Camelia, the Perl 6 bug

IRC log for #parrot, 2009-08-29

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:01 darbelo Using a lower -j value and avoiding -dp sounds fine to me.
00:02 darbelo I'll eventually find another way to reach a 14+ load average :)
00:13 dalek parrot: r40847 | cotto++ | branches/pluggable_runcore (2 files):
00:13 dalek parrot: [profiling] more cleanup and removal of unnecessary variables
00:13 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40847/
00:30 Zak joined #parrot
00:33 dalek parrot: r40848 | cotto++ | branches/pluggable_runcore/src/runcore/cores.c:
00:33 dalek parrot: [profiling] a few further cleanups and explanatory comments
00:33 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40848/
00:36 mokurai joined #parrot
00:46 jrtayloriv Is there an easy way to generate a random alpha-numeric string of length "n" in PIR?
01:13 dukeleto joined #parrot
01:25 dukeleto 'ello
01:31 dukeleto seen fperrad
01:31 purl fperrad was last seen on purl 6 days, 11 hours, 46 minutes and 41 seconds ago, saying: <private message>  [Aug 22 13:37:56 2009]
01:31 dukeleto that's a bug
01:31 purl No, it's a feature.
01:31 dukeleto purl: you're right ;)
01:31 purl dukeleto: what?
01:33 joeri left #parrot
01:43 darbelo left #parrot
01:47 mokurai joined #parrot
01:52 TiMBuS joined #parrot
02:00 bubaflub joined #parrot
02:33 jrtayloriv Can I somehow unlink/delete a file through a Parrot IO object, or do I have to keep the path around and just use 'unlink filename'? Is there a way to get the filename from a Parrot IO obj?
02:34 Ryan52 gah a few months ago I was looking for allison on irc because I wanted to have a discussion with her, and then I changed my mind, and I've lurked the channel and seen her around a lot, and now I've changed my mind back, and she's not here :/
02:34 Ryan52 anybody know when she's usually around nowadays?
02:36 cotto Ryan52, you may have to lurk a while before she shows up.  You'll have better luck emailing her.
02:36 janus joined #parrot
02:36 cotto If it's a generic Parrot question, there's no harm in asking it here and seeing if anyone can answer it.
02:36 Ryan52 I *hate* email. =P
02:38 Ryan52 anyways, it's wrt the packaging of parrot and rakudo in Debian.
02:38 Ryan52 which I think is something I should discuss with her specifically, since from what I see she's already done a lot of work on it.
02:45 dukeleto Ryan52: you can leave a message
02:45 dukeleto msg Ryan52 this is your message
02:45 purl Message for ryan52 stored.
02:45 Ryan52 left #parrot
02:45 Ryan52 joined #parrot
02:45 Ryan52 .
02:45 Ryan52 ah.
02:45 Ryan52 how do I retrieve my message?
02:47 mokurai joined #parrot
02:52 Zak joined #parrot
02:54 jdv79 messages
02:56 cotto massages
02:56 purl You have no massages waiting.
03:02 dukeleto purl, i need a massage
03:02 purl dukeleto: what?
03:04 cotto Massages are great, but I'm not sure I'd turn my back on purl.
03:04 dukeleto cotto++
03:05 bubaflub_ joined #parrot
03:45 mokurai joined #parrot
04:02 mokurai joined #parrot
04:11 mokurai joined #parrot
04:30 mokurai joined #parrot
04:44 dalek TT #957 created by jrtayloriv++: [PATCH]: Remove need_ext from config/gen/crypto/digest_pmc.in
04:47 cognominal joined #parrot
04:48 quek joined #parrot
04:56 cotto jrtayloriv, you can nopaste tiny patches like that here in the future.
04:56 jrtayloriv cotto, OK -- sorry
04:57 cotto don't worry about it
04:58 Makoryu joined #parrot
05:04 dalek parrot: r40849 | cotto++ | trunk/config/gen/crypto/digest_pmc.in:
05:04 dalek parrot: [dynpmc] remove obsolete need_ext from digest dynpmc template, jrtayloriv++ for the patch
05:04 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40849/
05:04 dalek TT #957 closed by cotto++: [PATCH]: Remove need_ext from config/gen/crypto/digest_pmc.in
05:12 kyle_l5l joined #parrot
05:12 nopaste "jrtayloriv" at 71.188.232.187 pasted "PATCH: docs/debugger.pod ( I don't think this is true anymore ...)" (23 lines) at http://nopaste.snit.ch/17732
05:13 Makoryu Hey folks. Is there a Parrot implementation of J (or K, or anything else based on APL)?
