket numbers. And I don't know if it's worth converting other wiki pages right now. 16:27
alvis: Are you making parrotbug submit directly to github? 16:33
I am very much in favor of this apparent development. :-) alvis++ 16:34
alvis benabik: yep 16:35
benabik \\o/
I was just looking at parrotbug going "does this do anything useful"
aloha (parrot/parrot) Issues opened : 695 (This is a test) by ayardley : github.com/parrot/parrot/issues/695 16:36
alvis benabik: well, i remember thinking (when first starting out) about using it several times ... 16:37
so, i think new folks tend to look at for submissions.
benabik alvis: I hadn't noticed it's existence until I was looking at docs/submission.pod
dalek rrot/ayardley/documentation_revisions: 52c654b | alvis++ | docs/revisions/docs/mans/vmspec/ (2 files):
Modifications to preface.pod and toc.pod
 16:42
alvis crap! i was on origin/github-links and it pushed my documentation_revisions as well. hmm, (git just really troubles me.) 16:44
benabik I think it defaults to pushing all matching refs.
aloha (parrot/parrot) Issues closed : 695 (This is a test) by ayardley : github.com/parrot/parrot/issues/695 16:46
alvis benabik: yeah.
benabik I tend to do the verbose version: 'git push origin <branch>' 16:49
alvis benabik: i'll do that in the future. thanks. :) 16:50
Coke did someone give perljam commitbits without getting a parrot CLA from him? 16:53
moritz Coke: I kinda thought that PerlJam had a commit bit to the old parrot svn, so he must have had a signed CLA too, no? 16:56
PerlJam I did sign a CLA back in the day 17:00
Coke wonders if he still has access to the list of CLAs. 17:04
hurm. spreadsheet I have access to was last edited 757 days ago by particle. 17:05
and you're on that ancient one, fwiw. ;) 17:06
PerlJam I thought I was feeling a little like a grey beard 17:10
so ... what tickets *weren't* migrated to github? 17:12
Looks like anything tagged experimental or deprecated. 17:13
(at least) 17:14
benabik TT#2168 (Deprecate .file, .line, setline, and setfile) is gh-519 and is a depreciation... 17:15
I think only open tickets got moved. 17:17
dalek rrot/github-links: 0654676 | duff++ | api.yaml:
Change trac links for those that exist on github
dukeleto benabik: yes, only open TTs were migrated 17:18
17:24 eternaleye_ joined
NotFound Coke: sorry, I just assumed he had a CLA without asking. 17:25
17:25 PacoAir joined 17:26 wendar joined 17:27 Timbus joined, elmex joined
dukeleto so, am I the only person that see the Timer PMC tests hanging forever? 17:28
It is driving me insane.
Which is why I converted the tests from PASM to PIR yesterday, during my travel delirium.
17:28 ruoso joined, krunen joined, jlaire joined
benabik dukeleto: Do they run doing make test? If so, then it's not hanging for me. 17:29
17:29 jevin joined, betterworld joined, ascent joined, Hunger joined
dukeleto benabik: yep, they are in test and coretest targets 17:29
benabik: it only hangs on OS X Lion for me 17:30
benabik dukeleto: It runs on my Lion MBP
dukeleto benabik: with g++, not sure if I have tried with gcc
benabik: which compiler you using?
benabik gcc
dukeleto also, i read src/scheduler.pmc and cried
cried, cried, cried.
it is amazing the Timer PMC works at all
dukeleto has to shutdown for landing
17:31 ttbot joined, arnsholt joined 17:32 Maddingue joined 17:38 silug joined
alvis msg not_gerd Please see gist.github.com/1627750 17:41
aloha OK. I'll deliver the message.
17:50 bluescreen joined
alvis benabik/moritz: do either of you want to rewrite submissions.pod (and update its wiki) or should I? 17:50
benabik alvis: If you have the tuits, go for it. 17:53
benabik has to get some "real" work done.
17:56 fperrad joined 17:57 alester joined 17:58 Timbus joined, atrodo joined
benabik dukeleto: Seems to work under g++ for me. :-/ 17:58
17:58 aloha joined
alvis benabik: ok. it's just a'bit more involved; thought it'd free you two up a'bit. 17:59
17:59 jlaire joined 18:00 jevin_ joined, Maddingue joined 18:01 zby_home joined, Hunger joined 18:02 p6eval joined, arnsholt joined 18:04 krunen joined 18:08 silug joined
alvis benabik++ # for trac -> github issues fixes 18:09
PerlJam++ # for the catch!
