Parrot 0.8.1 "Tio Richie" Released | parrot.org | 24 TT | 648 RT
Set by moderator on 10 December 2008.
00:14 gmansi joined 00:40 TiMBuS joined 01:05 Whiteknight joined 01:15 tetragon joined
dalek r33906 | pmichaud++ | assoc: 01:42
: [rakudo]: Update postcircumfix, remove unused get_pmc_keyed from Protoobject.
diff: www.parrotvm.org/svn/parrot/revision?rev=33906
01:45 MariachiElf joined
Andy We need a Compiler Squad. 02:00
leonardo-m.livejournal.com/73732.html
Just to exercise everything.
I AM GONNA DO SO MUCH PARROT SHIT WHEN THIS BOOK IS DONE.
I miss writing tests and fixing compiler warnings. 02:01
An honest man's work.
dalek r33907 | pmichaud++ | trunk: 02:04
: [rakudo]: Merge associative branch into trunk.
: Rakudo now supports slicing on hashes.
diff: www.parrotvm.org/svn/parrot/revision?rev=33907
r33908 | pmichaud++ | assoc:
: Delete obsolete branch (now merged to trunk).
diff: www.parrotvm.org/svn/parrot/revision?rev=33908
Whiteknight Andy, what book are you working on? 02:06
Andy oreilly.com/catalog/9781934356265/ 02:07
02:10 ChrisDavaz joined
Tene So, how would I catch just warnings, but not errors, in Perl 6? 02:11
Tene checks S04, maybe
HTTP is pretty broken for me right now.
Whiteknight how is warn() implemented? exception? 02:12
Tene It will be.
Right now it just prints to stderr
That's what I'm going to work on next, but I want to know what it looks like in Perl 6 before I implement in Parrot.
Whiteknight once it's exception-based, it should be easy enough to trap it. 02:13
Tene Right, I know how to do it on the Parrot level just fine.
pmichaud CONTROL is used to trap warnings, I think. 02:17
Tene Also, what exception type should Perl 6 warnings be?
Parrot exception type?
pmichaud might to make one up, if one doesn't already exist.
Tene If it's CONTROL, is it CONTROL_WARN? 02:18
02:18 jimmy joined
pmichaud it'd be okay with me to create a CONTROL_WARN 02:19
Tene Or is it just: no 'warnings'; or something in a block? 02:20
pmichaud to catch the warning, it'd be CONTROL { when <something> { ... } }
where <something> is the way we distinguish warning exceptions from other types 02:21
I don't know if that's specced. 02:22
Tene Hmm. 02:23
pmichaud (the <something> part, that is.)
but yes, it'd be in S04.
or perhaps in the test suite.
Tene So, if Parrot automatically resumes on WARNING and NORMAL exceptions, the default exception severity should probably be ERROR, not NORMAL like it is now. 02:26
jimmy seen bernhad 03:08
purl I haven't seen 'bernhad', jimmy
jimmy seen berney 03:09
purl I haven't seen 'berney', jimmy
cotto you mean barney?
jimmy seen barney
purl barney was last seen on #parrot 15 days, 12 hours, 8 minutes and 25 seconds ago, saying: will be back tomorrow evening, if there is internet access in Berlin [Nov 29 15:00:47 2008]
jimmy yes
he was online,but not here.
cotto I'm certain he was online more recently that 15 days ago. 03:20
s/that/than/
jimmy who updated pipp with bernhard account? 03:37
cotto: languages\\pipp\\t\\php\\string.t has double copyright and $id$, i don't think it is need to create a patch . 03:42
s/is//g 03:43
jimmy is having lunch 03:44
cotto thanks, jimmy 04:01
04:02 elmex_ joined
dalek r33909 | cotto++ | trunk: 04:04
: [pipp] remove redundant unnecessary copyright and svn id lines
: jimmy++ for noticing
diff: www.parrotvm.org/svn/parrot/revision?rev=33909
04:51 gaurav joined 04:53 gaurav joined
dalek r33910 | cotto++ | trunk: 05:10
: [embed] fix a memory leak (CID #181)
diff: www.parrotvm.org/svn/parrot/revision?rev=33910
05:10 gaurav joined
Tene Okay, I have auto-resume working. 05:12
Let's see if tests pass.
