Parrot 0.9.0 | parrot.org/ | 500 unresolved RTs left.
Set by moderator on 7 February 2009.
00:09 AndyA joined
dalek rrot: r36427 | NotFound++ | trunk/src/debug.c:
[debugger] use readline for interactive input (testing)
 01:10
rrot: r36428 | NotFound++ | trunk/src/debug.c:
[debugger] delete a diagnostic message commited by mistake
 01:13
01:14 Fayland joined 01:15 Fayland_logger joined 01:35 leto_ joined
Coke I have tried to get my mactop PPC up and running to smoke parrot, but it won't boot. :| 01:37
you can hear the drive spin up, but after a software update, it won't actually boot. suck. 01:38
01:40 kid51 joined
Coke kid51: any tips on getting a mactop PPC to boot? 01:47
(drive spins up, but screen stays dark. last thing I did on it was a software update that covered probably a year of updates.) 01:48
01:53 braceta_ joined
kid51 Coke: No, I've never experienced that. There is some key that you can hold down during reboot that is the equivalent of Win32 Safe mode -- but while a Genius Bar guy had to do that once, I myself have not. 01:54
Try this: support.apple.com/kb/HT1564 01:55
support.apple.com/kb/HT1455?viewlocale=en_US 01:56
I don't think that the software update per se explains your problem. Last year I had some corrupted data somewhere and had to reinstall the OS from scratch -- then do a year's worth of software updates. Very time consuming, but ultimately successful. 02:00
Coke at this point, I think my display is borked. safe mode doesn't work, resetting the display doesn't work, resetting the PROM doesn't work, booting from CD... nothing. 02:05
kid51 Bummer. Have to take it into the shop -- assuming it's worth it at its age. 02:10
02:10 TiMBuS joined
Coke it's only about 3years old. 02:22
HEY! it's booting.
02:25 leto_ joined 02:32 ask joined
tewk So is there a good solution to aliasing perl5 package names? ie java import, or shorten long package names? 02:55
kid51 In what context would you want this? 03:00
Infinoid XRL::18uYxM2R 03:03
Coke doesn't aliasing imply typeglobs?
or is that just for inside a given namespace? 03:04
Infinoid Yeah, I'm not sure that works for namespaces themselves
Coke build is broken. trying to find where it snapped. 03:06
kid51 Coke: src/debug.c ? I just got a failure there on Darwin PPC -- but I'm not getting it on Linux i386 at the same revision. 03:08
Coke failing on darwin/x86
opened a ticket. doing the bisect now. 03:09
kid51 purl paste
purl paste is (see: nopaste) or like glue but a little safer to sniff. or nopaste.snit.ch:8001/ or scsys.co.uk:8001/ anywhere shadowpaste is or mmm, sticky paste or You there! Eating the paste. or <see> 2 girls, 1 paste or App::Nopaste
nopaste "kid51" at 71.247.49.57 pasted "Build failure at r36428 on Darwin/PPC" (6 lines) at nopaste.snit.ch/15534 03:11
Coke ayup 03:12
trac.parrot.org/parrot/ticket/287
was borked in 36427 03:16
Coke finally upgrades svn on his mac. 03:17
nopaste "kid51" at 70.85.31.226 pasted "rev 36427 likely source of build failure" (64 lines) at nopaste.snit.ch/15535 03:18
kid51 But why would this have failed on Darwin but not on Linux. 03:19
?
Coke NotFound: ping.
<shrug> different versions of gcc? Iunno. 03:20
kid51 I think those warnings in src/dynext.c have been there for some time and are not germane.
Confirmed. 03:22
Coke it's the error that's a problem, yes.
I just pasted in the trail end of my build, that's all.
03:22 Tene joined
kid51 Hypothesis: We've long had problems with config::auto::readline on darwin. While I no longer get garbage in the Configure.pl output when I run that step, others (may) still do. This was most frequent configure error reported last June at buildfest at YAPC; we still have open tickets. And 'readline' is reference in NotFound's commit message. 03:24
In any event, we don't require 'readline' in order to build Parrot. So it may be an error to depend on it in our C source code. 03:25
Reject 2nd sentence in last post. 03:28
Perhaps he should have been using some Parrot-mangled readline instead of a bare 'readline' 03:29
nopaste "kid51" at 70.85.31.226 pasted "Where the string 'readline' occurs in our C source code: fns src | grep .c$ | xargs grep -n readline" (51 lines) at nopaste.snit.ch/15536 03:31
kid51 Extending Coke's compiler hypothesis: Note wordspace between '&' and 's' in this source code: 03:34
Parrot_PCCINVOKE(interp->debugger,
stdin, readline,
"S->S", prompt, & s);
Coke that's not it. 03:36
it's the assigning to stdin on the line it says the error is on.
run that through gcc -E and you'll see stdin is really a macro.
changing the var name, trying again...
kid51 As in 'gcc -E src/debug.c' ?
Coke I ended up doing a make, grabbbing the "compiling with" line, tweaking it to add a -E and changing xx.c to src/debug.c 03:37
that runs it through the C preprocessor, so that's more like what the compiler really sees. 03:38
then you can see that stdin has been replaced with (&_<some stuff>)
running test suite..
and with this patch, I will have pretty much exceeded the limit of my c-fu. 03:39
kid51 paste patch so I can try it on PPC? 03:40
nopaste "coke" at 72.228.52.192 pasted "Index: src/debug.c ===========" (27 lines) at nopaste.snit.ch/15537 03:41
Coke given how the c compiler works on darwin, i'd be very surprised if that didn't fix it.
all tests pass here. committing... 03:42
I'll leave the ticket open until you report back... 03:43
kid51 And dare I ask: How *does* the C compiler work on Darwin? (cue chromatic responding, "Badly") 03:44
Coke there are a lot of platform specific macros that hide a lot of the differences from you. 03:45
kid51 I applied your patch, then hit 'make' -- forgetting to hit 'make clean' first. But 'make' picked up right there as if there had never been a problem.
