Parrot 0.9.0 | parrot.org/ | 468 RTs remain
Set by moderator on 11 February 2009.
kj well i probably did something wrong? 00:00
rurban I thought the logic was better, somehow
I thought e.g. checking for .parrot_current_rev is better than DEVELOPING
00:02 elmex joined
rurban I wondered because my todo tests for native_pbs gave wrong results because I'm still developing 00:02
00:02 TiMBuS joined
rurban okay, bed is calling. bye! 00:03
kj thanks a bunch for your help!
allison: sorry to bother again, but I don't understand the PAUSE 00:04
rurban NotFound: t/pmc/fixedintegerarray.t fails
NotFound rurban: not for me 00:05
rurban it could be the gcov instrumentation
allison kj: PAUSE?
purl PAUSE is the Perl Authors Upload SErver
rurban I'm profiling right now
allison oh, uploading to PAUSE
kj allison: yes I can't make any sense of it
I need to assign myself co-maintainership
allison we should delete those instructions from the guide
NotFound The non-trivial and correct embed test now works both with c and c++ :) 00:06
allison but, for this month one of the CPAN authors can do the CPAN upload
00:06 rdice joined
kj does that imply I can skip that? because the release message already indicates that it'll be on CPAN 00:06
rurban release is done and here he comes
allison kj: yes, I can upload to CPAN (or chromatic) if the tarball is already up on the ftp site 00:07
kj its on ftp-osl...
allison kj: and you ran the script to trigger the mirroring? 00:08
kj i did
allison kj: then it'll show up soon
rurban pause.perl.org/pause/authenquery?A...il_logfile has no parrot yet
00:08 AndyA joined
allison kg: ah, already there 00:09
rurban: that's because it hasn't been uploaded to PAUSE yet
NotFound kj: nice title, I forgive you for not use 'Negative NaN' ;) 00:10
00:10 particle joined
kj NotFound: heh. which was a good option as well :-) 00:10
but you should just hear the intro music that comes with this :-) 00:11
moderator Parrot 0.9.1 Released 00:11
moderator Parrot 0.9.1 Released | parrot.org/
moderator Parrot 0.9.1 Released | parrot.org/ | 458 RTs left! 00:12
NotFound kj: I'll try to make a version on the NDS Korg DS-10 00:13
kj what's that?
purl that is, like, due to the echo service not running on BinGOs's box
NotFound kj: synthesizer/sequencer emulator for the Nintendo DS
kj oh very good :-)
cotto kj++ #gettin' 'er done
kj and we should all grow long hair :-)
cotto no, that is <reply> 00:14
00:14 Theory joined
cotto purl no, that is <reply> 00:14
purl okay, cotto.
jonathan kj++ # nice release name! 00:15
kj jonathan: thanks! 00:17
allison kj: uploaded to PAUSE. will take a while to process
kj allison: thanks! 00:18
how to update a small box on the side of a wiki page? 00:19
Whiteknight: ping # must know it :-)
purl I can't find # in the DNS.
Whiteknight kj:pong 00:20
kj hi Whiteknight; you're the wiki god over here
how do I add this little box on the wiki page with "quick facts"
Whiteknight which wiki, wikipedia?
kj eh, yes
Whiteknight what box do you need added?
kj well the box is there
but I need to edit the information 00:21
Whiteknight which page?
kj en.wikipedia.org/wiki/Parrot_virtual_machine
it's on the right
am I just overlooking something?
Whiteknight okay, no sweat. It's a template at the top of the page 00:22
kj aaah
i see!
Whiteknight it's marked {{infobox software
kj yues i see now
thx!
Whiteknight okay, excellent
00:29 gerd joined
Coke devel check - did that get straightened out? 00:32
(DEVELOPING) is not (normally) in the created tarball, but is always in svn.
kj Coke: DEVELOPING will not be included in the tarball; this is automatically, so "fool proof" 00:36
according to the rel.man.guide anyway
Coke right. just trying to figure out what rurban was concerned about. 00:37
00:38 verve joined
kj ok, I'm about to go home. news posted to reddit, slashdot, wikis updated, parrotblog 00:42
now checking out perlmonks.org
anything else I should do?
NotFound kj: take a beer 00:44
Infinoid kj++
NotFound kj++
I posted to barrapunto.com (spanish slashdot clone) 00:45
kj I'll go straight to bed :-), but thanks :-)
thank you all for help!! 00:46
dalek rrot: r36848 | coke++ | trunk:
Resolve TT #267 - perl6 has left the nest, and its new location is documented in the 0.9.1 release.
 00:47
kj good night all 00:50
Coke Can someone tell me when string_repeat was removed from the repository? 00:53
00:53 HG` joined
Coke ah. trac++ 00:54
chromatic Wasn't it renamed instead?
Coke yes, we were supposed to keep the old one as an alias until the release. 00:55
dalek rrot: r36849 | coke++ | trunk/DEPRECATED.pod:
TT #210 was resolved some time ago sans deprecation. Remove reference to closed ticket.