05:39 vaidyagi joined 05:40 vaidyagi joined, petdance joined
dalek r33911 | pmichaud++ | trunk: 05:47
: Forgot to commit MANIFEST earlier.
diff: www.parrotvm.org/svn/parrot/revision?rev=33911
05:51 tewk joined 05:53 adam_ joined 05:56 bacek joined
Tene So... count_eh is broken and untested 06:10
I'm done for the night, but someone needs to please file a ticket about this. 06:11
dalek r33912 | tene++ | trunk:
: [exceptions]
: * Nonfatal exceptions are automatically resumed, after printing the message
: * Default exception severity is EXCEPT_ERROR
diff: www.parrotvm.org/svn/parrot/revision?rev=33912
r33913 | tene++ | trunk: 06:14
: [rakudo]: Update warn() to use an exception.
diff: www.parrotvm.org/svn/parrot/revision?rev=33913
Tene Should the default handler there for nonfatals also print a backtrace?
nopaste "tene" at 166.70.38.237 pasted "trivial count_eh demo; segfault" (3 lines) at nopaste.snit.ch/14951 06:21
Tene someone please make ticket of that, if there isn't one already
dalek r33914 | tene++ | trunk:
: [NEWS]: updates
diff: www.parrotvm.org/svn/parrot/revision?rev=33914
r33915 | pmichaud++ | trunk: 06:28
: [rakudo]: spectest-progress.csv update: 249 files, 5101 passing, 0 failing
diff: www.parrotvm.org/svn/parrot/revision?rev=33915
Tene 100 new passing tests today? 06:30
Nice.
pmichaud yeah. Hard to know who to credit, so I credit everyone on this one. Moritz++ did a lot of test (un)fudging that probably helped. 06:31
we may see a similar gain in the next day or so, as I get the hash methods corrected. 06:32
06:32 gaurav joined
Tene Would it be a problem if i worked on loops refactor on my own tomorrow? Or too close to release? 06:34
pmichaud do it in a branch.
Tene Clever.
pmichaud never hurts to work in a branch. :-)
06:38 gaurav joined
Alias OK, building a Vanilla 6 Alpha I think is maybe good enough to let you guys play with, prior to real release tomorrow 06:50
06:51 Theory joined
Tene Alias: I hope to get control exceptions working tomorrow. It would be great to get that into whatever release you make public. 06:51
pmichaud if I get hashes cleaned up I can help on the exceptions. 06:53
Tene pmichaud: thoughts on .HLL_map Exception to Perl6Exception, so I can add methods like .resume() and such, to start exploring things for working out the spec? 06:55
Wait, I think we're already doing that.
Nevermind.
pmichaud yes, we are.
Tene In that case, adding some unspecced potentially-useful methods to Perl6Exception?
pmichaud that's fine.
watch the mailing list and #perl6 closely, though, because control exception handling (especially warn and return) is a current "hot topic" 06:56
Tene Good.
06:59 Theory joined
Alias Tene: I only build official tarballs, no exceptions 07:00
So if something is in the tarball release, it goes in
Attempting to patch of other modifying anything is just asking for maintenance nightmare
Tene Alias: I was confused by "prior to real release" 07:01
Alias Oh
I'm building Vanilla Perl 6 0.8.1 Alpha 2
Tene "Tomorrow" for me meant "Monday", not "Tuesday" when the release happens.
pmichaud Alias comes from the future. :-)
Alias Here in the future, it's already after business hours Monday evening
Tene Yes, I understand now, no problems.
If I can think of a good example for the autoresume and warn-uses-exception changes i just made, I'll blog about it. 07:02
pmichaud blog it anyway, if only to say "we now follow the spec"
Alias One thing I'd very very much like is to be able to run make test 07:03
I know chromatic calls then "stable" releases, but having to force tests really does say otherwise
pmichaud on the spectests?
Alias No, at all
make test fails on Win32 for 0.8.1
pmichaud Parrot make test fails?
Alias yes
pmichaud ah.
I don't know that we can do anything about it for the Dec release, but we can make that a critical milestone for Jan release (which I think it should be) 07:04
Alias The entire point of the test suite and skips is to decide which tests are so critical you would rather end users NOT use the package at all
If test fails on Win32, it says you'd rather people not use it on Win32 07:05
Alias ponders websites and what not 07:07
I'd very much like to do something similar to strawberryperl.com 07:08
Helps people to think of things as "real"
BTw, have you SEEN perl6.org, it's utterly horrid
07:08 davidfetter joined
pmichaud I don't know who owns that domain. 07:09
I have rakudoperl.com/.org/.net if we want to do something with those, though.
