Parrot 1.0 Released | parrot.org | 380 RTs left!
Set by moderator on 28 March 2009.
00:03 ilia joined 00:49 braceta joined 00:57 icicled joined 01:15 wayland76 joined 01:39 wickedsmoke joined 01:54 Khisanth joined 02:11 kid51 joined 02:30 Andy joined
dalek rrot: r37816 | jkeenan++ | trunk (2 files):
lib/Parrot/Ops2pm.pm: prepare_real_ops(): Deleting code which tests whether

codes. Not appropriate post 1.0. Cf.:
  trac.parrot.org/parrot/ticket/489.
src/ops/core.ops: Definition of load_language op was (presumably erroneously) tucked inside POD block. Pulled it outside POD. Cf.: TT 489, comments 4 and 5.		 02:33
02:41 janus joined
kid51 mikehh: r37816 should clean up the test failures you reported several hours ago. 02:47
02:48 Theory joined 02:49 ilia joined 02:59 jan joined 03:27 wickedsmoke left 03:31 Andy joined
mikehh kid51: trying it now 03:45
03:49 ilia joined
mikehh kid51: ok that works on the post-config tests - great 03:54
wayland76 Can someone remind me what the process is for getting patches applied? 03:55
I've posted the patch to a trac issue; is there anything else I need to do?
(for example, trac.parrot.org/parrot/ticket/503 ) 03:57
Coke wayland76: depends on the patch. asking about it here might prod someone to read it again. 04:11
I know jack about the spec file, however, so I'll leave it for someone else. sorry.
wayland76 ok, no worries. I'll bring it up here later when others are on. The other patch i was interested in asking about is... 04:12
trac.parrot.org/parrot/ticket/442
(no spec file stuff there :) ) 04:32
04:37 tuxdna joined 05:10 Theory joined
mikehh ok I am now passing all tests on parrot r37816 - perl Configure.pl --test PASS, make smoke PASS, make fulltest PASS 05:44
on Ubuntu Intrepid Amd64
however I am failing t/spec/S29-conversions/ord_and_chr.t on rakudo built on it 05:46
all the rakudo tests passed at 7af829f on r37813 but statred failing at 370dd76 on r37815 which also failed if I built against r37813 05:54
06:27 TiMBuS joined 06:51 tuxdna joined 07:58 wayland76 joined 08:25 masak joined
bacek good evening 08:57
I have a question about TT#190. Is scalar.pmc replacing Ref.pmc? (And why it's "scalar", not "Scalar"?) 08:58
09:30 elmex joined 10:26 alvar joined 10:39 wayland76 joined
wayland76 Hi all. I'm wondering if I need to do anything else to get some patches applied 10:47
The tickets are: trac.parrot.org/parrot/ticket/503 and trac.parrot.org/parrot/ticket/442 10:48
10:50 alvar joined 10:52 ruoso joined
dalek rkdown: 1432d7a | (Francois Perrad)++ | src/parser/grammar.pg:
more strict HTML grammar
 11:02
rkdown: 10c470f | (Francois Perrad)++ | src/ (4 files):
allow HtmlComment as HmtlBlock
shorten dalek's url is at xrl.us/bem37g
rkdown: b6b8c83 | (Francois Perrad)++ | src/ (3 files):
rename node Entity to Html
shorten dalek's url is at xrl.us/bem37i
shorten dalek's url is at xrl.us/bem37k
11:04 alvar_ joined 11:05 alvar_ joined 11:07 ruoso_ joined 11:31 amoc joined 13:01 particle1 joined 13:12 gryphon joined
dalek rrot: r37817 | coke++ | trunk/t/codingstd/c_function_docs.t:
[t] explode single test into one test-per-file-tested.

   as a single test so our test numbers weren't inflated. Now that codetest
   is separate, we can use the more natural idiom.
  - this gives us fine grained control over the todos.
  - explicitly list all the files known to fail
  -- new files held to the standard.
  -- files already passing won't backslide.
