Camelia, the Perl 6 bug

IRC log for #parrot, 2009-02-15

Parrot | source cross referenced

| Channels | #parrot index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:00 dalek parrot: r36738 | allison++ | trunk/tools/dev/install_files.pl:
00:01 dalek parrot: [install] Handle any compiler files installed in the main package.
00:01 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36738/
00:03 mikehh joined #parrot
00:04 eternaleye joined #parrot
00:05 mikehh Has anyone built parrot on AMD64 Linux with perl Configure.pl --optimize
00:06 rg i believe this is known to not work (segfault in miniparrot)
00:06 TiMBuS ive /tried/ to build it :D
00:07 TiMBuS i think its a segfault in PGE
00:08 mikehh my latest attempt got past PGE but failed in THE
00:08 TiMBuS ill give it a shot
00:08 AndyA joined #parrot
00:09 mikehh sorry that's TGE
00:09 mikehh it builds fine without the optimize
00:10 TiMBuS yep, tge
00:10 mikehh it failed with: FixedIntegerArray: Can't resize!
00:10 Limbic_Region joined #parrot
00:11 TiMBuS i jsut get a segfault =/
00:11 Theory joined #parrot
00:12 mikehh current instr.: 'parrot;P6metaclass;new_class' pc -1 ((unknown file):-1)
00:12 nopaste "rg" at 91.13.166.201 pasted "hints for openbsd to use shared libparrot" (24 lines) at http://nopaste.snit.ch/15626
00:13 mikehh make[1]: *** [TGE/Parser.pir] Error 1
00:13 TiMBuS your error is probably the same as mine except it gets caught
00:13 TiMBuS #0  get_attrib_index_keyed (interp=0x1417080, self=0x0, key=0x15261d0, name=0x141e0a8) at ./src/pmc/object.pmc:81
00:14 TiMBuS thats the backtrace of my segfault. im pretty sure 'self' being 0 is bad
00:14 mikehh the problem being of course that it comes almost at the end of the make
00:15 rg msg rurban openbsd 4.4/i386 is fine at r36736 and with this http://nopaste.snit.ch/15626 also builds shared libparrot (still all tests ok). feel free to commit that patch if you like.
00:15 purl Message for rurban stored.
00:19 mikehh It works fine without the --optimize --> http://smolder.plusthree.com/app/pu​blic_projects/report_details/18030
00:19 shorten mikehh's url is at http://xrl.us/bega2c
00:21 mikehh and it works fine with i38 Linux with the --optimize
00:22 mikehh i386 Linux
00:36 jonathan purl, rakudobug
00:36 purl rakudobug is mailto:rakudobug@perl.org
00:42 jonathan OH HAI
00:42 jonathan Anybody know anything about the following error:
00:42 jonathan Parrot::OpsFile: Unrecognized line: 'inline op rebless_subclass(in PMC, in PMC) :base_core {
00:42 jonathan '!
00:42 jonathan ops2c code generation failed (2304)
00:44 allison jonathan: I know someone else reported it on IRC earlier today
00:45 allison we don't have an opcode rebless_subclass
00:45 allison (I didn't look into it at all this morning)
00:45 jonathan allison: It's a Rakudo dynop.
00:46 allison jonathan: ah, makes sense
00:46 jonathan allison: Though it may be one you might want to ponder for core.
00:46 jonathan I dodn't feel strongly either way.
00:47 jonathan The code gen failure is...annoying though. Hmm.
00:48 allison jonathan: at the moment I'm considering ripping out dynoplibs.pl and replacing it with a makefile (only tangentially related)
00:48 jonathan Was anything deprecated in op syntax recently?
00:48 allison jonathan: no
00:48 jonathan Oddness.
00:48 purl hmmm... oddness is that the icons are on top of the google ads
00:49 allison jonathan: was the dyn ops file changed recently?
00:49 jonathan Not that I'm aware of.
00:49 jonathan I didn't change it. And it's sufficiently terrifying that I doubt anyone else will have. ;-)
00:54 rurban jonathan: ot's a crlf issue
00:54 rurban d2u fixes it
00:54 jonathan Oh, ouch.
00:54 jonathan But that'd explain the issue some more.
00:54 rurban the regex is not crlf aware
00:55 rurban I pastes the line, see the irclogs. no time for a tciekt. tommorrow
00:56 jonathan OK, thanks.
00:56 jonathan rurban++
00:58 Whiteknight allison: I swear I'm not trying to be argumentative with all these emails, I just want to get it Right so long as we are in there monkeying with calling conventions stuff
01:02 dalek parrot: r36739 | whiteknight++ | branches/rename_pccinvoke/src/call/pcc.c:
01:02 dalek parrot: [rename_pccinvoke] registering the sigobject. It was being unregistered later, but never registered. I added this to double-check that the GC knew it existed.