 00:59
01:30 particle joined
dalek rrot: r36850 | whiteknight++ | branches/rename_pccinvoke/src:
[rename_pccinvoke] update calls in src/io/api.c where S is in the results but not in the args. All tests pass (I ran them twice)
 01:44
Whiteknight i'm going to solve this issue if it kills me
and then I'm going to be all like "oh noes, it killz me! LOL"
Infinoid Whiteknight++ # but please don't die
Whiteknight what kills me is that the results from a call are fine, even though there's all sorts of crazy pointer nonsense with them. It's the args, which are all firmly grounded, that are getting lost 01:45
Infinoid++ # is generous with karma, and generosity is infectious! 01:46
Infinoid \\o/ 01:49
cotto there. 02:10
best to take care of this sort of thing before someone opens a ticket 02:11
cotto waits for dalek...
dalek tracwiki: v1 | cotto++ | ParrotQuotes
tracwiki: first version with a dozen or so quotes
tracwiki: trac.parrot.org/parrot/wiki/Parrot...ction=diff 02:12
shorten dalek's url is at xrl.us/begm4u
Infinoid Since dalek is now tracking 18 rss feeds, I thought it best to increase its polling interval from 30 seconds to 3 minutes...
cotto no complaints from me 02:13
Infinoid oh hey, I'm in a few of those 02:14
nice collection, cotto++
dalek tracwiki: v2 | Infinoid++ | ParrotQuotes 02:19
tracwiki: trac.parrot.org/parrot/wiki/Parrot...ction=diff
shorten dalek's url is at xrl.us/begm5v
02:27 eternaleye joined 02:28 mikehh joined 02:53 ilbot2 joined
moderator Parrot 0.9.1 Released | parrot.org/ | 458 RTs left!
02:54 moritz joined
Infinoid #165 (Disable perl6-internals@perl.org mailing list.) is listed under the "website" component, even though it seems more like a mlm sort of thing. 02:56
allison: Is the same thing going to happen to parrotblog.org, then? 02:58
allison Infinoid: yes, not really a website item, though it's connected to a website item to update the old references to the old mailing list
Infinoid Well, I can nag the powers that be about the list stuff, no problem. 02:59
allison Infinoid: that's the general idea, though I've been impressed at the external traffic pulled in by parrotblog.org
Infinoid If the content is all there on parrot.org and parrotblog redirects to parrot, I would think that shouldn't change 03:01
Might have some direct links broken by the redirect, I suppose
allison Infinoid: yeah, redirecting the url will take care of it.
Infinoid Ok, I'll see about crossposting stuff to tracwiki and/or backdated news posts if it seems appropriate 03:02
allison Infinoid: okay, thanks! 03:03
Infinoid: did you get editor access on parrot.org?
Infinoid I'm not really sure how to tell. I can click on a news post and I don't get an edit link, if that's relevant 03:05
I do have a "Create content" in the menu in the upper right, though
allison Infinoid: okay, just granted your account editor role
you should see edit buttons on every page, and have the ability to create all types of content
Infinoid Oh, there it is. I refreshed and got a bunch more menu items
Ok, thanks 03:06
allison also, you can create aliases to pages, so they have both the regular URL and a "fake" URL that matches the old website
rg quick question the release reminded me to ask: the freebsd port file has a note not to build as root "because of 'perldoc' issue". does anyone have an idea what that might be referring to? 03:07
Infinoid rg: might be referring to coding.derkeiler.com/Archive/Perl/c...01862.html 03:12
shorten Infinoid's url is at xrl.us/begm97
Infinoid (perldoc drops privileges, and thus might no longer have read access to the user's home directory) 03:13
rg thanks. 03:16
Util wakes from nap; sees Editor role has been added to his parrot.org account. 03:18
++allison
allison Util: thanks for the help
03:18 mikehh joined
dalek tracwiki: v51 | jimmy++ | WikiStart 03:19
tracwiki: add ParrotQuotes</a>
tracwiki: trac.parrot.org/parrot/wiki/WikiSt...ction=diff
shorten dalek's url is at xrl.us/begnao
GeJ Good morning everyone 03:27
Infinoid allison: Is it a good idea to move static pages from the old sites onto the tracwiki, or will page aliases not work from www.parrot.org to trac.parrot.org? 03:31
hi GeJ
Util Hi, GeJ 03:33
03:35 janus joined
Infinoid Util: If you want to help with the website update stuff, I think TT #349 should make a good coordination point. Feel free to take ownership of some of those tickets :) 03:36
What's the current status of docs.parrot.org? Is the work there as simple as making some pod2html output and rsyncing it into some webroot from a cron job? 03:39
From reading #305, that's what it sounds like. I'll toss together a cron script and attach it to the ticket 03:42
allison Infinoid: it'll be built using the 'make html' target from the most recent release 03:53
GeJ heya Infinoid, Util.
allison Infinoid: the trac wiki is good for development coordination. www.parrotcode.org pages should generally become www.parrot.org pages 03:54
Infinoid ok, cool
allison Infinoid: some pages from the old SocialText wiki should also become www.parrot.org pages
Infinoid that's the TPF wiki? or is there yet another site? 03:55
allison Infinoid: yes, that's the TPF wiki
Infinoid great, that's TT #350
allison Infinoid: because it was so difficult to modify the old www.parrotcode.org, people often ended up creating wiki pages for website content 03:56
Infinoid I've just made a script which cleans and updates a parrot checkout, configures, does "make html" and rsyncs the result into $WEBROOT/ (just to nudge the whole docs.parrot.org process along) 03:57
And it works, and it emits no output. Perfect for running from cron once an hour
allison Infinoid: cool 03:58
Infinoid Ok, I'll probably start moving pages across tomorrow. But for now, food time! 04:02
Thanks for all the answers 04:03
04:19 rurban_ joined
Util Infinoid: TT#350 pieces will be my first target. No real time until Monday evening. 04:26
After a few pieces of TT#350, I will look at the whole tickets in TT#349, and start the cycle of taking ownership and resolving its member tickets (one at a time).
++Infinoid for the make/rsync script; I had had the same thought, because I think Parrot will benefit greatly from more "live" online docs.