Coke you only need make clean if you want to start over.
make will continue if possible.
and since we just changed a straight C file, nothing oddly generated, or with missing dependencies, that's fine.
kid51 Understood. This will take up to 10 minutes more to complete build. 03:46
03:48 janus joined
kid51 On a different topic: I got a perlcritic.t failure tonight for a 'flag comment'. But those comments have been present in the file in question for years, without causing a perlcritic.t failure. Why? 03:53
Coke: Commit away. 'make' successful Darwin PPC. 03:54
# Perl::Critic found these violations in "/topdir/config/gen/makefiles/dynpmc_pl.in": 03:55
# Flag comment 'XXX' found at /topdir/config/gen/makefiles/dynpmc_pl.in line 66
# Flag comment 'XXX' found at /topdir/config/gen/makefiles/dynpmc_pl.in line 206
03:59 TiMBuS joined, Theory joined 04:09 Whiteknight joined
kid51 Hmm, dalek is dropping svn commits again. Coke's commit in 36430 didn't get displayed. 04:21
Infinoid hmm 04:24
Infinoid doublechecks his latest rss runloop update 04:25
kid51 Correction: None of 36428 36429 36430 displayed.
First of those was Coke's; others mine.
04:31 eternaleye joined
Infinoid has the bot just gone catatonic? 04:31
04:42 gravity joined
kid51 Someone said one of your commits failed to be displayed the other day. 04:46
kid51 must sleep
purl $kid51->sleep(8 * 3600);
Infinoid reverts parrot-one-rev-at-a-time.patch 04:48
04:50 sjn joined 04:56 dalek joined
Infinoid dalek: stop eating commits pls, kthx 04:56
dalek allison@perl.org | Debian/Ubuntu chroot Environment Setup: 04:57
link: www.perlfoundation.org/parrot/index...ment_setup
shorten dalek's url is at xrl.us/beesjm
06:05 TiMBuS joined 06:14 Andy joined 06:18 TiMBuS joined 06:44 rurban__ joined 07:58 davidfetter joined 08:05 eternaleye joined 08:53 rurban_ joined
rurban_ TT#276 update: mingw still fails, but I found something: trac.parrot.org/parrot/ticket/276#comment:5 08:53
09:13 Zaba_ joined 09:30 alvar joined, TiMBuS joined 09:44 barney joined
dalek tracwiki: v48 | barney++ | Languages 09:51
tracwiki: trac.parrot.org/parrot/wiki/Langua...version=48
09:54 iblechbot joined
janus meh, there's some kind of runtime path glitch: rafb.net/p/AP2w3m25.html 10:03
with 0.9.0.1 that is, so that doesn't necessarily is very recent information, but anyway... just in case 10:04
the actual path where it is installed is /usr/local/runtime/parrot/library/PCT.pbc 10:05
10:18 rurban_ joined 10:53 Zaba joined
dalek rrot: r36431 | barney++ | trunk:
[dynpmc] Resurrect the Pair PMC, as a dynpmc, as Eclectus uses it
 10:58
rrot: r36432 | barney++ | trunk/config/gen/makefiles/languages.in:
[languages] Add helpers to clone and pull shakespeare
 11:14
rrot: r36433 | barney++ | trunk/config/gen/makefiles/languages.in:
[languages] Mention up-rakudo and co-rakudo in help message
rrot: r36434 | barney++ | trunk/config/gen/makefiles/dynpmc_pl.in:
[config] align some '='
rrot: r36435 | barney++ | trunk/config/gen/makefiles/languages.in:
[shakespeare] clone into directory shakespeare
 11:23
rrot: r36436 | barney++ | trunk:
Let svn ignore languages/shakespeare
 11:26
rurban_ TT#276 consting: see www.cocoabuilder.com/archive/messag...2/6/198093 11:47
shorten rurban_'s url is at xrl.us/befg3i
dalek rrot: r36437 | barney++ | trunk:
[dynpmc] Reenable Pair as a dynamic PMC
rrot: r36438 | NotFound++ | trunk:
[cage] avoid some gcc warnings and add more protection to some casting macros
 11:53
NotFound rurban_: what's the relation between TT #276 and Objective C on Mac OS? 11:57
rurban_ nothing. it's just showing to cast from const to non-cast vice versa. 11:59
but it's not explaining why my PMC*const <> PMC* where both pointers have the same hex value 12:00
NotFound rurban_: Have you seen my last comment on the ticket?
