Parrot 4.2.0 "Ornithopter" | parrot.org | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC
Set by moderator on 21 March 2012.
00:27 kid51 joined 00:46 benabik joined
dalek sella: 48b83e8 | Whiteknight++ | src/unstable/net/ (5 files):
[Net] Rewrite request content handling to support POST requests (untested). Basic methods exist to set form data and file uploads with a request.
 00:48
sella: 7be362c | Whiteknight++ | s (9 files):
[Net] Update the request protocols to send status information back to a UserAgent. Add a new SimpleHttp UserAgent to start encapsulating basic HTTP requests
sella: afd4720 | Whiteknight++ | src/unstable/net/ (3 files):
[Net] Implement request methods for the SimpleHttp user agent
sella: 63fb8df | Whiteknight++ | s (4 files):
[Net] Refactor SocketFactory to be more general. It can create all sorts of sockets, not just tcp ones.
 00:49
benabik ~~
whiteknight hello benabik 01:04
benabik 'alo
dalek rrot/m0: 93f878c | nbrown++ | src/m0/perl5/m0_assembler.pl:
the size of the M0_META_SEG should include the size of the header (12)
 01:09
rrot/m0: 0e5c911 | dukeleto++ | src/m0/perl5/m0_assembler.pl:
Merge pull request #751 from nbrown/93f878cea015f56e9de3e41c1bc9a295454b34f1

The size of the M0_META_SEG should include the header
aloha (parrot/parrot) Issues closed : 751 (The size of the M0_META_SEG should include the header) by nbrown : github.com/parrot/parrot/issues/751 01:10
benabik whiteknight: What's up? 01:12
whiteknight benabik: not much. Just hacking on Rosella while running tests for my remove_sub_flags branch
bacek_at_work whiteknight, cotto, I would like to have some feed back on github.com/parrot/parrot/commit/7d...0e0cabf4b3 01:23
(not urgent because looks like $reallife will keep me distracted for some time)
whiteknight bacek_at_work: I saw it earlier today. I like it very much
bacek_at_work: It has potential to simplify lots of logic, such as resetting/reusing a CallContext 01:24
cotto looks
bacek_at_work whiteknight, good. Feel free to fire your $EDITOR and jump on this branch :)
whiteknight bacek_at_work: I have to finish my own branch first! but yes, I would love to help
bacek_at_work whiteknight, yes, you are right about logic. Current one is... incomprehensible.
whiteknight VTABLE_morph was the worst possible idea
cotto bacek_at_work: at a quick glance, I like where that's going. 01:25
bacek++
dalek sella: 92ff725 | Whiteknight++ | src/unstable/net/ (3 files):
[Net] Move the SocketFactory accessor to be more clear about what it is and who can use it
 01:57
sella: 3d22fcb | Whiteknight++ | s (7 files):
[Net] Rewrite Request and Response. They don't share a common subtype anymore, since the functionality is so different.
01:58 jashwanth joined
jashwanth Good morning,whitenight 01:58
whiteknight hello jashwanth 02:03
jashwanth: I'm actually going to bed now. I'll talk to you later
02:19 Hunger joined 05:45 eternaleye joined 05:54 eternaleye joined 05:55 eternaleye joined
moritz \\o 06:08
06:55 fperrad joined
dalek rrot: 783b189 | petdance++ | src/ (2 files):
removing unnecessary and noisy casts
 06:56
07:12 jashwanth joined 07:26 mj41 joined 08:25 lucian joined 08:45 mj41 joined 09:30 aloha joined 10:14 jsut_ joined
dalek rrot/pcc_reorder: 0e76dec | bacek++ | / (2 files):
Borrow some stuff from CallContext
 11:05
rrot/pcc_reorder: 2798736 | bacek++ | src/call/signature.c:
Borrow signature free and reset from CallContext
rrot/pcc_reorder: bd0312a | bacek++ | / (2 files):
Stylish reorder of perldoc
rrot/pcc_reorder: 8e1c54f | bacek++ | / (2 files):
Borrow pushing of positional parameters
rrot/pcc_reorder: 6458748 | bacek++ | src/call/signature.c:
Fix few typos
rrot/pcc_reorder: fec6761 | bacek++ | / (2 files):
Implement pushing of named parameters to Signature
rrot/pcc_reorder: 722fcbb | bacek++ | / (2 files):
Borrow fetching of positional arguments
 11:41
rrot/pcc_reorder: bc20be3 | bacek++ | / (2 files):
Implement fetching of named parameters and fix compilation errors
rrot/pcc_reorder: d3630a9 | bacek++ | / (2 files):