Even with always having a fresh Parrot checkout around, I keep finding myself wishing for Parrot's version of perlcabal.org/syn/ .
Night, all 04:31
04:40 galf joined
Infinoid Util_away: Great, thanks 04:43
05:25 rhr joined 05:56 Andy joined
cotto Infinoid++ #finding the funny 05:56
karma foo 06:04
purl foo has karma of 18
cotto ++foo
karma foo
purl foo has karma of 18
06:08 eternaleye joined
dalek rrot: r36852 | rurban++ | trunk/config/auto/arch.pm:
[cage] Specialize for solaris cpuarch: i86pc => x86_64
 06:18
06:18 eternaleye joined
dalek rrot: r36853 | rurban++ | trunk/config/auto/cpu/x86_64/auto.pm:
[cage] typo in comment: whenn
 06:22
06:27 Tene joined 06:40 Theory joined 06:59 uniejo joined
dalek rrot: r36854 | rurban++ | trunk/src/packfile/pf_items.c:
[core] fix compilation on Solaris/SPARC as reported by Andy Dougherty
 07:37
rrot: r36855 | rurban++ | trunk/config/init/hints/solaris.pm:
Fix TT #317. Use -R on solaris instead of -Wl,-rpath=, even with gcc.
 07:52
08:19 iblechbot joined 08:20 masak joined 08:27 integral joined
dalek kudo: 842ef38 | (Moritz Lenz)++ | Test.pm:
Import Ovid's Test.pm with improved diagnostics.

dies_ok 'some_string_here' that should really have been eval_dies_ok. That already caught some instances in the test suite.		 08:39
shorten dalek's url is at xrl.us/begnvf
09:21 bacek joined 09:47 PantheraPardus joined 10:09 riffraff joined 10:26 bacek joined 10:30 gaz joined 11:04 kj joined
dalek kudo: af2cace | jnthn++ | build/Makefile.in:
[build] Comment lines must have # at start of line to keep nmake happy.
 11:25
shorten dalek's url is at xrl.us/begn4v
dalek kudo: 044ba5e | jnthn++ | build/Makefile.in:
I have some issues building the fakeexecutable here; making test and spectest just depend on a perl6.pbc (which is what they actually *need*) for now.
 11:29
shorten dalek's url is at xrl.us/begn5d
12:15 jan joined 12:20 rurban_ joined 12:31 rg1 joined
dalek kudo: 5d74999 | jnthn++ | src/parser/ (3 files):
Support :: for declaring anonymous classes etc that can inherit from something. This incorporates the recent STD.pm change for this.
 12:39
shorten dalek's url is at xrl.us/begoir
13:03 jimmy joined
dalek kudo: 6663565 | jnthn++ | src/classes/Object.pir:
Fix initialization of parent attributes; resolves RT#61488.
 13:05
shorten dalek's url is at xrl.us/begonn
13:13 Gerd joined
jimmy good evening, masak. 13:19
masak 你儽, jimmy
jimmy 儽啊 13:20
masak ꈑåœØ学习äø­ę–‡. 13:22
jimmy ä½ äø€äøŖäŗŗåœØēœ‹ä¹¦å—ļ¼Ÿ 13:24
ęˆ‘åˆšę‰åœØäø‹č½½č‹±ę–‡čµ„ę–™å­¦č‹±ę–‡ļ¼Œå‘µå‘µ
masak åƹ,äø€äøŖäŗŗ. 13:25
good luck with the English!
jimmy 你也äø€ę ·ļ¼ŒēŽ°åœØäø­ę–‡čæ›ę­„很åæ«å§ļ¼Ÿ 13:26
masak well, I don't know about that... :) 13:27
I'm making _some_ progress, surely. 13:28
but hardly rapid. there's just too much of everything.
å¤Ŗ多äŗ†.
sjn loves having an UTF-8 terminal, finally being able to see all those symbols I don't understand :) 13:29
jimmy You too. Is your Process of Studing Chinese very fast?
szbalint too 13:30
jimmy sjn: using chatzilla
masak jimmy: no, not particularly. but there is some progress.
sjn gnome-term + screen + irssi here
szbalint sjn: my winning combo too
jimmy s/Process/progress/
masak everybody: this has been a UTF-8 test. thank you for your collaboration. 13:31
jonathan sees question marks everywhere
moritz sses some empty boxes where the fonts don't have the appropriate characters 13:32
masak be glad you're not the ones studying Chinese, then :)
masak uses Emacs, excellent both for viewing and typing hanzi 13:33
jonathan transliterates and says "ni hao"
jimmy jonathan: 你儽 13:35
masak more question marks... 13:36
jimmy what does ā€œmore questions mark" mean? 13:37
jonathan ??
:-)
szbalint are there chinese equivalents of smileys btw?
jonathan jimmy: I'm unable to see Chinese characters displayed here, they just show as ?? instead. :-| 13:38
jimmy szbalint: yes. the same as english.
sjn ā˜ŗ
(that was a smiley, jonathan)
szbalint interesting :)
jimmy just said 'ni hao' to jonathan 13:40
:) 13:43
:(
-)
dalek kudo: d2d0d84 | jnthn++ | src/parser/ (2 files):
Support use of does inside packages.
 13:44
shorten dalek's url is at xrl.us/begopi
dalek kudo: 7c5baa4 | jnthn++ | Configure.pl:
Turns out we don't need $slash in the Configure.pl for Windows to work - the $exe was the important thing.