01:02 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36739/
01:06 allison Whiteknight: lol :) well, neither am I
01:06 Whiteknight I really want to get some kind of consistent implentation, but I'm starting to feel like it's a hopeless cause
01:07 allison Whiteknight: well, you're seeking for consistency between to very different things
01:08 allison I completely agree that we need a way to get at the currently invoked code object
01:08 allison but that's not at all the same thing as the current invocant
01:08 jonathan allison: That is through interpinfo op or ParrotInterpreter PMC, no?
01:08 allison trying to collapse the two concepts will always cause confusion
01:08 jonathan Agree.
01:08 Whiteknight right, but then what's our behavior going to be for ".sub 'invoke' :vtable :method"? The invoke vtable override is the thing I'm most concerned with right now
01:09 jonathan Though I really don't want to lose implicit self...
01:09 Whiteknight or even without the :method, if that causes too much confusion
01:09 jonathan Whiteknight: I think for that, the Best Thing would be if we could give it the self parameter it expects.
01:09 jonathan Whiteknight: But that means we need to be able to "insert" an extra argument at the start of the argument set.
01:09 allison that is, self is the first parameter
01:10 jonathan allison: Right.
01:10 jonathan I did want to patch this. Ages back.
01:10 Whiteknight yeah, it's a huge mess
01:10 allison jonathan: unless we just mark that first parameter as the invocant
01:10 jonathan Then I went and look at inter_call.c...and half an hour later decided I'd prefer a beer.
01:10 allison jonathan: yes, it's a mess
01:10 jonathan allison: Basically though, I couldn't see a way to do it.
01:10 Whiteknight My next idea, which is probably better but more intensive, is to not pass "self" as a parameter at all, but to get it in a different way
01:11 allison Whiteknight: this may become easier once we rip out the current code
01:11 jonathan Whiteknight: That'd really get in the way of things.
01:11 kid51 joined #parrot
01:11 allison Whiteknight: actually, that's the current problem, that self isn't being passed as a parameter
01:11 jonathan Whiteknight: A *lot* of things rely on being able to invoke something without caring if it's a :method or not.
01:11 Whiteknight jonathan, there would still be an implicit "self", IMCC would just generate it in a different way internally
01:11 allison Whiteknight: if self were passed as a parameter, all this would go away
01:12 jonathan Whiteknight: No, I mean calling it as $P0($P1) and $P1.$P0() and it being the same call.
01:12 allison the first Pi parameter, as in the new pcc_invoke
01:12 Whiteknight when it is passed at all, it is passed as a parameter
01:12 NotFound We can't have usable functors if they can't be called as functions.
01:12 NotFound And if we can't overriding invoke in pir objects is useless
01:13 allison Whiteknight: but in the calling conventions it's just "current_object", not passed at all
01:13 allison interp->current_object
01:13 jonathan allison: Huh?
01:13 Whiteknight allison: but from PIR it's passed as the first set_param
01:13 jonathan allison: Oh, maybe...
01:13 jonathan :-)
01:13 jonathan allison: OK, I'll just agree with you on "it's a mess". :-)
01:13 Whiteknight now, if we unified it so that PIR acted more like PCCINVOKE, that would be better
01:13 allison Whiteknight: in the new code we added in the new pcc_invoke, it is passed as the first argument
01:14 Whiteknight allison: are you talking about calling from C or calling from PIR?
01:14 allison Whiteknight: curiously, PCCINVOKE is just emulating how PIR calling conventions work
01:15 allison Whiteknight: in which?
01:15 purl somebody said in which was the 3/5 part?
01:15 allison Whiteknight: when I say "new pcc_invoke" I'm talking about calling from C
01:16 Whiteknight allison: I'm talking about calling from PIR, where the invocant is passed using set_param like all the other parameters, and where "self" is treated as the first get_param
01:16 Whiteknight now if PIR kept the invocant separate from the list of parameters, we wouldn't be in this mess
01:17 NotFound Is a Pir2Pir ;)
01:17 jonathan Whiteknight: Thing is, PIR can't always know that...
