Parrot 4.7.0 "Hispaniolan" | parrot.org | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC
Set by moderator on 4 September 2012.
00:19 rurban joined
dalek rrot/native_pbc2: 4aaf1cf | rurban++ | / (4 files):
test with precision (numcmp)

All tests pass now, just __float128 has problems printing certain numbers properly. E.g. __float128 is probed for %.17Lg fmt, which prints 3.8 as 3.7999999999999998		 00:43
rurban_mobile Looks like I accidently merged threads into master, not vice versa. Should I revert this merge? 00:44
sorear quick, force push before anyone else notices 00:57
rurban_mobile It was a few days ago, kid51 already report via email :) 00:58
reported
Strange errors on strange platforms...
benabik Darwin/Intel is not really a strange platform these days. 00:59
rurban_mobile I'd rather leave it in to get more exposure.
benabik Darwin = "OS X"
rurban_mobile Strange for me, because I don't have t anymore :)
benabik Ah.
rurban_mobile I bought a ppc g4 powerbook, but gave away my mac air. 01:00
benabik That's not a good trade. ;-)
rurban_mobile nope, but it made my wife happy.
benabik Ah. See a MacBook Air for a happy wife is a very good trade. :-D 01:01
rurban_mobile But how I can repro the threads bugs now? 01:02
At least I got the numbers (in native_pbc2) almost fixed now. 01:03
benabik I'd offer shell access to my machine, but it's a laptop and has highly intermittent internet access.
benabik needs an old Mac mini.
rurban_mobile not today. I just revived my broken harddisc on my fast system.
benabik woo!
rurban_mobile Now I have a 1TB HD, with plenty of room for new VMs 01:04
benabik I found it very interesting that my new copy of Parallels had "Install Mountain Lion from the Recovery Partition" as an option for a new VM.
rurban A license trick probably 01:06
You may legally only do it from a Recovery Partition
benabik Well, the recovery partition is really the only place most people will have the installer. 01:07
I have a disk image, but that's because I knew to extract it.
rurban I also want to create a hackintosh
sorear benabik: extract it? you mean from the Recovery Parititon? 01:14
benabik sorear: From "Install Mountain Lion.app"
01:26 kid51 joined
kid51 seen whiteknight? 01:29
aloha whiteknight was last seen in #perl6 1 days 2 hours ago joining the channel.
kid51 seen cotto?
aloha cotto was last seen in #parrot 8 days 5 hours ago saying "everything old is new again".
kid51 seen dukeleto?
aloha dukeleto was last seen in #parrot 7 days 8 hours ago saying "somebody embedded perl 5 into nodejs: npmjs.org/package/perl".
cotto hio 01:31
01:53 lucian joined 02:14 kid51 joined
MikeFair hio 02:14
02:52 rurban_mobile joined 03:17 woosley joined 03:18 woosley left 04:23 rurban joined 05:08 rurban joined
rurban_mobile blogs.perl.org/users/rurban/2012/09...vived.html (without graphics yet) 05:20
moritz rurban_mobile++ 05:33
benabik rurban_mobile+=
rurban_mobile In native_pbc2 there is still a small problem on big-endian. native_pbc is stable. 05:35
05:35 rurban joined 06:05 rurban joined 07:02 brrt joined 07:07 rurban joined 08:19 lucian joined, Psyche^ joined
dalek kudo/nom: f3e1c3f | jonathan++ | src/Perl6/Grammar.pm:
Liberalize checking of whether we are in a proto.
 09:06
kudo/nom: e067a34 | jonathan++ | src/ (3 files):
First cut of more complex proto bodies. Caveats: optimizer not aware of the non-onlystar case yet, and can't put {*} in nested blocks in the proto.
kudo/nom: e1bd5c9 | jonathan++ | src/ (4 files):
Mark onlystar protos.
kudo/nom: 3b9fc15 | jonathan++ | src/Perl6/Optimizer.pm:
Don't optimize away the proto in the non-onlystar case.
kudo/nom: 15bdbb4 | jonathan++ | src/Perl6/Actions.pm:
Auto-generated proto also needs to be marked as onlystar.