  -- now have list of files to address.		 13:36
13:46 ujwalic joined 13:56 PacoLinux joined
dalek rrot: r37818 | coke++ | trunk (2 files):
[t/docs] fix function signatures, file now passes
 13:59
masak how come there are pages like this on the parrotcode.org site? www.parrotcode.org/docs/pdd/pdd04_datatypes.html 14:01
Coke I'm guessing PDD04 used to exist but was removed from the repository. 14:02
moritz masak: probably because parrotcode.org is moved to parrot.org, and nobody really cares about the old site
Coke the old site was fragile that way. 14:03
14:03 ilia joined
Coke tracking file renames and deletions was a manual process. I can fix PDD04. 14:03
moritz just make it redirect to the new site :-)
masak aye.
it's important still because search engines still point to the old site. 14:04
Coke but pdd04 /doesn't exist/
there's nothing to redirect.
I think infinoid was working on a master plan for redirecting. Eventually all requests to parrotcode.org will be automatically redirected to parrot.org; in the meantime I can setup individual redirects for pages, and still do maintenance on the old site. 14:05
Infinoid that "automatic redirect" will be at the DNS level
Coke k. in the meantime, if you want me to zorch pdd04, I can. 14:06
Infinoid (and it's allison's master plan :))
sure. maybe bounce that to docs?
Coke sure. 14:07
masak thanks.
Infinoid Thanks for the reminder btw... I should copy another year of back release newsposts over 14:08
14:08 rdice joined
Coke updated docs/pdd index, removed index file, added http redirect for the file to docs/ 14:08
looks like 14 is also fubar. 14:09
14 is now numbers, was bignum. 14:10
pdd14 now renamed to match the file in the repo- old name forwards to new name. 14:13
masak: feel free to open tickets for the parrotcode.org stuff, though it's mainly stopgap until infinoid finishes his tasks. =-)
masak: fixed all the pdd issues I just found.
masak nod. thanks.
I checked some other pages -- it looks better than last time I complained, a week or two ago. 14:14
Coke Anyone with commit bits looking for a simple task to occupy their time, lemme know. =-) 14:15
14:15 tuxdna joined
dalek kudo: 766b10d | pmichaud++ | docs/spectest-progress.csv:
spectest-progress.csv update: 339 files, 8039 passing, 0 failing
 14:46
shorten dalek's url is at xrl.us/bem4p2
14:49 Theory joined 14:56 wayland76 joined 14:59 mj41 joined 15:07 Psyche^ joined 15:11 Xrblackwe joined
dalek rrot: r37819 | coke++ | trunk (2 files):
[t/docs] fixup doc function signatures
 15:15
15:19 Khisanth joined 15:20 hudnix joined 15:21 rg joined 15:27 PerlJam joined
Coke particle1: ping 15:31
particle i'm here, instead 15:45
Coke particle: can you document floatval_divide_by_zero ? 15:48
in src/datatypes.c (the others have docs with the wrong function sig in that file; that function is missing documentation entirely.) 15:49
particle probably, this week. i need a reminder, though.
ticket me!
Coke there's already an overarching ticket. plenty more to clean up here before I open tickets for each func missing stuff. 15:50
I'll get back to you.
particle ok, thanks! 15:51
Coke yay, macport bug closed. 15:57
user had to upgrade their xcode.
(compiler toolset)
particle easy.
16:19 Khisanth joined 16:24 davidfetter joined
ujwalic Coke: I am unable to use Rakudo from installed parrot 1 on Windows ... I tried PARROT_RUNTIME too .. any help 16:25
moritz ujwalic: unmodified rakudo doesn't work with installed parrot 16:26
ujwalic I used both setup files
16:27 msmatsko_ joined
ujwalic setup-parrot-1.0.0.exe 16:27
setup-parrot-1.0.0-rakudo-15.exe
Coke I'm not a rakudo guy, sorry. moritz is a better guess. =-) 16:28
ujwalic any thing on Tcl
can I use ParTcl with Parrot 1
Coke partcl has been broken for some time. Now that I have a working macport, it's next on my list of big ticket items to fix. 16:29
ujwalic moritz: will setup-parrot-1.0.0-rakudo-15.exe work with setup-parrot-1.0.0.exe
ok 16:30
dalek rrot: r37820 | coke++ | trunk (2 files):
[t/docs] fixup function sigs, add boilerplate for new un-doc'ed function.
