Parrot 2.11.0 Released | parrot.org | Log: irclog.perlgeek.de/parrot/today | Goals: write GettingStartedWithPlumage, review html documentation, test HLLs, review deprecations | no major changes after Tuesday, only critical fixes and documentation updates after Friday
Set by moderator on 12 January 2011.
00:18 nwellnhof joined 00:20 kid51 joined
dukeleto ~~ 00:23
I am seeing lots of failing tests in master. Anybody else? 00:26
smolder.parrot.org/app/projects/smoke_reports/1 has all kinds of diverse-colored smoke 00:27
Coke Aye, me too. 00:32
nwellnhof it seems that some tests should be skipped when configuring with --without-extra-nci-thunks 00:40
smolder.parrot.org/app/projects/tap...m/2922/188
dukeleto yep. that is new_parrot_plain doing it's job 00:42
kid51 linux/i386 PASS smolder.parrot.org/app/projects/rep...tails/2924 00:44
nwellnhof t/src/embed.t and t/src/atomic.t are known issues, aren't they? 00:45
dukeleto nwellnhof: embed.t is a g++ casting issue most likely 00:46
nwellnhof: atomic.t is a test that has probably failed for a long time with those build flags, but the smoker just recently notified everybody about it 00:47
dukeleto plans make the next smoke machine a sparc openbsd box in the compile farm 00:48
s/plans make/plans to make/
nopaste "kid51" at 192.168.1.3 pasted "Where 'make' throws warnings on a system where 'make test' PASS" (16 lines) at nopaste.snit.ch/27850 00:51
kid51 On my linux/i386 box, make test PASSes. Then, when I say --optimize=-O3 (simply for a change of pace), I get PASS but I also get a TODO PASS on t/src/embed.t test #3: Minimal embed, create multiple interps without giving 1st interp to Parrot_new 00:56
00:58 mtk left
mikehh kid51: see tt #1937 00:59
dukeleto: see nopaste.snit.ch/27834 01:00
01:02 mtk joined
kid51 darwin/ppc: make test PASS 01:10
01:17 dmalcolm left
kid51 dukeleto: In smolder.parrot.org/app/projects/rep...ails/2929, your 'Configure args' shows first --cc=cc and then later --cc=g++ 01:17
01:24 whiteknight joined
whiteknight good evening, #parrot 01:27
dalek rrot/packfile_write_api: 50acdfb | Whiteknight++ | / (2 files):
remove unused function imcc_run_pbc
 01:36
rrot/packfile_write_api: 31792fa | Whiteknight++ | / (2 files):
Add a new API function to serialize a packfile PMC into a format suitable for writing a .pbc file. This code stolen in part from compilers/imcc/main.c:imcc_write_pbc
rrot/packfile_write_api: 68f8286 | Whiteknight++ | t/src/embed/api.t:
stub out a test for the new API function
kid51 mikehh: I've got that t/configure/065 thing. 01:37
01:37 Yuki`N joined
dalek rrot: d750eeb | jkeenan++ | t/configure/065-git_describe_from_cache.t:
localizing $@ inside subroutine caused 'require' to fail outside the subroutine.
 01:41
kid51 Yuki`N ping 01:42
dalek TT #1967 created by whiteknight++: Deprecate IMCC compreg 01:49
TT #1967: trac.parrot.org/parrot/ticket/1967
kid51 msg dukeleto If you get a sparc/openbsd build going, can you look at trac.parrot.org/parrot/ticket/271 ? Thanks. 01:57
aloha OK. I'll deliver the message.
dalek TT #1132 closed by jkeenan++: t/steps/auto/frames-01.t: Failures following pcc_reapply merge 02:06
TT #1132: trac.parrot.org/parrot/ticket/1132
TT #395 closed by jkeenan++: optimize dynpmc makefile deps 02:22
TT #395: trac.parrot.org/parrot/ticket/395
rrot: b76fe7f | mikehh++ | t/configure/063-git_describe_to_cache.t:
correct documentation
 02:32
rrot: d808617 | nwellnhof++ | / (2 files):
[t] Fix t/src/embed.t

Make test 3 fail with optimized builds (TT #1937). Hopefully with other failures (TT #1964).