 13:49
shorten dalek's url is at xrl.us/begopt
dalek tracwiki: v52 | jimmy++ | WikiStart 13:54
tracwiki: add Parrot 0.9.1 released news
tracwiki: trac.parrot.org/parrot/wiki/WikiSt...ction=diff
shorten dalek's url is at xrl.us/begorx
jimmy karma jimmy 14:00
purl jimmy has karma of 57
Coke_afk (chinese) instead of question marks, I get: \\u263a
Coke which is no less enlightening.
jimmy does not support utf-8. 14:02
nopaste "jonathan" at 85.216.157.73 pasted "patch - if anyone with Parrot commit bit will put in for me ;-)" (13 lines) at nopaste.snit.ch/15658
jonathan Coke: Any change my commit bit could be fixed? ;-) 14:03
*chance
Trac username is "jonathan"
Infinoid jonathan: done, r36856 14:09
jonathan Infinoid: thanks 14:10
14:11 Whiteknight joined
dalek rrot: r36856 | Infinoid++ | trunk/config/init/hints/mswin32.pm:
Apply patch from jonathan++:
 14:11
Coke jonathan: checking. 14:14
jimmy karma Infinoid 14:20
purl infinoid has karma of 489
dalek kudo: 9e72b9d | jnthn++ | build/gen_metaop_pir.pl:
Generate min= and max= meta-ops. Patch courtesy of bacek++.
 14:24
shorten dalek's url is at xrl.us/begouh
Coke jonathan: you currently have "developer" access. 14:29
do you have a "confirm your email" message pending?
(ah, that's trac only. you're also in the "parrot_commiters" group. 14:30
... that should be parrot_committers, AIGH.
jonathan I can never spell that right either. ;_) 14:31
Coke (it's spelled wrong in the permission system, not just here. ) 14:36
anyway, your account looks ok to me.
14:38 Tene_ joined
jonathan Coke: OK, hmm. 14:45
dalek kudo: 5c7da38 | pmichaud++ | build/PARROT_REVISION:
Bump PARROT_REVISION to include compiler VM size fix in r36856.
 14:46
shorten dalek's url is at xrl.us/begow8
dalek kudo: c18fafa | jnthn++ | (2 files):
Copy libparrot.dll into place for Win32, so perl6.exe works.
shorten dalek's url is at xrl.us/begoxa
dalek kudo: 876c093 | jnthn++ | build/PARROT_REVISION:
Merge branch 'master' of git@github.com:rakudo/rakudo
shorten dalek's url is at xrl.us/begoxc
jonathan meh...fail.
jonathan should always pull before he pushes 14:47
Infinoid pull --rebase will get rid of that "Merge branch 'master'" noise
jonathan OK, will try and remember to do that in the future.
14:48 gryphon joined 14:49 dalek joined
pmichaud fwiw, I don't think I should have to remember the "--rebase". Does the extra merge hurt? 14:51
(and istr there was a potential downside to using the --rebase)
jonathan pmichaud: AFAICT it just makes for more noise here. 14:52
pmichaud so I'm thinking maybe dalek could be smart enough to suppress the noise.
rather than making 'n' committers remember some weird command.
jonathan Aye.
My preferred solution too. ;-)
Infinoid The merge commit doesn't hurt, it just doesn't do anything
The only downside I know of rebase is a loss of synchronization if you are using distributed repositories, which rakudo isn't
pmichaud does the --rebase have the potential to hurt?
Infinoid There's a way to set git up to do --rebase by default. 14:53
jonathan Infinoid: I guess the commit message is always like "Merge branch 'master' of git@github.com:rakudo/rakudo" so should be easy to detect/filter out in dalek.
Infinoid Heh. if that's your preferred solution, I'll implement it. But if I then go check out a copy of your latest c18fafa commit, it won't have pmichaud's commits 14:54
pmichaud Infinoid: why would it not have my commits?
Infinoid pmichaud: because he didn't rebase, and therefore, he's forked history with that commit 14:55
That merge message is what git emits when it cleans up after that
pmichaud then just suppres the "Merge branch 'master'" message and not the commit. 14:56
i.e., still report the 876c093 commit, just don't report the text after that.
or mangle the report to say "merge branch master" as part of the commit message.
i.e., as part of the first line. 14:57
i.e., rakudo: 876c093 | jnthn++ | build/PARROT_REVISION: (Merge branch 'master'...)
jonathan pmichaud: Are you wanting to discuss/dig into Prelude stuff soonish? 14:58
pmichaud anyway, the tools need to support the developer's modes of working, not vice versa.
jonathan: we can discuss settings, yes. :-)
Infinoid Absolutely. That's why I suggested setting it up to do --rebase by default
jonathan tries to drill the new terminology into hs head.
pmichaud Infinoid: it would help more to know _how_ to do that, and to understand what that means. 14:59
anyway, I have bigger fish to fry right now.
jonathan: here or on #perl6? 15:00
Infinoid pmichaud: "git config branch.master.rebase 1"
jonathan Let's mvoe to #perl6
pmichaud #perl6 probably makes more sense.
Infinoid pmichaud: Anyway, sorry for bothering you guys with trivial stuff :)
15:33 Theory joined 16:00 clunker3_ joined
rurban Who's the JIT specialist here? chromatic? 16:06
16:15 Andy joined
particle svn blame 16:15
purl svn blame is just like p4 annotate, only better
16:30 leto_ joined
leto_ mornin' 16:30
my favorite current rakudo bug: 16:31
> say NaN*1i
NaNNaNi
rurban what is li?
purl i heard li was DOS boot, fdisk.exe /mbr and pray or Long Island or Long Island. or Long Island or Long Island or Very Long Island. or Long Island or okra
jonathan Complex thingy. 16:32
PerlJam rurban: 1 (the number one) i
pmichaud it's a one, not an 'ell'
rurban one plus i, okay
pmichaud actually, one times i
jonathan Hmm, si that not maybe the correct answer? ;-)
rurban mul Nan, Complex
leto_ it happens for any number, not just one
how would I go about fixing this?
pmichaud is it truly a bug?
