Parrot 3.10.0 "Apple Pi" | parrot.org | Log: irclog.perlgeek.de/parrot | #parrotsketch meeting Tuesday 19:30 UTC
Set by moderator on 2 December 2011.
dalek itor: b1effac | NotFound++ | / (2 files):
use a context menu in example pizarra
 00:08
itor: 01d7998 | NotFound++ | examples/winxed_repl.winxed:
Merge branch 'master' of git@github.com:NotFound/Guitor
00:23 Nol888 joined
aloha (parrot/parrot) Issues opened : 166 (Humbly suggesting improvements to workflow document) by gitster : github.com/parrot/parrot/issues/166, 169 (crow.pir should fill in digest information) by leto : github.com/parrot/parrot/issues/169, 170 (Release manager guide is too vague about 00:47
..how to update docs.parrot.org) by leto : github.com/parrot/parrot/issues/170, 173 (Script to cross-compile Parrot with qemu) by leto : github.com/parrot/parrot/issues/173, 174 (Fix and test .macro_local or remove it from our docs) by leto :
..github.com/parrot/parrot/issues/174, 175 (Sorting benchmarks) by leto : github.com/parrot/parrot/issues/175, 176 (GCC Compile Farm smokers are fragile) by leto : github.com/parrot/parrot/issues/176, 177 (Look at using the "restrict" keyword on pointers.) by petdance :
..github.com/parrot/parrot/issues/177, 182 (Segfault + seemingly crazy memory gobbling) by jnthn : github.com/parrot/parrot/issues/182, 183 (-t4 no longer gives sub-level output) by coke : github.com/parrot/parrot/issues/183, 185 (Memory corruption error) by coke :
..github.com/parrot/parrot/issues/185, 188 (plan(0) not supported) by leto : github.com/parrot/parrot/issues/188, 189 (Build failure in alarm.c on Win32) by jnthn : github.com/parrot/parrot/issues/189, 190 (Debian package missing osutils.pbc) by Benabik :
..github.com/parrot/parrot/issues/190, 193 (load_bytecode on Windows) by Vidmich : github.com/parrot/parrot/issues/193, 199 (windows build broken) by coke : github.com/parrot/parrot/issues/199, 201 (Work-directory references in installed files) by he32 :
..github.com/parrot/parrot/issues/201, 202 (Rewrite tools/dev/nci_thunk_gen.pir to avoid dependency of pcre.) by bacek : github.com/parrot/parrot/issues/202, 203 (Opcode new_si is deprecated.) by bacek : github.com/parrot/parrot/issues/203
(parrot/parrot) Issues opened : 166 (Humbly suggesting improvements to workflow document) by gitster : github.com/parrot/parrot/issues/166, 169 (crow.pir should fill in digest information) by leto : github.com/parrot/parrot/issues/169, 170 (Release manager guide is too vague about
..how to update docs.parrot.org) by leto : github.com/parrot/parrot/issues/170, 173 (Script to cross-compile Parrot with qemu) by leto : github.com/parrot/parrot/issues/173, 174 (Fix and test .macro_local or remove it from our docs) by leto : 00:48
..github.com/parrot/parrot/issues/174, 175 (Sorting benchmarks) by leto : github.com/parrot/parrot/issues/175, 176 (GCC Compile Farm smokers are fragile) by leto : github.com/parrot/parrot/issues/176, 177 (Look at using the "restrict" keyword on pointers.) by petdance :
..github.com/parrot/parrot/issues/177, 182 (Segfault + seemingly crazy memory gobbling) by jnthn : github.com/parrot/parrot/issues/182, 183 (-t4 no longer gives sub-level output) by coke : github.com/parrot/parrot/issues/183, 185 (Memory corruption error) by coke :
..github.com/parrot/parrot/issues/185, 188 (plan(0) not supported) by leto : github.com/parrot/parrot/issues/188, 189 (Build failure in alarm.c on Win32) by jnthn : github.com/parrot/parrot/issues/189, 190 (Debian package missing osutils.pbc) by Benabik :
..github.com/parrot/parrot/issues/190, 193 (load_bytecode on Windows) by Vidmich : github.com/parrot/parrot/issues/193, 199 (windows build broken) by coke : github.com/parrot/parrot/issues/199, 201 (Work-directory references in installed files) by he32 :
..github.com/parrot/parrot/issues/201, 202 (Rewrite tools/dev/nci_thunk_gen.pir to avoid dependency of pcre.) by bacek : github.com/parrot/parrot/issues/202, 203 (Opcode new_si is deprecated.) by bacek : github.com/parrot/parrot/issues/203
(parrot/parrot) Issues opened : 166 (Humbly suggesting improvements to workflow document) by gitster : github.com/parrot/parrot/issues/166, 169 (crow.pir should fill in digest information) by leto : github.com/parrot/parrot/issues/169, 170 (Release manager guide is too vague about
..how to update docs.parrot.org) by leto : github.com/parrot/parrot/issues/170, 173 (Script to cross-compile Parrot with qemu) by leto : github.com/parrot/parrot/issues/173, 174 (Fix and test .macro_local or remove it from our docs) by leto :
..github.com/parrot/parrot/issues/174, 175 (Sorting benchmarks) by leto : github.com/parrot/parrot/issues/175, 176 (GCC Compile Farm smokers are fragile) by leto : github.com/parrot/parrot/issues/176, 177 (Look at using the "restrict" keyword on pointers.) by petdance :