rurban_ it's because the compiler marks in a different segment, though they are not. This is my idea
not yet
I tried to fix PMC_IS_NULL by copying the const ptr to non-con st and comparing this then 12:01
nopaste "rurban" at 143.205.212.10 pasted "TT#276 idea which does not work" (29 lines) at nopaste.snit.ch/15539 12:03
NotFound rurban_: I see a potential problem in that piece of code: item is checked for PMCnullness *after* using it. 12:11
Accessing item->vtable->base_type when item can be NULL is wrong 12:12
rurban_ The problem is pmc_register, I have to get to the last line of pmc_type, Parrot_get_datatype_enum(interp, name) 12:14
NotFound I'll fix the item problem now 12:16
dalek rrot: r36439 | NotFound++ | trunk/src/pmc.c:
[core] avoid dereferencing a pointer before checking it for nullness
 12:22
12:41 mberends joined
dalek rrot: r36440 | barney++ | trunk/lib/Parrot/Pmc2c/Pmc2cMain.pm:
[perl] slight beautifications
 12:41
rrot: r36441 | barney++ | trunk/lib/Parrot/Pmc2c/Pmc2cMain.pm:
[build] Search also in 'src/dympmc' for PMC dump files
rrot: r36442 | barney++ | trunk/config/gen/makefiles/dynpmc_pl.in:
Add reference to TT #290
 12:42
lectus: 58bd15f | (Bernhard Schmalhofer)++ | config/makefiles/root.in:
export DYNPMC_INCLUDE, so that the Pair PMC is found
 12:48
shorten dalek's url is at xrl.us/befg7g
12:53 rurban_ joined
dalek lectus: d15a006 | (Bernhard Schmalhofer)++ | (19 files):
Give up on unified languages testing for eclectus.
 13:04
shorten dalek's url is at xrl.us/befg7p
dalek lectus: 93aba97 | (Bernhard Schmalhofer)++ | eclectus.pl:
Need to include riaxpander
 13:11
lectus: f0a2cd3 | (Bernhard Schmalhofer)++ | .gitignore:
Add .gitignore
shorten dalek's url is at xrl.us/befg7t
shorten dalek's url is at xrl.us/befg7v
dalek rrot: r36443 | NotFound++ | trunk/lib/Parrot/Pmc2c/Attribute.pm:
[cage] avoid regex duplications in Parrot::Pmc2c::Attribute
 13:13
13:17 rurban_ joined
dalek lectus: 3a57a24 | (Bernhard Schmalhofer)++ | docs/eclectus.pod:
Mention how to run the generated NQP-Files
 13:17
shorten dalek's url is at xrl.us/befg77
rurban_ with r36439 still parrot-mingw>parrot t\\dynpmc\\dynlexpad_1.pir => Null PMC access in get_integer() 13:18
13:28 Whiteknight joined
dalek lectus: 2f9ddae | (Bernhard Schmalhofer)++ | (34 files):
Remove the useless '$Id$' svn tags
 13:28
shorten dalek's url is at xrl.us/befg8h 13:29
NotFound rurban_: according to the documentation of Parrot_load_lib that test is wrong because it can return NULL/PMC_NULL or throw an exception, not always return something that can be checked as boolean. 13:40
But the documentation of Parrot_load_lib is a mess, and the code does not match it. 13:41
nopaste "NotFound" at 213.96.228.50 pasted "Attempt to saenify Parrot_load_lib" (35 lines) at nopaste.snit.ch/15540 13:43
NotFound rurban_: try this patch
rurban_ NotFound: still same Null PMC access in get_integer() 13:51
mingw only 13:52
NotFound rurban_: at what instruction fails? At the 'unless lib ...' ? 13:54
rurban_ No, at pmc_register, pmc_type 13:56
still the same wrong PMC_IS_NULL check 13:57
NotFound Inside the 'lib = loadlib "dynlexpad" ' instruction?
rurban_ In my latest comment in the ticket I mixed up teh macro with the function call. In fact I was using the function call, and added my tmpconst workaround there 13:58
inside Parrot_lib_dynlexpad_load only
Sorry, I have to leave the net now, go back onto my bus 13:59
14:11 ascent joined 14:12 masak joined
mberends hi masak 14:13
dalek rrot: r36444 | NotFound++ | trunk/src/pmc/namespace.pmc:
[cage] make get_pointer_keyed_str in the Namespace pmc more robust about mistaken NULL values
 14:16
lectus: 6b41b34 | (Bernhard Schmalhofer)++ | t/pmc/boolean.t:
Add test script t/pmc/boolean.t
 14:17
shorten dalek's url is at xrl.us/befg9v
masak hey there, mberends. 14:20
mberends Well done on proto! I'm wondering where it will lead to. I think others are leaving the Rakudo 'install' topic, so that may be your destiny. 14:21
masak we'll see. 14:22
I just want to provide a good default way for people to set things up in case I and others decide to go crazy and split out modules up into separate projects. 14:23
mberends how seriously did you mean 'write it myself' ?
masak mberends: oh, people are welcome to contribute, of course.
just bear in mind that I'll probably scrutinize commits and revert them if they don't conform to my goal image of the thing :) 14:24
an ask-first-shoot-later approach might be appropriate for anyone who wishes to add to the scripts right now.
mberends: however, comments like the ones you added are highly appreciated. 14:25
(and don't need permission)
mberends this won't become CPAN, but could become a short term substitute. I was considering a sort of 'GitPAN' interim design with a catalog of what everyone has in their repos.
masak exactly.
consider this project a prototype implementation of such a project. 14:26
14:26 Whiteknight joined
mberends I'd like to extend it in that direction. 14:26
masak you're welcome to do that.
btw, I'm feeling slightly OT here in #parrot.
maybe move discussion to #perl6? :)
mberends ok
14:43 rurban__ joined
dalek lectus: 6162ceb | (Bernhard Schmalhofer)++ | (35 files):
Move the scheme based tests into t/scheme
 14:54
shorten dalek's url is at xrl.us/befhbz
dalek lectus: 60d67aa | (Bernhard Schmalhofer)++ | (3 files):
Steal t/harness from Pipp. Prepare for smolder testing
shorten dalek's url is at xrl.us/befhb3
14:57 rhr joined 15:01 kid51 joined 15:12 gravity joined
dalek lectus: 45349c7 | (Bernhard Schmalhofer)++ | (3 files):
Add smoke testing.
