#parrot Parrot 0.6.1 "Bird of Paradise" Released | parrotcode.org
Set by moderator on 22 April 2008.
00:03 sjansen joined
dalek chromatic@wgz.org | Concurrency Tasks: 00:20
link: www.perlfoundation.org/parrot/index...ency_tasks
shorten dalek's url is at xrl.us/bjqpy
01:34 Theory joined
dalek r27226 | jkeenan++ | trunk: 01:39
: Applying patch submitted in RT 53270: Rename/refactor _handle_mswin32()). Those subroutines are now superseded by Parrot::Configure::Step::Methods::_add_to_libs().
diff: www.parrotvm.org/svn/parrot/revision?rev=27226
01:54 particle joined 02:01 irclogbot joined
moderator #parrot Parrot 0.6.1 "Bird of Paradise" Released | parrotcode.org
02:03 particle joined
PerlJam Is anyone using the copy of parrot installed on feather? 02:05
or, more to the point, if I installed the latest and greatest parrot on feather, would anyone complain? 02:06
kid51 Are you saying that you'd like to do 'make install'?
PerlJam aye 02:07
well, someone already did "make install" for an older version of parrot. I'd just like to update it.
$ parrot --version 02:08
This is parrot version 0.4.5-devel (r13933) built for i386-linux.
kid51 I know that others have advised against developers doing 'make install'. I don't recall the exact arguments why.
True, that *is* old. 02:09
PerlJam because it didn't quite work and the installed copy would occasionally get in the way of the developer's copy was the way I remember it.
kid51 Perhaps if you posted to the list and said, "I'm gonna do this in 24 hours unless someone tells me why not to."
I remember it similarly 02:10
PerlJam I may just do that.
particle[ventus] PerlJam: just do it.
parrot 0.4.x is USELESS.
PerlJam well, time for me to put kids to bed, be back in a little while.
avar particle[ventus]: Oddly I remember claims as to it being useful for something when it was the release version:) 02:16
particle[ventus] avar: times change :) 02:17
dalek r27227 | jkeenan++ | trunk:
: Changes made in the way steps tests are called appear to obviate the need for a pass() to keep Devel::Cover happy.
diff: www.parrotvm.org/svn/parrot/revision?rev=27227
tetragon My config/auto/readline.pm patch still applies... 02:21
kid51 tetragon: Did you see rt.perl.org/rt3/Ticket/Display.html?id=52904 02:22
tetragon That's why I'm hesitant about its results
The darwin changes were for temporary flags
tetragon misreads the number, thinking its one of the other recently mentioned ones that affects readline 02:23
kid51 parrotvm.org runs on feather, correct? Is that why it takes so long to come up?
No, you're correct. That link was for the tracker ticket. Did you see: rt.perl.org/rt3/Ticket/Display.html?id=53452 02:24
tetragon I don't encounter that, but I use my architecture patch
I can't get it to build the i386 side, so I apply a patch to stop the fat build 02:25
kid51 As I suggested last week, Allison is the person who, besides yourself, will be most likely to ferret out bugs on OS X.
tetragon Anyway, I'll be back within the next 1.5h
kid51 must sleep 02:26
purl $kid51->sleep(8 * 3600);
02:30 particl1 joined 02:45 Maddingu1 joined 02:52 particle joined 02:59 mire joined 03:10 Zaba_ joined 03:12 tetragon joined 03:13 particl1 joined
tetragon reviews the tickets linked to by the traking ticket 03:13
Aww... it doesn't refer to all of the tickets. I don't see a link to the glibcism in Parrot_print_backtrace that stops 10.5 users from getting a useful backtrace 03:16
03:20 particle joined
tetragon From what I've seen, I think a fair number of the tickets in it (and ones referred to by them) could probably be resolved with the architecture patch I use 03:21
(The one that removes '-arch foo' from the flags returned by Perl 5)
03:23 particle joined 03:34 Ademan joined 04:08 tetragon joined 04:16 particl1 joined
Infinoid opbots, trust japhb 04:18
slavorg Ok
clunker Ok
Infinoid opbots, trust tetragon
slavorg Ok
clunker Ok
Infinoid opbots, trust cotto_home 04:19
slavorg Ok
clunker Ok
Infinoid opbots, trust cotto_work
slavorg Ok
clunker Ok
Infinoid (the above people all seem vaguely human.)
