www.parrotcode.org/ | Last release: 0.7.1 "Manu Aloha"
Set by moderator on 17 September 2008.
Whiteknight I found it, it changed names in the move 00:06
00:09 AndyA joined 00:16 bacek joined 00:18 bacek_ joined
dalek r31329 | Whiteknight++ | pdd27mmd: 01:01
: [pdd27mmd] Refactor out the argument passing part of Parrot_pcc_invoke_sub_from_sig_object. Need to perform a similar refactor on Parrot_PCCINVOKE still.
diff: www.parrotvm.org/svn/parrot/revision?rev=31329
01:51 Zaba joined, cognominal joined 01:59 nopaste joined 02:00 Zaba_ joined 02:09 Ontolog joined 02:37 braderunna joined 02:40 tetragon joined 02:42 Theory joined 04:40 tetragon joined 04:54 raiph joined
raiph I'm a newbie to building sw on *nix; I've built parrot 0.7.1 r 31328 on debian Lenny/sid; How do I "Verify parrot is built with a shared libparrot"? 05:01
PerlJam I'm pretty sure that's the default build. 05:02
raiph Right. "# Verify parrot is built with a shared libparrot (this is the default)." But how do I verify? something in myconfig? 05:07
cotto_home ldd ./parrot 05:13
libparrot should show up 05:14
raiph bingo. thanks 05:15
cotto_home np
thanks for playing!
raiph kinda like listing dlls on windows? anyhow, i'm happy :) 05:16
cotto_home yup
05:48 Zaba joined
dalek r31330 | tene++ | exceptionmagic: 05:50
: Create a branch to work on exception enhancements.
diff: www.parrotvm.org/svn/parrot/revision?rev=31330
r31331 | tene++ | exceptionmagic: 06:17
: [exceptionhandler.pmc]
: * Add a min_severity and max_severity attribute
: * Add methods to set and fetch {min,max}_severity
: * Add checks for min_severity and max_severity in EH's can_handle method
diff: www.parrotvm.org/svn/parrot/revision?rev=31331
r31332 | tene++ | exceptionmagic:
: [exceptionhandler]
: * Add a test of min_severity and max_severity
diff: www.parrotvm.org/svn/parrot/revision?rev=31332
Tene is pleased. 06:25
Wait...
purl: karma tene
purl tene has karma of 215
Tene Oh, that's a very nice number.
43*5. not too many factors.
06:27 barney joined 06:31 uniejo joined 07:11 cosimo joined
dalek r31333 | moritz++ | trunk: 07:46
: [rakudo] fix prototype of Any.split(Str), and implement the limit argument.
: Ontolog++, patch courtesy by Chris Davaz
diff: www.parrotvm.org/svn/parrot/revision?rev=31333
07:58 johbar joined 08:35 tetragon joined 08:48 TiMBuS joined
dalek r31334 | kjs++ | trunk: 08:54
: [pirc/new] various bug fixes and improvements. $XY = <op> is now illegal if arg.type is IN/INOUT
: * fix some bugs
: * implement operand direction check: $S0 = print is now illegal.
: * code/documentation improvements
diff: www.parrotvm.org/svn/parrot/revision?rev=31334
09:10 Whiteknight joined 09:34 iblechbot joined 09:42 Zaba joined 10:05 tetragon joined 10:13 gmansi joined 10:17 tetragon joined 10:18 gaz joined 10:19 bacek_ joined 10:22 Zaba joined 10:44 bacek_ joined 10:47 tetragon joined 11:08 Whiteknight joined 11:09 bacek_ joined 11:12 tetragon joined 11:29 bacek_ joined 11:40 Ontolog joined 11:41 Administrator joined 12:04 tetragon joined 12:17 tetragon joined 12:37 AndyA joined 13:08 gryphon joined 13:21 purl joined 13:31 Zaba joined 13:34 PacoLinux joined 13:44 Ontolog joined 13:50 jhorwitz joined, masak joined 13:54 xiaoyafeng joined 14:46 cjfields joined
dalek r31335 | julianalbo++ | trunk: 14:51
: fails when opening a directory in PIO_unix_open RT#57776 cognominal++
diff: www.parrotvm.org/svn/parrot/revision?rev=31335
14:51 cjfields left, cjfields joined 14:52 hercynium joined 15:03 Zaba joined 15:17 Zaba_ joined
dalek r31336 | Whiteknight++ | trunk: 15:28
: [Book] update chapter 6 of the book to also include debugging as well as testing.
