Parrot 2.10.1 Released | parrot.org | Log: irclog.perlgeek.de/parrot/today | Long live Git! github.com/parrot/parrot | git clone git://github.com/parrot/parrot.git
Set by moderator on 18 November 2010.
davidfetter heh 00:00
anyhow, i'd like to make a PL/OSDCLANG 00:01
OSDCLang being "trivially isomorphic" to BF
it's just a matter of changing the 8 symbols of BF 00:02
so i guess i'd try PL/BF 1st, then try to morph it into PL/OSDClang 00:03
NotFound "modern, flexible unicode aware and twitter enabled turing machine" ? 00:05
00:10 theory left
dalek rrot/tt532_headerizer_refactor: 8128e25 | jkeenan++ | t/tools/dev/headerizer/ (4 files):
Test additional branches and conditions in
 00:19
00:36 silug joined 00:58 bluescreen_ joined, bluescreen left
Coke (NS issues) - partcl_old avoids this by putting things into a RPA container and then dereffing them to avoid having the subs be treated specially. 01:08
davidfetter NotFound, if you like ;) 01:11
NotFound davidfetter: I don't like Turing ;) 01:12
NotFound thinks about writing that lyric, to be sung with the 'I don't like mondays' music. 01:14
Definitely, i must go to bed. I'm daynightmareing. 01:15
davidfetter o/` i like tu-ring and i can-not lie o/` 01:27
01:28 kid51 left
whiteknight NotFound: can you put in the deprecation notice for readline? I can start drafting a replacement if you want 01:35
actually, let's discuss it at #ps on tuesday 01:38
02:04 marcio_ferreira left 02:10 davidfetter left 02:14 shockwave joined
whiteknight msg bacek can you take a look at TT #1561? I don't think any of our types autovivify nested aggregates, so I think we can close the ticket 02:21
aloha OK. I'll deliver the message.
bacek_at_work whiteknight, src/pmc/hash.pmc, line 101 02:23
it's still auto-vivify
whiteknight ah, that's what I was missing 02:24
any of the array types do it?
bacek_at_work fpa at least 02:25
lin 496
line
actually, killing bloody Keys will help.
whiteknight blarg 02:26
yes, killing Keys will be wonderful
bacek_at_work TT #903 :) 02:27
whiteknight I'm going to fix Hash right now 02:28
bacek_at_work fix 903 instead :) 02:29
whiteknight one thing at a time 02:30
bacek_at_work just kill Keys.
We can cleanup unused code in Hash/Arrays later
whiteknight is there a deprecation notice for keys ? 02:31
02:36 rurban_ joined 02:39 rurban left, rurban_ is now known as rurban, marcio_ferreira joined
bacek_at_work whiteknight, no idea... Probably no. 02:39
whiteknight bacek_at_work: can you put one in? 02:40
chromatic We've talked about deprecating them for a long time.
whiteknight we can talk about it at #ps if necessary
bacek_at_work chromatic, yes. But we didn't decided (yet?)
whiteknight I'm hugely in favor of the idea
chromatic I'd like to remove all of the buggy, ad hoc pseudopolymorphism they imply. 02:43
I wouldn't mind Key if it weren't always "one of a string, numeric value, NameSpace, or something else".
plobsing but users might want to use floating point numbers in their namespaces ;-) 02:44
chromatic That's fine. 02:45
As long as they convert it to a Key first.
02:50 marcio_ferreira left
bacek_at_work chromatic, than we should have StringKey, IntKey, etc. With definition like "pmclass StringKey extend RPA {}" :) 02:52
chromatic I don't think so. 02:54
I think you create a key from primitive and it's always a Key.
You never get a primitive value out of a Key.
bacek_at_work chromatic, you do. For navigating aggregates. 03:03
chromatic I think that's a mistake. 03:04
The current design of Key is very fuzzy about its value.
That, to my mind, is its biggest flaw. 03:05
bacek_at_work chromatic, trac.parrot.org/parrot/wiki/ParrotQ...version=15 03:08
chromatic Oh good, I worried before reading that that you were catching my opinion in a state of flux. 03:10
03:12 shockwave left, whiteknight left
dalek rrot: 5c6db56 | Whiteknight++ | src/pmc/hash.pmc:
don't autovivify nested hashes. TT #1561
 03:13
rrot: 73ceeb1 | Whiteknight++ | config/auto/ipv6/test.in:
Merge branch 'master' of github.com:parrot/parrot
03:45 Andy joined
Andy Do we intentionally not have .gitignore 03:45
cotto Andy, we have one 03:46
Andy in the root? 03:47
cotto yes
Andy wow 03:53
I wonder what happened that I didn't have one.