TT #1937 closed by nwellnhof++: t/src/embed.t - TODO passed: 3 in optimized builds 02:39
TT #1937: trac.parrot.org/parrot/ticket/1937
dukeleto ~~ 02:51
02:54 whiteknight left
dalek TT #735 closed by jkeenan++: path for installed header files 02:55
TT #735: trac.parrot.org/parrot/ticket/735
rrot: 2053ac1 | nwellnhof++ | include/parrot/atomic.h:
Fix TT #1945

Triggered by --without-threads		 03:00
TT #1945 closed by nwellnhof++: A parrot with all external libraries disabled fails t/src/atomic.t 03:11
TT #1945: trac.parrot.org/parrot/ticket/1945
dukeleto nwellnhof++ 03:18
03:24 Kristaba left
kid51 nwellnhof++ is on a tear! 03:25
Coke Could you git pull on your Darwin/386 and re-smoke? Thanks. 03:26
linux/i386: make fulltest PASS, configured with --optimize=-O3 03:27
03:27 contingencyplan_ left
dalek TT #1963 closed by jkeenan++: src/string/api.c: Throws warnings under two 'make' targets 03:28
TT #1963: trac.parrot.org/parrot/ticket/1963
03:56 mikehh left
dalek TT #1959 closed by nwellnhof++: Configure flag for --without-threads 04:00
TT #1959: trac.parrot.org/parrot/ticket/1959
dukeleto sets up the env for an openbsd-sparc parrot smoker 04:01
nwellnhof down to 569 tickets 04:03
dalek Heuristic branch merge: pushed 31 commits to parrot/nwellnhof/platform_src by nwellnhof 04:26
04:28 mikehh joined
kid51 darwin/ppc make fulltest PASS 04:31
04:31 kid51 left
dalek Heuristic branch merge: pushed 64 commits to parrot/nwellnhof/unicode_filenames by nwellnhof 04:43
plobsing is anyone else getting failures in rakudo t/spec/S05-mass/properties-derived.rakudo ? 04:47
nwellnhof plobsing: iirc that one can happen with certain libicu versions 04:56
are the failures something with 'ideographic'? 04:57
dukeleto starts building parrot on openbsd-sparc 04:58
nwellnhof dukeleto: it would be awesome if you could test nwellnhof/platform_src on openbsd 04:59
dukeleto nwellnhof: ok, i am seeing if master compiles now :)
05:13 Yuki`N left
plobsing nwellnhof: my failures are in 3 tests described as 'match unrelated ((Lu + Ll + Lt + Lm + Lo + OtherAlphabetic) 05:14
not sure if that's the same thing
nwellnhof i think it's different icu versions disagreeing about unicode character classes 05:15
plobsing wow. just read dan's blog article 'WCB: Loadable opcode libraries'. uncannily predicts dynop-mapping 5 years before it happened. 05:18
dukeleto gmake: *** [compilers/opsc/gen/Ops/Compiler/Actions.pir] Segmentation fault (core dumped) 05:24
that is what happens on my openbsd box
blarg 05:26
This GDB was configured as "sparc64-unknown-openbsd4.6"..."/home/leto/git/parrot/parrot-nqp.core": not in executable format: File format not recognized
Sometimes, life is hard.
Parrot seemingly needs too much memory to compile on this machine, or my ulimits are too low. 05:29
05:49 gbacon joined 05:50 rurban left 05:51 rurban joined 05:56 nwellnhof_ joined 05:59 nwellnhof left, nwellnhof_ is now known as nwellnhof 06:13 nwellnhof left 06:25 fbrito joined
dukeleto is close to submitting a smoke report from a sparc machine, if smolder doesn't barf 06:27
** POST smolder.parrot.org/app/projects/pro...d_report/1 ==> 502 Bad Gateway (1s) 06:46
Well, "make test" passes on debian sparc64
06:58 fbrito left
cotto ~~ 07:00
dukeleto cotto: wazzup 07:02
cotto: smolder is down
cotto: in happier news, "make test" passes on linux sparc64 07:03
cotto dukeleto, great. Is it on the list of known-working platforms? 07:06
In other news, I get to be the guy who advocates git at $dayjob. 07:09
moritz plobsing: 22min... I thought the difference was larger 07:12
07:24 AzureStone left 07:27 AzureStone joined
dukeleto cotto: I was that guy, twice now. 07:28
cotto: three times if you count the Parrot conversion :)
cotto and I do
actually, advocacy came up from a healthy chunk of developers 07:29
dukeleto cotto: I am going to go ahead and say that the format of the PLATFORMS file is insane
cotto: i would like to say that linux sparc64 works in that file, but the format makes me want to cry. I have no clue how to add this simple fact to the file 07:30
cotto: PLATFORM still talks about JIT
moritz btw rakudo intermittently segfaults on parrot master 07:31
sometimes even on startup
dukeleto moritz: :( 07:32
moritz: when did this start?