09:26 mdupont joined
dalek kudo/nom: 4e07a76 | (Gabor Szabo)++ | / (2 files):
change most of the Test.pm functions to return their truth value and test this in the sanity test
 09:36
kudo/nom: 61807a6 | (Gabor Szabo)++ | t/01-sanity/99-test-basic.t:
enable some commented out sanity test
kudo/nom: 792c782 | (Gabor Szabo)++ | / (2 files):
fix the rest of the Test functions to return their truth and test them
kudo/nom: ab08176 | (Gabor Szabo)++ | / (8 files):
merge from upstream and add the additional test cases
kudo/nom: 51435bd | moritz++ | / (2 files):
Merge branch 'nom' of github.com/szabgab/rakudo into nom
kudo/nom: 2ee2701 | moritz++ | / (13 files):
Merge branch 'nom' of git://github.com/rakudo/rakudo into nom
kudo/nom: 0af9e7b | (Gerhard R)++ | src/binder/container.c:
Rename typename to type_name, which isn't a C++ keyword
 09:59
kudo/nom: 8e2134e | (Gerhard R)++ | src/ (3 files):
Add pointer casts to quiet warnings
kudo/nom: 2f06eb0 | (Gerhard R)++ | src/binder/ (2 files):
Add parens to quiet precedence warnings
kudo/nom: 1007273 | moritz++ | src/ (5 files):
Merge pull request #80 from gerdr/gerdr/warnings

Quiet some gcc warnings
11:24 JimmyZ joined 12:19 mtk joined 12:23 JimmyZ joined 12:32 PacoAir joined 12:57 mdupont joined 13:42 whiteknight joined
whiteknight good morning, #parrot 13:42
14:18 bluescreen joined
rurban hi whiteknight 14:40
14:49 benabik joined
rurban With a54a581d2392a885c70093349382b5da696ec4dd I accidently merged threads into master, not vice versa. Should I revert that? 14:49
The commit message is also misleading. kid51 saw some darwin/intel smokes and tadzik found problems also 14:50
whiteknight threads is definitely not ready to merge into master 14:51
so if that's what you did, we need to revert it
the email i sent to parrot-dev a few days ago explains the last problem we're seeing
I think I have a solution for that problem, I'll try to hack it together soon 14:52
tadzik \\o/
whiteknight the first solution I considered was to tell tadzik to stop using closures
tadzik :> 14:53
whiteknight but, I decided that probably wouldn't work in the long term
tadzik that's not the thing
we can't use say() in Perl 6
whiteknight why not?
tadzik because it's lexically scoped :)
iiuc
whiteknight I don't suppose you'd take it seriously if I asked everybody to stop using lexicals :) 14:54
tadzik :>
moritz whiteknight: you cannot do anything serious in Perl 6 without using lexicals 14:55
whiteknight moritz: I know. I'm playing
moritz all built-in classes are lexical
PerlJam whiteknight: maybe not enough smileys? :) 15:07
benabik Reverting a merge in git is a bit tricky. Although I guess the tricky bit is merging it again, where the easiest method is to revert the revert and then merge the new commits in. 15:12
whiteknight benabik: tricky or not, threads is very much not ready to be merged yet 15:17
and the last problem that I'm aware of is a real doozie 15:18
benabik whiteknight: Understand. But I can help get the gitting right easier than I can help fix threads right now. :-D
whiteknight right, so let's make sure that threads isn't merged or that it quickly gets un-merged, then I'll start hacking together a fix 15:19
rurban Yes, I found that reverting the merge is not so easy.
whiteknight if we need to scramble history a little bit, we can do that 15:20
benabik It can be done without scrambling history.
Bad merge is c54a58? 15:21
rurban yes. with git reset --hard it is easier than with revert -m0
benabik revert -m0 wouldn't work. Parents are numbered from 1. 15:22
And I think you want -m2
rurban and the last commit belongs to threads. Got it now. 15:23
benabik Yeah. -m2 gives me conflicts in alarm.[ch], but nothing large. 15:24
Any further threads work needs to base off of a54a581
Which is where the branch is, thankfully.
rurban The conflict was the last wrong commit. cherry-pick into threads. 15:25
benabik We just have to remember to revert the revert before merging it again. 15:26
dalek rrot/threads: 67b2071 | jkeenan++ | / (4 files):
Add ASSERT_ARGS() to two files; re-run 'make headerizer.'
 15:27
rrot: b6584ae | rurban++ | src/dynpmc/ (2 files):
dynpmc/os: actually build this dummy pmc for loadlib "os" backcompat
rurban Ands now really merge master into threads... 15:28
benabik Well threads it at a merge commit of threads and master.
rurban I just had to scramble the last master commit, because it belonged into threads. 15:31
But I find it interesting that all my machines (linux, darwin/ppc, windows) smoked very well with threads. That gives me confidence 15:33
dalek Heuristic branch merge: pushed 44 commits to parrot/threads by rurban 15:42
rurban smoke on threads (with latest master) looks good 15:43
15:45 dmalcolm joined 15:49 benabik_ joined 16:22 benabik joined 16:27 benabik_ joined
whiteknight rurban: Yes, besides this last problem threads does look good 16:29
if I can fix tadzik's test case, I'll be ready to merge
rurban Good. I still want to fix one bad mutex lock and add get_num_cpus() to assign the right number of default threads. 16:31
whiteknight okay, awesome
I'll be working on it this weekend.
rurban I'll probably be faster.
whiteknight ok
16:34 benabik_ joined 16:36 alester joined 16:41 benabik_ joined
benabik_ rurban++ 16:41
17:12 mvorl joined 18:01 benabik joined
mvorl I'm on my my way generating a makefile that's usable under VMS. In the makefile generated by Configure.pl I see a line 18:10
include/pmc/pmc_fixedintegerarray.h: src/pmc/fixedintegerarray.c
without an action. How is include/pmc/pmc_fixedintegerarray.h supposed to be generated? 18:11
Coke that's just listing one of the deps. 18:14
(seeing if I can find the rule) 18:16
mvorl: gist.github.com/3716373 - I didn't see the rule, but hopefully that helps. 18:20
18:24 contingencyplan joined
dalek kudo/nom: 7973660 | jonathan++ | src/ (4 files):
Make {*} showing up in a nested block or thunk rather than directly in the proto work.