japhb Infinoid: thanks! 04:20
tetragon I am human, just don't try to disassemble me
Infinoid: Have you tested with the backtrace patch on win32? 04:21
Infinoid tetragon: I don't think my "objdump -d" could handle your file format anyway... 04:22
no, I was focusing all day on a sleep() issue. I can test it now if you like... which patch specifically?
tetragon objdump? No objdump here. Must be a GNUism
japhb Anyone else have any thoughts on my patch to split call_list.txt into pieces?
Infinoid yeah, part of gnu binutils. objdump -d disassembles objects
tetragon For that, I use otool 04:23
Infinoid japhb: if it helps, I don't have a problem with it. but I think the whole call_list.txt scheme is horrifically ugly, and I am hoping that if I remain in denial of it, it will go away
japhb Infinoid: I wish ... because that would imply that I got my wish from several months ago, that we were in the NCI JIT happy place. 04:24
Infinoid here tewk comes to save the day...
tetragon Infinoid: And it's the patch in 53402. You were going to try on the weekend, but you had to wait until a hard reboot due to Windows Explorer crashing. 04:25
Infinoid oh, awesome. thanks for reminding me
japhb Infinoid: is tewk the one doing the GSoC project?
Infinoid yes;
-;
japhb excellent 04:26
Now if he could be done tomorrow, that would be great .... ;-)
Infinoid pours some compressed time into tewk 04:27
anyway, in the meantime, any scheme that makes it more manageable gets a ++ from me 04:28
japhb Hmmmm ... in the mean time, maybe I should just continue coding the OpenGL stuff as if my patch was accepted already, and hope that it will be. :-)
04:28 Psyche^ joined
japhb cool beans. Just need to get chromatic to yay or nay my response to his comment, and then I can push for commit. 04:28
04:31 konobi_ left
Infinoid ooh. 0.6.1 is now in portage! 04:33
04:36 particle joined 04:38 Zaba joined
dalek r27228 | chromatic++ | pdd25cx: 04:56
: [PDD23] Ported Exception PMC to PDD 17 attributes. Fixed tests to match.
: Because there's no exit exception yet, the Exception PMC temporarily has an
: exit_code attribute.
diff: www.parrotvm.org/svn/parrot/revision?rev=27228
chromatic@wgz.org | Concurrency Tasks: 05:01
link: www.perlfoundation.org/parrot/index...ency_tasks
shorten dalek's url is at xrl.us/bjqpy
Patterner emerges parrot 05:26
hmmm... it has insecure RUNPATH's.... 05:36
05:58 Piper joined
Piper Hi there. I am Piper. I am now publicly logging this channel. If you don't want to be logged, please leave now. 05:58
06:11 uniejo joined 06:13 desertmax joined 06:19 IllvilJa joined
dalek r27229 | fperrad++ | trunk: 06:40
: [uuid]
: - add the comment :
: CURRENTLY, UNUSABLE. NEED A SOURCE OF ENTROPY.
diff: www.parrotvm.org/svn/parrot/revision?rev=27229
06:41 Psyche^ joined 06:45 Sartak joined 06:54 Sartak joined 07:03 iblechbot joined 07:20 cout joined 07:22 jisom joined 07:37 Zaba_ joined 07:45 Zaba joined 08:23 ruoso joined 08:27 desertmax_ joined 09:02 IllvilJa joined 09:03 contingencyplan joined 10:10 allison joined 10:33 wknight8111 joined 10:46 rdice joined 11:00 Zaba_ joined 11:10 kid51 joined 11:30 tetragon joined
Coke Infinoid: (in portage) which portage, now? 12:24
(I hadn't got the macports working for 0.6.1 due to the static build issue.) 12:26
odd. I just got some emails regarding the static build issue. I opened that ticket a week ago. 12:27
tetragon Static build issue? 12:29
Coke rt#52956 12:31
the static build is no longer static. 12:32
that is, trying to build something that doesn't depend on libparrot.so still depends on it.