Trying to figure right way to add *~ to it. 03:59
plobsing Andy: if that's the way you want to manage your editor, why not make it a global gitignore entry? 04:09
Andy It is.
plobsing I'm missing somethign. If its global, why do you need it in the project-local gitignore as well? 04:10
Andy I don't know. 04:11
I am getting ~ files in my git status
OK, hold on 04:14
what do you mean when you say "a global gitignore entry"
Do you mean ~/parrot/.git/info/exclude? 04:15
I never realized until just now that .gitignore was per-directory
plobsing: Can you shed light here? 04:18
plobsing git can have global settings as well as local settings 04:21
I'm fairly sure gitignore can work the same
Andy ok, Got that
plobsing reading man gitignore, you can specify any file to be a global ignore file by setting the core.excludesfile value 04:24
it works here for me. asdf~ isn't showing up in git status 04:25
Andy OK, specifcially what do you have set up
AHA 04:26
It has to be a separate file
plobsing in ~/.gitconfig, [core]\\nexcludesfile = ~/.gitignore
Andy My setting was core.excludes = *~ etc 04:27
04:41 davidfetter joined
cotto seen whiteknight 04:55
aloha whiteknight was last seen in #parrot 2 hours 14 mins ago saying "I'm hugely in favor of the idea".
chromatic citeseerx.ist.psu.edu/viewdoc/downl...p;type=pdf 06:04
cotto It's hard not to like that kind of emphasis. 06:06
plobsing emphasis? I read that as scare-quotes 06:07
06:08 Andy left
chromatic I ain't a-scared of no quotes. 06:09
cotto 'The Structure and Performance of "Efficient" Interpreters' would have a different and more suspicious meaning. 06:12
plobsing I'm always suspicious when someone tries to sell me something as more efficient. 06:14
chromatic After a few hundred tuning commits? Your words hurt me! 06:15
cotto plobsing, I've got a very efficient bridge you might be interested in.
06:49 chromatic left 07:10 contingencyplan left 07:35 bacek joined 07:44 bacek left 07:56 bacek joined
bacek aloha, humans 07:57
cotto aloha, bacek 07:59
bacek hi cotto
Did I miss something valuable over weekend?
cotto sunscreen, apparently 08:00
my weekend just ended a few seconds ago
goodbye, weekend 08:01
bacek yeah... Sunscreen... It wasn't bright idea to not reapply it every few hours. 08:03
Any screaming from Rakudo team about broken GC? 08:04
moritz for which revision?
cotto didn't hear anything
cotto doesn't hear a lot of things 08:05
moritz hasn't built rakudo on a new parrot for a while
bacek moritz, starting from f0aa28e4faa98850720196535463e19261dc4510 - "Merge gc_ms2_faster_list"
moritz will try now
cotto sleeps
'night
bacek night cotto 08:06
moritz, parrot should be about 30% faster :)
moritz that's good news :-)
bacek on "make" and "make spectest" in rakudo
moritz, indeed :) 08:07
moritz compiling core.pm takes lots of memory, but then I have lots of memory too 08:09
so far 903M virtual memory
971M was peak
bacek it should take about same amount as before 08:10
moritz, any luck with building rakudo? 08:25
moritz bacek: jup. The first attempt failed, but nuking the install location helped
bacek makes sense :)
moritz it died with Can't load oplib '
or so
rather strange
sorear moritz: sizeof(void*)?
moritz sorear: no idea... it's a 64 bit linux on amd64 08:26
bacek sorear, nope. Looks like "make install" in parrot doesn't update symlink for .so to .so.<version>
afk # going to fetch some dinner. Will be around in about 1 hour 08:27
08:29 fperrad joined
moritz a rakudo spectests just died with a memory trace (or whatever that thing is called) 08:30
rand.rakudo 08:31
08:38 PacoLinux left
sorear core dump? 08:46
stack trace?
dalek rrot: fea9faa | fperrad++ | / (2 files):
[smolder] swap sha1 & git_describe
 08:47
moritz core dump probably 08:48
dalek rrot: 72ef570 | fperrad++ | config/gen/makefiles/root.in:
[makefile] SVN is gone
 08:53
08:59 baest joined
bacek moritz, it's... bad. linux/amd64? 09:12
moritz bacek: yes
bacek: I've run two complete spectests now, in each 2 different files failed
bacek ok. I'll try to reproduce it.
otoh, make spectest passed on linux/ia32 before I merged branch. 09:13
which one is second?
moritz bacek: one of the trig ... tanh.t 09:16
bacek moritz, ok. I'll focus on these two
moritz not very reproducible, I'm afraid 09:17
bacek moritz, "echo 0 > /proc/sys/kernel/randomize_va_space" helps a lot to reproduce it 09:18
and "echo 0 > /proc/sys/vm/vdso_enabled"
from root shell of course :)
moritz -su: /proc/sys/vm/vdso_enabled: No such file or directory 09:19
bacek maybe it's 32-bit specific
what about randomize_va_space?
moritz that exists 09:20
bacek can you rerun spectest?
moritz yes
09:26 jasonmay left 09:46 jasonmay joined 10:01 plobsing left
bacek moritz, sigh. I can't reproduce any of your failures. 10:22
But I'm on linux/ia32
moritz :( 10:23
bacek trying to "fix" GC memory limit and rebuild rakudo. 10:24
will take some time due 16MB limit :)
dalek TT #938 closed by bacek++: Parrot svn could use a trusted-root-signed-cert 10:33
TT #938: trac.parrot.org/parrot/ticket/938
TT #1076 closed by bacek++: Smoking Git Branches
TT #1076: trac.parrot.org/parrot/ticket/1076
10:36 rurban_ joined
bacek msg plobsing You've got few deprecation tickets for *ManagedStruct. Any progress? Or I can just jump on it and remove deprecated behaviour? 10:37
aloha OK. I'll deliver the message.
10:39 rurban left, rurban_ is now known as rurban 10:50 dip joined 11:07 zby_ joined
mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#1254) fulltest) at 72ef570 - Ubuntu 10.10 i386 (g++-4.5 with --optimize) 11:47
dalek rrot: 5f29da5 | bacek++ | examples/pir/sudoku.pir:
Use hierarchical namespace instead of flat. Part of TT#1486.