Add couple of last required functions
 11:59
12:12 whiteknight joined
whiteknight good morning, #parrot 12:14
moritz LOL IT'S WITENIGHT! 12:17
or so :-)
12:24 benabik joined
benabik G'morning, #parrot 12:24
whiteknight hello moritz 12:39
hello benabik 12:40
benabik 'lo whiteknight
13:05 PacoAir joined 13:25 ascent joined 14:26 jashwanth joined 14:50 hercynium joined 14:55 hercynium joined 15:08 dmalcolm joined 16:10 alester joined 16:11 jsut joined 16:16 myhrlin joined
nine Hello, #parrot 16:34
benabik o/ nine 16:37
whiteknight hello nine 16:38
dukeleto ~~ 16:49
good localtime()
alester WHO LOVES THE NEW -Wuseless-cast flag? 16:55
ME ME ME
nine whiteknight: I think getting threads to work on Windows wouldn't be that hard. Since I resurrected the abstraction macros for Windows as well the only pieces missing are a replacement for POSIX alarms and for pipes. With threading available, the timer stuff is quite simple and I'm sure there's some simple API function for having a thread wait till notified. 17:04
dukeleto alester: useless cast? Haven't heard of that before. 17:06
nine: that sounds awesome
alester Brand spankin' new in dev version of GCC 4.8
See my last couple of commits. My favorite was the cast of 0 to (int)
17:11 lucian joined
dukeleto alester: just in case 17:27
alester Kinda like setting variables twice, just to be sure. 17:29
OK, lunch time. I'm headin' to Tacos El Norte on 120 if anyone's interested. I'm buyin'!
whiteknight nine: okay, that's good to know. 17:33
dukeleto: ping 17:52
17:57 alvis joined
nine Damn C90. So many chances for consting stuff lost because I have to declare variables too early. 17:57
whiteknight nine: preaching to the choir 17:58
nine: the only reason we're stuck on C90 is that microsoft's damn compiler is stupid and won't upgrade to a better standard
so in the interests of portability, we're stuck 17:59
nine whiteknight: if it weren't for jnthn++ I'd say just drop Windows support and deliver more awesome code with less work
benabik Parrot loses much of its appeal if it doesn't work on Windows. 18:00
(IMHO)
nine unfortunately
whiteknight it's not a matter of not working on windows. Parrot does build very well with mingw and icc on windows 18:02
nine So how many people actually build Parrot on Windows with Microsoft's compiler?
whiteknight the only person I know who does it regularly is jnthn. I think particle used to also 18:03
I used to do it sometimes back when I still had windows installed 18:04
mostly to fix bugs
benabik tends to use Strawberry
whiteknight I found it was very difficult to have strawberry and activeperl on the same machine and be able to compile with either without clobbering PATH and other evil nonsense 18:05
small nit, but definitely hurt my ability to have multiple compilers running in an automated way on that platform 18:06
benabik only needs one compiler working: whatever will compile Parrot. :-D
whiteknight I've got gcc, g++, icc and clang installed on my main computer, and used to smoke them all regularly 18:07
benabik arg, I need to write my GSoC proposals. Blah. 18:37
Figured I'll write something that goes from Bytecode -> basic assembler and one that's PACT.Packfile -> CFG -> Complex assembler. That way if someone wants to do the lower level, we just have to agree on an API for the midpoint. 18:38
18:40 icebattle joined
cotto long running meeting: might not make #ps today 18:51
benabik meetings--
dukeleto whiteknight: pong 19:13
whiteknight dukeleto: I've been talking to a student about LAPACK bindings for PLA. Is that the kind of thing you might be willing to mentor/co-mentor?
dukeleto whiteknight: sure
19:13 contingencyplan joined
whiteknight I'm hoping to end up with LAPACK bindings and fallback implementations of a few decompositions, in case LAPACK isn't available to accelerate things 19:14
sorear Infinoid: ping 19:18
whiteknight sorear: I haven't seen him around in a long while 19:29
benabik seen Infinoid
aloha Infinoid was last seen in #parrot 44 days 15 hours ago joining the channel.