05:14 jrtayloriv at least, I always seem to have parrot_debugger when I make parrot
05:15 jrtayloriv Makoryu, http://www.google.com/search?q=parrot+apl
05:16 Makoryu jrtayloriv: I'm aware there's an APL implementation. The names of APL derivatives are a bit harder to google for, however ;)
05:24 michel__ joined #parrot
05:25 jrtayloriv Makoryu, sorry about that then -- I thought you meant you were just looking for anything related to APL
05:26 cotto joined #parrot
05:29 mokurai joined #parrot
05:29 Makoryu jrtayloriv: I like the idea of APL, but... The encoding issue can be such a pain
05:49 dukeleto 'ello
05:50 jrtayloriv howdy
05:51 nopaste "jrtayloriv" at 71.188.232.187 pasted "[Patch]: Minor doc fixes" (77 lines) at http://nopaste.snit.ch/17733
05:52 jrtayloriv dukeleto, Think those are reasonable?
05:53 quek left #parrot
05:56 dukeleto jrtayloriv: i think it would be better if you mentioned that parrot_debugger is built with the normal "make" command, but if you want to compile *only* the parrot_debugger, you can say "make parrot_debugger"
05:56 jrtayloriv ok I'll fix that
06:01 dukeleto jrtayloriv: and my other question would be, did the "kill unionval" branch land?
06:01 dukeleto i see it at r40726
06:02 jrtayloriv dukeleto, Sorry, I don't understand what you mean by land.
06:03 jrtayloriv I am using r40849
06:03 dukeleto jrtayloriv: i meant, was it merged into trunk? and then I found out that it was, in rev 40726 :)
06:03 dukeleto so yes, your patch is a necessary update to the docs
06:05 nopaste "jrtayloriv" at 71.188.232.187 pasted "[Patch]: doc fixes" (82 lines) at http://nopaste.snit.ch/17734
06:05 jrtayloriv dukeleto, how about that (re: the debugger bit)
06:08 dukeleto jrtayloriv: the debugger stuff looks fine, but I think the indenting of the pdd17 stuff will break the rendering of the POD
06:09 jrtayloriv You mean I should just add spacing to make the PMC struct line up w/ the old one?
06:10 dukeleto jrtayloriv: yes. the indentation level of POD effects how it looks. run perldoc on the file to take a look
06:10 jrtayloriv yes -- it did break the formatting ... and I'll remove the /*properties*/ comment from the metadata field as well for readability since it's described right below anyway
06:16 nopaste "jrtayloriv" at 71.188.232.187 pasted "[Patch]: doc fixes" (85 lines) at http://nopaste.snit.ch/17735
06:16 jrtayloriv I also added a better description (I think) of the _metadata field for struct PMC
06:24 jrtayloriv dukeleto, re: #955, why do you think the tempfile version of 'open' should be a dynop, rather than a builtin?
06:26 Zak joined #parrot
06:26 jrtayloriv The only criteria I found for judging was that a built-in should be "sufficiently simple and fundamental, but at the same time are very common" (http://search.cpan.org/~arandal/parrot-​1.0.0/docs/book/ch12_opcodes.pod#Dynops) ... but that's kind of fuzzy.
06:26 dukeleto jrtayloriv: i didn't say it should. I asked if it should, since I assume that other people more experienced would know the pro/cons. If it was not deemed core-worthy, then it would probably be a dynop. If it was core-worthy, then a builtin.
06:28 dukeleto jrtayloriv: i like your patch, except it looks like you have deleted the only mention of the deprecation of _synchronized. perhaps it would be better to leave it there
06:29 jrtayloriv OK -- didn't mean to do that actually. I'll fix it.
06:33 nopaste "jrtayloriv" at 71.188.232.187 pasted "[Patch]: doc fixes" (85 lines) at http://nopaste.snit.ch/17737
06:45 dalek parrot: r40850 | dukeleto++ | trunk/docs (2 files):
06:45 dalek parrot: [t] Update docs regards building parrot_debugger and that UnionVal has been removed from the PMC struct, jrtayloriv++
06:45 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40850/
06:48 dukeleto jrtayloriv: nice work
06:49 jrtayloriv dukeleto, thank you for helping me clean it up.