01:17 Whiteknight jonathan: But what you can do is prepare all the information for every case, and have the callee only read the information that makes sense to it
01:17 allison Whiteknight: no, it's not
01:19 Whiteknight allison: compilers/imcc/pcc.c:390-398.  the invocant is unshifted onto the list of parameters
01:20 Whiteknight no, actually that's where it's pulled off the list in the callee. It's passed on line 791
01:21 allison Whiteknight: src/ops/object.ops:79 the object passed to callmethodcc gets saved as interp->current_object, which is what the rest of the system uses
01:21 Whiteknight that's a separate system
01:22 Whiteknight $P0($P1) and $P1.$P0() can be used interchangably for the most part, and both will put $P1 in "self" if $P0 is a method
01:22 Whiteknight because in both cases, $P1 is passed as the first parameter
01:22 allison Whiteknight: what you're looking at in imcc is only how the 'self' parameter is dummied up
01:23 allison Whiteknight: it's not actually used in invocation
01:23 Whiteknight right, I don't care how the invocant is handled internally, I only care about getting the "self" keyword to be where people expect it to be
01:23 allison $P1.$P0() is callmethodcc
01:24 allison Whiteknight: okay, but self isn't actually a regular parameter, it's a hack
01:24 Whiteknight I know the ops are different, but the result is the same between the two. The sub $P0 is called, and the value $P1 is put into "self"
01:24 allison (and a pretty icky hack at that)
01:24 Whiteknight exactly! Hence my desire to fix it and make it all more consistent!
01:24 allison the way to make it consistent is to make the invocant a regular parameter
01:24 Whiteknight it currently only sort-of works in an arbitrary handful of situations
01:25 Whiteknight I disagree, I think the way to make it consistent, especially considering all the various use cases, is to keep the invocant separate from the list of parameters
01:26 jimmy joined #parrot
01:26 Whiteknight because if we try to call a :method as a regular subroutine, we're going to get an error about having too few parameters passed
01:26 allison Whiteknight: it is separate currently
01:26 Whiteknight not so far as the "Self" keyword is concerned
01:26 Whiteknight everywhere else, yes
01:27 allison Whiteknight: so, remove the 'self' hack
01:27 allison Whiteknight: it's not used at all for invocation
01:27 jonathan allison: I'm with you on the self parameter being a normal parameter.
01:27 jonathan allison: But we do need a way to fix the invoke vtable method.
01:28 allison jonathan: yes, but there are deeper issues there
01:28 Whiteknight jonathan: The way to fix "invoke" is to make "self" not treated like a normal parameter
01:28 jonathan allison: Well, in theory what we *want* to do is easy.
01:28 Andy joined #parrot
01:28 jonathan allison: Just unshift something onto the argumetns list.
01:28 jonathan allison: In reality, that's actually a pain in the arse to achieve at the moment...
01:29 Whiteknight jonathan: it's very easy to unshift the parameter. It's harder to decide *when* to unshift it
01:29 allison jonathan: the problem is that currently the invoke vtable function is what sets up the interpreter environment for execution.
01:29 eternaleye joined #parrot
01:29 Whiteknight the call to $P0() should if it's an Object calling the invoke :vtable, but shouldn't when $P0  is a normal sub
01:30 allison jonathan: that is, overriding it from PIR doesn't make sense (with its current semantics)
01:30 jonathan Whiteknight: The majority of the time it's what we want.
01:30 Whiteknight jonathan: what is what we want?
01:30 jonathan Both of you: overriding invoke in PIR is a special case whichever road you take.
01:30 Whiteknight (I'm losing track of what "it" is)
01:31 Whiteknight jonathan: we can make it more consistent, but there are changes that need to be made to support that
01:31 Whiteknight and there are several different paths we can pursue
01:31 allison jonathan: yes, so we need to figure out how people are actually using it, or are likely to use it and support that
01:31 Whiteknight allison: there is no "actually using it" because it's borked beyond usability right now
01:31 jonathan Whiteknight: If we want $P0.$P1() to be like $P1($P0) - which we really do and it is now - then IMO we want to have the invocant as the first parameter.
01:31 allison jonathan: but, it's always going to be either a prefix or suffix to a real low-level invoke implementation
01:31 Whiteknight jonathan: I don't think I want the two to be the same.
01:32 jonathan Yes, I agree with Whiteknight - every time I wanted to or coulda used vtable invoke, it was too broken to be useful.
01:32 Whiteknight In the first case $P0 is "self". in the second case, $P0 is the first positional parameter
01:32 allison jonathan: er, I don't want $P0.$P1() to be like $P1($P0) to be the same
01:32 Whiteknight I don't want it either, but it is now and that's confusing
01:33 allison one has a Pi argument signature (marked as an invocant) and the other has a regular P signature
01:35 jonathan allison: Erm. They are currently interchangable.
01:35 jonathan allison: If you break that, you're going to cause headaches.
01:35 dalek parrot: r36740 | jkeenan++ | branches/update_pod:
01:35 Whiteknight jonathan: there's more headaches to be had if we let people treat two very different cases in the same way
01:35 dalek parrot: Refactor more code out of two test files and into Parrot::Test::Pod::Util.
