Parrot 4.6.0 "Wild Parrots of Telegraph Hill" | parrot.org/ | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC
Set by moderator on 17 July 2012.
00:35 jevin_ joined 00:36 eternaleye joined, simcop2387 joined, dngor joined 00:37 rblackwe joined, autark joined, perlite joined 00:39 LizM joined, tokuhirom joined, plobsing joined, benabik joined, schm00ster joined, rurban joined, Hunger joined, cotto joined, Coke joined, knewt joined, sorear joined, atrodo joined, NotFound joined, smash joined, pmichaud joined, ttbot joined 00:40 arnsholt joined 00:42 Maddingue joined, woolfy joined, Patterner joined, awwaiid joined, wendar joined, alvis joined, Timbus joined, contingencyplan joined, Khisanth joined, preflex joined, aloha joined, ingy joined, cxreg joined, szbalint joined, Tene joined, nine joined 00:44 slavorg joined 01:53 preflex_ joined 02:36 Psyche^ joined 05:44 fperrad joined 05:49 kurahaupo joined 05:51 brrt joined 05:53 jsut joined 05:59 brrt joined
moritz rurban: on www.perlmonks.org/?node_id=982343 you have '4.0.6' in the title, instead of 4.6.0 06:09
06:32 kurahaupo joined, brrt1 joined 06:34 brrt joined 08:57 preflex_ joined 09:20 Patterner joined
dalek Heuristic branch merge: pushed 20 commits to nqp/toqast by moritz 09:48
kudo/toqast: 85177c8 | pmichaud++ | src/core/Rat.pm:
Fix relational operators for (Int,Rat).
kudo/toqast: 3337df6 | pmichaud++ | src/core/Str.pm:
Refactor Str.match to be lazier and accept a few more options.
kudo/toqast: 24d833a | pmichaud++ | src/core/Str.pm:
Restore :x(*) handling suggested by spectests.
09:52 lucian joined 10:07 ligne joined 11:08 JimmyZ joined 11:55 lucian joined 12:06 schmooster joined 12:25 zhutingting joined
JimmyZ 晚上好 12:27
zhutingting hi muren~ 12:28
12:40 zhutingting left 12:42 zhutingting joined 12:59 PacoAir joined
dalek p: c1197a5 | moritz++ | src/ops/nqp_dyncall.ops:
[dyncall ops] mark in PMCs as invar

this reduces the size of the generated C code by a factor > 4, and the PMCs cannot reasonably be constants anyway
13:26
13:29 JimmyZ joined 13:42 whiteknight joined 13:43 wknight8111 joined
rurban moritz++ thanks. updated 13:55
wknight8111 good morning, #parrot 13:59
rurban++ on the release! 14:00
rurban good morning whiteknight. Did you see what I announced for you to do:) Targets for 4.7 threads and io_cleanup1
wknight8111 rurban: I caught some of the emails. I AM going to work on those two branches 14:01
and coke's rm_pasm branch too
rurban That's why I announced it. I hope we can make it this time.
Coke someone want to block SMK from here at #ps? 14:07
*and
wknight8111 hmmm, I don't know where "whiteknight" is signed on from, and I don't have ops with this account 14:09
JimmyZ SMK? 14:11
Coke spammer in #ps 14:12
14:13 zhutingting joined
wknight8111 Coke: While I'm thinking about it, what's the status of the rm_pasm branch? I haven't looked at it in a while 14:13
JimmyZ had been watching not_gerd's m0 and he really likes it. 14:14
Coke nor have I . did a recent merge from master, and it doesn't build now, needs more love.
(the move of pbc_to_exe to winxed needs to be updated to a non-pasm world) 14:15
I'd be slightly concerned about users.
probably needs a writeup about moving from PIR to PASM. (esp. for includes of old constant PASM files) 14:16
once it's building and running and testing, we still have the issue of IMCC still having lots of guts that work on PASM, even if we don't call them from anywhere.
... do we just leave that because m0? maybe. probably not worth the cleanup. 14:17
wknight8111 maybe what I can do is update the IMCC .include loader to search .pasm and .pir file extensions, for compatibility
Coke that's fine. I don't really have a horse in this race, I just wanted to move it along. 14:21
if there's a good way to remove most of PASM but still keep people moving, awesome.
wknight8111 ok, cool 14:27
if we can get any of those three branches in this month, I'll consider it a success 14:31
14:31 zhutingting joined
wknight8111 I also really need to take a look at GC and destructors, since those things seem to be a hot topic for Rakudo 14:31
14:34 dmalcolm joined
rurban agreed. There is something fishy 14:35
14:46 benabik joined 15:08 alester joined 15:57 kurahaupo joined 16:07 lucian joined 16:15 jashwanth joined 16:20 zby_home joined
dalek rrot: eb328cf | rurban++ | t/pmc/ (2 files):
[GH #799] cygwin sometimes hangs at alarm and task