 11:58
rrot: 49db65f | bacek++ | runtime/parrot/library/SDL.pir:
Add loadlib 'io_ops' to make SDL.pir buildable again. Doesn't work anyway due other failures and absence of any tests.
rrot/gc_ms2_sf: 36140f8 | (James E Keenan (Jim))++ | DEPRECATED.pod:
Correct spelling error only.
rrot/gc_ms2_sf: e36b0b5 | dukeleto++ | CREDITS:
Update CREDITS in preparation for git conversion
rrot/gc_ms2_sf: 0792632 | dukeleto++ | / (3 files):
[tools] Make the author map code an actual file and remove it from CREDITS
rrot/gc_ms2_sf: 6e51038 | (James E Keenan (Jim))++ | /:
Creating branch to explore possibility of adding 'osvers' during
bacek oh shi...
rrot/gc_ms2_sf: 6893eec | (James E Keenan (Jim))++ | config/ (5 files):
First pass at determining 'osvers' during configuration.
rrot/gc_ms2_sf: 7c75f1d | petdance++ | compilers/imcc/cfg. (2 files):
flagging more functions as pure
rrot/gc_ms2_sf: 6d65d4b | petdance++ | compilers/imcc/optimizer.c:
removed an unused variable, and consted a pointer
rrot/gc_ms2_sf: 405f614 | (James E Keenan (Jim))++ | tools/dev/mk_author_map.pl:
[codingstd] Add copyright notice and svn id tag.
rrot/gc_ms2_sf: 2bf0285 | mikehh++ | tools/dev/mk_author_map.pl:
fix so perlcritic will pass - it did not like -
rrot/gc_ms2_sf: 67409ed | NotFound++ | t/pmc/string.t:
test String get_string_keyed vtable
rrot/gc_ms2_sf: 5d62dba | cotto++ | CREDITS:
remove extra a.k.a. info from CREDITS
rrot/gc_ms2_sf: 3047796 | (James E Keenan (Jim))++ | t/pmc/hashiterator.t:
Correct spelling error in comment.
rrot/gc_ms2_sf: 32499f2 | petdance++ | include/parrot/hash.h:
localizing a loop var
rrot/gc_ms2_sf: be9f977 | petdance++ | src/gc/gc_ms.c:
rmeove unused var
rrot/gc_ms2_sf: fc4b06d | petdance++ | src/debug.c:
skip_whitespace is now a pure function
rrot/gc_ms2_sf: 4c28659 | petdance++ | src/gc/ (2 files):
shimmed up an argument
rrot/gc_ms2_sf: 7dbeeea | petdance++ | src/gc/string_gc.c:
removed unused vars
rrot/gc_ms2_sf: d91c2c5 | moritz++ | CREDITS:
[CREDITS] remove my middle initial, I nearly never use it
rrot/gc_ms2_sf: 3205a9d | NotFound++ | t/pmc/string.t:
test String get_integer_keyed and get_pmc_keyed vtables
rrot/gc_ms2_sf: 03356fe | (James E Keenan (Jim))++ | config/ (5 files):
Merge tt1841_osvers branch into trunk. Adds 'osvers' to configuration data.
rrot/gc_ms2_sf: a3995b7 | (James E Keenan (Jim))++ | config/init/hints/freebsd.pm:
Attempting to deal with tinderbox build failure report. Commenting out assignment to 'osvers'.
rrot/gc_ms2_sf: c1ca60c | (James E Keenan (Jim))++ | config/init/hints/freebsd.pm:
Trying to track down source of build failure on FreeBSD.
rrot/gc_ms2_sf: cb00f06 | (James E Keenan (Jim))++ | config/init/hints/freebsd.pm:
Again have to comment out assignment to 'osvers' to prevent build error reported by taptinder.
rrot/gc_ms2_sf: bca19ea | petdance++ | / (4 files):
adding compiler annotations to functions and args
rrot/gc_ms2_sf: 17db4d0 | (James E Keenan (Jim))++ | docs/dev/parrot_api.pod:
Correct POD formatting errors.
bacek msg chromatic Sorry, I fucked up your banch badly... I did merge it with master and totally forgot about it....
aloha OK. I'll deliver the message.
rrot/gc_ms2_sf: ad36671 | (James E Keenan (Jim))++ | config/init/hints/freebsd.pm:
Correct syntax error spotted by doughera++; TT #1841.
rrot/gc_ms2_sf: 075907c | (James E Keenan (Jim))++ | lib/Parrot/Headerizer.pm:
Correct spelling error. Improve some POD formatting.
rrot/gc_ms2_sf: 3994d4d | (James E Keenan (Jim))++ | config/init/hints/darwin.pm:
Correct erroneous argument to /usr/sbin/sysctl.
 12:00
parrot/gc_ms2_sf: 7094c7d | petdance++ | src/exceptions.c:
parrot/gc_ms2_sf: consting
12:00 dalek left
bacek And I killed dalek... 12:02
12:03 cognominal left 12:04 cognominal joined 12:07 dalek joined 12:13 ligne joined 12:34 fperrad_ joined 12:36 fperrad left, fperrad_ is now known as fperrad 12:42 contingencyplan joined
Coke how do I get a listing on github something like "svn log --stop-on-copy ^/branches/codestring-deprecation-prep 13:09
?