Alias perl6.com/
Even worse
Tene ... ew. 07:10
Alias No insult intended, but I'm trying to avoid using "Rakudo" anywhere
pmichaud none taken.
Tene perl6.com -- Jason Duke
Alias Although I shall of course say it's using Rakudo 07:11
Tene perl6.org -- Daniel Wright
Alias Both sites look like they mean well
No actual squatting
Tene blogs.gurulabs.com/stephen/2008/12/...tions.html 07:15
Alias pmichaud: So could you do platform-specific skips for the failing Win32 tests you expect to fail, if indeed they aren't important 07:17
pmichaud Alias: at the moment I have no way of knowing which tests those are 07:18
Alias I can give you a list :)
pmichaud also, I don't muck around too much in Parrot's test suite -- I tend to focus on Rakudo and PGE
but yes, istr there's a way to do platform-specific skips.
Tene Alias: if you could post a list to parrot-dev@lists.parrot.org that would be great 07:19
pmichaud or even as a ticket (trac.parrot.org)
Tene sleep, dream of exceptions 07:20
Alias OK, off to read my book while build finishes, back in an hour or so 07:29
07:32 UltraDM joined 07:37 adu joined 07:46 iblechbot joined
Alias Build seems to have worked 07:48
Alias looks for somewhere to upload to 07:49
OK, uploaded 08:02
08:03 ffwonko joined
Alias strawberryperl.com/download/vanilla...lpha-2.exe 08:03
In about 15 minutes, the zip version should also be available 08:04
strawberryperl.com/download/vanilla...lpha-2.zip
cotto but does it run on Linux?
(sorry)
Alias That's ok
moritz Alias++ # packaging effort
Alias Next time, s/Linux/Gentoo/ (funnier)
cotto strawberryperl.com/download/vanilla...lpha-2.rpm 08:05
Alias Wouldn't be too hard
People keep asking me about a Mac version
I keep asking for a 17 inch powermac 08:06
or mac book pro, or something shiny
It's not really that hard, it's mostly just work...
Anyways
zip lands in 10 minutes ish
Let me know how it goes, I'll be afk for a few hours or so readiung 08:07
08:12 adu joined 08:15 Hadi joined 08:30 alvar joined 08:33 ffwonko joined 08:45 Theory joined 08:56 ffwonko joined 08:59 ffwonko joined 09:06 alvar joined 09:32 allison joined 09:37 cotto joined 09:38 kj joined 09:51 Hadi left 09:52 donaldh joined 10:31 tomyan joined
jimmy vanilla-perl6-0.8.1-alpha-2.exe? ah, perl 6 alpha came out? 10:31
donaldh jimmy: more likely to be an alpha package of 0.8.1 for your platform 10:48
jimmy donaldh: thanks. 10:50
11:03 Theory joined 11:28 gaz joined
Alias Anyone got a chance to look at the Vanilla 6 alpha? 11:36
11:44 Lorn joined 12:10 gmansi joined 13:31 ffwonko joined 13:45 tetragon joined 14:02 iblechbot joined 14:09 gryphon joined 14:16 Wknight8111 joined 14:17 ruoso joined 14:28 PacoLinux joined 14:30 ffwonko joined
pmichaud Alias: after installing vanilla perl6, I get: 14:37
(eek, can't copy/paste in windows)
kj pmichaud: can't copy/paste in command shell? 14:38
pmichaud having to learn how to do it :-) 14:39
nopaste "pmichaud" at 72.181.176.220 pasted "results of "perl6 -e 'say "hello world"' in vanilla perl6" (4 lines) at nopaste.snit.ch/14954
kj open properties window of the shell window
pmichaud I got it :-)
kj then in 'options', klik 'quickedit mode'
ok
14:39 ff-wonko joined 14:45 AndyA joined
dalek r33916 | pmichaud++ | trunk: 15:07
: [rakudo]: Refactor Array assignment.
diff: www.parrotvm.org/svn/parrot/revision?rev=33916
15:12 ruoso joined 15:24 jhorwitz joined 15:35 ruoso joined
dalek r33917 | pmichaud++ | trunk: 15:53
: [rakudo]: Refactor Scalar coercions.
diff: www.parrotvm.org/svn/parrot/revision?rev=33917
16:07 gaurav joined 16:08 gaurav joined
dalek r33918 | Whiteknight++ | trunk: 16:11
: [Book] Add more information about some of the pmclass flags "abstract", "has_ro", "is_ro", and "is_shared". Many of these are almost completely undocumented elsewhere, so my descriptions here might not be completely accurate.