01:35 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36740/
01:36 allison jonathan: we can have an option to declare subs that conflate method and sub calls, but shouldn't be the default
01:37 jonathan allison: Why?
01:37 jonathan If your argument is that the invocant is the first parameter (which I agree with fully), why break this?
01:39 Whiteknight the invocant should not be the first parameter, it's a distinct thing
01:39 allison and, third perspective: the invocant is the first parameter, and it's marked in the signature as an invocant
01:40 allison jonathan: because not all languages are as lazy as Perl about what is and isn't an invocant
01:40 allison jonathan: and, it's easier to build in the distinction and relax it for Perl than to try to build the distinction on top of an inherently sloppy invocation pattern
01:41 jonathan allison: So you're saying that if I declare soemthing without an invocant and invoke it as $P0.'foo'() then it'd be an error?
01:41 jonathan (in which case I can see the value in this)
01:42 allison yes, you'd be passing an unexpected parameter
01:42 jonathan OK.
01:43 jonathan So long as we can keep being able to have $P0.$P1() *and* $P1($P0) work for the same thingy, you're not going to upset me too much. :-)
01:43 Andy Are any of you git studs?
01:43 jonathan (Even if we have to add some extra annotation and explicitly declare self to make that happen...)
01:43 jonathan Andy: I'm more of a git donkey...
01:44 Andy I can't figure how to get main trunk stuff for rakudo
01:44 jonathan That's strange, because git is the messiah and all...
01:44 jonathan Andy: I think you want:
01:45 jonathan git clone   git://github.com/rakudo/rakudo.git
01:45 Andy right I have that checkout
01:45 Andy but I want to apply current changes into my branch
01:46 jonathan Oh
01:46 jonathan git pull
01:46 purl i think git pull is slow.
01:46 jonathan purl: Of course it's not!
01:46 purl jonathan: excuse me?
01:46 jonathan Excuse you.
01:46 Andy but that pulls it from my branch
01:47 jonathan Andy: Oh?
01:47 purl Oh are you getting paid better than them?
01:47 jonathan Hmmm...
01:47 jonathan For me that pulls it from the main repo...
01:47 jimmy purl: no, git pull is not slow.
01:47 purl okay, jimmy.
01:47 jonathan I didn't figure out branching under git yet.
01:47 jonathan I only just got it worked out in svn before we moved to git. It'll take me at least a year yet. ;-)
01:48 jonathan Andy: Well, if you want "head", there's always getting a new clone...
01:48 Andy sorry not branch, fork
01:49 * Andy goes back to the git book
01:49 * jonathan wishes he could upload the git book into his brain...
01:49 jonathan I don't doubt there are things about git that are better than svn.
01:49 jonathan But the learning curve is sure steeper.
01:51 Tene Andy: git checkout mybranch; git pull master; should work
01:52 Tene if i understand what you're asking for
01:53 * Tene bbl
01:55 * jonathan -> sleep
02:00 Infinoid Andy: how was your checkout created?
02:01 TiMBuS Andy, pull from the rakudo git url
02:02 Infinoid If you've got a branch (non-master) with some local revisions, but you're still working in a repository that was originally created with "git clone", you probably want: git fetch; git rebase origin
02:06 Infinoid I use that a lot; it's the way I normally work.
02:06 Andy I made my own fork
02:06 Andy I checked out that form
02:06 Andy fork
02:07 Andy http://github.com/petdance/rakudo/tree/master
02:07 Andy I want to get the stuff that's been in http://github.com/rakudo/rakudo/tree/master since I started my fork
02:08 dalek parrot: r36741 | whiteknight++ | branches/rename_pccinvoke/src/io/api.c:
02:08 dalek parrot: [rename_pccinvoke] still getting intermittent test failures, so I'm reverting the update to src/io/api.c for now, which kills the errors definatively
02:08 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36741/
02:10 dalek parrot: r36742 | jkeenan++ | branches/update_pod:
02:10 dalek parrot: Re-code pod_todo.t to use Parrot::Test::Pod::Util.  Test for presence of
02:10 dalek parrot: Pod::Simple in Parrot::Test::Pod::Util rather than in each test file.
02:10 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36742/
02:18 dalek parrot: r36743 | jkeenan++ | branches/update_pod:
02:18 dalek parrot: Change name of Parrot::Test::Pod::Util to Parrot::Test::Pod.
02:18 purl dalek: that doesn't look right
02:18 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36743/
02:26 eternaleye joined #parrot
02:31 Andy AHA
02:31 Andy I have to do a "git remote add"
02:35 dalek parrot: r36744 | jkeenan++ | branches/update_pod:
02:35 dalek parrot: To avoid repeating code, move some auxiliary functions from individual test files to new module Parrot::Test::Pod::Utils.