*sigh*. it feels like I am never going to catch up with the break-fix work in partcl-nqp. 13:10
13:25 Patterner left 13:35 Psyche^ joined, Psyche^ is now known as Patterner 13:45 whiteknight joined
whiteknight good morning, #parrot 13:47
mikehh hi whiteknight 13:50
whiteknight hello mikehh, how are you today? 13:53
mikehh whiteknight: passable, passable - and yiou? 13:54
whiteknight doing well, actually. Basically got a full night's sleep 13:55
mikehh excellent 13:56
whiteknight yeah. it's a rare occurance
mikehh rakudo: 8ddf7f2 - build ok, spectest_smolder #1262 (some failures - two tests did not complete (passed what they ran) on parrot (RELEASE_2_10_1-423-g72ef570) - Ubuntu 10.10 i386 (g++-4.5 with --optimize) 14:02
p6eval rakudo : OUTPUTĀ«===SORRY!===ā¤Confused at line 22, near "8ddf7f2 - "ā¤Ā»
mikehh oops, shouldn't have put the : there 14:03
dalek TT #1151 reopened by whiteknight++: Exception thrown from constructor leads to oddness
TT #1151: trac.parrot.org/parrot/ticket/1151
whiteknight ...I don't udnerstand that at all. Coke replied to the ticket without reopening it. Why does my reply to him reopen the ticket? 14:31
Coke whiteknight: did you reply via email or ticketing system? 14:32
whiteknight email
Coke me too.
odd.
I didn't see a note from you reopening the ticket, just the notification that you reclosed it. 14:33
whiteknight yeah. Oddness
whatever, it is dead 14:34
Coke ah. my reply never hit trac.
due to the bug in trac that forgets to set the "reply-to" properly.
whiteknight trac-- 14:35
Coke if it's not fixed, how do you know it's the same issue?
whiteknight I traced through it last night in the debugger. It's the inferior runloops problem
There are lots of ways to trigger it 14:36
dalek TT #1151 closed by whiteknight++: Exception thrown from constructor leads to oddness
TT #1151: trac.parrot.org/parrot/ticket/1151
whiteknight it just goes to show how fragile our current architecture is
but you're right, on the surface they absolutely do not look like the same issues 14:38
Coke when closing tickets as duplicates, since trac doesn't link the tickets, it's probably a good idea to note in the remaining what the duplicate was. 14:39
(done for this one.)
since we'll want to have test for /both/ ways.
whiteknight I like that idea. It does make sense 14:41
I also think that's a good feature for Trac to add. Maybe we should request it
14:42 bluescreen_ left
Coke aloha, msg plobsing - I tried the codestring free branch on partcl, and it didn't /seem/ to generate any additional failures. 14:50
aloha Coke: OK. I'll deliver the message.
whiteknight Coke: did you see those two tickets I worked on for you last week? 14:52
Coke whoever had 100% failures on partcl the other day probably needs to fix their install directory.
whiteknight: Yes, thank you.
whiteknight Coke: I haven't had an opportunity to do so, and maybe don't know the best use-cases, but if somebody can put together short tests we can merge those issues and everybody will be happy 14:53
atrodo whiteknight++ # I enjoy reading your blog 14:54
whiteknight atrodo: thanks! 14:59
15:01 allison_ joined 15:06 Andy joined
whiteknight I only have two more features to implement before I think the embed_api branch will build and run: pmc thaw (to get a config hash from the binary data in parrot_config.c) and an error message reporting mechanism 15:20
15:32 tadzik left 15:33 PacoLinux joined
whiteknight The thaw will probably be called Parrot_api_pmc_deserialize() 15:34
I'm not sure whether I want that function to take a STRING* or an unsigned char*. I am leaning towards the later 15:35
atrodo I would say the latter 15:36
whiteknight yeah, from the embedder's perspective, they have byte arrays but not strings 15:37
it seems like a waste to have to wrap it up in a temporary string for only that purpose
atrodo Exactly
15:37 tadzik joined
whiteknight I mean, I'll have to do the wrapping internally, but it saves the user from having to repeat the same steps 15:38
atrodo whiteknight> And that's exactly how an embedding API should treat it's users 15:44
if you have to do it every time, you should do it for the user
15:50 allison left 15:52 allison_ is now known as allison 15:58 dmalcolm joined
NotFound whiteknight: I didn't close those tickets because they depend on finalize, and finalize has still experimental status. 16:01
whiteknight NotFound: I didn't close the main ticket. #833 is the ticket, all the rest were duplicates 16:03
when finalize is not experimental (I am going to talk about it tomorrow at #ps), we can close #833
NotFound Fine for me.
whiteknight: about that, set_label and get_label also should be out of experimental. 16:04
find_codepoint too
find_codepoint replaces functionality formerly in CodeString and we are about to kill CodeString. 16:05
16:07 silug left
Coke whiteknight: regarding tickets in parrot for things not parrot - seems reasonable to have a ticket on our radar to track a ticket elsewhere. (did this for macports occasionally.) 16:10
whiteknight Coke: okay, that works for me if we actually have a ticket in the elsewhere tracker for us to follow 16:13
A vague to-do ticket in our system for somebody else's software is not worthwhile
Coke agreed.
anyone here use perlbrew? 16:14
NotFound There is some metatracker software?
