"Parrot 0.6.4 "St. Vincent Amazon" Released | parrotcode.org/ | 15/648/80 new/open/stalled tix | logged in irclog.perlgeek.de/parrot/today".
Set by moderator on 15 July 2008.
00:00 bacek joined 00:10 AndyA joined 00:11 DietCoke joined
DietCoke ... oh, sure he's gone! 00:13
00:14 japhb joined 00:35 Theory joined 00:39 TiMBuS joined
Tene Man, I really wish I could spend all week working on this hll stuff. 00:55
I guess I'll have to spend all next week doing it instead.
Looks like I need to get hllcompiler to deal with namespaces and class objects internally a bit more. 00:56
It uses 'new' in a few awkward places.
new-p-s looks in the current hll for a class 00:57
moderator "Parrot 0.6.4 "St. Vincent Amazon" Released | parrotcode.org/ | 15/648/80 new/open/stalled tix | logged in irclog.perlgeek.de/parrot/today 01:31
japhb (The log link was broken because of spurious trailing punctuation) 01:32
01:36 Andy joined
japhb chromatic: ping 01:38
Whiteknight chromatic isn't even here, I dont think 01:39
japhb sigh
Probably enjoying himself at OSCON, the lucky bastard
(Or home for the night, as I recall he lives within driving distance)
dalek r29687 | jkeenan++ | parallel: 01:49
: Specify exact number of tests in plan.
diff: www.parrotvm.org/svn/parrot/revision?rev=29687
Whiteknight either or, he's a lucky bastard 01:50
dalek r29688 | jkeenan++ | parallel: 02:02
: Specify exact number of tests in plan.
diff: www.parrotvm.org/svn/parrot/revision?rev=29688
pmichaud Tene: I'm planning ot upgrade P6object and HLLCompiler soon. This Saturday, if not sooner.
my goal for this week is to get precompiled libraries and (hopefully) .hll features working in pct and rakudo 02:03
dalek r29689 | jkeenan++ | parallel: 02:04
: Refactor configuration step classes to move hard-coded lists of required files
: and similar data to the step object's data structure. Write basic tests of
: step classes, e.g., can we locate the prerequisite files and directories.
diff: www.parrotvm.org/svn/parrot/revision?rev=29689
r29690 | tewk++ | gsoc_nci: 02:12
: [gsoc_nci] MERGE 29656:29689
diff: www.parrotvm.org/svn/parrot/revision?rev=29690
02:23 Andy joined
dalek r29691 | coke++ | trunk: 02:28
: [tcl] skip a lot more tests we can't pass yet.
diff: www.parrotvm.org/svn/parrot/revision?rev=29691
04:38 timbunce joined
bacek perl6.past: my ($a,$b)=(1,2); 05:04
nopaste "polyglotbot" at 193.200.132.146 pasted "perl6 past paste" (168 lines) at nopaste.snit.ch/13638
bacek perl6.pir: my ($a,$b)=(1,2); 05:06
nopaste "polyglotbot" at 193.200.132.146 pasted "perl6 pir paste" (11 lines) at nopaste.snit.ch/13639
05:14 Psyche^ joined 05:34 Theory joined 05:40 teknomunk joined 05:46 particl1 joined
pmichaud ping 05:47
particl1 pong
pmichaud hiya!
particl1 hey there
saw you and paula walk in, but didn't catch you 05:48
pmichaud I was looking for you -- didn't see you
particl1 i was in the back center
pmichaud oh well :-)
05:49 Andy joined
pmichaud the talks were excellent, as usual :-) 05:49
do we need to discuss anything for tomorrow's talk(s)? 05:50
particl1 probably 05:58
i'm looking over the slides, while somewhat distracted
i've been distracted most of the day, in fact 05:59
pmichaud ah. where are you now?
particl1 doubletree 06:02
you at the red lion?
pmichaud la q
la quinta
particl1 oh yeah 06:03
i'm pretty uninterested in the keynote tomorrow, you? 06:06
pmichaud let me check sched
particl1 en.oreilly.com/oscon2008/public/sch...2008-07-23 06:07
pmichaud I'm not interested in the 9:30a keynote. I might see the 8:45 keynotes, but it wouldn't kill me if I missed them 06:10
particl1 it's 9:30 that doesn't interest me, either 06:11
perhaps we meet then, outside the main ballroom?