diff: www.parrotvm.org/svn/parrot/revision?rev=31336
15:32 Wknight8111 joined
Wknight8111 have two commits by me come through here in the past 5 minutes? 15:32
NotFound Wknight8111: r31336 15:33
moritz one
Wknight8111: in case of doubt, use the irc logs... irclog.perlgeek.de/parrot/today
Wknight8111 I wasn't sure how often the irc logs are updated 15:34
That weird 200 OK handshaking bug confuses the hell out of tortoiseSVN apparently
cognominal someone may be interested to review and possibly apply [perl #59202] AutoReply: [BUG] [PATCH] Perl 6 code "class A::A {}; class A::A {}" crashes parrot
moritz Wknight8111: every line that ilbot2 sees updates the logs
Wknight8111 ok, noted 15:35
cognominal there is still problem doing it a third time, but that is probably an execption handling problem
from analyzing crashes, I will finish by knowing some parts of the parrot C code. 15:36
Wknight8111 (knowing some parts of the parrot C code)++
pmichaud #59202: do not apply patch 15:41
Wknight8111 okay, back to work 15:42
cognominal pmichaud??
purl pmichaud is www.pmichaud.com/ or "Patrick R. Michaud" <mailto:pmichaud@pobox.com> or in charge of toaster experiments
cognominal what is the problem?
pmichaud Parrot does not treat '::' as a class separator. 15:43
so it's wrong to patch it to thinking that it is.
cognominal ok, but one should output some error message. There is still the same problem with A instead of A::A 15:44
pmichaud sure, but that's a rakudo bug, not a parrot one. 15:45
NotFound cognominal: I think a more general solution will be better than checking for specific types. Will take a look on it.
pmichaud I grant that Parrot should not crash in that instance, and to that extent it's a Parrot bug.
cognominal ok 15:46
NotFound Unfortunately, there is still a lot of ways to crash parrot.
pmichaud however, patching Parrot to recognize '::' as a classname separator is definitely the wrong approach. In general, we should not be patching Parrot with Perl-specific behaviors.
cognominal and there is an exception bug if one does class A {} after the error message (when the patch is applied 15:47
pmichaud Rakudo's class generation code needs to be testing if a class already exists. 15:48
currently it leaves that for Parrot to handle, which is the reason we get the various exceptions and crashes
15:48 Hinrik joined
cognominal good 15:49
NotFound get_string() not implemented in class 'ResizableStringArray' --> What ResizableStringArray is used?
cognominal I mean, good that the problem is understood
pmichaud notfound: the split opcode returns a ResizableStringArray
moritz pmichaud: replied to your Str.trans worries - took me a deep dive into the test suite to figure out how to do it 15:50
pmichaud NotFound: so the namespace that gets passed to build the class ends up being a ResizableStringArray
NotFound So the error message in that case is correct. 15:51
pmichaud moritz: I was trying to do it from PIR, though.
At any rate, .trans is still definitely broken.
moritz aye
cognominal nopaste? 15:52
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/
pmichaud I can cheat at .perl with this in the meantime, though.
moritz don't you have a general-purpose escaping function somewhere already?
cjfields what is borken w/ .trans?
nopaste "cognominal" at 82.67.232.89 pasted "addition to /languages/perl6/src/builtins/control.pir" (8 lines) at nopaste.snit.ch/14142 15:53
cjfields (missed that bit)
moritz cjfields: look at pmichaud's last mail in RT #59068
basically you have to know where exactly to put parenthesis to get it to do something useful
on the Perl 6 side 15:54
cjfields ok
NotFound moritz: we have string_esape_string, but I think there is no way to call it from pir 15:55
cjfields Just to note, .trans is a bit off the spec now (this implementation was just prior to spec changes). 15:57
cotto_home NotFound, escape is an opcode wrapper around that function
pmichaud Parrot's escape opcode is all wrong for .perl method on Str 15:59
.trans should be able to work with a hash argument, but doesn't.