 16:31
moritz ujwalic: I know nothing about those installers 16:32
ujwalic moritz: can PARROT_RUNTIME support more than one directory paths 16:33
> perl6 16:35
"load_bytecode" couldn't find file 'PCT.pbc'
current instr.: '' pc 743 (src\\classes\\Object.pir:20)
called from Sub 'parrot;Perl6;Compiler;main' pc -1 ((unknown file):-1)
16:36 rdice joined
ujwalic if I set PARROT_RUNTIME 16:37
16:37 ilia joined
ujwalic [C:\\wrk] 16:37
> set PARROT_RUNTIME=F:\\APP\\DEV\\Parrot\\1.0.0\\lib\\parrot\\library
[C:\\wrk]
> perl6
compiler returned NULL ByteCode 'config.pir' - No such file or directory
current instr.: 'parrot;PCT;HLLCompiler;init' pc 27 (src\\PCT\\HLLCompiler.pir:27)
called from Sub 'parrot;P6metaclass;register' pc 763 (runtime\\parrot\\library\\P6object.pir:390)
called from Sub 'onload' pc 18 (src\\PCT\\HLLCompiler.pir:19) 16:38
called from Sub 'parrot;PAST;Compiler;onload' pc 1184 (src/PAST/Compiler.pir:46)
called from Sub 'parrot;PCT;__onload' pc 2 (PCT.pir:16)
called from Sub '' pc 743 (src\\classes\\Object.pir:20)
called from Sub 'parrot;Perl6;Compiler;main' pc -1 ((unknown file):-1)
Coke nopaste?
clunker3 pasta.test-smoke.org/ or paste.husk.org/ or nopaste.snit.ch:8001/ or rafb.net/paste or poundperl.pastebin.com/ or paste.scsys.co.uk/
purl nopaste is at nopaste.snit.ch/ (ask TonyC for new channels) or rafb.net/paste or poundperl.pastebin.com/ or paste.scsys.co.uk/ or App::Nopaste or tools/dev/nopaste.pl or at www.extpaste.com/ or paste.scsys.co.uk (for #catalyst, #dbix-class, #moose and others)
Coke clunker3--
16:40 jhorwitz joined
moritz is there any compelling reason for clunker3 to be in here? 16:42
nopaste "ujwalic" at 59.93.72.249 pasted "Rakudo 15 and Parrot 1 not working on Win32" (33 lines) at nopaste.snit.ch/16002 16:43
Infinoid moritz: Not that I know of. 16:51
16:53 flh joined
Coke question: should our functions docs name the parameter, or just the type? 16:54
src/nci_test.c goes for just the type, which, upon reflection, seems slightly saner.
Infinoid There are some cases where the name helps the clarity of the description, and some cases where it's perfectly obvious 16:55
Coke ok. leaving it in for now. =-)
Infinoid I don't think the name needs to be a hard requirement, but I can think of some cases where it helps
16:58 ilia joined 17:07 dalek joined 17:12 flh joined 17:29 flh joined 17:30 barney joined
diakopter weird; dalek can't join irc.perl.org 17:31
17:31 dalek joined
diakopter oh, speak of the devil. 17:31
purl Eeeeek! Where?!?!
diakopter purl: shut it. 17:32
purl Make me!
diakopter you asked for it...
17:50 cognominal joined 18:04 ilia joined
barney www.isgithubdown.com/ 18:05
18:08 darbelo joined
Infinoid frantically keeps hitting refresh on that page 18:56
19:01 tuxdna left 19:02 Andy joined 19:05 ilia joined
particle i need to classify some api functions. anyone know of a simple tool for tagging? preferably with a textfile backend. 19:08
Coke your question makes no sense to me, so no. =-) 19:14
19:14 PacoLinux joined
particle i have a textfile with api function names and descriptions. i want to classify them using tags. 19:15
i figure there's some wiki engine or similar that will allow me to tag and search-via-tag
Coke is this for parrot? 19:16
particle it'd be nice to have a text-based backend, so i can modify without the web browser
no, it's for $work
Coke what language?