 15:15
shorten dalek's url is at xrl.us/befhc4
15:15 guru joined
guru Is anybody thinking actively about Windows installer programs for perl 6 / parrot? They will be needed soon. 15:18
NotFound guru: what do you mean by "thinking actively"? Is an euphemism for "working"? 15:19
guru "Thinking" should precede "doing"; sometimes it can eliminate "working" 15:22
NotFound I seriously doung that an installer can be done just by thinking about it.
doubt
guru There are installers for many open-source programs; if they can be adapted, it would save work. 15:24
15:28 mikehh joined 15:33 tetragon joined 15:47 guru joined 15:48 guru joined 16:04 Whiteknight joined 16:07 rhr joined
diakopter guru: see cpansearch.perl.org/src/ADAMK/Perl-...12/Changes 16:08
(mention of Parrot)
guru I started looking around, and this seems to be a good prospect: www.jrsoftware.org/isinfo.php 16:17
There's also nsis.sourceforge.net/Main_Page 16:24
16:32 Zaba joined 16:35 Tene joined
pmichaud hello, all 16:45
masak pmichaud: hello!
pmichaud I'm _finally_ caught up on sleep. :-) 16:46
masak :)
boy, am I glad to see you!
pmichaud got some tickets that need attention? ;-)
masak pmichaud: indeed. hold on. 16:47
pmichaud (I'm at the Frozen Perl hackathon now so might get distracted helping others here)
masak irclog.perlgeek.de/perl6/2009-02-04#i_885182
pmichaud masak: reviewing 16:48
purl reviewing is best done inebriated
masak purl, no reviewing is <reply>
purl OK, masak.
pmichaud I think #63004 is fixed already, isn't it? 16:49
masak checking.
yes. 16:50
it wasn't when I made that list.
pmichaud yes, that was actually a parrot bug.
masak aye.
pmichaud (that has been fixed since the ticket was filed)
masak pmichaud: also, moritz++ reported being unable to 'make spectest' when Parrot is inside Rakudo. 16:51
pmichaud I think that might've been fixed also (problem was slashes in makefile) 16:52
masak ok.
pmichaud but I'll check that too.
oh, it might be a path problem with parrot.
masak mhm.
I don't have that setup, so I can't confirm it.
pmichaud (i.e., can't find parrot because parrot isn't at ../../parrot)
I'll do that here shortly.
dalek rrot: r36446 | NotFound++ | trunk:
[debugger] drop 'debugger' member from interpreter, use pdb->debugger instead
 16:56
pmichaud it may take me a while to find/fix bugs dealing with objects/protoobjects (e.g., #62894); jonathan++ has made quite a few changes since I last looked at it and I'll have to catch up. 17:05
we need to rewrite Rakudo's t/harness, I think. 17:11
dalek rrot: r36447 | Infinoid++ | trunk/src:
[cage] Switch some hard tabs for spaces to pass t/codingstd/tabs.t.
 17:12
pmichaud preferably to not use Parrot's t/harness 17:13
NotFound Infinoid: sorry, my fault 17:14
Infinoid no biggie, free karma for me :) 17:16
Coke_afk tries to work on the one parrot task he's allotted to himself this week. 17:24
Coke will probably have an easier time than that AFK guy.
Infinoid NotFound: I'm updating split-encoding-charset-labeling-from-init.patch according to your function naming suggestions
NotFound Infinoid: good :) 17:25
dalek kudo: 6aaec33 | (Patrick R. Michaud)++ | (2 files):
Temporarily fix makefile construction to still work with Parrot.
shorten dalek's url is at xrl.us/befhmg
NotFound Infinoid: IMO you can commit it right now, there is no point in waiting if allison is aware of the intention.
Infinoid I'm hoping someone can confirm that I haven't broken darwin and win32, before I do that 17:28
test reports are welcome :)
NotFound rurban alraeady tested, isnt'it? 17:29
And Darwin... well, the best adepated will survive X-)
Infinoid true, rurban tested that *plus* the no-encoding/charset-globals patch 17:30
speaking of which, have you made any more progress on your patch? I'm still not sure which approach is the right one
NotFound Infinoid: no, I was waiting for solving this issue first 17:31
Infinoid well, guess I'd better check it in then 17:32
NotFound And I think we also must clarify the point of the default encoding and charset first. 17:33
dalek rrot: r36448 | Infinoid++ | trunk:
[core] Split the labeling of charsets and encodings until after the init
 17:34
NotFound "We" the parrot's people, of course
Infinoid I think that is a separate issue
Infinoid sees if he can get his patch working on jit 17:37
rurban TT #276 solved on mingw. my fault. I was linking to the wrong lib 17:46
I have to verify what patch exactly solved it, but I'm pretty sure it works now, after reverting parts of the mingw doublequote patch 17:47
dalek rrot: r36449 | NotFound++ | trunk/src/runops_cores.c:
[debugger] fix debugger tracing
17:54 jdv79 joined
rurban I accidently linked against blib/lib/libparrot.lib (static) but forgot to update that lib by disabling the ar step it in the Makefile. 18:02
NotFound rurban: you must commit sepukku 18:03
18:03 Theory joined
NotFound Mmmm... better don't do that, keep working on parrot ;) 18:03
Infinoid got "make test" working on linux/x86 without disabling jit in config... trying testj now 18:10
dalek kudo: fb66ff6 | (Moritz Lenz)++ | t/harness:
[t/harness] fix testing if not checked out as languages/rakudo
 18:11
shorten dalek's url is at xrl.us/befhoe
jdv79 the cpan install of parrot wont work with Perl6::Conf, right? 18:12
Infinoid the Perl6:: namespace on CPAN is for perl5 equivalents of perl6 features (or things inspired by perl6) 18:14
so, no.