cotto moritz, eek. spectest_regression looked fine yesterday
dukeleto cotto: ok, i am done complaining. I just added it to PLATFORMS 07:35
dalek rrot: c6caa41 | dukeleto++ | PLATFORMS:
JIT does not exist right now
rrot: 20adce6 | dukeleto++ | PLATFORMS:
[doc] Add update about linux sparc64
dukeleto cotto: i think PLATFORMS needs to go in the same direction as "deprecations as data"
cotto dukeleto, I like that.
It's not especially sane as-is 07:36
too much brain-based parsing required
dukeleto the sad thing is that PLATFORMS is not easy to read, as a human or machine. And that sucks.
cotto all the drawbacks of a machine readable format with none of the benefits 07:37
dukeleto cotto: i've already notified OSUOSL that smolder is giving a bad gateway error
whenever smolder comes back up, we will have a second smoker 07:38
07:38 theory left
cotto dukeleto, how will that work wrt where stuff gets submitted? 07:38
dukeleto cotto: i don't understand your question
cotto What's the point of having two smolder servers? Won't reports only get sent to one? 07:39
dukeleto cotto: two smokers 07:40
cotto I guess we can make the smolder target submit twice
ah
dukeleto cotto: two machines submitting smoke reports
cotto that's great then. dukeleto++
dukeleto cotto: one debian x86_64 and one debian sparc64
cotto: i setup stuff on an openbsd sparc machine, but it coredumped when parrot-nqp was compiling something 07:41
cotto x64 -> meh. sparc64 -> nice.
but more machines is generally great
dukeleto, sounds not entirely unlike GeJ's machine 07:42
dukeleto cotto: each smoker will be compiling parrot 16 ways with both gcc and g++
07:43 jsut_ joined
dukeleto cotto: each "flavor" gives Configure.pl various flags, such as --optimize or --without-icu 07:43
cotto: and i have a few combinations of flags
cotto That's a lot of smoking. Nice job putting those machines to work.
dukeleto cotto: thanks. the tools that I am building this on top of will be useful for many projects
cotto: and i hope that having proper smokers will allow us to make bigger changes with certainty that we aren't breaking lots of shit 07:45
cotto: HLL testing is the next big step
cotto dukeleto, very much so. I'm really looking forward to the day when it's routine to find out that we broke Rakudo before the Rakudo guys (mainly moritz++ ) do. 07:46
ideally even to find and fix bugs before they know they're there
dukeleto cotto: indeed 07:47
07:47 jsut left
cotto moritz, are the segfaults you're seeing with --optimize? 07:48
moritz cotto: yes 07:52
cotto sees if he can repro 07:54
though I definitely won't be up by the time it finishes 07:55
dalek TT #791 closed by cotto++: Crash on PBC execution under Windows Server 2003 08:05
TT #791: trac.parrot.org/parrot/ticket/791
cotto backs away slowly
dalek rrot/pbc-backref: a8215c9 | jimmy++ | src/packfile/api.c:
consting
 08:09
cotto moritz, should spectest_regression catch the segfaults? I'm not seeing any. 08:32
Dang. I'm seeing some spectest_regression failures. 08:38
08:44 fperrad joined, mtk left 08:51 Patterner left 08:52 Psyche^ joined, Psyche^ is now known as Patterner, mtk joined 09:00 chromatic left 09:25 mj41 left 09:44 cogno joined 10:24 cogno left, cogno joined 10:33 gbacon left 10:40 cogno_ joined 10:41 cogno left 11:07 mj41 joined 11:08 cogno_ left
dukeleto This looks interesting: github.com/jeffreykegler/Marpa 11:15
11:20 jsut joined 11:24 contingencyplan_ joined 11:25 jsut_ left 11:44 Kulag left 11:47 Kulag joined 11:58 Kulag left, Kulag joined 12:04 rurban left 12:06 rurban joined
bacek rakudo: ("a", ("^" if Bool::False), "b").join("").say; 12:11
p6eval rakudo d5b9d9: OUTPUTĀ«abā¤Ā»
bacek rakudo: pir::load_bytecode("nqp-setting.pbc"); ("a", ("^" if Bool::False), "b").join("").say;
p6eval rakudo d5b9d9: OUTPUTĀ«aNil()bā¤Ā»
bacek yay...