02:35 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36744/
02:40 dalek parrot: r36745 | jkeenan++ | branches/update_pod/lib/Parrot/Test/Pod.pm:
02:40 dalek parrot: Eliminate trailing whitespace.
02:40 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36745/
02:41 dalek parrot: r36746 | jkeenan++ | branches/update_pod/lib/Parrot/Test/Pod/Utils.pm:
02:41 dalek parrot: Add coda.
02:41 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36746/
02:48 dalek parrot: r36747 | jkeenan++ | trunk/src/call:
02:48 dalek parrot: Add *.obj to svn:ignore list.  Cf.:
02:48 dalek parrot: https://trac.parrot.org/parrot/ticket/327.
02:48 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36747/
03:18 Theory joined #parrot
03:35 TiMBuS Andy, its possible to pull from any repo without having to 'remote add' it.. if i wanted to merge my copy from your branch i'd just use "git pull git://github.com/petdance/rakudo.git master"
03:35 janus joined #parrot
03:35 Andy yeah, but I want to do it all the time.
03:35 Andy but thank you.
03:36 Andy now I can just "git pull main HEAD"
03:36 TiMBuS just letting you know so you dont alias a whole bunch of remote repos :)
03:37 mateu left #parrot
03:44 Andy ok, thanks
04:05 dalek parrot: r36748 | jkeenan++ | failed to fetch changeset:
04:05 dalek parrot: Remove branch whose line of development didn't pan out.
04:05 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36748/
04:19 rurban_ joined #parrot
04:28 dalek parrot: r36749 | allison++ | trunk/DEPRECATED.pod:
04:28 dalek parrot: [cage] Add deprecation notice for crufty build scripts.
04:28 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36749/
04:32 eternaleye joined #parrot
05:25 eternaleye joined #parrot
05:45 mikehh joined #parrot
06:10 Tene joined #parrot
06:15 eternaleye joined #parrot
06:27 eternaleye joined #parrot
07:49 Ademan joined #parrot
08:34 iblechbot joined #parrot
08:43 cognominal_ joined #parrot
08:45 mberends joined #parrot
09:14 alvar joined #parrot
09:37 barney joined #parrot
10:14 dalek parrot: r36750 | rurban++ | trunk/config/init/hints/mswin32.pm:
10:14 dalek parrot: [cage] uppercase cc on MSWin32 possible
10:15 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36750/
10:26 dalek tracwiki: v50 | barney++ | Languages
10:27 dalek tracwiki: https://trac.parrot.org/parr​ot/wiki/Languages?version=50
10:28 rurban barney: nqp is stable, Kea-CL misses several new changes
10:29 rurban c99 is good and stable
10:30 dalek tracwiki: v51 | barney++ | Languages
10:30 dalek tracwiki: https://trac.parrot.org/parr​ot/wiki/Languages?version=51
10:34 dalek tracwiki: v52 | rurban++ | Languages
10:35 dalek tracwiki: https://trac.parrot.org/parr​ot/wiki/Languages?version=52
10:37 dalek tracwiki: v53 | rurban++ | Languages
10:37 dalek tracwiki: https://trac.parrot.org/parr​ot/wiki/Languages?version=53
10:39 dalek tracwiki: v54 | rurban++ | Languages
10:39 dalek tracwiki: https://trac.parrot.org/parr​ot/wiki/Languages?version=54
10:44 * barney back from brunch
10:47 Ademan_ joined #parrot
10:48 rurban hi barney: https://trac.parrot.org/parrot/ticket/339 agreed?
10:50 barney centralising compiler detection for Win32 sounds sane to me
10:51 rurban ccache gcc or ccache cl or CL failed for me
10:52 rurban the importlib case e.g. fails currently on those cc's.
10:54 barney never tried compiling with ccache
11:05 NotFound Parrot_find_language - Find the magic language token for a language, by language name. - return 0; ----> Thats magic!
11:06 dalek parrot: r36751 | fperrad++ | trunk/tools/dev/install_files.pl:
11:06 dalek parrot: [codingstd] remove hard tabs
11:06 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36751/
11:14 barney Can I choose -0 as magic language token?
11:16 dalek tracwiki: v11 | barney++ | LeaveTheNest
11:16 dalek tracwiki: https://trac.parrot.org/parrot​/wiki/LeaveTheNest?version=11
11:16 rurban barney: joking or serious?