 18:30
18:51 lucian joined 18:56 bluescreen joined
dalek kudo/nom: 0f2e95a | moritz++ | src/core/Str.pm:
remove unused code, TimToady++
 18:56
19:02 alvis_ joined 19:24 chekkaa joined
mvorl Coke: thanks. 19:41
dalek kudo/nom: e03a22f | jonathan++ | src/binder/multidispatch.c:
Implement accept first matching constrained candidate rule for multi-dispatch.
 19:47
kudo/nom: 0e36b2d | jonathan++ | docs/ChangeLog:
Add multi improvements to the ChangeLog.
 19:57
chekkaa Hello Parrot Team, 20:01
We at the OSL were hoping to roll out some Drupal 6 upgrades to all of our Drupal 5 sites soon, including yours, parrot.org. If you were planning to upgrade it yourselves soon, please let us know your plans. If not, let's discuss an upgrade plan. We can provide assistance if necessary.
Thanks!
benabik Hah. Wasn't someone saying we needed Drupal6 for some service or another? 20:02
20:04 mvorl joined
chekkaa that too 20:06
benabik I'll message the list, but I don't think we're using anything terribly complicated right now and I doubt there'll be much objection to the change. 20:07
Nor do I think anyone has a plan to do it ourselves in the near future. 20:08
Hm. The os dynpmc still doesn't seem to install. :-/ 20:16
20:22 perlite joined
rurban benabik: Really? Looks like I really have to test it now. 20:32
benabik would like to have Rakudo back. 20:34
rurban just started installing it. let's see 20:38
benabik You don't have to install Rakudo to see the issue. There's no os.bundle in /usr/local/lib/parrot/4.7.0-devel/dynext 20:39
rurban chekkaa: Was it Coke who did our drupal site?
benabik: No, I'm installing parrot 20:40
benabik (I would imagine that would be os.so in Linux.)
rurban: Ah.
Coke rurban: no. we got drupal pre-installed, SFAIK
I have admin privs but have little else to do with it.
allison tweaked a style sheet or two, as I recall
rurban Who can assist chekkaa? If something goes wrong? 20:41
benabik doesn't have admin bit or many tuits.
rurban benabik: You are right. it's missing... bluish
benabik I am trying to pay vague attention to class otherwise I'd dig up the original "remove os dynpmc" commit to dig out the build bits. 20:42
cotto hi chekkaa. you still here? 20:43
rurban benabik: I'll have it figured out soon.
MANIFEST.generated was missing it 20:52
20:53 szbalint joined
rurban I think I had a patch before which ensured that all generated files (and only those) appeared there. Years ago. 20:53
That's why I forgot to put it there also. I assumed MANIFEST.generated was already generated. 20:56
dalek rrot: c36a7c5 | rurban++ | MANIFEST.generated:
dynpmc/os: and now actually install it

Sigh		 20:59
rurban benabik++ for pointing it out
Now I have to dig out my old patches somewhere. 21:03
21:11 travis-ci joined
travis-ci [travis-ci] parrot/parrot#546 (master - c36a7c5 : Reini Urban): The build was broken. 21:11
[travis-ci] Change view : github.com/parrot/parrot/compare/b...6a7c5cfec4
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/2440895
21:11 travis-ci left
benabik Looks like some VMs were power cycling. No big deal. 21:11
dalek rrot: 216ce80 | rurban++ | MANIFEST.generated:
[CAGE] remove duplicate os.bundle in MANIFEST.generated
 21:22
chekkaa Sorry, I was at lunch. 21:26
I can set up a test site before upgrading production
that way we can hopefully sort out any problems beforehand
@rurban, @cotto 21:27
rurban chekkaa: Please ask Coke. He got perms 21:29
cotto howdy 21:30
chekkaa Coke: mind if I go ahead and start on the test instance? 21:34
and if it works out, push it to production?
cotto What are you doing about the theme? Are you going to port our customizations or just use stock Drupify? 21:42
chekkaa cotto: I can convert the theme 21:44
cotto great 21:49
allison Coke: aye, I started with a stock theme and modified 21:50
22:11 sivoais joined
Coke rurban: I have admin on the drupal instance. I assume chekkaa has root on the box. he doesn't really need to confirm with me. 22:18
chekkaa: please go ahead and just try not to take down production. ;)
and I see cotto has replied on list, so follow his lead. 22:29
22:35 whiteknight joined
whiteknight good evening, #parrot 22:42
chekkaa Coke: all right, thanks 23:06
23:08 lucian joined 23:26 mvorl_ 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!