jonathan waves to everyone at Frozen Perl
Infinoid isn't at Frozen Perl but he waves back to jonathan anyway :) 18:15
jonathan Going to be back around properly and rakudo hacking on Tuesday. :-)
Infinoid cool. 18:16
18:17 rurban_ joined
jdv79 Well Perl6::Conf requires a parrot so that's not true anymore:) 18:17
Infinoid ok, looks like you're right, that module is written in perl 6 18:18
NotFound: I think the removal of Parrot_*_encoding_ptr and Parrot_*_charset_ptr will require an entry in DEPRECATED.pod and waiting for a release cycle, so I guess there's no hurry to get this checked in
jonathan hopes to fix various masakbugs. ;-) 18:19
NotFound Infinoid: we can replace them with "nice" macros until deprecation cycle finished.
18:20 geof joined
Infinoid ah, that's true (assuming a variable named "interp" is available in all the places where the macros are used) 18:20
NotFound Or we can keep it and make sure we synchronized it with the interpreter insides. 18:21
Infinoid in other words, we can fix everything *except* the issue we're trying to fix? :) 18:23
jdv79 is it languages/perl6 or is it languages/rakudo?
NotFound That's the beauty of deprecation cycles :D
moritz jdv79: tha latter
Infinoid jdv79: rakudo is transitioning at the moment, perl6 is the old version
jdv79 in the svn repo i don't see rakudo
NotFound Infinoid: you mean the non-working version? 18:24
Infinoid yeah, it has recently left the nest, and I'm not sure that fact is documented yet
moritz jdv79: perlgeek.de/blog-en/perl-6/where-ra...lives.html
Infinoid cool. moritz++
NotFound: I mean the existence of the globals themselves 18:25
NotFound Infinoid: there's no way of make something non-existent and keep compatibility with his existence 18:26
jdv79 thanks
NotFound Well, maybe in a quantum computing system is possible X-)
jdv79 will perl6 in the parrot repo go away?
Infinoid eventually, yes 18:27
rurban NotFound: How to solve the TT#276 problem then when we keep the globals?
NotFound Using the macros and accepting the possible problems caused to some people. 18:28
Infinoid getting rid of the globals will require a deprecation cycle, so it can't happen until the next release 18:29
NotFound That is, don't keeping the globals.
Infinoid (or something equivalent to the globals)
NotFound Then we can't solve the problem until that.
rurban fine for me as long 1.0 will have that fixed. or is the cycle 2 releases?
Infinoid it's just adding it to DEPRECATION.pod the day before the release and ripping it out the day after, I think 18:30
rurban AFAIK the linkage against static libparrot caused myproblems, that I cannot prove it. It worked ok on 0.9.0 18:31
Infinoid which, by the way, if 1.0 is next month, that means we'd better find everything we don't want in 1.0 and deprecate it now
NotFound rurban: maybe adding some platform specific qualifier to PARROT_DATA will solve the problem. 18:33
dalek kudo: 3d92f3f | (Francois Perrad)++ | (3 files):
improve .gitignore files
 18:35
kudo: 37e1492 | (Francois Perrad)++ | .gitignore:
ignore *.patch
shorten dalek's url is at xrl.us/befhpo
shorten dalek's url is at xrl.us/befhpq
Coke Infinoid: while I applaud that sentiment, I don't think we have the tuits to rip out the things that are /already/ on the list. 18:39
dalek rrot: r36450 | allison++ | trunk/config/gen/makefiles:
[core] Reverting rejected TT #262.
 18:41
Whiteknight Coke: RT#37578 is resolved. :unique_reg fixes the issue 18:42
pmichaud s/fixes the issue/provides a workable substitute for now/ 18:43
Infinoid Coke: hmm.. I don't plan on adding my bits to DEPRECATED.pod until I've finished polishing the replacement and have staged patches ready 18:44
will that help keep the problem from getting worse?
dalek rrot: r36451 | moritz++ | trunk:
[rakudo] Rakudo left the nest - so remove most of languages/perl6/ and change
 18:47
rrot: r36452 | allison++ | trunk/src/runops_cores.c:
[doc] Adding documentation for runops cores.
 18:53
rrot: r36453 | moritz++ | trunk:
[rakudo] remove a left over perl6.pir file
 18:54
rrot: r36454 | allison++ | trunk/docs/book/ch12_opcodes.pod:
[doc] Moving runcore documentation from book to core project documentation.
 18:55
rrot: r36455 | allison++ | trunk/docs/book:
[doc] Reorganizing the first two chapters of the book for better flow. Removing
 18:56
Whiteknight ECANNIBALIZEBOOK
rrot: r36456 | allison++ | trunk/docs/book:
[doc] Removing the pseudopod reference codas, as we no longer use psuedopod for
 19:06
purl the book is blog.jrock.us/articles/Catalyst%20Book.html or Philippe "BooK" Bruhat, aka Mr Pink or mailto:book@mongueurs.net or awesome or a dino or a black hat
rrot: r36457 | allison++ | trunk/docs/book/README:
[doc] Updating book outline, and notes on required chapter file meta-information.