Infinoid Coke: gentoo portage 12:51
purl well, gentoo portage is inherantly broken tho
Infinoid thanks, purl.
purl de nada Infinoid
13:02 gryphon joined 13:15 rdice joined 13:24 jhorwitz joined 13:55 iblechbot joined 14:11 Andy joined 14:24 particle[ventus] joined 14:42 wknight8111 joined
Infinoid msg tetragon I tested your RT#53402 patch on mingw. No change in functionality (backtrace* isn't detected), but it fails a couple of codingstd tests (c_indent.t and cuddled_else.t). 14:58
purl Message for tetragon stored.
nopaste "Infinoid" at 96.238.213.50 pasted "r27227: Failing tests on mingw" (155 lines) at nopaste.snit.ch/12852 15:16
Infinoid does anyone here have an example of a platform where PARROT_HAS_THREADS is not defined? (I'm in need of such a platform to test a patch on.) 15:17
Coke could you not test this by doing perl Configure -DPARROT_HAS_THREADS= ? 15:19
(or whatever the way to spell that is?) 15:20
Infinoid maybe. hmmm
I'm curious if the test case for RT #53458 fails under MSVC. Under mingw, it looks like the COND_TIMED_WAIT macro returns immediately... anyone around with msvc who can test this? 15:25
particle i may be around in 20m or so, gotta eat first and do some $work 15:28
Infinoid cool.
particle btw seems to be --optimize build problems with my linux setup
Infinoid ah, that makes sense. I don't build with that 15:29
particle try it and see
15:29 davidfetter joined
particle segfaults :( 15:29
with -j or without
dalek r27230 | coke++ | type_ids: 15:31
: [build]
: Give more diagnostics.
diff: www.parrotvm.org/svn/parrot/revision?rev=27230
Infinoid nice gc segfault compiling json 15:33
particle ah, yay! you get it too. 15:36
Infinoid yeah, you were right, its --optimize 15:48
Infinoid wonders if --optimize works on i686
15:50 NotFound joined
NotFound Hello. 15:50
Infinoid hi 15:52
15:57 Theory joined 16:00 sjansen joined
dalek r27231 | allison++ | pdd25cx: 16:01
: [pmc] Fixing a couple of bugs in generated PMC attribute accessors found while
: working on Exception PMC.
diff: www.parrotvm.org/svn/parrot/revision?rev=27231
Infinoid NotFound: hi, I just saw your NF_53112.patch (I had missed it previously), testing it now 16:13
though I don't see how it can do any harm... if it breaks on a non-gettext system, the existing code would have broken too 16:14
jhorwitz anyone seen the post to the list about my plumhead/perl6 problem? it's been a day now and i still don't see it in e-mail or in the archive. 16:18
Infinoid jhorwitz: sorry, it doesn't look like I've received anything 16:21
particle jhorwitz: was it from a subscribed addr?
jhorwitz yes
particle you've probably (and rightly) been blacklisted.
i guess you're evil. 16:22
jhorwitz who's running the server, robert?
particle yep
jhorwitz k
particle er, maybe ask
can't hurt to email both
jhorwitz yep
i'll paste my problem here
nopaste?
purl nopaste is at nopaste.snit.ch/ (ask TonyC for new channels) or rafb.net/paste or paste.husk.org/ or poundperl.pastebin.com/ or paste.scsys.co.uk/ or don't bother me while I'm eating
nopaste "jhorwitz" at 96.245.16.45 pasted "perl6/plumhead problem" (22 lines) at nopaste.snit.ch/12853 16:23
"jhorwitz" at 96.245.16.45 pasted "perl6/plumhead backtrace" (59 lines) at nopaste.snit.ch/12854 16:24
dalek r27232 | infinoid++ | trunk: 16:27
: [gettext] Followup to RT #53112:
: * apply patch from NotFound++ to use macros PARROT_TEXTDOMAIN and
: PARROT_BINDTEXTDOMAIN, to avoid polluting the namespace and hopefully
: avoid future problems of this type.