whiteknight NotFound: I agree (get_label, set_label, find_codepoint, etc)
dalek TT #581 closed by whiteknight++: make segfaults with PGE.pbc on fedora (r38365)
TT #581: trac.parrot.org/parrot/ticket/581
TT #628 closed by whiteknight++: Win64 build fails with 64-bit CL
TT #628: trac.parrot.org/parrot/ticket/628
TT #642 closed by whiteknight++: voting system for parrot foundation
TT #642: trac.parrot.org/parrot/ticket/642
whiteknight those should all be brought up tomorrow
Coke anyone here have admin privs on feather? 16:16
whiteknight I have some kind of login on feather, but I'm long-since forgotten what privs I have or even how to log in 16:17
Coke answers his perlbrew question (no, the cpanm that perlbrew installs doesn't do anything magic like share your INC dir across perl installs) 16:22
Coke regens trac.parrot.org/parrot/wiki/ParrotFaces 16:23
cotto_work ~.~ 16:30
dalek tracwiki: v5 | coke++ | ParrotFaces 16:31
tracwiki: rerun tools/dev/faces.pl
tracwiki: trac.parrot.org/parrot/wiki/ParrotF...ction=diff
PerlJam Parrot faces or potential serial killers? You decide. 16:45
NotFound I don't understand why serial killers have such reputation. Surely parallel killers are a lot worse. 16:47
whiteknight certainly a lot more efficient
and more scalable 16:48
Coke aloha, msg chromatic - whoops, the qw() issue was not fix. my build script just unconditionally used a 5.12 perl. 16:52
aloha Coke: OK. I'll deliver the message.
dukeleto 'ello 16:59
dalek rrot: f92a58e | NotFound++ | t/pmc/sub.t:
change get_addr to get_label in a Sub test, TT #1857
 17:01
TT #1861 created by coke++: ipv6 config probe too noisy 17:03
TT #1861: trac.parrot.org/parrot/ticket/1861
dukeleto whiteknight: i see you've forked parrot.git again on github. Any reason? 17:09
17:10 Kulag left, Kulag joined
dalek rrot: e9cbce2 | Coke++ | / (8 files):
perl 5.13 deprecates use of bare qw() in a for;
 17:10
nopaste "Coke" at 192.168.1.3 pasted "can't push to github." (7 lines) at nopaste.snit.ch/26073 17:11
Coke ... but I see my commit showed up? wtf?
any clues? 17:12
PerlJam Coke: Were you pushing to the codestring-deprecation-prep branch ? 17:13
Coke nope.
master.
I didn't commit anything to that branch.
PerlJam I read that as you have commits in that branch and your git is setup to push all branches (or all matching or something) by default. The commits in that branch were rejected. 17:14
Coke (can you push to just a single branch? I'm just doing "git push")
if I "git checkout codestring-deprecation-prep", "git log", I don't see any commits from me.
17:14 theory joined
PerlJam you can specify the branch to push to: git push origin master 17:14
Coke yes, but I don't want to do that. I just want to do "git push". =-) 17:15
(plus, my change to master already got pushed)
PerlJam you can also tell git that when you run "git push", it should only push the current branch (I don't remember the option for that right ofF) 17:16
oh, push.default looks like
Coke all of that is nice, but fails to address my problem.
PerlJam so you probably want to do: git config --global push.default current 17:17
Coke PerlJam: and how will this address the issue that my local branch of codestring-deprecation-prep is f'd? 17:18
I suspect deleting that branch locally is more helpful.
PerlJam you can delete it or reset it back to some known good SHA1
Coke eh. by the time I figure that out, I can just do another (*&#@$ clone. processing.... 17:19
dukeleto Coke: PerlJam++ is giving you good info
dalek tracwiki: v8 | NotFound++ | ParrotDeprecationsFor3.0 17:20
tracwiki: trac.parrot.org/parrot/wiki/ParrotD...ction=diff
dukeleto Coke: learning is *fun* ...
Coke git--
dukeleto Coke*i
dukeleto just rotated Coke by pi/2
Coke: you can use svn if you want, we won't stop you.
Coke thanks, that's helpful. ;) 17:21
dukeleto Coke: seriously. Github has svn access for parrot.git ...
Coke which, last I checked, doesn't allow for commits.
dukeleto Coke: you are wrong about that
Coke has that changed since the initial rollout?
dukeleto Coke: it is read-write
Coke but, assuming I want svn is also unhelpfuul. 17:22
thanks, though.
dukeleto Coke: yes, it has. It is read-write. But don't expect the Parrot tools/dev/* scripts to work with svn
Coke: because they don't
Coke: if you have questions about git, please ask in here or email parrot-dev. Many people probably have your question, and many people will learn by you asking a question 17:23
Coke dukeleto: christ.
dukeleto Coke: what is your issue? Why the poopy-pants?