signal delivery is a bit too slow and unreliable on cygwin most of the time
16:22
p: 344b6ca | moritz++ | src/ops/nqp.ops:
[nqp ops] invar annotations

reduces size of generated C code by a factor 2.5
17:02
17:41 davidfetter joined 18:10 zby_home joined 18:14 tuxit joined 18:22 davidfetter joined 18:43 ilbot2 joined
moderator Parrot 4.6.0 "Wild Parrots of Telegraph Hill" | parrot.org/ | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC
18:44 ligne joined 18:46 szbalint joined, schmooster joined, Maddingue joined, woolfy joined, awwaiid joined, wendar joined, alvis joined, contingencyplan joined, Khisanth joined, aloha joined, cxreg joined, Tene joined, nine joined, Util joined, Timbus joined, Hunger joined, eternaleye joined, kurahaupo joined, benabik joined, fperrad joined, LizM joined, tokuhirom joined, plobsing joined, schm00ster joined, rurban joined, cotto joined, Coke joined, knewt joined, sorear joined, atrodo joined 18:52 ingy joined 18:55 slavorg joined 19:32 lucian joined
rurban alvis: Cannot we rename "Parrot Documentation Releases (4.5.0 - 0.1.1)" to "Old Parrot Documentation Releases (4.5.0 - 0.1.1)" on parrot.github.com/ 20:09
20:20 travis-ci joined
travis-ci [travis-ci] parrot/parrot#484 (master - eb328cf : Reini Urban): The build was broken. 20:20
[travis-ci] Change view : github.com/parrot/parrot/compare/4...328cf43932
[travis-ci] Build details : travis-ci.org/parrot/parrot/builds/1896722
20:20 travis-ci left 20:29 whiteknight joined
whiteknight I cleaned out my laptop fan of dust. There was quite a lot in there 20:32
So I hope I'll be able to do CPU-intensive things like various builds and tests that I wasn't able before 20:33
also, I'm hoping to be able to run a VM at more than a crawl
msg dukeleto: The original goal was to definitely use PLA, because it handles allocation, indices, and some basic arithmetic. I know jashwanth was running into troubles with PLA/BLAS linkage 20:34
aloha OK. I'll deliver the message.
Coke wonders if we know if anyone faild midtersm yet 20:36
20:58 dmalcolm joined
whiteknight okay, I *think* I've got the algorithm outlined to fix .readline() in io_cleanup1 21:25
along the way it does add a cool new feature: the ability to use multi-character delimiters in FileHandle readlines
21:35 lucian joined
rurban but we only need \\n and \\r\\n 22:02
benabik rurban: \\r\\n is multi-character 22:03
rurban do you have a way to tell web apps to use .readline with \\r\\n delims easily?
benabik: sure, but mosty the implementations are just hacks for those two. mac classic with \\r is gone 22:04
benabik We had a system that claimed to work with arbitrary strings. So whiteknight++ went ahead and made it actually work. This is not a bad thing. 22:05
rurban I thought it was a cleanup to make it smaller and faster, not more general for alien visitors. 22:06
ehr, unicode folks
benabik I think regularity was high on the list too. 22:07
So given Socket.readline(delimiter), it made sense to create a generic "read a line broken by [delimiter]"
rurban I thought about .setdelim("\\r\\n") or such. 22:08
or as optional argument to the object creation
parrot maintainers: you can now freely update parrot on the old but still good freecode system freecode.com/projects/parrot/ticke...223f9111f5 22:15
22:30 drift joined 23:35 kid51 joined 23:51 zhutingting joined
whiteknight rurban: We can change the concept of "line" to be any delimited text 23:57
rurban: So we can read data from any delimited text, no matter what the delimiter is
If anything, the method name "readline" is wrong, because now it's "read_until_delimiter"