06:55 jrtayloriv dukeleto, in the file src/ops/io.ops for the op "open", what does the :filesys_open mean?
07:00 cotto The best I can come up with is that those flags had some intended purpose that was (mostly) never implemented.
07:01 jrtayloriv cotto, I think I found something related --> perldoc Opcode
07:01 jrtayloriv something to do w/ "opcode masks"
07:05 cotto jrtayloriv++.  That explains where those came from, even if they're not used extensively in Parrot.
07:06 jrtayloriv cotto, Yes -- lib/Parrot/Op*.pm are also providing a lot of info -- just do --> grep -rn ':flag' lib/Parrot/
07:15 chromatic joined #parrot
07:30 jrtayloriv Anyone else here having trouble accessing parrot.org ? I can't get to trac or svn ...
07:31 kyle_l5l jrtayloriv, www. seems to be timing out here.
07:33 treed I'm having iffy access to it.
07:33 treed One machine times out at corv-car1-gw.nero.net
07:33 treed the other got through
07:33 treed (Different ISPs)
07:40 jrtayloriv oh well it's about bedtime for me anyhow -- good night all
07:40 cotto jrtayloriv, what's your timezone?
07:41 jrtayloriv cotto, US/Eastern
07:42 cotto yup.  bedtime
07:42 jrtayloriv ;)
07:43 cotto they must be doing some maintenance at osu's osl where parrot.org is hosted
08:12 quek joined #parrot
08:31 HG` joined #parrot
08:52 quek left #parrot
09:06 dalek decnum-dynpmcs: r180 | darbelo++ | trunk/src/pmc/dec (2 files):
09:06 dalek decnum-dynpmcs: Let's leak some memory to get our tests passing again. Aren't order of
09:06 dalek decnum-dynpmcs: destruction bugs fun?
09:06 dalek decnum-dynpmcs: review: http://code.google.com/p/decnu​m-dynpmcs/source/detail?r=180
09:23 TiMBuS does rakudo have a simple way to convert a regex result from a hierarchy of Match objects into just a hash?
09:24 TiMBuS its weird to pass around a big match tree (plus i cant seem flatten it into arguments for subs)
09:38 darbelo joined #parrot
09:39 darbelo Damm. dalek is slooooooow today. I commited r180 before going to bed.
09:46 szbalint joined #parrot
10:37 iblechbot joined #parrot
10:48 MoC joined #parrot
11:25 MoC Hrm, parrot svn down?
11:27 mikehh I'm not getting a response at the moment
11:27 darbelo MoC: looks like all of parrot.org is down right now
11:34 mikehh darbello: decnum_dynpmcs - make test PASS no probs
11:35 mikehh dammit I keep doubling the l
11:35 mikehh darbelo: decnum_dynpmcs - make test PASS no probs
11:36 mikehh 0I must be thinking of bello or something :-}
11:37 mikehh still no response :-{
11:38 mikehh I'm a waiting on an svn update
11:42 darbelo mikehh: I expected that. The bug is still there, it will segfault on calling the destroy VTABLE, I just hid it by leaking the memory.
11:44 darbelo there's a comment explaining it in decnumcontext.pmc
11:49 darbelo mikehh: It seems you are the only one testing this so I should give you a heads up: I'm thinking of rewriting the Configure script in pir, so 'perl Configure.pl' will change into 'parrot Configure.pir' in the near future.
12:11 quek joined #parrot
12:15 mikehh darbelo: an interesting thought - not sure how practical it is - all the other configure scripts use Configure.pl and it might be wise to maintain compatability - but go for it
12:16 mikehh darbelo: I looked at http://speleotrove.com/decimal/ foir the first time in a while and it has been updated a lot since I last looked
12:18 mikehh darbelo: in fact spent far too many hours perusing it
12:18 darbelo mikehh: It's an experiment, someday I'd like to depend on just parrot and a few basic tools, no perl at all.
12:19 mikehh that's the general idea for around 2.5/3.0 or something
12:20 darbelo mikehh: Yeah. But it will be easier to manage for parrot if somebody attempts it first on a smaller scale.