NotFound string_escape_string does pir escape. Maybe it needs a more descriptive name, and same for the opcode.
pmichaud $str.trans(%hash)
here's the version of .perl that I have now, which only partially works. 16:00
cjfields (re: .trans) spaces are no longer supposed to be ignored in ranges, and regex matching isn't implemented (though I can work on adding that in now).
pmichaud: ok
nopaste "pmichaud" at 76.183.97.54 pasted "version of Str.perl based on .trans (doesn't quite work)" (14 lines) at nopaste.snit.ch/14143 16:01
NotFound By the wat, I don't think (UINTVAL) ~0 is a good way of saying MAX INTVAL
16:01 Zaba joined
moritz pmichaud: you call it with two lists, not with a Pair of lists 16:02
pmichaud moritz: hmmm?
moritz oh wait, the infix=> should make a pair out of it
cjfields Yup 16:03
pmichaud that's exactly what [ ... ] => [ ... ] is doing in your example :-)
the only difference (which shouldn't be a big difference) is that the key and value are lists instead of arrays
when I use my version of .perl, it works once and then fails
moritz so maybe the caching doesn't work quite right? 16:04
pmichaud I think that .trans might be modifying its argument.
nopaste "pmichaud" at 76.183.97.54 pasted "running .perl using .trans" (6 lines) at nopaste.snit.ch/14144 16:05
pmichaud yes, that appears to be what is happening. 16:09
nopaste "pmichaud" at 76.183.97.54 pasted ".trans modifies its argument" (12 lines) at nopaste.snit.ch/14145 16:10
pmichaud ....opening ticket.
16:11 kj joined
cjfields pmichaud: I'll take a look. I think I know where the problem is. 16:13
Tene pmichaud: I implemented exceptionhandler severity-filtering in exceptionmagic branch last night. Any suggestion on API for type-based? 16:16
pmichaud tene: methods.
purl methods is probably a Hash, you find that. then what breaks?
Tene pmichaud: yes. which methods, though?
I'm thinking .add_type, .set_types([1,2,3]), and maybe something for lookups? 16:17
pmichaud .add_type sounds too generic
Tene maybe just .types(1,2,3), and .types() to get the list of types?
See, that's why I asked. :)
moritz in a German perl forum somebody asked which perl 5 module could be used to embed parrot 16:20
pmichaud I think that anything I suggest is likely to be rejected by allison, so I'd prefer to not expend the effort.
Tene Ah, okay. I'll just do whatever an dlet her clean up after me. :)
moritz neither Inline::Parrot nor Parrot::Interpreter seems to work 16:21
should they?
16:25 Lorn joined
nopaste "cjfields" at 128.174.127.121 pasted "patched .trans" (11 lines) at nopaste.snit.ch/14146 16:26
cjfields pmichaud: added a patch to #59204 which fixes the argument modification; will work on Hash issue 16:27
16:28 clunker3 joined
moritz cjfields++ # helping, and helping fast! 16:28
16:28 Theory joined
cjfields moritz: it was a simple fix, actually. 16:30
one other thing (this popped up te other day)... 16:31
moritz cjfields: nonetheless, response time matters
cjfields Do we need to redefine .'ACCEPTS' in terms of .match? Acc. to S03 this appears to be for Regex 16:32
16:33 sjansen joined
moritz which .ACCEPTS? 16:34
cjfields I would think for Regex, but we don't have a Regex yet do we?
moritz we don't a Regex type, no 16:35
they are just compiled to 'Block's
cjfields That's what I thought.