purl somebody said language was for being a pedantic twit
particle 270 api methods with no rhyme or reason
php, of course
:(
Coke I have you beat. =-)
pmichaud PmWiki can do it. 19:17
particle in the hopes you'd say that, i've just installed it :)
pmichaud its backend isn't pure text, but it's easy to extract the text.
there's also an ImportText recipe that allows pages to be built from pure text.
particle perfect.
purl perfect is the enemy of good enough.
particle now, to configure apache...
coke: yeah, you "win" ;) 19:18
i put pmwiki under htdocs 19:19
do i need to do anything more than tell it to run php docs there?
it=apache
pmichaud you'll need to change permissions on the dir... but PmWiki will tell you about that when you run it. 19:21
particle ok, i see first i need mod_php....
19:22 Tene joined
particle thought i had a different apache distro :( 19:22
pmichaud you need php, yes :-)
cotto Wow. It looks like everybody gets to work with crappy php code.
pmichaud hey, my php code isn't crappy!
particle it's an amazingly crappy language.
cotto pmichaud, I didn't mean to imply that. 19:23
pmichaud :-)
particle oh, look, they have a split function. lemme read the docs.
pmichaud particle: also preg_split
particle hrmm, it says preg_split will be faster.
(me reads preg_split)
which says... explode is faster if you're splitting without regex.
SO WHY HAVE SPLIT?!?! 19:24
pmichaud backwards compatibility.
purl backwards compatibility is probably overrated. or important
cotto If you're asking for a reason, you're probably using the wrong language.
pmichaud lol
I think that's one of the best PHP quotes I've ever heard.
particle is there another language with user-editable online docs? 19:26
should we do that for parrot? ;)
cotto Ooh. Wiki docs would be shiny. 19:27
particle yeah, so poor examples of coding parrot could be forever enshrined. 19:28
cotto I mean an actual wiki, not comments a la php.net.
pmichaud PmWiki's docs are in a wiki. :-) 19:29
cotto not for long...
darbelo particle: You available? I'm the guy that mailed you about BigRat for SoC. 19:33
particle i'm half-available :)
what's up?
purl The Canadian Dollar
particle true, CAD is up two cents on the USD today 19:34
darbelo I'm looking for more details on what parrot needs out of the Big* PMCs 19:36
particle we've been leaning towards ibm's bignum library... decnumber, maybe? 19:37
ja: www.alphaworks.ibm.com/tech/decnumber
i'm almost certain it's licence-compatible with parrot.... 19:38
19:38 amoc joined
pmichaud Ideally Rakudo would like to see Integer PMCs be able to handle arbitrary ints of any size. 19:39
darbelo It is, it has the same license as ICU.
Coke ditto partcl.
particle Rationals of Unusual Size
one of the great things about decnumber is its comprehensive, data-driven test suite.
should make it a snap to convert to parrot tests 19:40
pmichaud if decnumber already has a test suite, I'm generally willing to stipulate that it passes its tests without having to duplicate them all in Parrot.
particle then, we need arbitrary size rational and irrational number support
pmichaud afk for a bit (kids from school) 19:41
particle pmichaud: the test suite conversion would be to check that the parrot implementation works
otherwise, yes, agreed.
darbelo Is it desirable retaining the current GMP BigInt for systems tha don't have decNumber? 19:43
particle i'm not sure of the portability characteristics of gmp or decnumber
i'm happy to settle on the more portable one, and forget the other
since decnumber is ansi c, it's probably the more portable 19:44
cotto Would that make a good GSoC project?
particle it doesn't begin with 'G', which is a plus
yes, it'd be an excellent projcet 19:45
which is why darbelo is here :)
darbelo from what I've seen that's the case. But gmp is more "available" in pre-packaged form.
cotto finally notices the obvious.
particle darbelo: we can package decnumber, if need be.
may have to, for windows... haven't researched that yet 19:46
packaging shouldn't be a gsoc concern, though
the main things are: 19:48
making sure parrot can link to decnumber (preferably on linux/osx/windows, x86/x64, optimized/unoptimized builds)
making sure the tests pass (at least a reasonable subset, full decnumber implementation not strictly necessary)
parrot sprintf can deal with decnumber 19:49
pdd and other docs are updated to reflect new reality
configure subsystem correctly probes for decnumber and errs if missing
i think that's it.