12:20 mikehh still no response from www.parot.org :-{
12:21 jrtayloriv joined #parrot
12:21 mikehh darbelo: that's why I suggested - go for it, but I would recommend you retain the Configure.pl as an option
12:24 darbelo Depending on my mood I might make it a simple compatibility wrapper "system('parrot', 'Configure.pir')" :)
12:24 mikehh darbelo: that way you can say - Hey look my Configure.pir produces EXAXTLY thye same results as Configure.pl
12:24 mikehh EXACTLY
12:27 darbelo OTOH my Configure.pl has diverged a bit from the original implementation.
12:29 mikehh BTW I like the way the partcl Configure.pl reports the revision you used - I think some of the others (like rakudo) could use this - it certainly help me as I have a lot of different versions around
12:30 mikehh helps
12:33 mikehh anyway as I can't update parrot at the momennt I think I will take dog out and go to the store - bbl
12:44 jrtayloriv joined #parrot
13:20 jrtayloriv joined #parrot
13:24 payload joined #parrot
13:31 quek left #parrot
13:46 JimmyZ joined #parrot
13:47 joeri joined #parrot
14:07 JimmyZ parrot.org is too slow.
14:07 jrtayloriv JimmyZ, It's down right now for a lot of people.
14:08 jrtayloriv someone earlier said that they were able to access it via one ISP, but not another.
14:20 Psyche^ joined #parrot
14:25 joeri left #parrot
14:32 JimmyZ ...
14:49 jhorwitz joined #parrot
15:01 jhorwitz just got here...anyone else having issues w/ {svn,www}.parrot.org?
15:04 jhorwitz i'll send a message to the OSUOSL admins
15:08 theory joined #parrot
15:16 whoppix joined #parrot
15:23 jhorwitz ah, things appear to be back online
15:38 jrtayloriv joined #parrot
15:47 yath ehm
15:47 yath compilers/pirc/src/pir.l says "DO NOT EDIT THIS FILE"
15:47 yath and refers to compilers/pirc/new/pir.l
15:48 yath but, eh, there aren't that many .l files in parrot...
15:49 yath or is this meant to be to included in the generated c file?
15:51 yath hm, seems so
16:02 payload parrot.org is down?
16:02 payload hmm :/
16:03 payload can someone put a recent clone on gitorious.org? :D
16:03 jrtayloriv joined #parrot
16:03 jhorwitz payload: i have the admins looking into it now (parrot.org).
16:04 jrtayloriv jhorwitz, seems to be working for me now -- thanks
16:05 jrtayloriv (both www and svn)
16:05 jhorwitz yeah, trac was still down a few mins ago.
16:05 jhorwitz everything else seems fine
16:07 mikehh All tests PASS (pre/post-config, smoke, nqp_test, fulltest) at r40850 - Ubuntu 9.04 amd64 (g++)
16:10 dalek parrot: r40851 | NotFound++ | trunk (3 files):
16:10 dalek parrot: [core] more refactoring of pmc destruction
16:10 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40851/
16:10 dalek parrot: r40852 | NotFound++ | trunk/tools/dev/pbc_to_exe.pir:
16:10 dalek parrot: init stacktop in fakecutables
16:10 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40852/
16:11 iblechbot joined #parrot
16:17 mikehh rakudo (7666e92) builds on parrot r40850, make test / make spectest (up to r28110) PASS - Ubuntu 9.04 amd64 (g++)
16:24 mikehh partcl r647 builds on parrot r40850 - make test - same 6 tests fail but all subtests PASS - Ubuntu 9.04 amd64 (g++)
16:26 dalek decnum-dynpmcs: r181 | darbelo++ | trunk/src/pmc/dec (2 files):
16:26 dalek decnum-dynpmcs: Convert DecNum and DecInt to use auto_attrs.
16:26 dalek decnum-dynpmcs: review: http://code.google.com/p/decnu​m-dynpmcs/source/detail?r=181
16:29 darbelo dalek is not being a speedy bot today, I'll tell you that.