16:36 Hinrik joined
cjfields perl6: my $x = regex {\\d+}; say $x.WHAT; 16:36
polyglotbot OUTPUT[Blockā¤]
pmichaud eventually ACCEPTS will be redefined in terms of .match for Regex 16:47
but that will need to wait until we have the ability to bless or morph subs into a different type
(i.e., it's a parrot limitation at the moment.)
although istr that DietCoke had some way of doing it for Tcl -- perhaps I could add something similar to P6object. 16:48
16:50 sjansen joined
dalek r31337 | julianalbo++ | trunk: 16:54
: add 'complex' function to pirric
diff: www.parrotvm.org/svn/parrot/revision?rev=31337
16:55 cotto_work joined
dalek r31338 | coke++ | trunk: 17:09
: [tcl] generate the right error message when you can't find a dynamic command; add a test to avoide regression .
: One step closer to figuring out why tcltest.tcl fails to run.
: (code.google.com/p/partcl/issues/detail?id=7)
diff: www.parrotvm.org/svn/parrot/revision?rev=31338
17:40 ruoso joined
dalek r31339 | pmichaud++ | trunk: 17:47
: [rakudo]: Fix so that .trans doesn't modify its argument (RT #59204)
: * Patch courtesy Chris Fields (cjfields++)
diff: www.parrotvm.org/svn/parrot/revision?rev=31339
jonathan pmichaud: Leaving for Russia tomorrow. 17:49
May get a little hacking in, while in Vladivostok. ;-)
pmichaud jonathan: okay, please have a good trip. I'll try not to have changed the entire world on you :-)
but some parts of the world will likely change. :-) 17:50
dalek r31340 | julianalbo++ | trunk: 17:51
: chnage Complex.sqrt method for better precission
diff: www.parrotvm.org/svn/parrot/revision?rev=31340
jonathan Well, hopefully for the better. ;-) 17:52
Especially parsing stuff. :-)
NotFound Now sqrt(-1) is 0+1i :)
jonathan would be very impressed to return to protoregexen. 17:53
pmichaud that's definitely within the realm of possibility :-) 18:03
is there an easy way to convert integer to hex in PIR (short of using sprintf)? 18:05
NotFound pmichaud: Using sprintf. 18:07
paste? 18:08
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 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
nopaste "NotFound" at 213.96.228.50 pasted "Integer to hex" (7 lines) at nopaste.snit.ch/14149 18:09
pmichaud NotFound: yes, I was looking for a way to do it without sprintf. 18:13
but thanks.
Tene Another option would be to reimplement sprintf in pir! 18:15
NotFound Converting to hex is just some arithmetic, no need to reimplement a full sprintf. 18:16
particle1 does pir not have sprintf? 18:17
pmichaud it does, but it seems wierd to have to create an Array just to convert an int to hex
*weird
18:17 Zaba joined
NotFound Wait, there is a get_as_base method in the Integer pmc. 18:22
nopaste "NotFound" at 213.96.228.50 pasted "Integer to hex 2, the revenge of the hex" (7 lines) at nopaste.snit.ch/14150 18:25
pmichaud oh, that's good to know. 18:28
NotFound Somenone here loves retro Basics and has played with pirric? 18:33
dalek r31341 | pmichaud++ | trunk: 18:35
: [rakudo]: Fix Str.perl to better escape things (RT #59068).
diff: www.parrotvm.org/svn/parrot/revision?rev=31341
jhorwitz pmichaud: i want the PIR class "Foo::Bar" to be available to rakudo as "Bar". i'm pretty sure i can use protoobjects for this, but not exactly sure how. ideas? 18:39
18:39 timbunce joined
jhorwitz er, Foo;Bar to be proper.... :) 18:40
NotFound export? 18:42
purl export is a bashism or a kshism
18:49 Ivatar joined 19:00 timbunce left
pmichaud jhorwitz: it's part of the export mechanism 19:00
oh, you're talking about PIR classes.