seen moritz 12:12
aloha moritz was last seen in #parrot 4 hours 19 mins ago saying "cotto: yes".
bacek msg moritz 'pir::load_bytecode("nqp-setting.pbc"); ("a", ("^" if Bool::False), "b").join("").say;' is main reason for failures in nqp_pct branch. At lease one of them. 12:13
aloha OK. I'll deliver the message.
bacek msg moritz try same code without loading nqp-setting
aloha OK. I'll deliver the message.
moritz re 12:19
bacek moritz, aloha
moritz aloha: greet bacek :-) 12:20
bacek moritz, I'll try to find out how nqp-setting can affect rakudo. But it can take some time.
dalek TT #1968 created by bacek++: Loading of nqp-setting brakes rakudo. 12:26
TT #1968: trac.parrot.org/parrot/ticket/1968
Coke msg dukeleto: I am also running a smoker on darwin x86 12:30
aloha OK. I'll deliver the message.
Coke embed.t is better, but still failigng: 12:37
Failed tests: 30, 37
bacek moritz, erm... Main culpit is RPA.join from setting. I have no idea why and how to fix it. 12:42
12:48 cogno joined
bacek moritz, looks like rakudo's dispatcher bug. At least I couldn't find what's wrong on parrot side. 12:50
tadzik hello Parrots 12:52
moritz hello zebra
bacek: sure, could be.
12:53 whiteknight joined
bacek seen jnthn 12:54
aloha jnthn was last seen in #perl6 11 hours 52 mins ago saying "night o/".
bacek msg jnthn Can you take a look at 'pir::load_bytecode("nqp-setting.pbc"); ("a", ("^" if Bool::False), "b").join("").say;' Looks like a bug in rakudo's dispatcher. It calls RPA.join from nqp-setting instead of Any.join from rakudo. 12:55
aloha OK. I'll deliver the message.
12:55 cogno left
bacek plobsing++ # pbc-backref is awesome. 12:56
and time to recharge for some useless robot. 12:57
night
msg cotto Can you please put broad pir deprecation notice into DEPRECATED? Something generic like "current PIR syntax is deprecated. We will tell you about new syntax in near future" 12:59
aloha OK. I'll deliver the message.
13:00 cogno joined
whiteknight I don't think NQP uses an HLL namespace, so when you load it, it loads into the current HLL namespace, causing conflicts 13:09
13:09 [hudnix] left
whiteknight I'm having that same issue in kakapo 13:09
13:14 hudnix joined
moritz bacek: deprecation notices without running alternatives aren't useful for the user 13:26
13:27 cogno left
moritz and since many HLLs have quite large bodies of PIR code, just saying that there will be an alternative in future doesn't help at all 13:29
for example rakudo has 9.5k lines of PIR hand written PIR code 13:30
wtf are supposed to do without that, if PIR is deprecated and we have no alternatives yet?
13:45 kid51 joined 13:48 rurban_ joined 13:50 rurban left, rurban_ is now known as rurban
Coke agreed. 13:54
"PIR is going to change" gets a big fat -1 from me.
msg bacek "PIR is going to change" gets a big fat -1 from me. (see backscroll about this time.)
moritz++
aloha OK. I'll deliver the message.