06:12 uniejo joined
pmichaud okay, let's do that. 06:12
before the keynotes I need to double-check that my laptop is working
(on the projectors)
particl1 good idea 06:13
06:18 barney joined
dalek r29692 | jhorwitz++ | trunk: 06:21
: [NCI] add missing NCI signatures for mod_parrot (r346)
diff: www.parrotvm.org/svn/parrot/revision?rev=29692 06:22
pmichaud doh! I was hopoing to get interpolated namespaces working for jhorwitz
*hoping
it's been a lousy week, hacking-wise 06:23
particl1 he says he packed some tuits if you need them 06:25
pmichaud heh. I might be able to have it working tomorrow night.
dalek Tim Bunce | Inter-HLL Mapping Notes: 06:29
link: www.perlfoundation.org/parrot/index...ping_notes
06:39 iblechbot joined
dalek r29693 | pmichaud++ | trunk: 07:17
: [rakudo]: spectest-progress.csv update: 102 files, 1799 passing tests
diff: www.parrotvm.org/svn/parrot/revision?rev=29693
r29694 | fperrad++ | trunk: 07:30
: [Lua]
: - add gl & glut
diff: www.parrotvm.org/svn/parrot/revision?rev=29694
07:32 Ademan joined
dalek bernhard.schmalhofer@gmx.de | Pipp: 07:33
link: www.perlfoundation.org/parrot/index.cgi?pipp
07:34 masak joined
dalek r29695 | fperrad++ | trunk: 07:34
: [Lua] OpenGL
: - some methods
diff: www.parrotvm.org/svn/parrot/revision?rev=29695
07:53 timbunce joined
moritz how do I change the title of a ticket in RT? 08:04
08:07 uniejo joined 08:13 Debolaz joined 08:22 rblackwe joined
barney moritz: click on 'Grundlagen' or 'Basics', in the non-public ticket-view 08:31
moritz barney: thanks 08:34
08:34 gmansi joined
barney is heading for a short vacation in www.soelden.com/ 08:37
08:40 uniejo joined 09:06 purl joined 09:07 uniejo joined 10:02 donaldh joined
dalek r29696 | fperrad++ | trunk: 10:05
: [ICU] update with version 4.0
diff: www.parrotvm.org/svn/parrot/revision?rev=29696
Patterner still has only 3.8... 10:12
10:51 ruoso joined
dalek Daniel Ruoso | Inter-HLL Mapping Notes: 11:01
link: www.perlfoundation.org/parrot/index...ping_notes
11:03 Whiteknight joined 11:11 uniejo joined
dalek Daniel Ruoso | Inter-HLL Mapping Notes: 11:16
link: www.perlfoundation.org/parrot/index...ping_notes
Tene pmichaud: Yes, that's what I was trying to do. I didn't get very far, though. 12:48
12:50 masak joined 13:06 pmurias joined
pmurias if i want to add a new backend to rakudo i should add it to the PCT? 13:07
moritz at what stage would your backend replace the old one? 13:10
PAST? or POST?
pmurias POST 13:11
i plan to make rakudo emit sm0p instead of pir
would it be possible to do this withought writing pir? 13:17
moritz you could write PASM or PBC instead ;-) 13:18
pmurias seriously it would be possible to write it in p6 or an other *structured* high level language? 13:22
dalek r29697 | pmichaud++ | trunk: 13:28
: [pct]: Add 'pow' opcode signature to PAST::Compiler
diff: www.parrotvm.org/svn/parrot/revision?rev=29697
pmurias pmichaud: do you think it is feasible to write a PCT backend in perl6? 13:30
13:33 gryphon__ joined 13:35 davidfetter joined
pmurias rakudo: multi foo($a) {say $a};foo(1); 13:35
moritz pmurias: on #parrot that's called 'perl6:', but polyglotbot doesn't update its repo very often 13:38
pmurias perl6: say 4
polyglotbot OUTPUT[4ā¤]
pmurias kp6: say 4
moritz: other implemenations are not tollerated here? 13:39
;)
* tolerated
moritz pmurias: not a question of tolerance ;-) polyglotbot hosts parrot-based languages, in #perl6 p6eval hosts perl 6 implementations 13:40
pmurias i see 13:41
perl6: multi foo($a) {}
polyglotbot OUTPUT[Statement not terminated properly at line 1, near "{}"ā¤current instr.: 'parrot;PGE::Util;die' pc 120 (runtime/parrot/library/PGE/Util.pir:82)ā¤called from Sub 'parrot;Perl6::Grammar;eat_terminator' pc 22126 (src/gen_grammar.pir:2813)ā¤called from Sub 'parrot;Perl6::Grammar;statementlist' pc
..21139 (src/gen_grammar.pir:2450)ā¤called from Sub '...