diff: www.parrotvm.org/svn/parrot/revision?rev=27232
Infinoid my mail server is having trouble connecting to mx.develooper.com 16:30
which explains why my bugs aren't hitting the list 16:31
cotto_work rt?
purl rt is just RT (bestpractical.com/rt) or (:rt3) or (: rt bugs) or Obra's trouble ticketing system or the first IBM RISC workstation (www.contrib.andrew.cmu.edu/~shadow/ibmrt.html) or the bombsquad or the Right Thing or very very capable and open-source or an application framework that bundles a ticketing system or obra's baby or SOOOO slow :-S or email mailto:perlbug-owner@perl.org for access
Infinoid yeah 16:33
jhorwitz aha. user error on my part. i sent to perl6.org, and forwhatever reason i haven't seen the bounce yet.
i'll resend to the right place this tim!
time
Infinoid :)
RT #53496 is a duplicate (I've merged with #53112)... Senaka sent that to parrotbug two weeks ago, on the 14th 16:34
16:35 kj joined
NotFound Infinoid: thanks. 16:48
It was my fault to sent it to a ticket already closed. 16:51
Infinoid and my fault for not opening a new ticket for the remaining issues :) 16:58
dalek r27233 | allison++ | pdd25cx: 17:01
: [pdd25cx] Merge in my port of Exceptions PMC to use core struct instead of
: array for attribute storage.
: - Keep 'type' attribute.
: - Use attribute access macros instead of direct struct manipulation so
: Exception PMC is subclassable from HLL.
diff: www.parrotvm.org/svn/parrot/revision?rev=27233
r27234 | allison++ | pdd25cx: 17:04
: [pdd25cx] Fix a few tests still using the old way of setting the exception message.
diff: www.parrotvm.org/svn/parrot/revision?rev=27234
allison@perl.org | Concurrency Tasks: 17:07
link: www.perlfoundation.org/parrot/index...ency_tasks
shorten dalek's url is at xrl.us/bjqpy
Coke allison: ping. 17:08
allison hi coke 17:09
Coke question regarding exceptions for you.
allison sure
Coke er, questions. figured I'd hit you pre-meeting if you're here. =-)
is there still room for HLL-specific information? I had been using [9] and [10] for two bits of info in tcl... 17:10
allison sounds good, go ahead
yeah, there's room
it's not currently spec'd so, what do you need?
Coke I'm using one for tcl's flow control type (ok, return, continue, break, error) 17:11
allison were those two integer data elements?
17:11 smashz joined
Coke the type is an int. (languages/tcl/src/returncodes.pir) 17:11
smashz hello everyone
Coke looks at that file and realizes he's down to a single slot. 17:12
smashz Coke: sorry, my jabber connection went wild
Coke I think I moved my tcl error message into [0].
another thing I need but never implemented is a place to hold a stack trace.
allison er, you were using the HLL file and line attributes for exception type?
17:13 barney joined
allison oh, no, the 'unused' spot 17:13
Coke hurm? no, my exception types are just those first ints.
I original was using two custom slots, apparently I am only using one now. =-) 17:14
(originally)
allison the general idea is that 'payload' is a PMC that can hold any arbitrary data that the particular exception needs
Coke Ok. if we're using exceptions for flow control in many languages, we might want to codify that a bit mroe.
allison so, you could stick flow-control ty;e in there
Coke (so a [return] in tcl means something to the calling perl6 method.) 17:15
allison or, we can make flow control type just an ordinary exception type
Coke we might want to make flow control type... right.
I'd recommend that.
allison is it an integer or string value?