Coke I DID that.
dukeleto Coke: and I appreciate that :) 17:27
17:32 bluescreen_ joined
dukeleto particle: please fill out the Google Code-In mentor profile, so I can assign you as the mentor of some tasks 17:34
aloha, msg particle please fill out the Google Code-In mentor profile, so I can assign you as the mentor of some tasks
aloha dukeleto: OK. I'll deliver the message.
dalek tracwiki: v24 | NotFound++ | ParrotDeprecations 17:36
tracwiki: trac.parrot.org/parrot/wiki/ParrotD...ction=diff
TT #1606 closed by NotFound++: [DEPRECATION] String.lower()
TT #1606: trac.parrot.org/parrot/ticket/1606
17:37 jsut joined 17:39 bluescreen_ left
dalek rrot: 0941496 | NotFound++ | DEPRECATED.pod:
remove find_lex note from DEPRECTATION.pod The change was done, there is doc in the wiki for 2.9 deprecations and TT #1207 is closed
 17:50
tracwiki: v4 | NotFound++ | ResolveExperimentals 18:08
tracwiki: trac.parrot.org/parrot/wiki/Resolve...ction=diff
TT #1846 closed by fperrad++: win32 compile fails on sysmem.c on mingw gcc 4.5 (w. workaround) 18:09
TT #1846: trac.parrot.org/parrot/ticket/1846
18:13 silug joined
dukeleto whiteknight: your tasks about changing function names need more detail. Like "clone parrot.git from github, create a branch, and make your changes on github, then send a pull request" 18:13
atrodo dukeleto> ping 18:25
18:37 rurban_ joined 18:39 rurban left, rurban_ is now known as rurban 18:41 bluescreen joined 18:44 elite joined
elite Hey guys 18:44
18:45 elite left
whiteknight hello elite 18:55
...fail
dukeleto: okay, I can add that
dukeleto: how do we turn tasks from private to student-visible? 18:59
dukeleto: www.google-melange.com/gci/task/sug...244420?s=0 19:02
if that is correct, I'll add it to all the tasks
it's worth mentioning that the tasks I've created have FAR more detail than most of the other tasks in that list 19:03
dalek TT #1862 created by doughera++: Configure.pl should actually test for gcc -fvisibility=hidden support 19:13
TT #1862: trac.parrot.org/parrot/ticket/1862
TT #1832 closed by doughera++: Build error
TT #1832: trac.parrot.org/parrot/ticket/1832
bluescreen good afternoon 19:19
cotto_work hi bluescreen 19:22
whiteknight hello bluescreen 19:23
bluescreen are you guys getting ready for thanks giving? 19:24
whiteknight yeah, we're trying 19:25
bluescreen I'm happy cause I will have less meetings...
NotFound bluescreen: Give me some money and I'll thank you ;)
bluescreen thats a tradeof of being offshore
NotFound: is the economic situtation as bad as eerybody says in spain? 19:26
NotFound bluescreen: mine is never as good as I want. 19:27
bluescreen well... that's true :D 19:28
whiteknight I still haven't earned a million dollars
NotFound That's the beauty of working in free projects: you don't need to care about raises. 19:29
bluescreen the worst part is that you probably spent one million dollar through your life
and still you're not driving ferraris 19:30
NotFound bluescreen: I just watch Fernando Alonso racing on them. 19:31
bluescreen good driver... have you seen him live?
NotFound No, only in TV
bluescreen are you still celebrating the world cup, still screaming "...vamos la furia..." 19:33
I'm still crying about it... I'm from Arg
NotFound I'm not a soccer fan. We are a unprotected minority in Spain
19:34 ligne left
NotFound Ugh, I'm getting used to call it 'soccer'. 19:34
bluescreen whiteknight: were you able to compile in embed_api branch... I can't finish compilation 19:35
Coke NotFound: I live in the US and I don't watch american football. I feel your lack of pain.
bluescreen Coke: can you do something else during thanks giving? 19:36
NotFound Coke: at least you have football, basket, baseball and hockey to choose from. Here they are monotematic.
bluescreen as far as I know is beer and football
whiteknight bluescreen: no, it doesn't work yet. Still problems with the config hash
I have some commits when I get home tonight that should help
bluescreen whiteknight: miniparrot's config hash?
whiteknight bluescreen: all config hashes 19:37
miniparrot doesn't actually have a config hash, it's empty. Miniparrot builds the config hash for real parrot
bluescreen right but the tool that creates parrot_config.c its broken for miniparrot
whiteknight ah yes, that's what I'm fixing today 19:38
I talked to shockwave yesterday. He wants to start using the new API next week
bluescreen excellent
NotFound Typical sport news in Spain: "Some random guy won Roland Garros, a few guys won some gold medals at something... Now that we're done with child games, go back to football"
whiteknight I think we can do it by then
bluescreen i got this: "error:imcc:No such file or directory in file 'runtime/parrot/library/parrotlib.pir' line 195" 19:39
whiteknight hmm...that's definitely a problem with the config hash 19:42
the config hash contains the library search paths
it isn't being set yet. I have to add that code tonight
bluescreen ok... you think that will fix it? 19:43
whiteknight yes, I think so 19:45
I'm expecting to be able to do complete builds by this evening
at least, complete builds of the parrot executable and maybe a few other things
bluescreen that's awesome
whiteknight some of the utility executables like pbc_merge, pbc_disassemble, etc will need to be updated 19:46
dalek TT #1193 closed by NotFound++: Setting PMC to NULL STRING
TT #1193: trac.parrot.org/parrot/ticket/1193
whiteknight NotFound++
closing tickets is always awesome
Coke CD SAN 19:47
ww 19:48
whiteknight you ever have one of those days where, if a piece of software was a person you would seriously consider murdering him? 19:56
NotFound Just send him a signal... 20:00
-9 for example.