 19:07
rrot: r36458 | allison++ | trunk/ports/debian:
[debian] Updated descriptions for all Debian Parrot packages.
 19:08
rrot: r36459 | allison++ | trunk/docs/book/appX_patch_submission.pod:
[doc] Moving patch submission guidelines to an appendix.
 19:09
rrot: r36460 | allison++ | trunk/docs/book/appendix.pod:
[doc] Deleting stub appendix that won't be used.
 19:10
Coke cries, as drupal wiki format NE trac wiki format 19:11
pmichaud "drupal wiki format" is really just html, I think.
Coke I am not looking forward to rewriting this article in html. 19:14
moritz Coke: trac renders the article. You could just rip off the rendered article 19:15
rurban allison: TT#276, So how should the required linkflags be added? this breaks solaris 64-bit and all user-defined linkflags. Do we drop solaris support?
Coke rurban: ISTR andy dougherty suggested the proper way to fix that. 19:16
(was it on the same ticket?)
rurban sorry, tt#262 of course. 19:17
NotFound Is in the mailing list 19:18
rurban yes, but the change to link was just added. so using linkflags is the correct way.
19:18 Tene_ joined
Whiteknight it's weird to see other people making lots of edits to the book! It makes me happy 19:19
rurban And breaking things on purpose on name wars is also not fine. 19:20
moritz speaking of the mailing list - does it now work to reply to tickets by mail, and have them recorded by trac?
Coke not sfaik. 19:21
Infinoid nope. trac.parrot.org/parrot/ticket/49 tracks that issue
you can create new ones from email, but replying doesn't work yet
moritz Infinoid: ok, thanks for the update
dalek rrot: r36461 | whiteknight++ | trunk:
[Core] rip out src/bignum.*. It's unused and when we do have a BigNum it will likely be a GMP-based PMC. Rip out associated tests for this too.AAA
 19:22
rrot: r36462 | allison++ | trunk/docs/project/ubuntu_packaging_guide.pod:
[doc] Some additions and cleanups to the Ubuntu packaging guide, made while
Coke pmichaud: there are 4 different ways to input data on drupal; only one of them does what I want, and that involves writing all the html by hand. ah well.
Infinoid Coke: but can you "view source" the trac page and cutpaste? 19:24
Coke I'm not USING trac. I'm using drupal.
dalek rrot: r36463 | NotFound++ | trunk/src/scheduler.c:
[core] less drastic way of handling recursive throwing
Infinoid oh, ok. I misunderstood, it sounded like you were trying to move one to the other
Coke no, I wrote content for one assuming the syntax of the other would work. my bad.
Whiteknight Coke, I'm sure there are some CPAN modules that can do the syntax conversion automagically 19:27
dalek rrot: r36464 | allison++ | trunk/docs/pdds/draft/pdd30_install.pod:
[pdd] Add install location for Parrot docs.
 19:29
rrot: r36465 | NotFound++ | trunk/MANIFEST:
update MANIFEST
 19:30
pmichaud is there a git equivalent to 'svn copy' (src and dest in working dir) 19:31
moritz pmichaud: I think you just copy the file, and git-add it 19:34
pmichaud: it recognizes the duplicity automaticaly
pmichaud that doesn't preserve the history though, does it?
(not that it's terribly important in this case... I'm just curious) 19:35
moritz it does, because each commit is a diff agains the whole tree and all of its history
pmichaud that's confusing.
if I have a.txt
and then I add b.txt
moritz (at least that's what I've been told, never tried it)
pmichaud and b.txt is the same as a.txt
then git causes b.txt to get a.txt's history? 19:36
moritz not sure.. where are the git experts? Infinoid? PerlJam? 19:37
Whiteknight what's the incantation to create a Parrot STRING from a C string?
dalek rrot: r36466 | allison++ | trunk/docs/pdds/pdd07_codingstd.pod:
[doc] Abandoning daft Perl 5-style documentation headings.
pmichaud Parrot_str_new (I think)
pmichaud wow, pbc_to_exe has gotten really slow. 19:42
Whiteknight pmichaud++ # it was Parrot_str_new, thanks!
pmichaud I don't remember it taking this long.
moritz pmichaud: btw all spectests pass for rakudo with the fake executable
pmichaud moritz: is pbc_to_exe slow for you also?
moritz pmichaud: yes, and it has been for weeks or months 19:43
pmichaud hmmm.
Infinoid checks how stgit implements cp 19:44
NotFound pbc_to_exe is not essentiallly doing an hex dump?
pmichaud it is. But it does it in parrot, and that can be slow.
it's a lot of string concats against a very long string.
I thought I had sped it up at one point
NotFound Concats? Will not be simpler to write to a handle? 19:45
pmichaud that might work 19:46
NotFound t/perl/Parrot_Distribution reports an error 19:47
Infinoid the docs indicate cp and add would work, but I can't verify that with blame or log 19:48
Coke www.parrot.org/news/twip-week-parro...y-7th-2009 19:49
shorten Coke's url is at xrl.us/befhvq
pmichaud it's not even a hex dump
it's just an integer dump :-)
but yes, it should be refactored.