diff: www.parrotvm.org/svn/parrot/revision?rev=33918
16:12 DietCoke joined
Coke now has power at home, work, and everywhere in between. 16:12
pmichaud yay 16:15
wb, Coke
16:24 hercynium joined 16:27 gmansi joined
Coke sees tcl is not on trac.parrot.org/parrot/wiki/Languages 16:28
Coke suggests we kill languages/LANGUAGES.STATUS and just use the wiki. 16:29
PerlJam well, it's a wiki ...
Coke PerlJam: someone went to great trouble to copy an existing set of documentation and leave a note to keep them in sync.
PerlJam: and tcl was already in the other file.
Coke is fixing it.
PerlJam weird
particle Alias: your claim of windows tests failing is baseless without a bug report 16:34
Coke hates wiki table editing.
gaurav PerlJam, Coke: oh, hey, that was me.
PerlJam gaurav: Do you have some aversion to tcl? Did Coke tick you off? :-) 16:35
gaurav heh, i *like* wiki table editing, which is partially why I did it. I'll add tcl to the list if you like.
PerlJam: No, I *like* Coke. I hate Pepsi. :) 16:36
I must have missed it while running all the tests
PerlJam gaurav: probably because tcl is no longer in the parrot repo?
16:37 gaurav joined
Coke gaurav: I just added it. 16:37
Thanks. 16:38
Coke opens a ticket on my suggestion. Feel free to vote on it if you care. 16:41
gaurav++ since that won't show up here.
guarav, have you been converting the old wiki, was that you? 16:42
16:42 ruoso joined
gaurav No, just the Languages thing (and a couple of edits to NewParrotDevelopers...) so far. 16:42
dalek r33919 | infinoid++ | trunk: 16:45
: [src] Fix the following warnings:
: src/events.c:808: warning: null argument where non-null required (argument 1)
: src/events.c:912: warning: null argument where non-null required (argument 1)
: src/exceptions.c:215: warning: ignoring return value of 'die_from_exception', declared with attribute warn_unused_result
: src/exceptions.c:295: warning: ignoring return value of 'die_from_exception', declared with attribute warn_unused_result
diff: www.parrotvm.org/svn/parrot/revision?rev=33919
r33920 | infinoid++ | trunk:
: [cage] Fix the following codetest failures:
: * t/codingstd/c_cppcomments.t
: * t/codingstd/c_operator.t
: * t/codingstd/trailing_space.t
diff: www.parrotvm.org/svn/parrot/revision?rev=33920
Coke is fperrad in here? 16:49
dalek r33921 | infinoid++ | trunk: 16:50
: [cage] t/codingstd/pdd_format.t fails on pdd23 with some line length
: errors, split the lines.
diff: www.parrotvm.org/svn/parrot/revision?rev=33921
r33922 | pmichaud++ | trunk: 16:52
: [rakudo]: Refactor hash construction and assignment.
diff: www.parrotvm.org/svn/parrot/revision?rev=33922
17:03 chromatic joined, barney joined 17:05 tomyan left 17:34 baest_ joined 17:53 khisanth_ joined 18:03 PerlJam joined 18:16 rdice joined 18:18 alvar joined 18:21 ffwonko joined 18:31 Debolaz joined 18:37 Hadi joined 18:44 Theory joined 18:48 jan joined
Tene pmichaud: ideas about including except_*.pasm yet? 18:50
I'm wanting to fix continue() in when{} blocks
dalek r33923 | coke++ | trunk: 18:54
: Allow this function to be used in class.pmc and avoid linkage issues.
: Patch Courtesy Andrew Dougherty <doughera@lafayette.edu>
diff: www.parrotvm.org/svn/parrot/revision?rev=33923
chromatic Ah, good. 18:56
Wknight8111 :( my first book proposal just got rejected 18:57
Tene pmichaud: possibly in PAST::Compiler.to_post?