18:10 dukeleto joined
alvis dukeleto: guess you safely landed :) 18:11
dukeleto alvis: +1 to rewriting submissions.pod
alvis: yes, safe and sound and rocking some tethering action in the taxi to $work :)
alvis thinks our bus-factor is too high. 18:12
dukeleto alvis: too low, you mean? 18:13
alvis: usually a high bus number is good
alvis: low bus numbers (like 1) are bad
alvis dukeleto: ha! yep! thinking obversly ;) 18:15
dukeleto: do we have a 'main' github email account? something to which, e.g., 'parrotbug' reports could be sent? 18:26
moritz no; just don't allow email bug submissions. The web exists. 18:27
alvis moritz: ah. so ... bad idea then. ok. 18:28
dukeleto alvis: parrotbug should just make a github issue 18:29
alvis: but that might be impossible. Not sure.
alvis: the simplest thing is that parrotbug should email parrot-dev 18:30
alvis: somebody will have to approve the emails, though
alvis: in any case, i am very +1 to making parrotbug actually useful
PerlJam Can you attach files to a github issue like you could with trac? 18:34
alvis dukeleto: hmm, ... i think on that ...
s/i/i'll/ 18:35
Coke moritz: we absolutely allowed email bug reports. 18:39
moritz Coke: did that ever work reliably with trac? 18:40
Coke yes.
every six months, someone would complain about it sucking that they had to authenticate, login to the web... and I'd point out that parrotbug still worked, and everyone would be very surprised. ;)
alvis well, since there seems to be a lack of consensus, i'll contact github support about it. they may already have a solution set. 18:47
Coke looks like email replies already work. 18:49
cotto ~~ 18:56
Coke !!
19:03 GodFather joined
dalek kudo/nom: aa1d8b6 | moritz++ | src/core/Range.pm:
optimize Range.roll, and Range.pick if the number of range elems is much larger than the number of elements to pick
 19:03
19:07 contingencyplan joined
cotto #ps in 10 19:21
#ps in now 19:31
19:42 mj41 joined
alvis cotto: sorry, i'm just now thinking about this: what do we do about trac.parrot.org/parrot/roadmap? ... 20:19
do we want to kill references to this and put up something on the github wiki instead? 20:20
cotto alvis: I suspect that it's mostly of historical interest now
alvis cotto: so, should I just remove the reference in 'release_manager_guide.pod'?
cotto alvis: yes 20:21
benabik Github issues does appear to have a milestone function...
Coke parrot has never really dealt well with planned milestones. 20:24
cotto indeed 20:25
Coke so unless you have someone clamoring for them and devoting time to them, it's probably not worth setting up new infrastructure for that.
cotto +1
benabik Fair enough. I just wanted to point out it was there if we wanted to use it. :-) 20:26
alvis i'm also gonna remove references to trac for bug reports, et. al. (unless there's an objection) 20:27
cotto alvis: please do
benabik Less remove, more point to github.
Coke the only place I'd leave them is in the test suite if they are to closed bugs. 20:28
alvis benabik: agreed. :)
dalek kudo/nom: 0deac24 | moritz++ | src/core/control.pm:
try to fix sleep() with non-Num argument
 20:31
kudo/nom: 2fd18f4 | moritz++ | src/core/Range.pm:
fix Range.{roll,pick}
cotto alvis: looks like the release is go. 21:02
dalek p: 08e07d7 | jonathan++ | src/Q (2 files):
First crack at getting backtracking into subrules working. Needs testing with Rakudo yet.
 21:03
alvis cotto: ok. really, i just need the release so i can hand-walk the changes to 'parrot.github.com' and 'parrot-docs3'/'parrot-docs4' ... 21:10
and, then, update 'release_manager_guide.pod' appropriately. so, thanks.
cotto whiteknight will be cutting it this evening, if all goes well 21:11
all is expected to go well
alvis whiteknight++
cotto++ # for making it all go! 21:12
cotto I just poked whiteknight++
21:22 perlite_ joined
dalek p: a2b60b7 | jonathan++ | src/QRegex/Cursor.nqp:
Prevent a crash when building CAPHASH if one of the cstack entries is there solely for backtracking into rather than serving as a capture.
 21:50
p: c9fd128 | jonathan++ | src/QRegex/Cursor.nqp:
Oops, I meant this...
 22:14
23:01 bacek joined 23:06 whiteknight joined
whiteknight Anybody who has additions to NEWS et al would do well to commit them soonish 23:09
23:14 donaldh joined
alvis msg dukeleto I think developer.github.com/v3/ -- specifically, section "<> Create an issue" at developer.github.com/v3/issues/ -- should do nicely for 'parrobug'. What do you think? 23:27
aloha OK. I'll deliver the message.
cotto alvis: the only problem with that is that we'd need to put the credentials in git 23:35
other than that, it'd be perfect
alvis cotto: There's an OAuth2 protocol I'm looking at. It may be a solution. See developer.github.com/v3/oauth/ 23:37
23:41 whiteknight joined
whiteknight always forgets how long fulltest takes 23:42
cotto whiteknight: selective memory 23:43
whiteknight changelog this month is going to be...underwhelming 23:48
tis the season for development to slow down, of course 23:49
dukeleto whiteknight: yeah 23:51
whiteknight: are we still pre-release?
whiteknight yes, I'm working on the release now. I'm making a branch 23:52
dukeleto whiteknight: can i still add to NEWS?
alvis: +1 to using the v3 issues api
whiteknight duke: I'll push my branch soon, so you can add to that
dukeleto alvis: one gotcha will be storing our authentication key in parrotbug
23:58 bacek 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!