Yes, this is possible: simply set 'name'=>'Bar' on the call to register or new_class
jhorwitz beautiful. i knew it would be simple. :) 19:01
dalek r31342 | coke++ | trunk: 19:09
: [tcl] when generating procs, declare what namespace they're in. It's not enough to put them into the right namespace; you have to declare it before the anon sub declaration, or the lookups -inside- that proc default to main.
: This fixes a paper-over fix done in a recent commit in this area where we discarded an unused parameter; turns out we needed to actually -use- it.
: Add a regression test.
: code.google.com/p/partcl/issues/detail?id=7
diff: www.parrotvm.org/svn/parrot/revision?rev=31342
19:16 cjfields joined
dalek r31343 | Whiteknight++ | trunk: 19:21
: [Book] A few small changes to chapter 3, more needed.
diff: www.parrotvm.org/svn/parrot/revision?rev=31343
jhorwitz pmichaud: worked great, thanks! 19:23
19:28 timbunce joined
dalek r31344 | kjs++ | trunk: 19:30
: [pirc/new] update the README file.
diff: www.parrotvm.org/svn/parrot/revision?rev=31344
19:32 timbunce left
dalek r31345 | Whiteknight++ | trunk: 19:34
: [Book] Adding in a few more logical section breaks to fill-out the chapter on subroutines. Will add content to them later
diff: www.parrotvm.org/svn/parrot/revision?rev=31345
r31346 | coke++ | trunk: 19:35
: [tcl] implement basic version of [auto_load] and test how it works with parray/info body (used in tcltest)
: code.google.com/p/partcl/issues/detail?id=35 19:36
: code.google.com/p/partcl/issues/detail?id=7
diff: www.parrotvm.org/svn/parrot/revision?rev=31346
confound why are we talking about whether any web stuff uses java? 19:40
dalek r31347 | Whiteknight++ | trunk: 19:45
: [Book] Add in a new chapter 6 (all the other chapters will be renamed +1) to cover the PIR standard library.
diff: www.parrotvm.org/svn/parrot/revision?rev=31347
NotFound confound: I don't even see the point of discussing market share. 19:46
19:48 johbar joined 19:54 rnn joined 19:55 rurban joined
dalek r31348 | Whiteknight++ | trunk: 19:55
: [MANIFEST] update MANIFEST to include new book chapter
diff: www.parrotvm.org/svn/parrot/revision?rev=31348
19:57 silug joined
particle does xlib require -dev pkg to work with parrot? 20:02
NotFound particle: if you talk about the Xlib module, it just requires a xlib library viewable by the dynamic loader. 20:04
20:05 donaldh joined
pmichaud ...pir standard library? 20:06
dalek r31349 | coke++ | trunk:
: [tcl] hackily allow 'auto_load ::parray' to work. With this, we can now parse tcltest.tcl with no errors.
diff: www.parrotvm.org/svn/parrot/revision?rev=31349
NotFound At least that is the assumption, have not tested yet to build a parrot that does not link with xlib.
I don't even know why parrot links with ncurses, xlib and stdc++ 20:08
And readline
purl readline is the default/dumb app
NotFound And crypt, and gmp, and pcre, and... 20:09
20:10 bacek joined
particle if you're building static parrot, you need to link them 20:10
NotFound Maybe, but I don't need static parrot. 20:13
particle they only need to be linked to static parrot anyway
NotFound And looks like I can't make a parrot build that does not link with all stuff without deinstalling all dev packages.
And life is too short to do that. 20:14
rurban we need more and better Configure.pl args 20:16
-Udefine -Adefines -Dkey:value and such 20:17
so you can skip xlib, ncurses, opengl, ...