16:32 cognominal it seems that parrot has problems to build with snow leopard
16:33 cognominal apparently, some test gave a crash. Too bad I dismissed the pop up
16:34 cognominal and downstream I get :
16:34 cognominal error:imcc:The opcode 'getattribute_s_p_sc' (getattribute<3>) was not found. Check the type and number of the arguments
16:34 cognominal in file 'compilers/pge/PGE/Exp.pir' line 1345
16:35 cognominal some deprceation messages fo glut too : AVAILABLE_MAC_OS_X_VERSION_10_2_AND_LATER​_BUT_DEPRECATED_IN_MAC_OS_X_VERSION_10_5
16:36 mikehh decnum_dynpmcs r181 builds on parrot r40850 - make test PASS - Ubuntu 9.04 amd64 (g++)
16:37 mikehh cardinal builds on parrot r40850 - make test same as r40842 - Ubuntu 9.04 amd64 (g++)
16:44 cognominal 0   test_45972                    0x0000000100000ee6 Parrot_memcpy_aligned_mmx_code + 6
16:47 ntgrl joined #parrot
16:58 yath is PAST and POST only used by the PCT or by parrot itself?
16:58 yath +also
16:59 jrtayloriv Why do parrot_string_t and buffer_t use 'void *' while struct PMC uses DPOINTER? Should parrot_string_t and buffer_t be using DPOINTER as well?
17:01 dalek TT #936 reopened by jkeenan++: "make install" doesn't install PCT
17:04 mokurai joined #parrot
17:04 dalek parrot: r40853 | jkeenan++ | trunk/MANIFEST.generated:
17:04 dalek parrot: Merge tt936_install_pct branch into trunk.  Changes metadata for PCT-related files so that PCT gets installed with 'make install' rather than 'make install-dev'.  See https://trac.parrot.org/parrot/ticket/936.
17:04 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40853/
17:06 HG` joined #parrot
17:08 cognominal apparently parrotbug ate my report :(
17:22 Whiteknight joined #parrot
17:25 dalek TT #936 closed by jkeenan++: "make install" doesn't install PCT
17:25 dalek TT #958 created by jrtayloriv++: [PATCH]: Doc fixes related to GC and UnionVal
17:26 kid51 joined #parrot
17:28 dalek parrot: r40854 | jkeenan++ | branches/tt936_install_pct:
17:28 dalek parrot: Branch has been merged into trunk and is no longer needed at HEAD.
17:28 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40854/
17:32 mokurai joined #parrot
17:38 Whiteknight hello #parrot
17:38 jrtayloriv hoody hoo
17:38 Whiteknight jrtayloriv: you're right, need_ext is obsolete now
17:39 jrtayloriv Whiteknight, yes -- cotto already applied the patch to remove it
17:40 Whiteknight awesome. jrtayloriv++
17:45 dukeleto 'ello
17:45 Whiteknight hello dukeleto
17:45 dukeleto cognominal: did you find a bug in parrotbug?
17:45 dukeleto Whiteknight: how goes it?
17:46 dukeleto msg chromatic refactor of your config_lib.pasm updating script: http://nopaste.snit.ch/17739
17:46 purl Message for chromatic stored.
17:47 chromatic joined #parrot
17:53 HG` joined #parrot
17:54 chromatic duk3leto, I think you mean     chomp( $revision = `git svn find-rev HEAD` );
17:55 yath hm. say, does anyone want to enlighten me (not having a degree in CS) why a register-based VM is better than a stack-based one?
17:55 treed http://www.sagecertification.org/eve​nts/vee05/full_papers/p153-yunhe.pdf
17:56 Whiteknight register-based requires fewer instructions to manipulate the order of things on the stack
17:56 yath treed: ha, thanks
17:56 dukeleto chromatic: it wasn't my refactor, just passing it along
17:56 treed (I actually have no idea if that paper comes to that conclusion.)
17:56 dukeleto chromatic: but I agree :)
17:56 Whiteknight and less logic pushings things and popping them
17:56 Whiteknight it's really not a "better" so much as a "different"
17:56 yath Whiteknight: but why did people build stack-based VMs in the first place?
17:56 yath +then
17:56 Whiteknight yath: easier to manage a single stack instead of dynamic register sets
17:56 treed (Just a PDF I found on the subject just now.
17:57 yath Whiteknight: ah, hm
17:57 dukeleto yath: the Factor language is probably the best modern stack-based vm. it makes sense, because it is a stack-based language
17:57 dukeleto yath: factorcode.org
17:58 chromatic Stack VMs are really easy to write.
17:58 yath dukeleto: hm, looks a bit like erlang :)
17:58 yath and what about optimization stuff?
17:59 yath i mean: why didn't, say, java, changed to using a register-based VM?