19:02 rurban joined
Infinoid perlcritic doesn't like it when files start with "#!/home/tewk/srcs/parrot/perl6". for a rakudo .pm file, is the hashbang even necessary? 19:05
Tene No. 19:07
Infinoid that said, perlcritic should probably just whitelist perl 6 files, as it can't understand them
particle yep
and within two months, we won't have to worry about most languages/ anyway 19:08
19:09 gmansi joined
dalek r33924 | chromatic++ | trunk: 19:10
: [t] Fixed an uninitialized value warning in test for Parrot::Test.
diff: www.parrotvm.org/svn/parrot/revision?rev=33924
19:17 cognominal joined
particle some of these tests are maddening 19:18
19:18 Andy joined
dalek r33925 | particle++ | trunk: 19:19
: [t] account for different name for null device on windows
diff: www.parrotvm.org/svn/parrot/revision?rev=33925
r33926 | Whiteknight++ | trunk: 19:22
: [Core] remove unreachable code from src/multidispatch.c:Parrot_mmd_find_multi_from_sig_obj()
diff: www.parrotvm.org/svn/parrot/revision?rev=33926
Infinoid should all perl 6 libraries contain a "use v6;"? if so, I can make perlcritic skip them based on that 19:25
particle no
there was discussion on that topic, but i can't remember the latest trend
Infinoid sadly, ext/SQLite3/DBDI.pm doesn't have much I can use to filter on
particle put it in an exclusions list, we already have one 19:26
chromatic Coke, ping.
particle might be nice to move it to ext/perl6/... or something
because we could easily also put python, perl 5, ruby modules in ext 19:27
Infinoid doesn't see an exclusions list in perlcritic.t
(excluding languages, yes, but not ext files)
its too bad, really. the "grep { $_->read !~ m/use v6;/ }" I added was working great for a while 19:28
particle ok, actually, v6 is still specced as preferred 19:29
dalek r33927 | particle++ | trunk:
: [t] use path-independent shell quotes from parrot's config
particle so change the files to include "use v6;"
dalek diff: www.parrotvm.org/svn/parrot/revision?rev=33927
r33928 | chromatic++ | trunk:
: [src] Delayed a string copy until absolutely necessary during an inplace chop.
diff: www.parrotvm.org/svn/parrot/revision?rev=33928
Infinoid so I can go around adding it to untagged perl6 files and make perlcritic ignore them?
particle yes 19:30
chromatic particle, r33925 could use File::Spec->devnull()
Coke chromatic: pong.
particle chromatic: you need to read the test...
chromatic Okay, that test is crack. 19:31
particle i don't have the energy to fight, just want to get it passing
nopaste "chromatic" at 69.64.234.10 pasted "Coke: Tcl Sub NameSpace PMC Patch" (30 lines) at nopaste.snit.ch/14957 19:32
tewk chromatic: t/pmc/key.t has a todo test that should pass w/ fix 19:33
19:33 cout left
chromatic tewk, it's not passing for me here. Hmm. 19:34
particle, I see what you mean.
tewk chromatic: my todo job was ugly, you have to uncomment the test 19:35
How should I todo a segfault style failure, skip? 19:36
Coke chromatic: with that, #10 is resolved. (if I run it against the parrot_io branch in partcl.)
dalek r33929 | particle++ | trunk:
: [t] un-todo bonus MSWin32 dynoplibs test
diff: www.parrotvm.org/svn/parrot/revision?rev=33929
chromatic Excellent. Let me see how Perl 6 and Pheme fare. 19:37
Coke chromatic++
tewk++
chromatic: perl6 seems ok 19:40
(OS X/intel_)
19:40 alvar joined
Coke chromatic: pheme test failure, but it looks like an odd syntax failure. 19:41
chromatic It should be a warning, not a failure.
I need to add operator precedence; I just haven't had time.
Coke ah, yes. just verbose.
otherwise ok.
Tene chromatic: did you see my warn() commit?
dalek r33930 | infinoid++ | trunk: 19:42
: [cage] More codingstd fixes.
: * t/codingstd/perlcritic.t (ignore perl6 files based on containing "use v6;")
: * t/codingstd/trailing_space.t (again)
diff: www.parrotvm.org/svn/parrot/revision?rev=33930
r33931 | chromatic++ | trunk:
: [PMC] Worked around TT #10 temporarily by skipping deep Sub-like introspection
: when using a PIR-level subclass of Sub. The right solution is making Sub's
: attributes into PMC ATTRs, but that's too big to do before tomorrow's release.
: Reclaimed a related test for the Key PMC.
diff: www.parrotvm.org/svn/parrot/revision?rev=33931
chromatic Tene, very nice.