Coke still getting failures on t/src/embed.t 13:55
14:01 Eclesia joined
Eclesia hi 14:01
whiteknight hello Eclesia
I'm getting a t/pmc/class.t failure on master 14:03
dalek rrot: 177577c | Whiteknight++ | t/pmc/class.t:
it doesn't make sense to test against any specific class ID number. Instead, test that inspect('id') returns the same value as VTABLE_get_integer
 14:11
14:12 ambs joined
kid51 has filed a ticket with osuosl re "Error 503 Service Unavailable" after hitting "Submit changes" in Trac tickets 14:19
whiteknight I was having some troubles with trac this morning too 14:23
kid51 has also filed ticket about being unable to submit Smolder report 14:27
Smolder has been down (not receiving new submissions) for over 13 hours 14:33
dalek rrot/pbc-backref: 56e94eb | jkeenan++ | t/configure/065-git_describe_from_cache.t:
Correct test file as it has been corrected in master.
 14:58
15:06 jsut_ joined 15:07 kid51 left 15:11 jsut left
Coke I swear I just submitted one. 15:16
dalek rrot/imcc_cleanups: 31792fa | Whiteknight++ | / (2 files):
Add a new API function to serialize a packfile PMC into a format suitable for writing a .pbc file. This code stolen in part from compilers/imcc/main.c:imcc_write_pbc
rrot/imcc_cleanups: 5e6374d | mikehh++ | docs/dev/profiling.pod:
fix pod to conform with make html

uc =head1 Name to =head1 NAME (this seems to be required) also Description to DESCRIPTION (only other =head1 in the pod) did not change copyright as this probably needs to be updated and
   do not want to show that this is up-to-date
rrot/imcc_cleanups: 68f8286 | Whiteknight++ | t/src/embed/api.t:
stub out a test for the new API function
rrot/imcc_cleanups: d750eeb | jkeenan++ | t/configure/065-git_describe_from_cache.t:
localizing $@ inside subroutine caused 'require' to fail outside the subroutine.
rrot/imcc_cleanups: b76fe7f | mikehh++ | t/configure/063-git_describe_to_cache.t:
correct documentation
Coke scrolled out of buffer, 'll check th enext one.
rrot/imcc_cleanups: d808617 | nwellnhof++ | / (2 files):
[t] Fix t/src/embed.t

Make test 3 fail with optimized builds (TT #1937). Hopefully with other failures (TT #1964).
rrot/imcc_cleanups: 2053ac1 | nwellnhof++ | include/parrot/atomic.h:
Fix TT #1945

Triggered by --without-threads
rrot/imcc_cleanups: c6caa41 | dukeleto++ | PLATFORMS:
JIT does not exist right now
rrot/imcc_cleanups: 20adce6 | dukeleto++ | PLATFORMS:
[doc] Add update about linux sparc64
rrot/imcc_cleanups: 8a7dab5 | Whiteknight++ | / (36 files):
Merge branch 'master' into imcc_cleanups
rrot/imcc_cleanups: 177577c | Whiteknight++ | t/pmc/class.t:
it doesn't make sense to test against any specific class ID number. Instead, test that inspect('id') returns the same value as VTABLE_get_integer
rrot/imcc_cleanups: b5e412d | Whiteknight++ | t/pmc/class.t:
Merge branch 'master' into imcc_cleanups
rrot/imcc_cleanups: 29c4a7f | Whiteknight++ | / (8 files):
re-remove include/parrot/imcc.h
Coke nope, bad gateway here too. 15:22
Coke leaves the smoker running. 15:23
msg dukeleto it'd be nice if we could save a failed smoke and then mass upload them later.
aloha OK. I'll deliver the message.
15:26 nwellnhof joined
dalek Heuristic branch merge: pushed 43 commits to parrot/exception_backtraces by Whiteknight 15:30
jnthn ohhai 15:40
tadzik o/
jnthn msg bacek the reason is almost certainly that something in Rakudo multiply inherits from RPA and Any, and what we're seeing is just the search finding what it normally would in an MI search. So not a bug as such. 15:41
aloha OK. I'll deliver the message.
jnthn msg bacek after the 6model switch, all the stuff that currently inherits from Parrot PMCs will stop doing so, because I'm not going to implement inherting from PMCs in 6model. 15:42
aloha OK. I'll deliver the message.
jnthn msg bacek In the meantime...tricky. I'll see if I can think of a workaround.
aloha OK. I'll deliver the message.
dalek Heuristic branch merge: pushed 258 commits to parrot/gci_orderedhash_coverage by Whiteknight 15:52
whiteknight msg cotto I created an RFC ticket to deprecate current PIR compreg (). Can you look at it? Also I would like to deprecate Eval PMC but trac is being obnoxious. thoughts? 16:02
aloha OK. I'll deliver the message.
whiteknight msg cotto #1967 and #1969 16:03
aloha OK. I'll deliver the message.