Coke then I can map the 5 tcl kinds (ok, 4) to the various TCL_RETURN, etc.
for me it's just an enum.
allison so, an integer
Coke well, a faux-enum.
pmichaud fwiw, rakudo will be using a capture as the payload on its return types
Coke yes.
pmichaud s/return types/return exceptions
Coke do we want to add a stacktrace ala java? 17:16
allison ok, I'll add a wiki task to expand the exception types enum
Coke (which I can then abuse in tcl.)
I can always manually muck with the stacktrace in the payload, of course.
Coke tries to find an example of what a tcl stacktrace looks like. 17:18
www.tcl.tk/man/tcl8.5/tutorial/Tcl36.html //Produces the following output 17:19
dalek allison@perl.org | Concurrency Tasks: 17:21
link: www.perlfoundation.org/parrot/index...ency_tasks
shorten dalek's url is at xrl.us/bjqpy
Coke (Tcl also has errorInfo and errorCode, which I think I can manage separately. I think.)
allison stacktrace as an attribute of the Exception PMC? or stacktrace as functionality in src/exceptions.c? 17:22
17:23 desertmax joined
Coke I'm not sure where we would store it except in the exception PMC. 17:23
smash allison: you lost me with that mail exchanges about debian packages, i have the packages for 0.6.1 lying around my laptop without destination :-P
Coke I could make due with an array of file/code/line tuples. (maybe a few more items, I'm not sure.) 17:24
we could provide standard "convert that to a string" convenience functions for general parrot use, and HLLs could roll their own.
allison oh, i see, you're talking about storing the stack trace so it can be accessed by the handler, right? 17:25
Coke yes.
then my top level error handler can print a nicely formatted HLL message.
(and I can muck with the data as it's passing up through my handlers.
allison smash: can you put them up in a place where I can download and test them?
smash allison: but of course, gimme a tick 17:26
allison smash: if they look good I'll load them onto alioth
smash: no rush, it's evening here, so my day ends soon after parrotsketch
smash and if they don't look good.. you'll tell me what i need to fix
allison smash: exactly
coke: yeah, I think the exception object is a sensible place to store that. It'll have to be a String PMC to start, since we don't capture the stack trace with any more detail yet, but it could be a more easily introspectable PMC later 17:27
smash allison: besides your testing, any reason i can'y upload them myself to alioth ? 17:28
allison smash: admin privs, and I need to set up the categories correctly, since we haven't uploaded any packages yet (quicker to do it myself than to explain, and once I do the first upload you'll easily see what to do next time) 17:29
smash allison: roger that 17:30
which files do you want? *.deb ?
i guess that's enough for installing 17:31
jonathan parrotsketch is meant to be now? 17:33
dalek allison@perl.org | Concurrency Tasks:
link: www.perlfoundation.org/parrot/index...ency_tasks
shorten dalek's url is at xrl.us/bjqpy
smash allison: nrc.homelinux.org/parrot/debian/ 17:34
Infinoid jonathan: I think its an hour from now
allison jonathan: yes, an hour from now 17:35
jonathan Oh?
Geck, I'm here now, but I can't be in an hour.
smash i'm off then, bbl
jonathan Let me put an early report in the channel. 17:36
allison 6:30pm UTC
Infinoid sounds good
jonathan Oh, we don't change the time with daylight savings?
That'd explain it.
allison jonathan: nope, too confusing with so many different daylight savings times 17:37
Infinoid apparently your DST switch happens later than my DST does :)
allison no DST at all here in ZA
avar rad
(wrong channel)
Infinoid no DST in AZ, either. it made life so much simpler when I was living in Phoenix 17:38
jonathan No DST sounds like a good idea. 17:39
pmichaud #ps is always 18:30 utc 17:42
jonathan Sure, it works for me, I'd just not factored in DST and such things.