PerlJam leto_: what's the right behavior?
jonathan Maybe NaN + NaNi is better ;-)
leto_ NaN * anything = NaN
NaN * (a complex expression ) = NaN 16:33
rurban check parrot for this mul type
leto_ I haven't asked Larry about this, that is just what seems intuitive
rurban Parrot_mul_n_p
pmichaud there's a good argument that Parrot is incorrect there, yes.
leto_ rurban: thanks, I will look at that
pmichaud i.e., it might be a bug in Parrot's Complex PMC and not in Rakudo.
rurban It shoudl be in src/pmc/Complex.pmc
I believe it's just a specification thing. What is the expected result? An exception or NaN? 16:34
Same for Inf * Complex, Inf + Complex, ... 16:35
leto_ rurban: yes, semantic for complex numbers are still up in the air, but they interest me so I will hack on them and add tests and hopefully get some clarification from Larry
rurban I dont think that Larry is an authoritative on complex math... 16:37
dalek kudo: 9c86a27 | jnthn++ | (4 files):
Break compile into two stages, the first compiling everything we did before, the second using that first stage to compile the Perl 6 setting and then making the final perl6.pbc. Also adds support for 'no Main' to suppress generating the Parrot :main method. Re-wrote Whatever's perl and ACCEPTS methods into Perl 6 as proof of concept.
kudo: 325cc2e | jnthn++ | (8 files):
Break compile into two stages, the first compiling everything we did before, the second using that first stage to compile the Perl 6 setting and then making the final perl6.pbc. Also adds support for 'no Main' to suppress generating the Parrot :main method. Re-wrote Whatever's perl and ACCEPTS methods into Perl 6 as proof of concept.
shorten dalek's url is at xrl.us/begpau
shorten dalek's url is at xrl.us/begpaw
PerlJam er ... why did 2 commits have the same message? 16:38
rurban Octave may be a good reference though
leto_ rurban: WelI, I can be authoritative on the math, but Larry can be authoritative about what semantics get settled on
jonathan PerlJam: Because I'm too stupid for git.
rurban www-old.cae.wisc.edu/pipermail/bug...05932.html 16:39
shorten rurban's url is at xrl.us/begpa8
jonathan PerlJam: I'm still not used to having to explicitly list what I want to commit...
rurban Looks like the expected result is NaN
leto_ rurban: i will see how Mathematica/Octave/Yacas/Matlab/Maple handle it
rurban better the C99 standard or IEEE 754 16:40
leto_ rurban: I don't know that either has complex numbers, but I will check 16:42
rurban amd64 jit with my new long double fixes get along quite well. I'm passing the first tests on 64bit 16-byte long double
The sentence: "any number (if integer, float, complex, or big) multipled with NaN must result in Nan" is enough. 16:45
leto_ jonathan: I don't know if you use OS X, but GitX has a very nice interface to dealing with adding files (called staging them) that shows you a list of what has been staged, what hasn't and the current diff, all in a feel-good OS X GUI-way
rurban With Inf/-Inf it will get tricky though :)
leto_ rurban: Where is that sentence from? 16:46
rurban from my head :)
leto_ rurban: can I cite that :) ?
Coke sees quevlar has re-opened his trac account.
jonathan leto_: No, Windows
rurban You can cite your mathematics teacher.
PerlJam leto_: how doe gitx differe from git gui ?
Coke misreads leto_ as leo. whoops.
rurban greetings from leo btw. I met him in vienna. he has no time at all. 16:47
PerlJam jonathan: see if you have git-gui. I think it's standard with git these days.
dukeleto PerlJam: I haven't used git-gui, it is like gitk ?
szbalint rurban: you've been in vienna when? :)
Coke rurban: Hi, leo! 16:48
szbalint oh, still
rurban vienna-bratislava workshop
PerlJam dukeleto: not really. It's a gui for staging commits. git-gui will launch gitk for some operations in fact.
dukeleto PerlJam: gitx and gitk are similar except that GitX has native OS X widgets/trackpad support and the "staging view" which is like a commit mode that lets you add files and see the current diff while you are writing the commit message
PerlJam: cool, I haven't messed with it
dukeleto does not = leo, sorry for the confusion 16:49
PerlJam gitk doesn't do repo changes, git-gui does.
rurban I had a lot of fun with leo and andreas kļæ½nig
dukeleto to give fair warning, GitX seems to have some memory leaks when opening huge repos
szbalint is he still smoking his hand rolled cigarettes? :)
jonathan vienna-bratislava was fun :-)
rurban no, not smoking 16:50
I think I will use dtrace to check jit on amd64. Was that Andy Armstrong who added it to perl5?
16:54 PantheraPardus joined
dalek rrot: r36857 | rurban++ | trunk/config/auto/arch.pm:
TT #256: TODO better cpuarch check
 17:03
kudo: 459a7c8 | jnthn++ | (2 files):
Re-work things a bit so we just have one perl6.pir, as suggested by pmichaud++.