pmurias perl6: multi foo($a) {...}
polyglotbot OUTPUT[Statement not terminated properly at line 1, near "{...}"ā¤current instr.: 'parrot;PGE::Util;die' pc 120 (runtime/parrot/library/PGE/Util.pir:82)ā¤called from Sub 'parrot;Perl6::Grammar;eat_terminator' pc 22126 (src/gen_grammar.pir:2813)ā¤called from Sub 'parrot;Perl6::Grammar;statementlist' pc
..21139 (src/gen_grammar.pir:2450)ā¤called from Su...
pmichaud pmurias: at some point I expect to rewrite PCT in Perl 6, yes 13:53
13:58 pmurias joined
pmichaud yes, it would be possible to create a PCT backend that emits something other than POST or PIR -- just need a replacement for PAST::Compiler 14:00
(and yes, that can probably be written in Perl 6. I don't know what Rakudo would be missing to be able to do that "today", though.)
14:01 uniejo joined
Tene Man, it sure would be nice if I could get the parrot rebuild to happen properly on feather3 14:03
I've tweaked it several times, and the 'svn up' is happening fine, but it doesn't actually rebuild. 14:04
Infinoid does it try? :)
Tene I'm not sure. 14:07
Infinoid can you redirect stdout/stderr to a temporary logfile?
pmichaud particl1: ping
Infinoid I'd offer to help, but I think I only have an account on feather1.
Tene I probably could. 14:08
Infinoid: if you're willing to debug it for me, I'll give you an account there. 14:09
Infinoid sure, I'm willing
Tene username?
purl username is the first entry?
Tene purl: forget username
purl Tene: I forgot username
Infinoid infinoid 14:10
purl you are Mark Glines <mailto:mark@glines.org>
pmichaud particl1: latest draft of slides is at www.pmichaud.com/2008/pres/oscon-rakudo/
(suggestions from all welcomed)
moritz "Rakudo Consulting Group" - is there such a legal entity? ;-) 14:11
pmichaud www.rakudoconsulting.com/ 14:12
moritz wow 14:13
pmichaud (short answer: "yes")
14:17 kj joined
pmurias any clients yet? 14:17
moritz pmurias: s/Plumhead/Pipp/ in your presentation 14:19
pmichaud oh, yes, thanks.
is chitchat far enough along to be mentioned?
moritz in the list of supported features, you could s/blocks/blocks and lambdas/ 14:20
blocks in p6 are far more than in most other languages
but then again, it should probably be listed in "mostly working" 14:21
pmichaud won't be long before I can get rid of those slides altogether, and just mentioned what doesn't work yet :-) 14:22
moritz ;)
pmichaud slides updating (slow hotel network connection) 14:24
afk for a bit 14:25
pmurias "nothing works right now as we are doing heavy refactoring..." ;)
14:27 iblechbot joined 14:30 timbunce joined
Tene pmichaud: chitchat isn't far enough along to be mentioned. The grammar has properly parsed everything I've thrown at it, but that's about all. 14:31
pmurias how can i see how many rakudo test pass 14:38
didn't mean to ask that :(
moritz perl tools/test_summary.pl 14:39
masak I'm glad that question was asked, because I didn't know about test_summary.pl before :) 14:56
moritz 'make spectest_regression' lists skipped tests as passes, I think 15:11
15:23 confound joined 15:39 Theory joined 15:45 timbunce joined 15:58 particl1 joined
pmichaud urgh. my week of technical oddity continues -- wireless is not working on my laptop :-( 16:01
Infinoid pmichaud: are you using the dreaded iwl4965 driver? 16:02
pmichaud no. 16:03
moritz Infinoid: I do. d'oh.