Anyways, gotta run...around tomorrow or certainly here all day Thursday 17:43
pmichaud jonathan: btw, ROADMAP has rakudo milestones, if you hadn't noticed it yet
jonathan pmichaud: I saw the commits, glanced over the first one, will read it properly soon, thanks for doing that. 17:45
pmichaud okay. We're thinking of allowing people to claim particular milestones 17:47
not exclusively, of course, but for someone to say "I really want to do this part" 17:48
I guess as pumpking I get to claim all of the parts nobody else wants to do :-)
Infinoid had hopes of getting a bunch of PDD13 stuff done last weekend, but got distracted by video games instead. *sigh*
pmichaud yeah, I got "Mario Kart Wii" yesterday. Evil. 17:49
particle last weekend video games, this weekend sailing... you've got a rough life, mark.
17:49 sjansen joined
particle ;) 17:49
avar pmichaud: I hate its battle mode
Infinoid poor poor me
pmichaud I haven't tried battle mode yet.
avar They ruined it 17:50
Infinoid particle: hey! how'd you know I was going to be racing? :P
pmichaud tis okay, I like the other modes well enough. And it's nice to have a game that Matthew can play with his sister and me and still have a good chance of winning :-)
barney /me stumbled over 17:58
sysinfo op.
Courtesy of Brad Gilbert
#
20:05 Friday\tparrot
sorry wrong paste
meant to mention en.wikibooks.org/wiki/Parrot_Virtual_Machine 17:59
18:00 grim_fandango joined 18:06 alin joined
Coke (mario kart wii) Hurm. I just picked up Assassin's Creed PS3. =-) 18:10
barney: is that new content or copied from elsewehre, OOC? 18:12
18:21 teknomunk joined
barney I think that whiteknight wrote some new content 18:21
18:22 alin left 18:23 AndyA joined
wknight8111 I try to add a little bit of content every day 18:25
18:28 chromatic joined
chromatic #ps in 2 18:28
18:34 Ivatar joined 18:54 ambs joined
dalek will@coleda.com | YAPC::NA::2008: 18:55
link: www.perlfoundation.org/parrot/index...pc_na_2008
18:55 mire joined
ambs seen merlyn 18:56
purl merlyn was last seen on #moose 39 days and 21 hours ago, saying: ... methodsandmessages.vox.com/library/...ltalk.html [Mar 20 15:09:20 2008]
davidfetter ambs, he's over on irc://irc.freenode.net/postgresql 19:01
(among other places)
ambs davidfetter, thanks, but nothing too important :) 19:06
davidfetter he's there right now :)
davidfetter chatting w/him atm
19:10 rgrjr joined 19:17 ruoso joined 19:19 IllvilJa joined
dalek r27235 | chromatic++ | pdd25cx: 19:19
: [Exceptions] Renamed "_message" to "message" globally.
diff: www.parrotvm.org/svn/parrot/revision?rev=27235
Coke tries to imagine a woofie-like currency based on tuits. 19:23
19:27 ambs joined
Coke er, whuffie. 19:28
dalek r27236 | bernhard++ | trunk: 19:30
: [Eclectus]
: Funny quoting. Pass t/vectors.t for 'gen_past_in_nqp' implementation.
diff: www.parrotvm.org/svn/parrot/revision?rev=27236
rgrjr Coke: Sorry, still don't get it. 19:31
Coke whuffie?
purl rumour has it whuffie is (: affero) or cory doctorow's word for 'global karma system'
Coke whuffie is also at Act: 1,3,10]
[#parrot]
purl okay, Coke.
Coke f*ck.
rgrjr thx. 19:32
chromatic zero or more 'f' characters followed by 'ck'
dalek r27237 | chromatic++ | pdd25cx:
: [Exceptions] Renamed "_severity" to "severity" globally.
diff: www.parrotvm.org/svn/parrot/revision?rev=27237
Coke whuffie is also at en.wikipedia.org/wiki/Whuffie
purl okay, Coke.
Coke whuffie?
purl whuffie is (: affero) or cory doctorow's word for 'global karma system' or at en.wikipedia.org/wiki/Whuffie
NotFound I can't build with g++, here is the error:
src/builtin.c: In function 'int find_builtin(const char*)':
src/builtin.c:145: error: declaration of 'int find_builtin(const char*)' throws different exceptions
src/builtin.c:100: error: from previous declaration 'int find_builtin(const char*) throw ()'
Coke chromatic: you gonna go through languages in that branch too? 19:33
chromatic Did.