Coke whiteknight: isn't there already a defined API for compreg compilers? 16:04
(which PIR doesn't use)
whiteknight not really
Coke er, IMCC
16:05 Eclesia left
whiteknight we were working on the design 16:05
jnthn whiteknight: Why deprecate those?
whiteknight jnthn: I want to make a better PIR compreg PMC
Coke ok. the ticket makes it sound like you're doing a completely new one. which means the one that's in progress and already used by stuff in compilers/ is getting tossed too.
jnthn DO NOT WANT
whiteknight why not?
jnthn At the moment HLL::Compiler objects can be stroed with compreg
whiteknight right 16:06
jnthn And thus we have a simple mechanism for storing something that implements an OO API
whiteknight I dont want to change any others, just the PIR one
jnthn Oh
Ah, sorry
whiteknight right now it's just an NCI
jnthn I thought you meant deprecate compreg itself
16:06 mberends joined
whiteknight oh hell no 16:07
jnthn OK then no objections. :)
whiteknight compreg stays
jnthn OK
whiteknight :)
dalek TT #1969 created by whiteknight++: Deprecate Eval PMC 16:14
TT #1969: trac.parrot.org/parrot/ticket/1969
16:21 nwellnhof left
whiteknight t/src/embed.t fails tests 30 and 37 when built with g++ 16:30
16:32 mtk left 16:39 mtk joined 16:41 JimmyZ joined 16:55 Coke left 16:58 Coke joined 17:08 mikehh_ joined, mikehh left, mikehh_ is now known as mikehh
mikehh opbots, names 17:08
dalek rrot: 81c993c | nwellnhof++ | t/src/embed.t:
[t] Fix remaining failures in t/src/embed.t with g++

Fixes TT #1964		 17:45
17:47 ambs_ joined 17:50 ambs left, ambs_ is now known as ambs
dalek rrot: a1299f9 | nwellnhof++ | NEWS:
Add a NEWS item
 17:51
17:52 nwellnhof joined 17:53 theory joined
dalek TT #1964 closed by nwellnhof++: t/src/embed.t failed 34 tests of 49. 17:53
TT #1964: trac.parrot.org/parrot/ticket/1964
18:03 JimmyZ left 18:12 contingencyplan_ left 18:13 contingencyplan joined 18:16 Coke left
dalek rrot/pbc-backref: 81c993c | nwellnhof++ | t/src/embed.t:
[t] Fix remaining failures in t/src/embed.t with g++

Fixes TT #1964		 18:18
rrot/pbc-backref: a1299f9 | nwellnhof++ | NEWS:
Add a NEWS item
rrot/pbc-backref: 622c3d1 | plobsing++ | src/packfile/object_serialization.c:
cast for c++
rrot/pbc-backref: 598107b | plobsing++ | / (8 files):
Merge branch 'master' into pbc-backref
18:33 ambs_ joined, ambs left, ambs_ is now known as ambs 19:06 chromatic joined
kthakore hi chromatic 19:11
dalek tracwiki: v7 | plobsing++ | PlobsingTaskList 19:48
tracwiki: pbc-backref is done
tracwiki: trac.parrot.org/parrot/wiki/Plobsin...ction=diff
tracwiki: v11 | plobsing++ | ParrotDeprecationsFor3.0
tracwiki: implicit :main was not removed
tracwiki: trac.parrot.org/parrot/wiki/ParrotD...ction=diff
rrot: e7cbbb1 | plobsing++ | NEWS:
unicode improvements are to more than just file IO
 19:50
dukeleto ~~ 20:00
20:02 Kristaba joined
dukeleto Coke: you idea about saving failed smoke reports is feasible, are you volunteering for that? ;) 20:02
dalek tracwiki: v29 | plobsing++ | ParrotDeprecations 20:04
tracwiki: implicit :main was not removed
tracwiki: trac.parrot.org/parrot/wiki/ParrotD...ction=diff
Heuristic branch merge: pushed 34 commits to parrot/leto/embed_grant by leto 20:18
plobsing coverage? 20:22
aloha: cover?