 17:04
kudo: 8fef096 | jnthn++ | build/Makefile.in:
Replace spaces in makefile with tabs.
rurban ah wrong ticket number. it should be #356
shorten dalek's url is at xrl.us/begpd9
dalek's url is at xrl.us/begpeb
rurban ah: src/packfile/pf_items.c:385: error: pointer value used where a floating point value was expected 17:15
17:18 rafl_ joined 17:29 alinbsp joined
rurban kj: all t/native_pbc/*.pbc files are broken 17:29
windows probably
or the pbc header updater is broken. regenerated files do convert fine 17:39
17:49 iblechbot joined 17:55 Tene joined
Coke Tene: You have an item in DEPRECATED.pod, methinks. 17:57
hurm. mebbe not.
dalek tracwiki: v4 | Infinoid++ | ParrotQuotes 18:00
tracwiki: Reformat things to look (hopefully) a little nicer.
tracwiki: trac.parrot.org/parrot/wiki/Parrot...ction=diff
shorten dalek's url is at xrl.us/begpky
dalek rrot: r36858 | NotFound++ | trunk/examples/nci/Xlib.pir:
[examples] update data types in Xlib.pir
rrot: r36859 | rurban++ | trunk/t/native_pbc:
Update wrong native_pbc files, broken with the release.
 18:13
Tene Coke: what is it? 18:15
purl it's it!
Coke Tene: I can't find it. 18:16
I probably mean the other $committer =~ m/^t...$/i; 18:17
18:18 barney joined
Tene tewk? 18:18
purl tewk is, like, Kevin Tew, mailto:tewk@tewk.com
Tene Coke: oh, there was that exceptions thing that allison was going to put in deprecated, I think. 18:22
18:32 PantheraPardus joined
rurban I think I just fixed the 64-bit align_16 issue with pbc reading. Stupid bug 18:37
dalek rrot: r36860 | coke++ | trunk:
TT #109 - remove File::Spec
 18:38
18:48 rdice joined 18:50 timbunce joined 19:06 sproingie joined
dalek rrot: r36861 | coke++ | trunk:
TT #107 - remove deprecated library Data::Escape (if you were using this, use the escape opcode instead.)
 19:07
Coke rdice: hey, Richard. 19:09
purl well, richard is going to Foxtons
Coke foxtons?
is a foxton a large unit of gambling?
rdice Coke, g'day mate. 19:12
Coke ... you're nawt australian! 19:13
rdice I can fake it, Bruce.
Bruce that's what sheila said.
dalek pp: 19f09f0 | (Bernhard Schmalhofer)++ | (3 files):
Generate to *.pbc in top dir, simplify installation
 19:14
shorten dalek's url is at xrl.us/begpwk
dalek pp: b196f39 | (Bernhard Schmalhofer)++ | .gitignore:
Add a .gitignore file
pp: 558e71e | (Bernhard Schmalhofer)++ | .gitignore:
ignore pipp_test_run.tar.gz
shorten dalek's url is at xrl.us/begpwn
shorten dalek's url is at xrl.us/begpwp
dalek pp: b6120ee | (Bernhard Schmalhofer)++ | .gitignore:
use complete pathes in .gitignore
shorten dalek's url is at xrl.us/begpwr
dalek pp: 82fdc2b | (Bernhard Schmalhofer)++ | src/phc/phc_xml_test2.xsl:
Merge branch 'master' of git@github.com:bschmalhofer/pipp
Bruce AIGH! DROWNING IN PIPP!
shorten dalek's url is at xrl.us/begpwt
dalek pp: 0dc6f36 | (Bernhard Schmalhofer)++ | .gitignore:
let 'git status' report *.pbc files that are not in topdir
 19:18
shorten dalek's url is at xrl.us/begpw7
rurban who came up the pbc alignment numbers? We have 16/8 = 2 on 64bit, 16/4 = 4 on 32bit. IN reality I would have choosen at least 16 or 32 for 64bit and maybe 8 for 32bit. 19:30
We user a smaller alignment on 64bit than on 32bit. 19:31
19:31 ron joined
dalek rrot: r36862 | fperrad++ | trunk/config/gen:
[config] use TEMP variables (1)
 19:34
Coke rurban; svn blame will probably help.
19:34 chromatic joined
Coke but I'm guessing that dates back to Dan. 19:34
19:34 _timbunce joined
rurban I'm guessing it was an error 19:35
Coke chromatic: you'll be happy, I'm updating a 3rd party library dependency in my app.
no living in the past here!
chromatic o/~ Ohhh, you're halfway there. o/~
rurban But at least I found why 64 + 32bit pbcs don't like each other
The ALIGN_16 is always absolut, but should be specific to the wordsize or the parrot which wrote the pbc. So we have a 64bit - 32 bit mismatch 19:36
stupid.
Coke it was probably also written in the absence of any 64 bit testing platforms. 19:37
rurban yes.
Coke chromatic: any chance you've poked at my favorite RT? 19:38
dalek rrot: r36863 | fperrad++ | trunk/config/auto:
[config] use TEMP variables (2)
rurban I looked for big bugs, but this is only a tiny one. sigh.
19:38 davidfetter joined
Coke (rt.perl.org/rt3/Ticket/Display.html?id=57088) 19:39
nopaste "rurban" at 93.82.86.82 pasted "64bit pbc problem cuase. but still not ready" (11 lines) at nopaste.snit.ch/15664
19:42 silug joined
chromatic The inferior runloop? I've poked at it. 19:43
rurban chromatic: I've re-enabled long-double for jit: at TT#352 is a patch 19:45
I've also taken the set_i_n rounding problem TT#353 (with patch), but this crashes. 19:46
nopaste "bacek" at 123.243.38.218 pasted "Big reduce metaop patch for pmichaud, jnhtn for review." (267 lines) at nopaste.snit.ch/15665 19:48
dalek rrot: r36864 | rurban++ | trunk/src/packfile/pf_items.c:
TT #308. cvt_num8_num12 tested ok, so cheap compiler casts work ok.