To the person, I mean. 20:01
whiteknight :)
bacek Good morning, humans 20:09
whiteknight good morning, bacek
bacek hi whiteknight
Coke рŠ¾Š±Š¾Ń‚! 20:17
davidfetter worker? 20:18
20:19 nwellnhof joined
Coke aloha, translate? 20:20
aloha Coke: Dunno.
nwellnhof I'm getting random segfaults running Rakudo spectest on Linux/i386. 20:21
At first I thought it's my notebook overheating. But then I saw moritz also got spectest failures. 20:22
It's mostly during the S32 tests.
But very hard to reproduce, about 3 failures per spectest run. 20:23
bacek aloha, translate to english рŠ¾Š±Š¾Ń‚! 20:24
hmm... Bad bot, no cookie 20:25
aloha, translate to english from russian рŠ¾Š±Š¾Ń‚!
aloha bacek: robot!
bacek a-ha!
aloha, translate to english from russian Š”Š¾Š±Ń€Š¾Šµ утрŠ¾, Š»ŃŽŠ“Šø
aloha bacek: Good morning, the people
bacek meh...
Coke aloha, translate to russian from english christmas
aloha Coke: ƑƐĀ¾ĆĀ¶ĆĀ“ƐĀµĆ‘ƑƐĀ²ĆĀ¾
Coke hee!
bacek Looks like double utf8 encoding. 20:26
GeJ Bonjour everyone.
bacek Bonjour GeJ, ca va? 20:27
Coke aloha: translate from english to bulgarian christmas
aloha, translate from english to bulgarian christmas
GeJ bacek: I've been better. And you? 20:30
bacek GeJ, same here
whiteknight msg NotFound can we close #299? It looks like you applied the patch a while ago and there have been no further complaints 20:31
aloha OK. I'll deliver the message.
jnthn resists temptation to see how aloha handles rude Russian :) 20:34
bacek aloha, translate to english from russian трŠæŠøŠ¼Š°Š½Š“Š°Š±Š»ŃŠ“сŠŗŠøŠ¹ Š°Ń€Ń…ŠøŠæŠøŠ·Š“рŠøт 20:35
aloha bacek: [trpimandablyadskiy] [arkhipizdrit
dalek TT #194 closed by whiteknight++: Review overriddable vtable interfaces
TT #194: trac.parrot.org/parrot/ticket/194
NotFound whiteknight: I think the attributes thing needs a rethinking and redesign, but that ticket is too specific for that purpose. We can close it, yes.
bacek jnthn, nope... She can't do it :)
jnthn bacek: Ah well, at least I have a phrase to ask my Russian teacher to explain to me next lesson now... :) 20:36
whiteknight Notfound: closed 20:37
jnthn oh...I just spotted the ŠæŠøŠ·Š“ in there...yeah, the meaning gets...clearer... :P
bacek jnthn, yeah. It was from my "third language" :) 20:38
NotFound bacek: you should ask google to add it to the list of interface languages 20:43
bluescreen Although I think I know the answer, are all the active tickets in Trac valid? Is there anybody garbage collecting them? 20:49
bacek volunteers welcome! 20:51
dalek TT #299 closed by whiteknight++: Access attributes of PMC from derived pir classes
TT #299: trac.parrot.org/parrot/ticket/299
TT #300 closed by whiteknight++: Bad .get_results in a exception handler gives infinite recursion
TT #300: trac.parrot.org/parrot/ticket/300
bluescreen bacek: I would do it, but I'm affraid I would do a bigger mess
rule of thumb... well maybe a chainsaw approach... is any ticket older than X months should be closed. As either not happening any more or won't be addressed any time soon 20:53
whiteknight bluescreen: I like that idea. Some other people will not like it 20:55
some issues are valid issues, even if they are old
some things are too vague or open-ended to ever be closed 20:56
eventually, you just need to say "no" and be done
some tickets, especially with the various HLLs and all get very hard to reproduce, so they aren't likely to ever be closed
well, unless some kind of "project manager" drops the hammer 20:57
er, "product manager"
bluescreen lol...
but... for new people like me is hard to pick something from the queue to work on 20:58
whiteknight EXACTLY!
exactly
we have 13 pages of tickets. If those tickets don't already have owners, they are not likely to get adopted and worked on 20:59
bluescreen managing trac is not as easy as one would think
whiteknight personally, I don't think trac should allow the user to leave the assignee field empty, or to select a "None" component 21:00
The user should have to select a component, and every component should have a default assignee
bluescreen it totally makes sense with the teams approach
it won't hurt also run an aging report once a while 21:01
whiteknight being able to do team management would be nice too, because you could assign the ticket to the team, all the team members could see it and work on it, etc
bluescreen: you any good with sqlite3? If you want to write a query I'll install it 21:02
bluescreen is there any web interface to add queries? 21:03
whiteknight yes, admins can add queries
We could close 20 tickets tomorrow if we said that PIRC wasn't going to be used, was going to be deleted, or was going to be extracted out into it's own repo
21:03 davidfetter left
whiteknight another 15 tickets if we moved Parrot::Interpreter into it's own project 21:03
moritz then you shouldn't close the tickets, but migrate them to the bug tracker of the new project 21:04
cotto_work whiteknight: we've all but said that already. 21:05
Coke (close old tickets because they are old) hell no.
cotto_work I hate to think that kj's efforts will be wasted, but they're not the direction we're moving in.