darbelo How up to date is pdd14? 19:52
particle it's slightly off reality
i don't think current NaN/Inf is properly documented
if not, that's my fault, and i can fix
darbelo It claims that all Big* pmcs have "the same internal representation, and differentiates between them only when performing rounding operations." 19:55
Should it stay that way? 19:57
particle if we're moving to decnumber as an implementation, then it should change to match that 19:59
s/an/the/
20:01 Khisanth joined 20:04 rdice joined 20:06 Khisanth joined 20:07 msmatsko joined, gryphon joined 20:08 ilia joined
darbelo If the PMC are dependent on an external library, how shold they handle exceptions behabior from the library? 20:10
Coke presumably the PMC will throw parroty exceptions. 20:14
particle yes
darbelo That is what I was thinking, but since the pdd14 draft claims that part of this "is yet to be decided" I figured I should check. 20:16
it also has a TODO block with "Conversion to and from Parrot strings" and "sprintf output of bignumbers". Any thougts on this? 20:21
Is that supposed to imply more than the "stringification" of the value of the BigNum? 20:27
particle there's also the other way around 20:29
$I0 = "-123456789E123456789"
Coke would that work? or are I registers stuck at native ints? 20:33
I would expect perhaps box $P0, "-123456789E123456789" to work there.
darbelo Telling apart the BigRats is going to require some sort of notation.
particle telling what bigrats apart? 20:34
coke: yeah, that should be $P0 = ..., you're right 20:35
darbelo A BigRats and BigNums (floats) cover the same kind of numbers, but have different semantics in some cases.
particle darbelo: ah, yes. i'm happy to live without BigRat, at least for now. 20:36
darbelo in non-big terms: is 0.5 a float or a rational?
particle 1/2 is rat, 0.5 is float, in my book
Coke there is no non-big rat, izzer?
particle no, we don't have a native rat type.
or, even a core Rat pmc type 20:37
darbelo if BigRats sringify as fractions (1/2 ) they could be confused with a division on the way back. 20:38
moritz they *are* a division, just a delayed one
pmichaud actually, rats are parsed directly now. 20:39
Infinoid mice, too
pmichaud and "on the way back" would be done with .perl
particle checks /topic
there's no .perl in parrot, pmichaud ;) 20:40
pmichaud anyway, I don't have any expectation that native ints have to support bigints.
particle no, i screwed that syntax up. big* are pmc-only 20:41
pmichaud particle: there's a .perl in *my* Parrot :-P
particle they should overflow native IREG and NREG
20:45 particle2 joined
dalek rkdown: e83a324 | (Francois Perrad)++ | markdown.pir:
doc : how use Markdown from Rakudo
 20:51
shorten dalek's url is at xrl.us/bem6a4
dalek rkdown: 66cf8ef | (Francois Perrad)++ | src/parser/grammar.pg:
clean up grammar by using <-[class]>
shorten dalek's url is at xrl.us/bem6a6
20:52 rdice joined 21:01 bsdz joined 21:09 particle2 joined
darbelo so, what is this .perl you were talking about? 21:32
21:33 ilia joined
particle2 in rakudo (perl 6 on parrot) there's a .perl method for stringification 21:34
like perl5's data::dumper
moritz rakudo: say (1, 2, 'foo').perl
where's that evalbot when you need it? 21:35
rg still dead it seems. it's almost a week now :(
moritz feather was rebooted 21:36
PerlJam who runs the bot? 21:38
moritz Tene, I think
21:41 polyglotbot joined
moritz rakudo: say 1 21:41
21:42 polyglotbot joined
darbelo Ah, I see. I was talking about Big* for Parrot, not Rakudo. 21:42
So, .perl is out for me.
rg rakudo: say (1, 2, 'foo').perl 21:43
moritz lolcode: VISIBLE 3 21:45
polyglotbot OUTPUT[Failed to parse sourceā¤current instr.: 'parrot;PCT::HLLCompiler;panic' pc 156 (src/PCT/HLLCompiler.pir:103)ā¤called from Sub 'parrot;PCT::HLLCompiler;parse' pc 589 (src/PCT/HLLCompiler.pir:356)ā¤called from Sub 'parrot;PCT::HLLCompiler;compile' pc 438 (src/PCT/HLLCompiler.pir:291)ā¤called from
..Sub 'parrot;PCT::HLLCompiler;eval' pc 753 (src/P...