pmichaud no -- my wireless was working fine in the hotel, but at oscon it's not working at all
moritz Infinoid: it works all fine with linux-2.6.26, but with that kernel my sound doesn't work ;)
Infinoid moritz: I've had to put my ifconfig/iwconfig stuff into a shell loop, so it retries until the connection succeeds 16:04
16:07 cbrandtbuffalo joined 16:38 Andy joined 16:39 ^conner joined
^conner anyone going to the erlang talk? 16:40
TimToady I wish I could 16:43
japhb Wow, that's a power staff listing: www.rakudoconsulting.com/staff.html 17:00
17:06 ruoso joined 17:07 particl1 joined
japhb pmichaud: when slide 3 is continued, "Parrot Foundation" disappears 17:12
pmichaud: slide 18 says "Jul 2008", slide 19 says "Jun 2008" 17:20
17:20 cbrandtbuffalo joined 17:26 Theory joined 17:30 gmansi joined 17:33 cognominal joined
mj41 pmichaud: smop "C-based VM" -> "C-based engine" ? 17:39
17:39 timbunce joined 17:41 iblechbot joined 17:43 Theory joined
cotto_work is this nqjit error correction for pmichaud's presentation? 17:45
17:52 particl1 joined 17:56 slightlyoff joined 18:01 Theory joined
japhb cotto_work: if we're late, pmichaud should have posted the slide deck earlier. :-) 18:07
Besides, he does editor inheritance of his slide deck, so our changes now should help him for the next time he gives this talk 18:08
cotto_work cool. Where are they posted? 18:09
particl1 www.pmichaud.com/2008/pres/ 18:10
18:18 Ron joined 18:23 Andy joined 18:30 cbrandtbuffalo joined 18:32 Ivatar joined 18:36 Theory joined 18:37 timbunce joined 18:43 Ron joined 18:54 kj joined, Andy joined 19:11 cbrandtbuffalo_ joined 19:21 kasei joined
dalek r29698 | moritz++ | trunk: 19:22
: [rakudo] added S05-metasyntax/unknown.t to spectest_regression, Auzon++
: +6 passes
diff: www.parrotvm.org/svn/parrot/revision?rev=29698
kasei wondering if anyone knows reasons why parrot would fail to configure on darwin (both HEAD and 0.6.4), or if my local config is to blame. 19:23
cotto_work no guarantees, but you could nopaste the output here 19:38
kasei during config, ccflags seems to get set starting with "64" (which causes cc to die) 19:42
something to do with -arch flag checking, i suspect, but didn't want to dig into it if it was a known problem. 19:43
19:46 rurban joined 19:48 apeiron joined
Infinoid kasei: if you could cutpaste your output into nopaste.snit.ch/, we'd love to take a look 19:50
kasei sure thing. let me re-run... (after clearing ARCHFLAGS, I got the build to continue)
nopaste "kasei" at 24.195.199.16 pasted "Configure.pl failure (apparently due to ARCHFLAGS)" (48 lines) at nopaste.snit.ch/13643 19:54
rurban well, having a c compiler would help 19:55
kasei i have a c compiler
rurban where is your gcc pointing to? 19:56
kasei like i said, clearing out ARCHFLAGS allows the build to succeed.
Infinoid weird.
rurban maybe it's at i386-apple-darwin9-gcc-4.0.1
kasei gcc is a series of symlinks that ends at /usr/bin/i686-apple-darwin9-gcc-4.0.1 19:57
which is, afaik, the default on OS X. 19:58
rurban so we need a cc 64 -g -pipe -fno-common -no-cpp-precomp -Wdeclaration-after-statement -I/usr/local/include -pipe -fno-common -Wno-long-double -I./include -c test.c --verbose
Infinoid sounds like we need to not pass the "64" 19:59
kasei is that necessary? "64" shouldn't be an arg to cc there, so cc's output is kind of beside the point.
Infinoid yeah, it's an invalid argument, we've got some string handling wrong in config/
rurban well, your ccflags start with 64... looks odd
kasei the ccflags start with 64 because of bad parsing of ARCHFLAGS (if I've understood what's going on correctly) 20:00
Infinoid so, what was your $ARCHFLAGS set to?
kasei "-arch i386 -arch ppc -arch ppc64 -arch x86_64" 20:01
rurban could you paste your perl -V also?
kasei everything before '-g' in the first Pre-check line
Infinoid oh man, OS X makes my brain spin
rurban that's a bundle for 4 archs?
kasei yes
Infinoid OS X compilers like to generate binaries for all possible platforms, all at once, in the same file 20:02
nopaste "kasei" at 24.195.199.16 pasted "perl -V" (62 lines) at nopaste.snit.ch/13644 20:03
rurban did you read the darwin hints file? there is the explanation
kasei no. is that config/init/hints/darwin.pm you're referring to? 20:05
Infinoid rurban: right, that's where the arch list comes from. but where does the 64 come from?
rurban =~ s/-arch\\s+$arch//g is wrong
first ppc is replaced, the ppc64, it shoudl be the other way round. 20:06
Infinoid interesting. reversing the order of @arches should fix that 20:07
rurban care to submit a bugreport? => @arches = qw(i386 ppc64 ppc x86_64);
kasei sure. link?
rurban ./parrotbug
Infinoid or just mailto:parrotbug@parrotcode.org 20:08
kasei sure. thanks for the help tracking that down.