 19:50
kudo: 75ea93f | jnthn++ | src/parser/grammar-oper.pg:
infix:<**> should be right associative; patch courtesy of bacek++.
 19:53
shorten dalek's url is at xrl.us/begp23
dalek kudo: a020241 | jnthn++ | src/parser/actions.pm:
Some slight tweaks to code generation of is also classses and when no Main is in effect.
shorten dalek's url is at xrl.us/begp25
dalek pp: 1eb8293 | (Bernhard Schmalhofer)++ | (193 files):
Remove a lot of useless $Id$ tags
 19:54
shorten dalek's url is at xrl.us/begp27
dalek rrot: r36865 | NotFound++ | trunk/t/src/embed.t:
[t] destroy interpreter in embed tests
rrot: r36866 | pmichaud++ | trunk/compilers/pct/src/PAST/Node.pir:
[pct]: Update PAST to generate unique subids across invocations.
 20:08
rrot: r36867 | whiteknight++ | trunk/docs/book/ch04_pir_subroutines.pod:
[Book] Committing two fixes to chapter 4. Thanks hexcoder++
 20:16
20:17 iblechbot joined 20:21 rurban_ joined 20:27 Whiteknight joined
GeJ Good morning everyone 20:28
Infinoid hi GeJ
Whiteknight hello GeJ 20:29
GeJ heaya guys, how's life treating you these days? 20:32
Whiteknight every day above ground is a good day
purl i already had it that way, Whiteknight.
20:33 rdice joined
Whiteknight I had a boss that used to say that all the time 20:48
of course, he was an alcoholic who cursed at nuns and purposefully made his shop handicapped-inaccessible 20:49
Whiteknight reconsiders where he takes his quotes from
szbalint haha 20:53
dalek kudo: 80fd5c5 | jnthn++ | (7 files):
Move a few methods on List and Pair over to the Perl 6 setting.
 21:01
shorten dalek's url is at xrl.us/begqac
Coke spends five hours tracking down something is basically "make sure you specify the encoding at every step."
Infinoid uck
dalek rrot: r36868 | fperrad++ | trunk:
[crypto] t/distro/test_file_coverage.t wants a test file per pmc.

See TT #348		 21:04
Coke now I have < 1 hour left to perform my original task for the afternoon. :) 21:05
21:09 leto_ joined 21:21 bacek joined 21:23 rdice joined
nopaste "bacek" at 114.73.132.155 pasted "Revice reduce metaop patch for pmichaud" (210 lines) at nopaste.snit.ch/15666 21:25
bacek pmichaud: s/revice/revised/...
pmichaud why are we having to do the same stuff for [//=] ? we already know its associativity. 21:29
also GETOPASSOC should return associativity, not true/false. or, if it's checking specifically for right associativity, it should be named something that indicates that. 21:30
bacek ISOPRIGHTASSOC? Or just ISRIGHTSSOC? 21:31
pmichaud I'd prefer if REDUCEMETAOP could dtrt with all the associativities. 21:32
not just right and left.
bacek hm. How other associative can change REDUCEMETAOP? Or you prefer to merge REDUCEMETAOP and REDUCEMETAOPCHAIN? 21:33
pmichaud merging sounds useful, if we can do it.
bacek (And don't quite understand question about //=...) 21:34
pmichaud I've never been too keen on having them as separate functions (although it may be a necessary distinction if they're radically different).
jonathan I forget why they were separate now; I know I added REDUCEMETAOPCHAIN.
pmichaud I think it was originally done as separate because it was shortest path to a goal. 21:35
bacek jonathan: ..CHAIN is shortcircuit
pmichaud but I try to not let too much of that spread throughout the codebase.
jonathan bacek: Ah. That would be one good reason...
bacek: But I think it was more than an optimization.
pmichaud i.e., I'm not a big fan of lots of copy+paste code, which is what this is starting to look like
(I'm also not a big fan of lots of itty-bitty helper functions to make the copy+paste approach look less like copy+paste :-) 21:36
many times copy+paste means we didn't figure out the underlying pattern correctly. 21:37
bacek pmichaud: No way! It's not my fault :)
21:40 IceGuest_7 joined
bacek pmichaud: I can simplify REDUCE if I will always build result_list. 21:40
It will be more clean but less efficient. 21:41
pmichaud why isn't triangle simply a matter of deciding whether to do a push?
i.e., why does it need its own code section? 21:42
bacek In REDUCEMETAOP or ..CHAIN?
pmichaud in either.
bacek ..CHAIN is shortcircuiting. And I have to build full result-list.
REDUCEMETAOP - to avoid building result_list if it not required. 21:43
pmichaud bacek: isn't avoiding the push the same as "avoid building result_list" ?
i.e., I don't see why we need two separate loops.
it would help if I could see the file after the patch in addition to just the patch 21:44
bacek There is only one loop in REDUCEMETAOP...
Ah.
pmichaud let me apply the patch and take a look, just a sec.
bacek no.
just a sec 21:45
dalek tracwiki: v61 | fperrad++ | Languages
tracwiki: update Parrot version to 0.9.1
tracwiki: trac.parrot.org/parrot/wiki/Langua...ction=diff
shorten dalek's url is at xrl.us/begqfo
dalek rrot: r36869 | allison++ | trunk:
[install] Installing appropriate generated files for running dynops

utility libraries for op generation.
pmichaud GROWL 21:49
$I0 = opfunc(a, b)
unless $I0 goto false
WRONG WRONG WRONG
purl NOW YOU MUST DIE!
dalek tracwiki: v62 | fperrad++ | Languages
tracwiki: add links to Smolder (Lua &amp; Pipp)
tracwiki: trac.parrot.org/parrot/wiki/Langua...ction=diff
shorten dalek's url is at xrl.us/begqge
jonathan takes a few paces back form pmichaud and purl 21:51
pmichaud i see that a lot in the codebase and it makes me cringe.
jonathan Why is it a fail, generally?
pmichaud 0 but True
purl 0 but True is a C exit code thing.
jonathan Ah.