18:00 treed (The paper does, in fact, come to that conclusion.)
18:00 yath okay, i'll read it then :)
18:00 yath thanks, treed
18:01 mberends joined #parrot
18:01 treed We found that a register architecture requires an average
18:01 treed of 47% fewer executed VM instructions, and that the re
18:01 treed sulting register code is 25% larger than the correpsonding
18:01 treed stack code. The increased cost of fetching more VM code
18:01 chromatic http://www.inf.puc-rio.br/​~roberto/docs/MCC15-04.pdf
18:05 dukeleto did trac's ssl cert change?
18:06 cotto looks like it expired
18:10 cotto Whiteknight, ping
18:12 dalek parrot: r40855 | cotto++ | branches/pluggable_runcore (2 files):
18:12 dalek parrot: [profiling] catch the edge case where the file changes on an inner runloop
18:12 dalek parrot: also, use more STRINGs internally, try (and fail) to abuse ->strstart less and print output filename on interp exit
18:12 dalek parrot: review: https://trac.parrot.org/parrot/changeset/40855/
18:15 payload joined #parrot
18:16 Whiteknight cotto: pong
18:16 cotto Whiteknight, interesting GC misbehavior - https://trac.parrot.org/parrot/bro​wser/branches/pluggable_runcore/sr​c/runcore/cores.c?rev=40855#L1093
18:17 Whiteknight okay, I'll look at it but am pretty busy today
18:17 cotto there's a very easy workaround, so don't worry about it too much
18:24 cotto It's very hard to reconcile laziness with avoiding ->strstart .
18:30 dukeleto cotto: the debugger uses ->strstart all over
18:30 dukeleto cotto: chromatic told me it was evil, but I don't quite understand why.
18:33 Whiteknight probably because strings are evil, at least as they are implemented now
18:33 jrtayloriv dukeleto, I saw something in the docs for pdd28_strings.pod that said: {{DEPRECATION NOTE: the C<char *> pointer C<strstart> will be removed. It
18:33 jrtayloriv complicates the entire string subsystem for a tiny optimization on substring
18:33 jrtayloriv operations, and offset math is messy with encodings that aren't byte-based.
18:33 jrtayloriv oops --sorry
18:34 cotto It'll be a huge pain if anyone gets ambitious enough to do a string refactor.
18:36 dukeleto cotto: so you are saying that ->strstart should be abolished from our codebase?
18:37 cotto we should be using something that allows encapsulation, e.g. Parrot_str_to_cstring in non-string code instead of depending on a STRING's internal structure.
18:38 cotto It wouldn't be so bad except that the value returned from Parrot_str_to_cstring has to be freed.
18:39 dukeleto cotto: perhaps a less carpal-tunnel-inducing pair of macros can be made to emulate the laziness of ->strstart ?
18:39 cotto (assuming that we don't like memory leaks)
18:40 cotto It'd be worth a shot.
18:43 mokurai joined #parrot
18:43 cotto It'd be nice if the return value from Parrot_str_to_cstring could be made GCable.
18:44 cotto and yes, I do also want a pony.
18:52 HG` joined #parrot
19:08 Zak joined #parrot
19:12 darbelo left #parrot
19:21 Whiteknight IF anybody is brave enough to do a proper strings refactor, it will be done in many many many small steps
19:22 kid51 Wasn't strings refactoring something simon was working on?
19:22 Whiteknight I know he was working on prototyping some of it in Perl6
19:23 Whiteknight I don't think his work extended beyond those prototypes though
19:23 kid51 What's in this branch?  https://svn.parrot.org/parrot/branches/string/
19:24 kid51 Sorry:  https://svn.parrot.org/parrot/branches/strings/
19:24 Whiteknight I believe it's just his P6 prototypes of the new strings functions
19:24 Whiteknight I don't think he did any C work at all
19:26 kid51 https://trac.parrot.org/parrot/browser has this log message:  simon:   First stab at char indexing of NFG strings.
19:27 kid51 No activity in 7 months.
19:29 Whiteknight kid51: I don't know what the status of that work is, I don't think the branch should be deleted without consulting allison first
19:30 Whiteknight I really just don't know enough about it
19:43 kid51 I wasn't suggesting deletion.
19:45 chromatic Removing _synchronize from the PMC struct gets a 4.768% performance improvement on the primes.pasm benchmark.