Infinoid kasei: does it fix anything when you reverse the arch order in that line?
rurban subject: [BUG] darwin hints with -arch ppc64
I'm sure that will fix it, please try that.
kasei yes, that fixes it. 20:09
rurban if you care you can also attach the patch then.
dalek r29699 | coke++ | trunk:
: [smolder] This test has been consistently failing on at least one netbsd
: system. Silence it for now.
: smolder.plusthree.com/app/public_pr...st_failure
diff: www.parrotvm.org/svn/parrot/revision?rev=29699
rurban coke is not at OSCON? 20:10
kasei: is that the default darwin9 perl? 20:11
kasei afaik, yes 20:12
why?
rurban because I wonder why nobody before reported that failure
it should have failed for everybody. 20:13
kasei the default system doesn't set the ARCHFLAGS to build for all four archs.
rurban hmm
kasei which can lead to building libraries that don't work with some of the system apps. 20:14
20:16 sjansen joined
kasei will ./parrotbug allow me to attach a file, or should I inline the diff? 20:18
moritz www.perlmonks.org/?node_id=699670 maybe somebody else from the rakudo crowd (jonathan, pmichaud, particle) wants to comment on that? 20:19
Infinoid kasei: I don't see a command line option for attachments. 20:20
rurban first submit and then go the link and attach it in the tracker. but since its so short you can also inline it.
Infinoid rurban: when I asked you for a fixed patch for pbc_to_exe_gen.pl, you sent me a patch which you said contained changes for a second ticket. what was the other ticket number? and are there additional changes I should incorporate before checking it in? 20:22
dalek r29700 | moritz++ | trunk:
: [rakudo] added S05-metasyntax/changed.t to spectest_regression, Auzon++
: +10 pass, +3 todo tests
diff: www.parrotvm.org/svn/parrot/revision?rev=29700
rurban moritz: I have both pugs and parrot-perl6 working fine under windows.
Infinoid: wait a sec... 20:23
purl hmmm... wait a sec... is purl one of those answer bots?
moritz rurban: with a 5.10er perl? (pugs, that is?)
rurban yes, 5.10
pugs is via cabal
moritz ok, I'll update my post
rurban I've packaged it cygwin, perl6 is removed since it was buggy, I cannot release pugs, because it's too hard to bootstrap automatically. 20:24
the new parrot update in a few days will again contain a fine perl6: parrot-perl6-0.6.4-2 20:25
moritz rurban: do you have an inofficial cygwin repo for pugs?
rurban sorry no. just a huge 15mb /usr/local/bin/pugs.exe
when I have more time I could make a package at my own setup site. 20:26
moritz no hurry, just being curious
rurban but I have so many ghc fixes to make, and I submitted some of them.
well, with the pre-compiled win32 ghc everything worked fine. but then I started to port it to cygwin. 20:27
and then I got stuck.
20:31 masak joined
rurban Infinoid: I have no ticket yet for the object vs libs order. I just played with it and asked the mailing list. the problem is that the windows linker is not so forgiving as the linux linker. 20:31
Infinoid ok, thanks. any objection to my checking it in? 20:32
masak are heredocs still '<<' in Perl 6? seems that would clash with hyper ops.
moritz S02:3023 20:33
masak thanks
rurban no, please check it in. better than nothing.
masak moritz: how implemented are they? 20:34
rurban About the design: I rather will change it sooner or later for --install to be the first arg, so that you can merge multiple pbc's into on exe sooner or later.
but for now it is consistent with the other tool which also takes --install as last arg.
Infinoid and as-is, it'll give people something to play with 20:35
moritz masak: I think TimToady++ added them to STD.pm the other day. I don't think rakudo does any heredocs (dunno about pugs)
rurban And one more thing: the limitation for no directory is annoying, it should use some pir match() code to put the installable_ to the basename only. but I haven't understood that part yet. 20:36
masak moritz: oki
20:36 Andy joined
masak matches the results of my experiments so far :) 20:37
rurban btw. re pbc merging. I could not merge the lua pbc's to a self-hosting exe.
20:37 timbunce joined
moritz masak: what are you experimental pugs results? 20:37
rurban is there a tricky merge order for the pbc's to follow? 20:38
Infinoid you know, I've done the majority of the work in the pdd13pbc branch so far, and I still don't understand .pbc files at all. :P 20:39
dalek r29701 | infinoid++ | trunk:
: [pbc_to_exe] Commit patches from Reini Urban in RT #57188. rurban++
: * Add an optional --install flag to pbc_to_exe.