Coke tries to kick off a windows build of rakudo to appease benabik. 19:38
benabik Huh?
Coke or whoever was asking in scrollback. 19:41
nine #ps now 19:49
whiteknight damnit 19:51
I can't remember that meeting to save my life
20:02 perlite_ joined
Coke is building rakudo now. (after --gen-parrot build nqp & parrot) 20:27
*built
make test has a single error, but everything otherwise ok on rakudo build on windows (which included a from-scratch build of parrot&nqp) 20:31
NotFound nine: task.pir is supposed to run ten seconds and stop? 20:32
Something bizarre is happening. It never stops. A translarion to winxed stop at the seconds when run with the winxed fakecutable, but compiled to pir and run with parrot does the same as the pir version. 20:36
nine NotFound: since the last commit one has to issue an explicit exit. Otherwise running tasks would keep Parrot from exiting 20:38
Just pushed a fix for tasks.pir
btw. dalek seems to be dead
NotFound Ah, that explains it, the winxed driver does an explicit exit. 20:39
nine it's well behaved then :)
What does such code look like in winxed? 20:40
nopaste "NotFound" at 192.168.1.3 pasted "example task translated to winxed" (27 lines) at nopaste.snit.ch/125668
nine nice 20:41
NotFound nine: you can give winxed a try, is fairly easy to do simple things if you know pir, and faster to write than pir.
nine NotFound: seems very much like that, yes
NotFound: does it use packfile constants for code like "p1.code = task;" or does it issue a getglobal? 20:42
NotFound In this case it use constants, because the function is known in scope. 20:43
nine too bad
Then chameneos.pir won't work in winxed. Calling packfile constants doesn't work yet. Using getglobal and then invoking the PMC does however. 20:44
NotFound No problem: var task; ${ get_global task, "task" };
benabik You can use get global explicitly, I think. 20:45
Yes, that was it.
NotFound++
nine :)
NotFound And to test with the constants, just comment out that line. 20:46
nine Ok, I'm sold
NotFound++ 20:47
NotFound Beating PIR in user friendliness is easy. 20:48
Now that I think about it, it's not really calling constants, it sets a register with the value from the constant table. That's what .const 'Sub' does. 20:55
cotto ~~ 21:06
nine Good night, #parrot! 21:08
NotFound msg whiteknight gist.github.com/2159106 fixed in github.com/NotFound/winxed/commit/...8d283ac72a 21:16
aloha OK. I'll deliver the message.
NotFound No dalek? 21:17
Coke he's been missing for a whil ein #perl6 also 21:23
NotFound msg whiteknight snapshot updated with the fix 21:51
aloha OK. I'll deliver the message.
21:54 preflex_ joined 22:18 whiteknight joined
whiteknight good evenin, #Parrot 22:21
NotFound++ # I would have fixed those myself! 22:22
NotFound whiteknight: the fastest way to diagnose the problem was fixing it. 22:24
whiteknight That's awesome. It's going to fix so much ugly Rosella 22:26
code
one day Rosella is going to be the prettiest library ever
NotFound whiteknight: If I can make to work some test putting together threads, Rosella http and Guitor, it will be a great parrot demo. 22:28
whiteknight NotFound: Yes, I want to use threads and Rosella to make an IRC bot some day
but we need threads for the async reads
NotFound Just downloading some file and showing it on the UI will be a nice start. 22:30
Even better, downloading and installing parrot modules. 22:31
whiteknight yes. I really want to make a pretty test harness UI with Guitor, I just don't have the tuits 22:36
maybe it would be a good GSOC project....
NotFound Not very pretty right now, Guitor lack features. 22:37
whiteknight Would those kinds of feature improvements make a good GSOC project? 22:48

Powered by Cro and the Raku® Programming Language.

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

Thank you!