moritz perl6: say 3
polyglotbot OUTPUT[3ā¤]
darbelo It might be good for inspiration. How does Rakudo handle rats? 21:46
rg my guess would be as well as parrot does ;)
PerlJam with traps full of cheese
pmichaud rakudo isn't doing rats yet. 21:48
what is t/benchmark ? That test takes _forever_. 21:49
rg does perl6 have a synopsis for rats he could look at?
pmichaud there's not a synopsis for it, no.
personally I'm not all that concerned about Rats as long as we have bigints 21:50
moritz they are mentioned in S02, iirc
rg it does indeed. it says Num may support any precision if it can be done portably and efficiently. 21:54
darbelo Yep. "You can produce a ratio by using C<< infix:<div> >> on two integers"
But I don't see a way to specify rat literals. 21:56
moritz 1 div 2 is a rat literal in Perl 6
pmichaud actually, I think 1/2 is a rat literal 21:57
I think 1 div 2 is still an operation.
moritz 1/2 is also an operation, and one that produces a Num
purl okay, moritz.
pmichaud no, 1/2 is a rat literal. 1 / 2 (note spaces) is an operation
purl okay, pmichaud.
pmichaud see STD.pm, under "number:rational" 21:58
moritz runs awaying crying
darbelo So it ends up as :"1" is an int, "1.0" is a num, "1/1" is a rat. 22:00
pmichaud yes. 22:01
(compilers are free to optimize down and interchange as long as semantics are improved)
and "4+6i" is a Complex
these are all of course Perl 6 semantics (not necessarily Parrot ones), and they're fairly recent, so they may still change 22:02
22:05 Maddingue joined
darbelo I know. But since Rakudo will use parrot's rats it makes sense to keep the semantics similar. 22:06
22:13 Maddingu1 joined 22:18 Maddingue joined 22:25 alvar joined 22:28 alvar joined 22:34 alvar joined
pmichaud when I do an "svn merge" between my pge1 branch and trunk, I get a bunch of files (that I didn't modify in the branch) that show property changes. Anyone encounter that before? 22:34
(I'm preparing to merge pge1 branch back to trunk.)
ehhh, I'll just use a diff. 22:36
rg that will kill your separate commit messages, but i guess it weren't that many. 22:37
pmichaud the "separate commit messages" are either already in svn, or won't be copied anyway by an svn:merge
it changes the file history a bit... but as you say, there aren't that many.
I'd rather deal with the devil I know (commit messages) than the one I don't (weird svn property changes that had nothing to do with what I was working on) 22:38
rg there should be a way to find out, but my svn foo is not that up to date lately :( 22:39
you could probably do an svn revert on the files you know you didn't touch 22:40
Coke_afk finds a few functions where the signature changed but the doc was not updated to follow. 22:52
Coke (most of them are just missing signatures)
22:55 Whiteknight joined
dalek rrot: r37821 | coke++ | trunk (13 files):
[t/docs] fixup more function signatures in docs

present (no docs) and catches a few cases where the actual signature changed, but the docs weren't kept up.		 23:01
rrot: r37822 | coke++ | trunk/src (9 files):
[distro] fixup copyright dates on recently edited files
 23:05
Tene PerlJam: I can give you access to feather3 to fix the bot if you'd like. 23:09
23:15 nopaste joined 23:17 wayland76 joined 23:26 Swaroop joined 23:28 bacek_ joined
wayland76 As it's a different time of day, I'll ask one more time... 23:37
Hi all. I'm wondering if I need to do anything else to get some patches applied
The tickets are: trac.parrot.org/parrot/ticket/503 and trac.parrot.org/parrot/ticket/442
GeJ Good mornin g everyone 23:40
wayland76 'morning :) 23:48
23:50 eternaleye joined 23:53 Swaroop left 23:55 ilia joined

Powered by Cro and the RakuĀ® Programming Language.

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

Thank you!