Coke (converting format) if I have to re-learn a new format, it's just not worth it. =-)
s/new//
NotFound I tend to add 'hex' without thinking :D 19:50
dalek rrot: r36467 | NotFound++ | trunk/t/pmc/exceptionhandler.t:
[test] unskipping a exception handler test, TT #154
Infinoid NotFound: I only had to change one place in src/jit/i386/core.jit and now JIT works (well, it works as well as it did before the patch)
testj fails 5 tests with or without the patch. That's a good sign, right?
NotFound Infinoid: very good!
Infinoid what problems were you running into? because this seemed like a pretty easy tweak, here
Coke I'm not entirely happy with the resulting TWIP, but I figure I can improve it as we go. Comments on the article will help me figure out what needs updating in future versions. 19:52
Infinoid I'm wondering if I've missed something.
dalek kudo: 46f5a49 | (Moritz Lenz)++ | Configure.pl:
Apply coding style fix by Fran�ois Perrad, fperrad++
Infinoid Coke_afk: reading
rrot: r36468 | whiteknight++ | trunk:
[PMC] add a prototype BigNum PMC, which is basically the BigInt PMC run through a few regexes. Compiles on my system but claims it can't find GMP, needs more work (but doesn't break anything as-is)
shorten dalek's url is at xrl.us/befhvw
Coke_afk (I only went back two weeks so I could include the repository move; if the length stays about the same, we should be stuck with more detail next week.) 19:53
19:55 allison joined
Infinoid basic: Have you had any luck with email2trac and handling replies/followup comments? Is there anything I can do to help? 19:58
nopaste "Infinoid" at 75.28.75.73 pasted "testj failures on linux/x86" (16 lines) at nopaste.snit.ch/15542 19:59
Whiteknight allison: If I wanted to work on some docs, where do you suggest I focus my effort? 20:02
I'm especially interested in TT #237 20:03
Infinoid Coke++ # good info and well presented
allison Whiteknight: for TT #237, I would suggest working on the PIR chapter(s) in docs/book/ 20:04
dalek kudo: cb3422f | (Moritz Lenz)++ | .gitignore:
.gitignore `perl6'
 20:07
shorten dalek's url is at xrl.us/befhwz
NotFound allison: We are trying to get rid of the global variables with pointers to charsets and encodings. Replacing them with macros will be an acceptable deprecation path?
rurban mingw: src\\pmc\\bignum.pmc:128: error: too few arguments to function `__gmpf_get_str'
too old libgmp I assume 20:08
cygwin the same error
Infinoid has no bignum.pmc 20:09
Whiteknight allison: Thanks! Thats where I've been focusing my recent efforts, and I'll just continue with it. 20:10
Infinoid ah, now I do. (updating is a wonderful thing)
Whiteknight urg, this stupid bignum pmc has been a pain in my butt
20:11 riffraff joined
allison Whiteknight: Many parts of docs/book will be integrated into the core documentation anyway, so it has multiple benefits 20:11
rurban cygwin: gmp-4.2.4, mingw: some gmp-4.? 20:12
Whiteknight allison: Do you like the work that's been done in docs/book/ already? I've been putting a lot of effort into it and can always use feedback
rurban: I'm going to kill that damn pmc for now
rurban which version had you?
there should be proper macros
dalek kudo: 71179f1 | (Patrick R. Michaud)++ | (5 files):
Merge branch 'master' of git@github.com:rakudo/rakudo
 20:13
shorten dalek's url is at xrl.us/befhxd
Whiteknight rurban: I just took it out of the listing. it shouldn't be part of the build anymore 20:14
dalek rrot: r36469 | whiteknight++ | trunk/src/pmc/pmc.num:
[PMC] take bignum out of circulation, it's not ready for prime time yet
rurban Whiteknight, it still builds 20:20
I mean same error. 20:21
It must be disabled in config/auto/pmc.pm or remove the .pmc and fix MANIFEST 20:32
dalek kudo: 2c026f7 | (Moritz Lenz)++ | src/parser/ (2 files):
Merge branch 'package-redeclaration' of git://github.com/chrisdolan/rakudo
 20:34
shorten dalek's url is at xrl.us/befhzj
dalek kudo: 1861489 | (Moritz Lenz)++ | .gitignore:
Merge branch 'master' of git@github.com:rakudo/rakudo
shorten dalek's url is at xrl.us/befhzm
allison Whiteknight: yes, lots of good work in docs/book/, much appreciated 20:35
20:40 jdv79 left
nopaste "rurban" at 212.183.63.49 pasted "bignum fix for whiteknight" (15 lines) at nopaste.snit.ch/15543 20:40
NotFound You can also be nice and set the eol-style svn property to free us of those pesky ^M X-) 20:41
nopaste "rurban" at 212.183.63.49 pasted "bigum fix take2" (16 lines) at nopaste.snit.ch/15544 20:44
20:47 mateu joined, Limbic_Region joined
rurban take2 works fine now 20:51
NotFound rurban: I think is simpler to #undef GMP in the file
rurban really? 20:52
purl no, not really
rurban it's still initialized then
moritz purl no, really is <reply>
purl okay, moritz.
NotFound rurban: just one file to touch, and in the meantime in test that it build without gmp
rurban I'also touched only one file :) 20:53
NotFound rurban: but the people that try to fix need to touch more, and reconfigure
rurban can you paste a patch and I'll confirm 20:54
NotFound I'll commit it. It's failing to build, anyway... 20:55
Whiteknight thanks rurban. I'm a config system newbie
rurban I've added just #undef PARROT_HAS_GMP /* not yet ready */ in the .pmc 20:56
unix2dos bignum.pmc would also be nice 20:57
dalek rrot: r36471 | NotFound++ | trunk/src/pmc/bignum.pmc:
[pmc] disable almost completely bignum
rurban sorry, the other way
dalek rrot: r36472 | allison++ | trunk/docs/pdds/README:
[doc] Removing stale and dusty notice about stale and dusty PDDs.