: * Add a rule to the rakudo Makefile to generate an installable perl6 binary.
diff: www.parrotvm.org/svn/parrot/revision?rev=29701
masak moritz: that pugs has'em
I just tried pugs -e 'say q:to/END/;\\nhi\\nEND' (where the \\n are real newlines), and that worked 20:40
moritz ok, cool 20:41
20:41 rurban_ joined
dalek r29702 | infinoid++ | trunk: 20:44
: [rakudo]
: Followup to r29701, add installable_perl6 to the $CLEANUP list.
diff: www.parrotvm.org/svn/parrot/revision?rev=29702
20:45 cbrandtbuffalo joined
jonathan hi all 20:45
masak hi jonathan
rurban Infinoid: we shoudl also think about general install targets for all makefiles. the group so's, the binaries, the pir, the docs.
masak I accidentally found myself writing `push @in, $_ while =$*IN` in rakudo, and it segfaulted 20:46
Infinoid rurban: and tests for whether installing them works, too
rurban: patches welcome :)
masak it works better with `for =$*IN` :)
rurban sure :) the installable_ tests are the hard part. for most langauges I already have the targets.
masak but I'm still thinking of submitting a rakudo bug 20:47
rurban Now I'm cherry picking which parts go where to in the DESTDIR
20:47 cbrandtbuffalo_ joined
masak does `... while =$*IN` have any discernably meaningful semantics in Perl 6? or should it perhaps give a warning? 20:48
rurban And I have pod2man orgies for all language binaries. This would be needed for debian and gentoo also.
Infinoid awesome. 20:49
rurban was there never the idea to use the new perl6doc format for the standalone utils/langs?
moritz rurban: perl6doc isn't even really specced 20:50
20:50 Theory joined
moritz rurban: Damian wrote up a spec which was heatedly discussed on p6l 20:50
Infinoid rurban: I'm looking at #57110 and #57112, your patches for ncurses and postgres. I don't have cygwin, so I can't test them, but ... they're just patches to try alternate names for the cygwin versions of those libraries, right? if so, it looks pretty safe.
rurban but I found a nice old implementation at CPAN
moritz rurban: and then he (Damian) that he needed some time thinking, and stalled it 20:51
rurban sure. no magic. pretty safe
but it's almost complete
moritz rurban: yes, he also wrote a module (which wasn't bad), but because of the non-finalness it never really gained popularity 20:52
rurban anyway, 'm fine with old perl5 docs. so I can use pod2html and pod2man.
re damian: I really have to bug him to give me comaint for some of his old B modules.
They need to be updated for 5.10 and 5.11 20:53
ByteCache for sure.
oops, that was simon. he is still around. easy 20:54
20:56 particl1 joined
jonathan masak: If you make it segfault, file a bug report. That is never the right thing. 20:58
dalek r29703 | infinoid++ | trunk:
: [runtime library] Apply patch from rurban++ in RT#57110.
: * Cygwin has renamed the ncurses libraries; fall back to trying the cygwin names if the normal library names fail.
diff: www.parrotvm.org/svn/parrot/revision?rev=29703
r29704 | infinoid++ | trunk:
: [runtime library] Apply patch from rurban++ in RT#57112.
: * Cygwin has renamed the postgres libraries; fall back to trying the cygwin names if the normal library names fail.
diff: www.parrotvm.org/svn/parrot/revision?rev=29704
masak jonathan: oki
masak files
20:59 Theory joined
rurban Infinoid: we've renamed the pg lib on purpose so that the mingw folks don't fall in our trap. or vice versa. 20:59
Infinoid that's a very messy problem... 21:00
rurban the real messy probloem is the versioning without hardlinks as on unix.
Infinoid and parrot may still fall into mingw's trap, given that it tries the normal names first 21:01
rurban oops. I have think about that.
Infinoid ...or will loadlib fail and return undef, and allow the pir to continue trying the other name?
rurban ...and I have to write slower... 21:02
I wanted to check that in library.c with simple stat() calls.
Infinoid it would be great if mingw builds and cygwin builds could both DTRT here
rurban but now I think I will take the path checking the importlib first, because that never has any numbers. 21:03
and from the importlib I easily find the correct dll. This works best for all windows builds. cygwin, mingw, msvc.
and the libs are always at a specified path, the dll's not so. 21:04
jonathan masak: things :-) 21:05
Infinoid If you always rely on driving directions from a 4 year old kid, you always end up at the ice cream shop. 21:06
that's what $PATH management in win32 reminds me of
jonathan pmichaud, another interested parties: will do Rakudo day on Friday this week, all being well.