Coke even tcl?
NotFound With gcc 4.2.3
chromatic Especially Tcl.
Coke Excellent.
chromatic Even dotnet, which still had, inexplicably, '$P0 = new .Exception'
Coke hasn't been run in a while, apparently. 19:34
particle i bet nobody has mono to test it 19:35
chromatic BASIC has similar things.
Coke I've had mono. I don't recommend it. 19:36
chromatic You didn't know there were patent problems with being really bored.
19:37 rgrjr left
Tene I had mono in December and January. 19:39
davidfetter you got .NET working on linux?!? 19:41
NotFound 'mono' in spain is slang for abstinence syndrome.
Tene I was really into c# back around '03
dalek r27238 | chromatic++ | pdd25cx: 19:42
: [Exceptions] Renamed "_type" to "type" globally.
diff: www.parrotvm.org/svn/parrot/revision?rev=27238
ambs lol
dalek chromatic@wgz.org | Concurrency Tasks:
link: www.perlfoundation.org/parrot/index...ency_tasks
shorten dalek's url is at xrl.us/bjqpy
dalek r27239 | chromatic++ | pdd25cx: 19:57
: [src] Cleaned up code per coding standards.
diff: www.parrotvm.org/svn/parrot/revision?rev=27239
chromatic@wgz.org | Concurrency Tasks: 19:58
link: www.perlfoundation.org/parrot/index...ency_tasks
shorten dalek's url is at xrl.us/bjqpy
dalek r27240 | chromatic++ | pdd25cx: 20:05
: [PMC] Added stacktrace attribute to Exception PMC.
diff: www.parrotvm.org/svn/parrot/revision?rev=27240
chromatic@wgz.org | Concurrency Tasks: 20:06
link: www.perlfoundation.org/parrot/index...ency_tasks
shorten dalek's url is at xrl.us/bjqpy
20:16 Zaba joined
Infinoid ah! the mingw32-doesn't-sleep issue (RT #53458) was caused by r24416... reverting that diff locally causes the sleep op to behave as expected. 20:36
particle[ventus] interesting. indeed. 20:38
Infinoid thinks the win32 version of COND_TIMED_WAIT needs a rework 20:39
dalek r27241 | chromatic++ | pdd25cx: 20:43
: [Exceptions] Renamed throw_exception() to run_handler() globally.
: Tidied up some Tcl ops while I was in there.
diff: www.parrotvm.org/svn/parrot/revision?rev=27241
chromatic@wgz.org | Concurrency Tasks: 20:45
link: www.perlfoundation.org/parrot/index...ency_tasks
shorten dalek's url is at xrl.us/bjqpy
ambs Mac, still failing t/examples/library.t test three. 20:46
21:13 slightlyoff joined
dalek r27242 | pmichaud++ | trunk: 21:35
: [rakudo]:
: * Add hyperops, reduction operators to ROADMAP.
diff: www.parrotvm.org/svn/parrot/revision?rev=27242
21:50 askie joined
ambs wonders if Ask has a child 21:51
ambs grins
well,good night
21:55 NotFound left 22:50 particle[ventus] joined
Infinoid in pir, N0 is supposed to be a number (meaning, FLOATVAL), right? 22:50
Coke yes. 22:54
though it's typically spelled $N0
Infinoid hmm, doesn't seem to make any difference whether I add a $ or not
nopaste "Infinoid" at 96.238.213.50 pasted "broken timestamp handling on mingw" (17 lines) at nopaste.snit.ch/12858 22:55
Infinoid the time(out INT) and time(out NUM) ops are virtually identical... yet N0 has an extra digit 22:56
Coke interesting, yah, that works fine on my platform.