Wknight8111 I'm getting test failures in t/compilers/pct/complete_workflow.t in trunk after an update and a realclean
can anybody else duplicate that?
tewk chromatic++ test reclamation 19:43
particle Wknight8111: win32? yes.
Tene Did anyone get alias's report of win32 test fialures?
Wknight8111 I didn't 19:44
particle nope 19:45
Coke Are they the normal test failures?
particle Wknight8111: i've fixed 4 win32 test failures today already, working on pct now
Coke *cringe*
Coke can try strawberry perl on his laptop tonight if there's time. 19:46
particle Wknight8111: if you have tuits for t/pmc/eval, go for it
Wknight8111 i probably don't till later tonight
particle r32058 | particle | 2008-10-20 16:17:09 -0700 (Mon, 20 Oct 2008) | 3 lines 19:51
[t] create utility module for parrot tests
~ add function for createing cross-platform tempfiles
~ convert tests to use it
i have absolutely no memory of this commit.
Tene complete_workflow.t passes fine for me
chromatic October Surprise
Wknight8111 this test has it in for me
this is almost the same test errors that were appearing in the calling_convention branch 19:52
chromatic Is it possible that your computer is haunted? 19:54
Tene Have you tried submerging it in holy water?
Infinoid hauntedness is a little too meta for my tastes
Wknight8111 it's two separate computers though, and three operating systems 19:55
chromatic Okay, then you're haunted.
Wknight8111 the test *knows* where I am and what I am working on, then then fails right there 19:56
dalek r33932 | Whiteknight++ | trunk:
: [Core] mark an undocumented function as being subject to RT#48260
diff: www.parrotvm.org/svn/parrot/revision?rev=33932
Coke I dislike RT#48260 markings. 20:03
Infinoid will test -rHEAD on strawberry perl, if "svn st" ever gets around to returning
Coke we -have- a test. Let's just use it.
20:04 mberends joined 20:06 elmex joined
chromatic I kind of like them, because I really dislike the 'RT#' all scrunched together, so whenever I tidy a file, I'm always tempted to write documentation to get rid of them altogether. 20:07
dalek r33933 | coke++ | trunk: 20:08
: Apply remainder of patch from Andy Dougherty that I thought was already applied.
diff: www.parrotvm.org/svn/parrot/revision?rev=33933
20:14 riffraff joined
Wknight8111 I like the RT#48260 markers too. They're easy to spot when you're working on a file, and easy to find all of them with an ack 20:15
Coke Wknight8111: (easy to find) so would be the codingstd test. 20:17
chromatic: I can just add the failing codingstd test back into 'make test'. =-)
chromatic Or we could review diffs and patches and reject the latter without docs. 20:18
Coke ah, crap. 'make test' with latest tcl & latest parrot is failing some tests, and hanging in t/cmd_expr.t
riffraff hi 20:19
Wknight8111 Coke there are, by my count, about 300 documented functions in the repo
adding a test for it would basically guarantee that test would fail forever
300 "un"documented functions
PerlJam Wknight8111: your patch is rejected because you didn't provide documentation :-) 20:20
Coke Wknight8111: there IS a test for it.
Wknight8111 I know there is a test, and that test will always fail
Coke the RT notes make the TEST PASS.
which is insanity.
chromatic Welcome, riffraff.
Wknight8111 or, almost always fail, in the foreseeable future
Tene riffraff: welcome
Coke sees tools/docs/func_boil* and wonders a) if that really helps, and b) why it doesn't work as adverted on the tin. 20:28
ah. it does work as adverted on the tin. It doesn't modify the files in question, though, which makes it <useful. 20:29
20:30 peters joined
peters Hi folks =) 20:30
nopaste "peters" at 83.183.129.106 pasted "URI (RFC 3986) grammar with tests" (312 lines) at nopaste.snit.ch/14958 20:31
dalek r33934 | particle++ | trunk:
: [t] slight refactoring for readability
diff: www.parrotvm.org/svn/parrot/revision?rev=33934
r33935 | particle++ | trunk:
: [t] whitespace cleanup
diff: www.parrotvm.org/svn/parrot/revision?rev=33935
peters could use a few more tests, but it's a good start =)
chromatic Very nice, peters. 20:32
Coke peters++
peters =)
particle peters: you should see if rakudo supports MAIN() well enough for you to use it for the tests 20:33
that way, if you run the file directly, it'll run the tests, but if you 'use' the file, it won't. 20:34
pmichaud has refactoring MAIN on his todo list.
peters particle: cool, I'll try that later 20:35
Tene pmichaud: objections to refactoring the controltypes hash initialization in pastcompiler to just store the ints instead of strings of the macros, if we're not going to get prelude with includes this release?