11:17 rurban I really want to have backtrace for cygwin/mingw
11:17 dalek parrot: r36752 | barney++ | trunk:
11:18 dalek parrot: [Pipp] Pipp is now at http://github.com/bschmalhofer/pipp
11:18 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36752/
11:19 dalek parrot: r36753 | barney++ | trunk/config/gen/makefiles/languages.in:
11:19 dalek parrot: [languages] add Makte-targets co-fun and up-fun
11:19 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36753/
11:22 dalek parrot: r36754 | barney++ | trunk/config/gen/makefiles/languages.in:
11:22 dalek parrot: [languages] Add co-pipp and up-pipp
11:22 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36754/
11:25 dalek parrot: r36755 | barney++ | trunk:
11:25 dalek parrot: Let svn ignore languages/pipp and languages/fun
11:25 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36755/
11:34 dalek bernhard.schmalhofer@gmx.de | Pipp:
11:34 dalek link: http://www.perlfoundation.​org/parrot/index.cgi?pipp
11:37 dalek parrot: r36756 | barney++ | trunk/NEWS:
11:37 dalek parrot: [docs] Pipp has no support for 'switch' yet
11:37 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36756/
11:48 masak joined #parrot
12:12 rurban NEWS: "Removed (incomplete) Bignum implementation" shouldnt that say "Disabled (incomplete) Bignum implementation"
12:13 rurban NEWS: �Better  "Disabled incomplete BigNum implementation"
12:19 dalek parrot: r36757 | fperrad++ | trunk:
12:19 dalek parrot: [TGE] fix load_bytecode in install tree
12:19 dalek parrot: (pb with 'compilers' in the path)
12:19 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36757/
12:19 rurban_ joined #parrot
12:30 TiMBuS joined #parrot
12:31 barney rurban: wasn't src/bignum.c removed?
12:38 rurban no. src/pmc/bignum.pmc was just disabled. And I got it almost working
12:38 rurban I'm just adding bigint support for bignum.
12:42 dalek parrot: r36758 | rurban++ | trunk/NEWS:
12:42 dalek parrot: BigNum wording
12:42 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36758/
12:43 rurban It never worked anyway. I make it work now.
12:45 barney rurban++
13:02 Whiteknight joined #parrot
13:11 Ademan joined #parrot
13:14 TiMBuS joined #parrot
13:14 particle1 joined #parrot
13:19 barney Pipp is now on github, http://use.perl.org/~Bernhard/journal/38480
13:28 Infinoid Cool, I'll add another rss parser.
13:33 dalek parrot: r36759 | whiteknight++ | branches/rename_pccinvoke/src/io/api.c:
13:33 dalek parrot: [rename_pccinvoke] update the Parrot_PCCINVOKE calls in src/io/api.c that don't use string or PMC args or results. All tests pass here
13:33 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36759/
13:33 barney Infinoid++
13:37 sjn joined #parrot
13:45 kid51 joined #parrot
13:50 dalek parrot: r36760 | barney++ | trunk/include/parrot/packfile.h:
13:50 dalek parrot: [codingstd] satisfy c_macro_args.t
13:50 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36760/
13:50 barney joined #parrot
13:53 * Infinoid installs SUPER.pm on feather
13:54 dalek parrot: r36761 | barney++ | trunk/include/parrot/packfile.h:
13:54 dalek parrot: revert 36760, didn't see the comment in the line above
13:54 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36761/
13:58 mikehh joined #parrot
13:59 Infinoid Yeah.  I already did that once, too.  Sounds like it's still a work in progress.
13:59 rhr joined #parrot
14:07 gravity joined #parrot
14:16 dalek parrot: r36762 | barney++ | trunk/t/codingstd/perlcritic.t:
14:16 dalek parrot: [codingstd] The list of languages for 'make codetest' is empty,
14:17 dalek parrot: add a dummy language, so that the filter doesn't degenerate
14:17 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36762/
14:40 Hinrik joined #parrot
14:41 Tene joined #parrot
14:48 rurban bignum works fine! I'm tuning the tests and interfaces now
15:07 Hinrik joined #parrot
15:21 rafl joined #parrot
15:21 rurban https://trac.parrot.org/parrot/ticket/280 has now bignum
15:34 Hinrik joined #parrot
15:48 dalek tracwiki: v55 | chrisdolan++ | Languages
15:48 dalek tracwiki: https://trac.parrot.org/parr​ot/wiki/Languages?version=55
15:49 dalek tracwiki: v56 | chrisdolan++ | Languages
15:49 dalek tracwiki: https://trac.parrot.org/parr​ot/wiki/Languages?version=56
15:52 dalek tracwiki: v57 | chrisdolan++ | Languages
15:52 dalek tracwiki: https://trac.parrot.org/parr​ot/wiki/Languages?version=57
16:00 rg infinoid: why is the change comment not displayed for wiki changes? and could you append &action=diff to the url?