aloha plobsing: Sorry, I don't know.
plobsing aloha: coverage?
aloha plobsing: coverage is cv.perl6.cz or tapir2.ro.vutbr.cz/cover/cover-results/
dukeleto trac seems to be really swimming in molasses today 20:34
dalek TT #673 closed by dukeleto++: Parrot 1.1.0 build on OS X with Fink ICU package 20:37
TT #673: trac.parrot.org/parrot/ticket/673
bacek ~~ 21:05
seen jnthn 21:06
aloha jnthn was last seen in #perl6 15 mins 18 seconds ago saying "bye bye o/".
bacek sigh
nopaste "bacek" at 192.168.1.3 pasted "Workaround for RPA.join" (7 lines) at nopaste.snit.ch/27901
jnthn bacek: o/ 21:07
bacek jnthn, aloha
see nopaste
It's same workaround as for .sort
jnthn We already have a workaround for .sort like that in Rakudo? 21:08
bacek jnthn, yes :)
jnthn bacek: ah, OK...
Well, guess they go away eventually :)
bacek jnthn, I hope so :) 21:09
jnthn Just put it in the same place as the other one.
bacek jnthn, I'm not Rakudo committer :)
jnthn bacek: Ah.
bacek src/core/Parcel.pm
21:10 jsut joined
bacek jnthn, what is ETA for 6model? 21:11
cotto ~~ 21:12
bacek cotto, hi
cotto hi bacek
jnthn bacek: Applying 21:13
bacek: In Rakudo? Not sure.
bacek jnthn, ok, thanks.
jnthn bacek: I expect to make the Rakudo branch to start working on it either late this month or early next month.
bacek: Need to finish up getting it into nqp-rx first. :)
But progress on that is going quite well.
bacek jnthn, so, in about 3 month?
cotto What kind of deprecations were you thinking of for PIR? Deprecating features without specifying them is not helpful to the users of PIR in its current state.
bacek cotto, bare function calls at least. 21:14
jnthn bacek: I'd like to have Rakudo master be on 6model by then.
bacek cotto, plobsing did want to change labels syntax.
cotto bacek, sure. Those can go. It sounded like you had some pretty significant changes in mind though.
bacek jnthn, excellent!
21:14 jsut_ left
bacek cotto, not at all. Just small crappy bits. To avoid ambiguity in syntax. 21:15
cotto bacek, that's fine, but we still need to list them. I do agree that PIR could use a cleanup. 21:16
bacek cotto, actually, I think we can wait till 3.3. I will finish PIRATE (hopefully) and we will have cleaned version of PIR syntax. 21:17
cotto bacek, that sounds good. 21:18
21:21 cogno joined 21:36 fperrad left 21:39 cogno left 21:44 nwellnhof_ joined 21:47 rurban_ joined, nwellnhof left, nwellnhof_ is now known as nwellnhof 21:49 rurban left, rurban_ is now known as rurban
dalek p-rx/nom: 99b6773 | jonathan++ | / (4 files):
A bit more stubbing in of multi-dispatch; this just always hands back the first candidate in the list to invoke for testing purposes. More interesting is that the invocation logic, passing along the same argument set, is in place now.
 21:58
22:06 PerlJam left 22:09 tadzik left, pmichaud left, Util left 22:16 bacek left 22:26 kid51 joined 22:34 ambs left 22:50 PerlJam joined 23:01 nwellnhof left
dalek p-rx/nom: ef0b76b | jonathan++ | src/ops/nqp.ops:
Fix a copy-paste fail.
 23:37
p-rx/nom: d0c72d1 | jonathan++ | src/ (2 files):
Get multiple dispatch working up to the point that it can do arity-based multiple dispatch. A lot of the stuff for type-based is stubbed in too, as is some incomplete error reporting. Still needs hardening, caching, etc.
23:57 bacek 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!