Coke chromatic: ../../parrot tcl.pbc t/cmd_break.t segfaults with latest of everything now. 20:36
pmichaud Tene: no object. If I manage to decide how I want to do prelude before the release we can (easily) switch back.
*objection 20:37
Coke (not saying it's your patch.)
particle pmichaud: any particular reason pct generates 8 extra spaces before every sub line? it's just making generated pir files larger and imo harder to read
pmichaud ...pct does this?
particle makes sure.
it happens in complete_workflow.t 20:38
nopaste "particle" at 76.121.106.245 pasted "a temp pir file generated from complete_workflow.t" (187 lines) at nopaste.snit.ch/14959
pmichaud oh. That's PGE.
particle yep, so it is 20:39
pmichaud it's because of the way codegen works in PGE. We can remove the spaces but it makes the code *much* harder to read.
Tene chromatic: if you're looking for an easy-to-reproduce segfault, try using count_eh
pmichaud because then one can't easily see what is PGE code and what is PGE generating. 20:40
particle ok, maybe a filter that removes X spaces when writing to a file or something
pmichaud at this point, readability of PGE source is more important than readability of what PGE generates.
chromatic Tene, is there a test file somewhere?
particle true enough, it's a minor nit
nopaste "tene" at 97.117.74.5 pasted "count_eh test for chromatic" (3 lines) at nopaste.snit.ch/14960 20:41
Tene chromatic: coutn_eh is completely untested
pmichaud ultimately I hope that PGE will generate POST, and then it'll be a non-issue :-)
particle :)
pmichaud I don't know that count_eh is going to be all that useful, fwiw. 20:42
at least, not as it exists now. 20:43
Coke tt #48 is also (ime) very reproducable. 20:50
what is the difference (on trac) between 'mac' and 'darwin' ? 20:52
20:54 hercynium joined
Tene peters: do all of those tests pass in your URI grammar? 20:54
mugwump maybe mac means MacOS 9-
Coke mugwump: which we don't support or plan to any time soon. I suspect it was just a copy over from RT. Deleting. 20:55
... or it's built into trac and I can't.
peters Tene: all 71 tests pass using r33858
mugwump mmm. and how's the Z80 port going? 20:56
Tene peters: very good work. :)
mugwump: hasn't been touched in ages
mugwump blast, was really looking forward to getting it running on my OS-X box
peters Tene: thanks, when grammar inheritance has been fixed I'll start on more URI schemes 20:57
nopaste "Infinoid" at 96.238.213.50 pasted "currentish failures on strawberry perl (takes a while on this box)" (29 lines) at nopaste.snit.ch/14961
Infinoid all tests pass on linux/x86-64 20:58
Wknight8111 w00t
particle Infinoid: there's a TT for ubuntu 64 failing configure. can you have a look? 20:59
Infinoid I can try. 21:01
particle Infinoid++
Infinoid segfault at VTABLE_get_string_keyed(INTERP, agg, key) looks awfully familiar though 21:02
Coke GAH
our new ticket handling sucks.
Can we get someone to cleanup parrot-tickets so we can't keep mis-posting there? 21:03
particle oh, i can't post there? uggh.
Infinoid drives home through the snow storm
particle coke: what version of trac are we running? 21:05
cotto 0.10.4, if the footer is telling the truth 21:06
particle that's pretty old 21:07
i believe it supports email2trac plugin, though...
subtrac.sara.nl/oss/email2trac/wiki/ 21:08
Coke particle: TT#49 21:11
-> 21:13
cotto Anyone have a clue what compilers/imcc/parser_util.c:multi_keyed does? It's only used in one place, which has been #if'd out for most of Parrot's life. 21:17
It looks like it could be removed without any pain. 21:18
chromatic No idea. 21:22
cotto should I remove it? 21:24
particle wednesday
chromatic No rush to remove it before the release.
cotto will do 21:25
21:35 davidfetter joined
riffraff in PGE how do I make a rule so that it matches a dot-terminated string? 21:41
I was thinking rule foo { <[^.]*> {*} } would have worked but it seems it doesn0t 21:42
pmichaud you mean a string ending in dot? / '.' $ / will work.
moritz the * should be outside the <[...]>
pmichaud oh, you want to match everything but a dot?