Coke we should not force our users to triage our tickets. 21:06
(though it would certainly help)
cotto_work Coke: agreed
Coke (getting parrot team to make decisions) amen.
whiteknight We have 55 open RFC tickets (I just created a new query to show them). Some date back to 2008 21:08
anything older than 2010 at this point should probably just be deleted if no resolution has been reached
cotto_work I'll bring up pirc at #ps.
Coke To my way of thinking, if there's an RFC, those should all go to cotto. 21:09
(as the new architect guy)
cotto_work I guess I am the new decider.
+1 21:10
whiteknight I'll help triage the list down some. Some of these RFCs are pointless or unnecessary now
cotto_work whiteknight: thanks 21:11
Does anyone who happens to be here see any advantage to keeping pirc around? 21:12
whiteknight dukeleto: can you take a look at #1822? I think we can just close it 21:13
cotto_work (I'll also ask at #ps, of course)
Coke let me counter with another question: is IMCC going to be around long term?
cotto_work Coke: the long-term plan is to replace it with PIRATE.
Coke then there's no point in having YA IMCC-like thing around, is there?
why replace it with something that's getting replaced, I mean.
cotto_work no advantage that I can see, but I'm not as all-seeing as I'd like 21:14
whiteknight: don't feel like you need to assign all the RFCs to me. I accept them as my responsibility and don't want you to have a bunch of busywork. 21:15
Coke trac.parrot.org/parrot/wiki/BranchDescriptions can probably just have a link to github's branch info. 21:16
atrodo isparrotfastyet.com/ # My weekend rathole 21:18
whiteknight cotto_work: If we need to have a PIR front-end, it should be PIRC
cotto_work whiteknight: why not PIRATE? 21:19
bacek atrodo++
cotto_work, we still need minimal version of C-based PIR compiler for bootstrapping. 21:20
cotto_work bacek: yes, but only until we can generate C from Lorito. 21:21
bacek cotto_work, which is... far far away from now 21:22
cotto_work Sure. In the meantime, the question is whether it's better to keep imcc on life support or to make pirc usable. 21:23
dalek TT #1770 closed by whiteknight++: Boolean PMC without Integer
TT #1770: trac.parrot.org/parrot/ticket/1770
TT #866 closed by bacek++: DEPRECATE (after review) the *bigint*, pow* and nextkey_keyed VTABLE ...
TT #866: trac.parrot.org/parrot/ticket/866
TT #1863 created by nwellnhof++: Parrot IO and encodings 21:24
TT #1863: trac.parrot.org/parrot/ticket/1863
bacek cotto_work, at least we have "imcc hero" now...
cotto_work you mean plobsing++? 21:25
If we're picking between the two, I chose imcc mainly because it works now. 21:26
NotFound The problem with having several pir compilers is to keep them in sync. PIR is amoving target.
whiteknight cotto_work: ah, that's a good point. I keep forgetting about PIRAGE 21:27
can we rename that damn repo to parrot/pirate?
cotto_work bacek: +1 if it's possible
+1 if it's not
bacek I have no objections to rename it 21:29
whiteknight it's not a big deal, it's for my sanity 21:30
to call it one name, but have the repo named something else
heading home now. Later 21:31
21:31 whiteknight left
bacek speaking of which. So we have ticket to separate "main.c" from imcc? 21:31
s/So/Do/ 21:32
21:37 bluescreen left
bacek cotto, we do need gsoc_optimizations merged to continue work on PIRATE 21:40
afk # $dayjob 21:42
21:49 bluescreen joined 22:05 plobsing joined
plobsing .wav 22:06
no objections on codestring-deprecation-prep merge. merging. 22:08
dalek rrot: e5cbd68 | plobsing++ | / (13 files):
Merge branch 'codestring-deprecation-prep'
 22:09
plobsing aloha: msg bacek the (Un)?ManagedStruct deprecations were in anticipation of NCI changes that didn't come. I'd still like to see the API of these modified to simplify them and reduce their scope, but that's not high on my agenda ATM. Take them if you want. 22:12
aloha plobsing: OK. I'll deliver the message.
dalek rrot: 53c40e3 | plobsing++ | DEPRECATED.pod:
remove deprecation notice for completed codestring deprecation
 22:14
22:20 rblackwe left, rblackwe joined, bluescreen left
dalek lek-poller: 847cbe7 | sorear++ | push.psgi:
Save copies of GitHub announce packets
 22:21
plobsing aloha: msg chromatic the efficient interpreters paper you pasted yesterday suggests "efficient" interpreters are dominated by indirect branching and op-dispatch costs (I'd call that inefficient, but that's just me). I thought the cgoto and cpgoto cores (which is what this paper essentially advocates) were removed as they were observed to not make a difference. Are you suggesting we reevaluate that decision? 22:25
aloha plobsing: OK. I'll deliver the message.
NotFound We can reevaluate the recomendation of read papers X-) 22:26
dalek TT #1633 closed by plobsing++: CodeString PMC deprecated 22:28
TT #1633: trac.parrot.org/parrot/ticket/1633
rrot: 93f58ba | NotFound++ | config/gen/platform/win32/pid.c:
win32 version of Parrot_getpid, TT #1564
 22:32
22:41 kid51 joined 22:53 fperrad left 23:03 bluescreen joined
dalek lek-poller: 395a68f | sorear++ | modules/local/karmalog.pm:
Add ad-hoc github link shortening
 23:34
23:34 whiteknight joined 23:42 bluescreen left 23:53 bluescreen joined

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!