16:00 Hinrik joined #parrot
16:00 rurban I justthought the same
16:02 dalek parrot: r36763 | jkeenan++ | trunk/t/codingstd/perlcritic.t:
16:02 dalek parrot: Clarify inline comment.
16:02 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36763/
16:02 nopaste "infinoid" at 96.238.213.50 pasted "parrotwikilog.pm plugin source; patches welcome." (104 lines) at http://nopaste.snit.ch/15627
16:03 Infinoid (sorry, I've got limited tuits at the moment and I want to finish updating github stuff first)
16:05 nopaste "infinoid" at 96.238.213.50 pasted "test.pl script for running plugin without bot" (45 lines) at http://nopaste.snit.ch/15628
16:06 rg ok, that's a big help. i'll see what i can do
16:06 * Infinoid is consolidating the github rss parser into a common base module, which is a big help because we're about to have 8 of these
16:06 dalek tracwiki: v58 | rurban++ | Languages
16:06 dalek tracwiki: https://trac.parrot.org/parr​ot/wiki/Languages?version=58
16:07 Limbic_Region joined #parrot
16:24 ask_ joined #parrot
16:28 Infinoid oh my, quite a lot of parrot languages on github
16:30 dalek parrot: r36764 | jkeenan++ | tags/tcif-26373:
16:30 dalek parrot: Branch corresponding to tag has been deleted, so tag may be deleted as well.
16:30 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36764/
16:30 masak quite a lot of Perl 6 projects on github, as well.
16:31 rurban I've put mine to google svn. I prefer svn so far
16:31 Hinrik joined #parrot
16:34 jan joined #parrot
16:40 Infinoid rurban: remind me about that, I've got a googlecode atom feed parser that I can consolidate into a common base module and then add a plugin for your language too
16:40 rurban later. I'll add bignum now
16:40 Infinoid when I get home, there are 7 more github language plugins I will add parsers for
16:40 Infinoid back later &
16:40 rurban most tests pass. just stringification is broken, markled as todo
16:41 nopaste "rg" at 91.13.171.120 pasted "tracwiki parser enhancement" (24 lines) at http://nopaste.snit.ch/15629
16:41 rg oh well. 5 seconds too slow
16:42 rg msg infinoid please check http://nopaste.snit.ch/15629
16:42 purl Message for infinoid stored.
16:43 Infinoid thanks, I will.  sorry, gotta run
16:51 eternaleye joined #parrot
16:56 dalek parrot: r36765 | rurban++ | trunk:
16:57 dalek parrot: TT #280: add bignum (again)
16:57 dalek parrot: - some stringification and overflow tests are failing: todo
16:57 dalek parrot: - bigint inheritence not yet enabled
16:57 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36765/
17:01 alvar joined #parrot
17:03 dalek parrot: r36766 | rurban++ | trunk/t/codingstd/c_macro_args.t:
17:03 dalek parrot: [cage] TRACE_PRINTF exception
17:03 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36766/
17:03 clunker3_ joined #parrot
17:08 mikehh joined #parrot
17:30 eternaleye joined #parrot
17:31 bacek joined #parrot
17:50 jq- joined #parrot
17:53 dalek parrot: r36767 | NotFound++ | trunk:
17:53 dalek parrot: [core] allow Parrot_compile_sub direct usage from embedding, modify embed.t to use it
17:53 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36767/
17:58 eternaleye joined #parrot
18:14 Hinrik joined #parrot
18:30 eternaleye joined #parrot
18:38 mberends joined #parrot
18:39 mberends joined #parrot
18:39 Coke .
18:40 ask_ joined #parrot
18:41 rurban seen kj?
18:41 purl kj was last seen on #parrot 21 hours, 12 minutes and 57 seconds ago, saying: Whiteknight: seems you got another reply on self :-)
18:42 rurban msg kj I'll commit some native_pbc's for you, already for 0.9.1. missing is darwin/ppc (kid51?) and 64-bit BE (not known)
18:42 purl Message for kj stored.
18:50 Hinrik joined #parrot
18:54 particle joined #parrot
18:54 dalek parrot: r36768 | rurban++ | trunk/config/gen/config_pm/myconfig.in:
18:54 dalek parrot: [cage] add longdoublesize to myconfig
18:54 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36768/
18:55 Tene_ joined #parrot
19:08 dalek parrot: r36769 | allison++ | trunk/tools/dev/gen_makefile.pl:
19:08 dalek parrot: [install] Adding a simple tool that generates a single makefile from the
19:08 dalek parrot: configuration data.