Yes
pmichaud "xyz"
numifies to 0 but boolifies to true. 21:52
jonathan *nod*
Coke ah. so $P0 = opfunc(a,b)\\n unless $P0 goto false would work.
jonathan $P0 = opfunc(a, b) / unless $P0 goto false should be fine.
pmichaud yes
Coke smacks jonathan.
jonathan But yes, good point.
pmichaud but the point is also that it makes the code itself simpler.
purl okay, pmichaud.
jonathan Coke: What'd I do to deserve that? ;-)
pmichaud because we don't need to then turn around and convert the integer back into a True/False value. 21:53
it's silly that we convert true/false results into integers just to turn them back into true/false again.
jonathan Aye.
pmichaud (in addition to the fact that it's an error)
jonathan I'm sure I've managed to screw that up at least once... 21:54
dalek tracwiki: v63 | fperrad++ | Languages 21:58
tracwiki: typo
tracwiki: trac.parrot.org/parrot/wiki/Langua...ction=diff
shorten dalek's url is at xrl.us/begqhy
pmichaud when bacek returns: the key feature about reducing chained ops is not that it short circuits, but rather that there's an implicit 'and' in the comparisons. 21:59
cotto fperrad++ #"redundance or duplication"
Coke jonathan: nothing, merely send duplication.
dalek kudo: d3cc978 | jnthn++ | (5 files):
Some more methods from Bool and Pair converted to Perl 6.
 22:00
shorten dalek's url is at xrl.us/begqik
jonathan pmichaud: Ah, that's it.
pmichaud: I was sure I'd done it for more than optimization... 22:01
22:01 Whiteknight joined 22:08 timbunce joined 22:14 bacek joined, TiMBuS joined
bacek pmichaud: sorry, my laptop battery just died... 22:14
22:17 bacek_ joined
pmichaud 21:59 <pmichaud> when bacek returns: the key feature about reducing chained ops is not that it short circuits, but rather that there's an implicit 'and' in the comparisons. 22:27
or, if there's not an implicit 'and', then we might need to make that clear in the specification for reduction of chained ops.
for example: [\\==] 3, 4, 4, 4 should probably be False, False, False, False 22:28
since 3 == 4 is false, 3 == 4 == 4 is false, etc.
dalek kudo: 444a4c8 | jnthn++ | (2 files):
Add Array.pm to the prelude. For now it just contains splice, written by cspencer++ and attached to RT#49173. Happily, this gets two more of the 99 problems spectests running.
 22:31
shorten dalek's url is at xrl.us/begqn3
bacek pmichaud: (sorry for delay... I at $dayjob ATM). github.com/bacek/rakudo/blob/55f81a...assign.pir with my version of REDUCEMETAOP 22:44
shorten bacek's url is at xrl.us/begqpa
dalek rrot: r36870 | NotFound++ | trunk:
[core] add functions Parrot_get_root_namespace and Parrot_PMC_get_pmc_strkey to extend.h
bacek And I'll try to read spec more carefully to understand why [\\==] implies explicit 'and'
Infinoid It's for short-circuiting, right? 22:59
Well, that's the effect, not the cause 23:01
dalek rrot: r36871 | NotFound++ | trunk/t/src/embed.t:
[t] compile and use a pir sub in embed
 23:18
Tene does $.foo not work for attr access in a class? 23:23
in rakudo?
jonathan Should work. 23:24
Surely we have tests for that.
Tene Pair.pm uses the self.foo
jonathan Could use $.foo too 23:25
I expect.
Tene is 'min' implemented?
jonathan Pair.pm was a port of the PIR. ;-=)
Yes.
Tene Ah.
jonathan Tene: If you have a commit bit, feel free to s/self.key/$.key/
Tene I should just check out rakudo from git and figure out if I have access
jonathan It's clearer.
Tene I was asking about 'min' for Array.pm in the setting 23:27
splice()
purl i think splice() is sexy...is there a Perl module that exposes that yet?
Tene purl: forget splice
purl Tene: I forgot splice
jonathan Tene: splice can probably be written more idiomatically. 23:28
sproingie smoke_languages fails with lua :(
src\\pmc_freeze.c:955: failed assertion '(int)io->image->bufused >= 0'
jonathan Tene: OTOH, we don't wanna get *too* idiomatic there.
Tene jonathan: is there an official or unofficial kind of policy on the balance there? 23:29
jonathan Or too clever, I guess.
Tene: Not yet, but it may come down more to, what's available.
Since you can only really use stuff classes that were declared at some point earlier. 23:30
Tene Right.
jonathan (Though is also lets you extend them some...)
I ain't got that ordering completely worked out yet, though. 23:31
Tene pmichaud: can I have commit privs on rakudo on github? 23:33
23:33 timbunce joined
Tene or am I just supposed to fork it on github and ask for pull requests? 23:34
23:37 gravity joined
chromatic Can you define custom operators in Rakudo right now without writing PIR? 23:54
jonathan chromatic: No. 23:59
chromatic Okay, all I needed to know. Thanks.

Powered by Cro and the RakuĀ® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!