(Been a bit of a week...by Friday I'll have had three trips to the doctors... :-|) 21:07
dalek r29705 | moritz++ | trunk:
: [rakudo] dynamic x_label_skip in tools/progress-graph.pl
diff: www.parrotvm.org/svn/parrot/revision?rev=29705
particl1 Infinoid: i *just* talked to somebody at microsoft who wants to help us solve any windows problems we have with parrot
Infinoid great. does that include cygwin?
rurban :)
particl1 including library loading w/ cygwin msys etc
Infinoid awesome! 21:08
purl That's not awesome, that's par for the course at best. Are you from NZ
particl1 purl, forget awesome
purl particl1: I forgot awesome
rurban we don't really have technical problems, do we?
particl1 purl, you're stupid
purl particl1: what?
rurban It just have to be specced and agreed on.
21:10 teknomunk joined
Infinoid If they could implement a standard versioning system for libraries so our 3 flavors could fetch the right flavors of library, we sure wouldn't complain. 21:10
while we're at it, see how they feel about putting libraries in /usr/lib/ :)
rurban again another version to version dll's? rather not
ms has the tradition to use only major without any - or _ or . 21:11
that's just fine.
cygwin/X uses -1.3.2.dll
Forcing to use ENV{LIB} for the MS case is also fine. Similar to the java classpath. 21:13
particl1 i don't know how much help we can get from ms, but it'd be nice to have them involved in the conversations with cygwin/msys folks
rurban No FS standards are also good standards
sure.
Infinoid heh. I was only joking with those suggestions. all I know is, if everything worked today, we wouldn't be talking about this. :) 21:14
rurban MS should rather help us with Bob Rogers' problems 21:15
particl1 which problems are those? 21:16
masak rakudo: class A; my A $a .= new 21:19
perl6: class A; my A $a .= new 21:20
polyglotbot OUTPUT[./parrot: error while loading shared libraries: libparrot.so.0.6.4: cannot open shared object file: No such file or directoryā¤]
rurban particl1: rgrjr.dyndns.org/perl/acsip/actions.html
masak polyglotbot: that's not good
anyways, should the above work?
on my computer, it gives the error "Method 'ACCEPTS' not found for non-object" 21:21
rurban particl1: in short pushaction 21:22
masak hm, and `class A {}; my A $a .= new` gives 'get_bool() not implemented in class 'A''
I think this calls for another bug report...
moritz masak: it tries to determine some kind of value for $a - see #perl6 21:24
jonathan masak: Does it work with new()
masak jonathan: nope
another error, tho'
jonathan That's not good.
masak indeed not
jonathan my A $a .= new() has worked in the past.
The fact it's stopped working means we're missing a test.
masak yes. 21:25
moritz jonathan: it didn't stop working
./rakudo -e 'class A {}; my A $a .= new; say 1'
1
jonathan Oh.
moritz it just can't print out a representation of $a 21:26
masak yes, I get that too
jonathan OH! So it's not .= that is...right.
masak but even on the command line I get this error
and then I'm not trying to print anything
moritz masak: in the REPL?
masak no, -e 21:27
perl6 -e 'class A {}; my $a = A.new()'
moritz it somehow treats the last expression spcially
masak also '.new' w/o parens, also '.= new'
moritz *that's* a subject for a bug report
masak files 21:28
this is fun! :>
moritz: I don't much use the REPL any more. we didn't see eye to eye on things. 21:32
21:33 slightlyoff joined 21:34 slightlyoff left
dalek r29706 | coke++ | trunk: 21:41
: [tcl] We are now, and have always been, targeting tcl 8.5.3.
diff: www.parrotvm.org/svn/parrot/revision?rev=29706
21:41 Theory joined 21:42 julian_ joined 21:54 teknomunk joined 22:01 japhb joined
dalek r29707 | coke++ | trunk: 22:10
: [tcl] After 'make tclsh', remove the intermediate tcl.c and tcl.o files on 'make clean'
diff: www.parrotvm.org/svn/parrot/revision?rev=29707
paco NotFound: people from #linux are requesting your attention .. 22:11
dalek r29708 | coke++ | trunk: 22:25
: [tcl] code.google.com/p/partcl/issues/detail?id=58
: Eliminate some more __foo style sub names
diff: www.parrotvm.org/svn/parrot/revision?rev=29708
22:37 teknomunk_ joined 22:38 Whiteknight joined 22:42 cotto_work joined
dalek r29709 | Whiteknight++ | gsoc_pdd09: 22:50
: [gsoc_pdd09] update to trunk r29708
diff: www.parrotvm.org/svn/parrot/revision?rev=29709
22:59 Limbic_Region joined 23:00 Andy joined
cotto_work coke, ping 23:00
23:07 ruoso joined 23:13 cognominal joined 23:27 DietCoke joined
DietCoke cotto_home: yes? 23:27
cotto_work is your windows build env working? 23:28
DietCoke I haven't checked in some time. Strawberry perl? 23:29
dalek r29710 | coke++ | trunk:
: [tcl] code.google.com/p/partcl/issues/detail?id=58
: Eliminate some more __foo style sub names
diff: www.parrotvm.org/svn/parrot/revision?rev=29710
cotto_work I just would like a patch tested.