19:08 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36769/
19:19 Hinrik joined #parrot
19:47 Andy joined #parrot
20:00 GeJ Good morning everyone
20:00 jrockway joined #parrot
20:06 dalek parrot: r36770 | rurban++ | trunk/t/native_pbc:
20:06 dalek parrot: [t] update release native_pbcs and test, darwin/ppc missing
20:06 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36770/
20:06 jrockway left #parrot
20:07 jrockway joined #parrot
20:07 dalek parrot: r36771 | rurban++ | trunk/tools/dev/mk_native_pbc:
20:07 dalek parrot: [t] more mk_native_pbc fixes
20:08 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36771/
20:09 * Coke accidentally hits submit instead of preview.
20:09 Coke "close enough"
20:09 Coke http://www.parrot.org/news/twip-wee​k-parrot-ending-february-14th-2009
20:09 dalek parrot: r36772 | rurban++ | trunk/src:
20:09 shorten Coke's url is at http://xrl.us/begcz4
20:09 dalek parrot: [core] add more needed 12 + 16 double byte transformers. switch trick by Infinoid.
20:09 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36772/
20:10 rurban Infinoid++ # for the transformer switches and codingstd cleanup
20:13 Andy joined #parrot
20:17 * rurban doing couple of smokes now
20:20 rurban_ joined #parrot
20:29 Andy joined #parrot
20:51 dalek parrot: r36773 | fperrad++ | trunk/tools/dev/mk_inno_language.pl:
20:51 dalek parrot: [inno-setup] LICENSE & .pbc are optional
20:51 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36773/
20:53 dalek parrot: r36774 | fperrad++ | trunk/tools/install:
20:53 dalek parrot: [install] move all languages in smoke_languages.pl
20:53 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36774/
21:01 dalek parrot: r36775 | fperrad++ | trunk/config/gen/makefiles/languages.in:
21:01 dalek parrot: [Pipp] bad copy/paste
21:01 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36775/
21:04 dalek parrot: r36776 | fperrad++ | trunk:
21:04 dalek parrot: [install] filename convention
21:04 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36776/
21:04 eternaleye joined #parrot
21:19 jrockway joined #parrot
21:20 zostay joined #parrot
21:25 Eevee joined #parrot
21:27 Theory joined #parrot
21:27 zostay joined #parrot
21:40 chromatic joined #parrot
21:49 jrockway_ joined #parrot
21:49 HG` joined #parrot
21:56 Coke chip: you're not the autocomplete I expected!
21:56 Coke chromatic: how goes?
22:01 chromatic Should have time today and tomorrow to work on anything necessary for the release.
22:01 chromatic Any suggestions?
22:01 purl Any suggestions are welcome.  (including ripping it out entirely :))
22:04 particle todo/skip review
22:06 riffraff joined #parrot
22:09 chromatic Will do.
22:32 soxet joined #parrot
22:33 rg coke: you might want to use irc.parrot.org (instead of irc.perl.org) in your twip. also, a link to the logs (http://irclog.perlgeek.de/parrot/) might be helpful.
22:38 dalek parrot: r36777 | fperrad++ | trunk/languages/c99:
22:38 dalek parrot: [c99] improve Makefiles
22:38 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36777/
22:39 Coke rg: I might, but there's no functional difference between them. =-) Logs are a good idea. I'll update my template now.
22:39 Coke we need nytprof for parrot.
22:39 Coke someone go make that happen.
22:41 rg i know, i just thought we were trying to separate from perl nameing
22:42 Coke allison is.
22:42 Coke changed in my template so I'll advert the new ones next week. Thanks.
22:43 Coke hurm. I didn't scan the parrotsketch this week. whoops.
22:43 Coke rg++
22:44 rg thanks, just trying to help ;)
22:55 mikehh joined #parrot
23:24 bacek_ joined #parrot
23:39 dalek parrot: r36778 | allison++ | trunk:
23:39 dalek parrot: [install] Some modifications to which files are installed, allowing certain
23:39 dalek parrot: build tools to run outside the build directory.
23:39 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36778/
23:42 dalek parrot: r36779 | allison++ | trunk/config/init/install.pm:
23:42 dalek parrot: [install] Calculate the versioned directory name once during the configure
23:42 dalek parrot: process, and save it.
23:42 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36779/
23:44 dalek parrot: r36780 | allison++ | trunk/lib/Parrot:
23:44 dalek parrot: [config] Make the configuration process smarter about reading the version
23:44 dalek parrot: information from a file. Allow more than one location for the file, so libs can
23:44 dalek parrot: be used from an install.
23:44 dalek parrot: review: https://trac.parrot.org/parrot/changeset/36780/
23:58 bacek joined #parrot

| Channels | #parrot index | Today | | Search | Google Search | Plain-Text | summary

Parrot | source cross referenced