19:46 kid51 chromatic:  On your modernperlbooks site, if I wish to make a comment I have to sign in with a Movable Type username and password.   Do you know if that is specific to your site or something more general to Movable Type?
19:46 chromatic I think it's specific to MT.  I know OpenID works; some people use that instead.
19:47 kid51 The problem is that it uses a "password recovery phrase" model to restore passwords.  I can remember neither the password nor the recovery phrase.
19:48 chromatic I can reset your password.
19:48 cotto I hate those.
19:48 Eevee joined #parrot
19:49 kid51 chromatic:  Please do!  thanks.
19:52 joeri joined #parrot
20:02 Eevee joined #parrot
20:06 beta joined #parrot
20:06 beta left #parrot
20:36 jrtayloriv Why do parrot_string_t and buffer_t use 'void *' while struct PMC uses 'DPOINTER *'? When is 'DPOINTER *' supposed to be used instead of 'void *'?
20:37 mikehh BTW svn update claims that the certificate for *.parrot.org has expired
20:42 mikehh - Hostname: *.parrot.org
20:42 mikehh - Valid: from Fri, 29 Aug 2008 16:54:34 GMT until Sat, 29 Aug 2009 16:54:34 GMT
20:47 dalek TT #953 closed by jkeenan++: [gc] t/op/copy.t failure in OpenBSD amd64
20:52 mberends joined #parrot
20:52 kid51 Whiteknight ping
20:53 Whiteknight kid51: ping
20:53 Whiteknight or pong
20:53 Whiteknight or whatever
20:53 Whiteknight mikehh: yes, the certificate is gone
20:54 Whiteknight or at least expired
20:55 kid51 Whiteknight:  are you in a position to review what jrtayloriv has submitted in https://trac.parrot.org/parrot/ticket/958?
20:56 jrtayloriv Whiteknight, just about to add one more minor patch to the ticket -- one sec
20:57 jrtayloriv there ya go
20:57 Whiteknight kid51: busy most of the day today, can't review it till tomorrow PM
20:57 Whiteknight (sorry, I'm in a meeting)
20:57 kid51 np
21:02 mikehh All tests PASS (pre/post-config, smoke, nqp_test, fulltest) at r40855 - Ubuntu 9.04 amd64 (gcc)
21:04 jrtayloriv joined #parrot
21:13 Zak joined #parrot
21:17 mikehh rakudo (7666e92) builds on parrot r40855, make test PASS / make spectest (up to r28114) 1 FAILure - Ubuntu 9.04 amd64 (gcc)
21:18 mikehh t/spec/S06-operator-overloading/sub.rakudo - Failed test:  45 - Parse errors: Bad plan.  You planned 44 tests but ran 45.
21:22 pmichaud rakudo:  say any('foo','bar').WHAT
21:22 polyglotbot OUTPUT[Parrot VM: Can't stat languages/perl6/perl6.pbc, code 2.␤main: Packfile loading failed␤]
21:29 Whiteknight kid51: I think I'm busy on Sept 12th
21:31 ruoso joined #parrot
21:32 MoC joined #parrot
21:32 mikehh rakudo (7666e92) builds on parrot r40855, make test PASS / make spectest (up to r28115) PASS - Ubuntu 9.04 amd64 (gcc)
21:36 mikehh partcl r647 builds on parrot r40855 - make test - same 6 tests fail but all subtests PASS - Ubuntu 9.04 amd64 (gcc)
21:45 theory joined #parrot
21:47 mikehh cardinal builds on parrot r40855 - make test - 3 FAILures due to - get_pmc_keyed() not implemented in class 'String' - Ubuntu 9.04 amd64 (gcc)
21:51 mikehh decnum_dynpmcs r181 builds on parrot r40855 - make test PASS - Ubuntu 9.04 amd64 (gcc)
21:54 mikehh lua faiuls to build - error:imcc:make_pmc_const: no such pmc - in file 'lua.pir' line 307 - make: *** [lua.pbc] Error 1
21:55 mikehh fails
22:19 s1n joined #parrot
22:38 Zak joined #parrot
22:42 rg1 joined #parrot
22:59 joeri left #parrot
23:18 Zak joined #parrot
23:30 rhr joined #parrot
23:49 Zak joined #parrot
23:58 * jrtayloriv picks his nose while nobody's looking ...

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

Parrot | source cross referenced