DietCoke Sure. Let me re-up and clean and test, and then test with the patch. URL?
cotto_work rt.perl.org/rt3/Ticket/Display.html?id=46681 (strerror_7.patch)
DietCoke ... That's a most annoying stomach cramp. Ow. 23:30
cotto_work to test, just run t/pmc/file.t
thanks!
DietCoke I think I'll test in the msvc world. More likely to squawk.
23:32 Theory joined
Limbic_Region salutations all 23:35
DietCoke hio
Limbic_Region IIRC, you do DreamWeaver development in your meatspace life right?
DietCoke ... ColdFusion. 23:36
Limbic_Region yeah, that's it
DietCoke (I understand you can actually do dreamweaver development, but for me, it's like 'vi'. =-)
dare I wonder why you ask? =-)
Limbic_Region well, I posted something on use.perl kind of as a joke 23:37
DietCoke I also have been known to be paid for C++, Java, ...
url?
Limbic_Region but also because I wanted to try and quantify how much free time I don't have anymore
use.perl.org/~Limbic+Region/journal/36996
23:37 particl1 joined
DietCoke ... wait a minute. COTTO is asking me to run tests on WINDOWS!? 23:38
DietCoke senses something terribly wrong with the universe.
Limbic_Region heh
23:38 timbunce joined
Limbic_Region well, the point of the post and the point of the question is trying to figure out if I am just missing something obvious 23:39
people who have more responsibilities than me are far more productive
but I can't seem to figure out how they do it
DietCoke I often work on parrot because it's the most interesting thing on my plate. This may say more about my plate than our parrot. 23:40
Limbic_Region so - you having a real job that wasn't paying for parrot development qualified for my "how do you do it" lead in
DietCoke AH!
I was trying to figure what the hell you were getting at. =-0
I work on parrot because I think it's shiny. That's all.
Limbic_Region right - but presumably you have a real life and a real job
DietCoke debatable. =-) 23:41
Limbic_Region and you don't just sit on the channel chatting (like me) - you actually produce a work product
Limbic_Region is just depressed is all
DietCoke is too, just not about parrot.
Limbic_Region well - I wouldn't mind not contributing to parrot if I was working on ANYTHING 23:42
but whatever - this is #parrot not #woe_is_me
Limbic_Region shuts up
DietCoke if you're interested in working more on parrot, I'd try to figure out how much time you have available. Sounds like "not much". Try to find an activity that fits into that time slot. 23:43
like "going through the bug queue", or "setting up a smolder client", or "fixing and enabling a new warning/codingstd test" etc.
plenty of small stuff that would be very appreciated. 23:44
particl1 dietcoke: do you use a "plate" with medifast?!?!?! ;)
DietCoke how the hell do you apply a patch with tortoise svn? 23:47
it brings up a merge tool for a trivial diff with no conflicts, and I don't see a big red "apply" button.
Whiteknight I used to know how to do it 23:48
there is an "apply patch" option somewhere
it's on the bottom of the pop-out list, I think
DietCoke I have a hovering window with the two file names in the patch.
no button.
if I right click, there's a "patch all".
that by itself doesn't change files on disk. 23:49
Whiteknight no no no. Right-click on the patch file in explorer
particl1 right click the files
Whiteknight in the menu, there should be an option to "apply patch"
particl1 patch all does change the files, at least once you exit the merge viewer
DietCoke particl1: that's what I"m saying, it doesn't. 23:50
I'll try Whiteknight's way, instead of picking "apply patch" on the directory and then selecting the patch.
particl1 that should open the merge viewer 23:51
iirc
DietCoke Yup. I end up at the same place.
P(*@#&$
particl1 which should Just Work
i also have a 'patch' util 23:52
DietCoke once I select "patch all", ... nothing happens. there's no option in the merge viewer to save.
Whiteknight you broke it 23:53
DietCoke ... it didn't like the single pane view, apparently.
23:54 cout joined
DietCoke cotto_home: fail. doesn't even build. 23:58
cotto_work oh no my patch asplode
Do you know what kills the build process? 23:59
Whiteknight a stiff breeze?

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!