moritz token foo { <-[.]>* \\. $ }
riffraff everything but a dot yes
pmichaud or if you want to match everything up to a dot (but not necessarily require the dot to be at the end of the string): / (<-[.]>*) '.' / 21:43
riffraff I nitially tried putting the * outside of the brackets but the parse three did not contain the data
pmichaud there aren't any captures present, so "no parse tree".
riffraff ah right :) 21:44
ah, it seems i got the negated character class wrong 21:48
pmichaud ...anyone besides allison have admin privileges to the parrot-tickets list? If so, we can disable all postings except from trac.
riffraff I must reread the apocalypse I'm afraid 21:49
pmichaud the apocalypse may be out of date in key areas. The synopsis is more up to date.
moritz aye, S05++ 21:50
pmichaud++ # slowly killing the $str ~~ Grammar::rule meme
pmichaud mostly I'm closing tickets. :-) 21:51
moritz pmichaud: I'm a bit concerned about RT #61308... 21:52
pmichaud how so?
moritz pmichaud: the test that i provided in the ticket doesn't work
pmichaud your test had 'xab' instead of 'xay' 21:53
moritz ah
I was too tired to find that yesterday night
no concern then ;-)
pmichaud Martin found it :-)
I didn't see it either.
moritz ah, right 21:54
22:00 Whiteknight joined
dalek r33936 | pmichaud++ | trunk: 22:01
: [pct]: Fix -e option to execute '' or 0 argument if given (RT #57018).
diff: www.parrotvm.org/svn/parrot/revision?rev=33936
jonathan pmichaud++ # fixing/closing lots of tickets 22:05
peters pmichaud++ # RT #61338 22:11
Whiteknight particle still here? 22:14
22:14 ff-wonko joined 22:15 Limbic_Region joined
particle ayep 22:16
Whiteknight particle, what was that test you said I should look at tonight? 22:17
particle i get a win32 failure with t\\pmc\\eval.t:10
Whiteknight ...great, need to reboot into windows 22:18
failure have a trac ticket?
particle iunno
Whiteknight brb 22:19
22:39 andrew_ joined 22:44 MariachiElf joined
dalek r33937 | particle++ | trunk: 22:56
: [t] increased use of parrot test library functions, portability, test description accuracy, documentation. also fixes six failing tests on MSWin32
diff: www.parrotvm.org/svn/parrot/revision?rev=33937
chromatic Was that the culprit? 22:57
particle yeah 22:58
a few bugs in that code. specifically, needed to s/nqp/pm/ in one place
secondly, windows wasn't looking in the proper dir for the input file to the compiler, so needed to use absolute path 22:59
the rest makes ugly code a little less ugly
particle makes test 23:00
23:04 Zaba_ joined
particle andrew_: as of r33938, all tests pass with msvc/win32 23:32
dalek r33938 | particle++ | trunk:
: [t] cannot 'rm' an open file on windows
diff: www.parrotvm.org/svn/parrot/revision?rev=33938
andrew_ thanks particle++. I was working on it then got sidetracked 23:33
wife started baking cookies, and I'm a sucker for cookie dough
particle ok, maybe now i'll turn my attention to making gifts for my nieces and nephews 23:34
jonathan make gift
wknight8111 what programming language are you using to make their gifts? :)
particle looks like we'll get parrot.org updated to trac 0.11.2 this week 23:35
wknight8111 what version are we at now?
jonathan looks forward to playing with it, to see if it's worth upgrading the installation he looks after
particle mostly i'm using acrobat, illustrator, and photoshop for the gifts :)
we're at 0.10.1 iirc
0.10.4 i mean 23:36
23:41 ruoso joined
wknight8111 I like what I've seen of padre so far, but wxWidgets can be such a pain in the ass to install 23:51
of course, it's no harder then Gtk2, installing that on Windows was almost impossible
GeJ What are the "Tier 1" languages in Parrot? Rakudo, tcl, pipp? others? 23:58
jonathan LOLCODE
chromatic Pheme 23:59
(because it's so fast to test)
dalek r33939 | pmichaud++ | trunk:
: [rakudo]: Remove obsolete 'hash' method from List.
diff: www.parrotvm.org/svn/parrot/revision?rev=33939
r33940 | pmichaud++ | trunk:
: [rakudo]: Make sure Protoobjects don't participate in list flattening.

Powered by Cro and the Raku® Programming Language.

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

Thank you!