 20:58
rurban typo
rurban Temporariliy disabled
NotFound Fix the code, not my ortography :P 20:59
rurban :) testing now...
dalek rrot: r36473 | allison++ | trunk/MANIFEST:
[cage] Manifest changes for added and removed doc files.
 21:00
NotFound BTW, that splint thing was not fixed geological ages ago? 21:02
rurban t/pmc/pmc 3 fails, t/perl/Parrot_distribution 19 also 21:08
mingw needs some dynpmc+dynoplibs fixes because of the static lib being picked up 21:12
blib_dir => '.' is shared essentially 21:13
blib_dir => '.' if shared essentially
solaris is broken (of course) 21:14
moritz rurban: we have some Makefile problems with Rakudo... it calls 'svn co svn.pugscode.org/pugs/t/spec', but on windows the double slash is substituted by a single slash 21:16
so it's not a valid URL anymore
is there an easy fix for that?
pmichaud I'm fixing it.
the difference is whether rakudo is subdir of parrot or vice-versa.
moritz ok 21:17
pmichaud that's why I want to get rid of the reconfigure.pl approach.
moritz pmichaud: on linux it doesn't seem to make a difference - with current 'master' branch it works in both setups for me 21:18
rurban svn co "url" maybe?
moritz (RT #63082 for your reference)
rurban: if pmichaud++ doesn't come up with a solution i'll try that, thanks 21:19
pmichaud I have a solution.
the original problem is that parrot's makefile builder converts // to /
purl Hmm. No matches for that, pmichaud.
rurban you can disable that
we need an exemption for http:// in the Parrot::Configure::Compiler 21:20
pmichaud it's not going to matter
because compilers typically won't be using Parrot's configure.
dalek rrot: r36474 | NotFound++ | trunk/src/pmc/bignum.pmc:
[pmc] more hacks to cleanly disable bignum
rurban parrot:Distribution: Missing languages/perl6/src/pmc
sorry Parrot_Distribution.t: Missing languages/perl6/src/pmc
NotFound rurban: that was failing before the bignum thing 21:22
dalek kudo: 577566e | (Patrick R. Michaud)++ | (2 files):
More fixes to Configure.pl and makefiles
 21:23
pmichaud 577566e may fix the makefile issues.
shorten dalek's url is at xrl.us/befh43
dalek kudo: 7ab01c5 | (Patrick R. Michaud)++ | (2 files):
Change Configure.pl to use build/Makefile.in.
purl dalek: that doesn't look right
shorten dalek's url is at xrl.us/befh45
dalek rrot: r36475 | allison++ | trunk/docs/pdds/pdd17_pmc.pod:
[doc] Redubbing "PMC" with a new backronym.
pmichaud (it may also introduce new ones)
Whiteknight rurban, what does that fix do that you posted above?
rurban Whiteknight, NotFound fixed it in svn already 21:24
Whiteknight oh, okay. Thanks
NotFound For some value of 'fixed' X-)
moritz rurban: I'm fixing the Parrot_Distribution.t failure now
rurban NotFound: did you test that r36474? looks odd removing the init function 21:25
dalek rrot: r36476 | moritz++ | trunk/t/perl/Parrot_Distribution.t:
[t] remove outdate reference to perl6 source code
rurban well, just one missing test in t/src/pmc.t
NotFound rurban: it builds and pass the test, I thnik we don't need more for a now 21:27
Otherwise, it take less time to make it really work
dalek rrot: r36477 | allison++ | trunk/docs/book:
[doc] Redubbing "PMC"s.
 21:28
GeJ Good morning everyone 21:36
Whiteknight good morning GeJ
GeJ heya Andrew 21:37
How's the week-end going?
Whiteknight it's going well, i'm sitting in my pajamas, in the dark, playing on my computer
21:53 Theory joined, rhr joined
dalek rrot: r36478 | whiteknight++ | trunk/docs/book/ch04_pir_subroutines.pod:
[Book]. Misc small fixes
 21:56
rrot: r36479 | NotFound++ | trunk/src/pmc/bignum.pmc:
[pmc] make a bit more robust the bignum thing
 22:02
22:31 ask joined 22:44 rurban__ joined
dalek rrot: r36480 | rurban++ | trunk:
[config] Enable make reconfig. RT #58034
 22:52
22:53 mikehh joined
mikehh t/native_pbc/integer.t is failing 23:07
Infinoid which platform, mikehh? 23:08
mikehh linux i386
moritz which revision? 23:09
purl which revision are you running?
moritz purl: no, which revision is <reply>
purl OK, moritz.
mikehh All 3 tests fail - smolder.plusthree.com/app/public_pr...ails/17866 23:10
shorten mikehh's url is at xrl.us/befihf
mikehh r36479
moritz works here, with 36480 on i386 linux 23:12
mikehh: care to open a ticket?
mikehh let ne investigate a bit further then ok 23:13
moritz and I'll go a bit further towards my bed ;-)
23:24 elmex joined
dalek rrot: r36481 | allison++ | trunk/src/ops/core.ops:
[cage] Removing TODO for load_bytecode to search library paths, because it's
 23:25
23:26 chromatic joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!