NotFound Well, there is Configure option to not use opengl, and as far as I know is the only component that may need xlib, but disabling it does not avoid xlib linkage.
donaldh I think the ncurses and xlib are just for PIR examples. 20:18
OpenGL doesn't require xlib.
rurban --without-<lib>
NotFound We have opcodes for runtime linking and NCI, there is no need to link such things at building time. 20:19
rurban or better --without-<feature>, where feature normally is the libname. this should be skipped in the runsteps 20:20
but AFAIK we do not link to xlib
just readline. xlib is loaded via nci (ffi) 20:21
NotFound rurban: ldd says yes
20:24 Zaba_ joined
NotFound By the way, the loadlib and dlfunc functionality need some clarification, because if you fail to load the library and don't check, but the library is already linked, the result is an incorrect code that works. 20:25
The reason is that a PMCNULL and an Undef are taken as a reference to the main program. 20:26
So whatever loadlib returns to flag error, may be valid. 20:27
Mmmm... I was, wrong, using all --without... options Xlib is not linked. 20:39
But ncurses and crypt still are. 20:40
And there is no --without option for pcre. 20:42
Xlib module still works, no confirmed there is no dependence other than the library availability at runtime. 20:43
s/no/so
cotto_work t/distro/file_metadata.t is sad 20:51
moritz aye
cotto_work and by "sad" I mean "not passing"
moritz couldn't we just have a script that fixes all meta data?
cotto_work the current test is one step away from doing that 20:52
rurban sure, here it is: use.perl.org/~rurban/journal/37441
a one-liner: perl t/distro/file_metadata.t 2| perl -ne'system (substr($_,3)) if /^#\\s+svn ps /'
dalek r31350 | moritz++ | trunk: 20:53
: [cage] make meta data tests happy
diff: www.parrotvm.org/svn/parrot/revision?rev=31350
moritz rurban++ 20:54
cotto_work is there any reason that test needs to be chatty?
(apart from test failures)
rurban to be readabale for humans.
and for this special test it was good to be that chatty
20:56 Zaba joined 20:57 Whiteknight joined
NotFound Building --without-opengl X libraries are not linked. Without that option, they are. 20:58
Last time I doed that tests must not be one of my best days, or something changed in between. 20:59
Tene purl: seen kjs? 21:04
purl kjs was last seen on #perl 91 days, 5 hours, 31 minutes and 6 seconds ago, saying: yo [Jun 23 15:33:30 2008]
Tene purl: seen kj?
purl kj was last seen on #parrot 3 days, 8 hours, 9 minutes and 36 seconds ago, saying: isop: you can also use print/say to check out what code is executed [Sep 19 12:55:03 2008]
Tene purl: seen tewk?
purl tewk was last seen on #parrot 10 days, 3 hours, 18 minutes and 22 seconds ago, saying: I thought there was a command you could privmsg to the log bots but I don't remeber what it is. [Sep 12 17:46:22 2008]
NotFound purl: seen purl? 21:06
purl purl was last seen on #sao-paulo.pm 3 days, 23 hours, 56 minutes and 58 seconds ago, saying: Kill them all; God will know his own. [Sep 18 21:09:59 2008]
cognominal no, that's Palin . Armageddon 2010 21:10
21:13 tewk joined
cotto_work xlibtest needs more features 21:13
I want mspaint for Parrot. 21:14
NotFound cotto_work: yes, I'm on it, but slowly. 21:15
dalek r31351 | coke++ | trunk: 21:16
: [tcl] un TODO a recently passing test.
diff: www.parrotvm.org/svn/parrot/revision?rev=31351
r31352 | coke++ | trunk:
: [tcl] a very very basic stub to get [package require tcltest] headed in the right direction
diff: www.parrotvm.org/svn/parrot/revision?rev=31352
NotFound Anyway, low level xlib usage is difficult, tk, gtk+ or qt will give faster results.
cotto_work I wasn't being very serious. 21:17
NotFound About mspaint, me also. About more fatures, I'm serious.
cotto_work It'd be cool to have a full application work in Parrot, but it'll happen eventually if Parrot is worth using.
21:18 rhr joined
NotFound November is a full application, it isn't? 21:19
cotto_work for some value of "full"
21:21 cotto_home joined
rurban_away off to sleep 21:44
21:52 Ademan joined 22:27 Limbic_Region joined 22:57 jan joined 23:06 Zaba_ joined 23:17 tetragon joined 23:29 contingencyplan joined 23:35 contingencyplan joined 23:58 revertTS 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!