I suspect a slipped digit in the conversion on your platform. 22:57
Infinoid I got to this after wondering if Parrot_floatval_time() is broken on mingw... unfortunately time_t is 32 bit on mingw, and all the time calculations done by the sleep routine end up with 0xffffffff values, which is why it never sleeps 22:58
so, where can I find this conversion you speak of? :)
23:05 chromatic joined
Infinoid chromatic: any advice on how to un-break float conversion on mingw32? nopaste.snit.ch/12858 23:06
chromatic Let me look.
Infinoid thanks. 23:07
23:08 tetragon joined
chromatic Hm. 23:08
Looks like a precision problem somewhere in the platform information.
Infinoid is this parrot's fault, or gcc's?
chromatic Depends on how big INTVAL is, I suspect. 23:09
Infinoid INTVAL_SIZE is 4 (long) 23:10
chromatic config/gen/platform/win32/time.c:40:# if INTVAL_SIZE <= 4 23:13
Infinoid _MSC_VER isn't defined, this is mingw32. 23:14
chromatic So much for random guesses based on ack. 23:15
What's in include/parrot/config.h on lines 85 and 86? 23:16
Mine's 4 and 8, respectively.
Infinoid same here
but you brought up a very good point, that I missed 23:17
"N0 = time" calls Parrot_floatval_time, which is quite a lot different from Parrot_intval_time
chromatic They should probably be similarish. 23:18
Infinoid the code in Parrot_cx_schedule_sleep and in win32's COND_TIMED_WAIT() both expect their return values to be compatible
and they very obviously aren't. so that's what I need to fix
thanks! 23:19
home & 23:20
chromatic You're welcome.
dalek r27243 | chromatic++ | pdd25cx: 23:21
: [Exceptions] Removed Python-specific exceptions from exception_type_enum.
: Ported all instances of these exceptions to Parrot-specific types globally.
diff: www.parrotvm.org/svn/parrot/revision?rev=27243
chromatic@wgz.org | Concurrency Tasks:
link: www.perlfoundation.org/parrot/index...ency_tasks
shorten dalek's url is at xrl.us/bjqpy 23:22
chromatic That was a big patch. 23:23
Coke chromatic: you're parrot-busy today.
+1
purl 1
chromatic I'm on vacation. 23:25
Tene chromatic++
Coke chromatic+=3.14 23:28
chromatic Yeah, you're just glad I made the Tcl ops slightly more readable. 23:29
Coke I'd be happier if you made tcl less than slightly more fast. 23:32
how long is your vacation? =-)
chromatic: I think there's some vtable level exception wonkery in there you might want to update, btw. 23:33
(where I do the equivalent of $P1[0] = $S0 for the message. 23:34
(but in C)
chromatic VTABLE_set_string_native(interp, exception, message) 23:37
Coke VTABLE_set_pmc_keyed_int(interp, exception, (INTVAL)0, $1); 23:39
23:39 Limbic_Region joined
chromatic Exception doesn't extend ResizablePMCArray anymore, so no keyed int, I believe. 23:40
Coke ... this is what I'm saying.
chromatic set_attr does work though.
Coke you updated the PIR variants in tcl, but not the C variants.
chromatic If you just want to set the message, use set_string_native.
Coke Right. I'm just telling you because you're on vacation and I'm still at work, nearly 3 hours late. =-)
I'll get to it in a day or two. 23:41
chromatic I'll update the wiki so I don't have to keep it in my braiiiiins.
dalek chromatic@wgz.org | Concurrency Tasks: 23:42
link: www.perlfoundation.org/parrot/index...ency_tasks
shorten dalek's url is at xrl.us/bjqpy
Coke danke. 23:44
Coke is reminded to buy zombie fluxx.
chromatic++
dalek will@coleda.com | Concurrency Tasks: 23:48
link: www.perlfoundation.org/parrot/index...ency_tasks
shorten dalek's url is at xrl.us/bjqpy
davidfetter i am dalek's shortened url 23:57
Infinoid I get cancer. I kill dalek.
Coke I am dalek's angst ridden send. 23:58
yay, enough cold fusion done for one day.

Powered by Cro and the Raku® Programming Language.

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

Thank you!