..github.com/parrot/parrot/issues/177, 182 (Segfault + seemingly crazy memory gobbling) by jnthn : github.com/parrot/parrot/issues/182, 183 (-t4 no longer gives sub-level output) by coke : github.com/parrot/parrot/issues/183, 185 (Memory corruption error) by coke :
..github.com/parrot/parrot/issues/185, 188 (plan(0) not supported) by leto : github.com/parrot/parrot/issues/188, 189 (Build failure in alarm.c on Win32) by jnthn : github.com/parrot/parrot/issues/189, 190 (Debian package missing osutils.pbc) by Benabik :
..github.com/parrot/parrot/issues/190, 193 (load_bytecode on Windows) by Vidmich : github.com/parrot/parrot/issues/193, 199 (windows build broken) by coke : github.com/parrot/parrot/issues/199, 201 (Work-directory references in installed files) by he32 :
..github.com/parrot/parrot/issues/201, 202 (Rewrite tools/dev/nci_thunk_gen.pir to avoid dependency of pcre.) by bacek : github.com/parrot/parrot/issues/202, 203 (Opcode new_si is deprecated.) by bacek : github.com/parrot/parrot/issues/203
(parrot/parrot) Issues opened : 166 (Humbly suggesting improvements to workflow document) by gitster : github.com/parrot/parrot/issues/166, 169 (crow.pir should fill in digest information) by leto : github.com/parrot/parrot/issues/169, 170 (Release manager guide is too vague about
..how to update docs.parrot.org) by leto : github.com/parrot/parrot/issues/170, 173 (Script to cross-compile Parrot with qemu) by leto : github.com/parrot/parrot/issues/173, 174 (Fix and test .macro_local or remove it from our docs) by leto : 00:49
00:49 aloha joined
aloha (parrot/parrot) Issues opened : 166 (Humbly suggesting improvements to workflow document) by gitster : github.com/parrot/parrot/issues/166, 169 (crow.pir should fill in digest information) by leto : github.com/parrot/parrot/issues/169, 170 (Release manager guide is too vague about 00:50
..how to update docs.parrot.org) by leto : github.com/parrot/parrot/issues/170, 173 (Script to cross-compile Parrot with qemu) by leto : github.com/parrot/parrot/issues/173, 174 (Fix and test .macro_local or remove it from our docs) by leto :
..github.com/parrot/parrot/issues/174, 175 (Sorting benchmarks) by leto : github.com/parrot/parrot/issues/175, 176 (GCC Compile Farm smokers are fragile) by leto : github.com/parrot/parrot/issues/176, 177 (Look at using the "restrict" keyword on pointers.) by petdance :
..github.com/parrot/parrot/issues/177, 182 (Segfault + seemingly crazy memory gobbling) by jnthn : github.com/parrot/parrot/issues/182, 183 (-t4 no longer gives sub-level output) by coke : github.com/parrot/parrot/issues/183, 185 (Memory corruption error) by coke :
cotto um...
aloha ..github.com/parrot/parrot/issues/185, 188 (plan(0) not supported) by leto : github.com/parrot/parrot/issues/188, 189 (Build failure in alarm.c on Win32) by jnthn : github.com/parrot/parrot/issues/189, 190 (Debian package missing osutils.pbc) by Benabik :
..github.com/parrot/parrot/issues/190, 193 (load_bytecode on Windows) by Vidmich : github.com/parrot/parrot/issues/193, 199 (windows build broken) by coke : github.com/parrot/parrot/issues/199, 201 (Work-directory references in installed files) by he32 :
..github.com/parrot/parrot/issues/201, 202 (Rewrite tools/dev/nci_thunk_gen.pir to avoid dependency of pcre.) by bacek : github.com/parrot/parrot/issues/202, 203 (Opcode new_si is deprecated.) by bacek : github.com/parrot/parrot/issues/203
(parrot/parrot) Issues opened : 166 (Humbly suggesting improvements to workflow document) by gitster : github.com/parrot/parrot/issues/166, 169 (crow.pir should fill in digest information) by leto : github.com/parrot/parrot/issues/169, 170 (Release manager guide is too vague about
bacek_at_work cotto, ignore it for now. It's for first run only
cotto is it import o'clock?
aloha ..how to update docs.parrot.org) by leto : github.com/parrot/parrot/issues/170, 173 (Script to cross-compile Parrot with qemu) by leto : github.com/parrot/parrot/issues/173, 174 (Fix and test .macro_local or remove it from our docs) by leto :
..github.com/parrot/parrot/issues/174, 175 (Sorting benchmarks) by leto : github.com/parrot/parrot/issues/175, 176 (GCC Compile Farm smokers are fragile) by leto : github.com/parrot/parrot/issues/176, 177 (Look at using the "restrict" keyword on pointers.) by petdance :
cotto ignores
aloha ..github.com/parrot/parrot/issues/177, 182 (Segfault + seemingly crazy memory gobbling) by jnthn : github.com/parrot/parrot/issues/182, 183 (-t4 no longer gives sub-level output) by coke : github.com/parrot/parrot/issues/183, 185 (Memory corruption error) by coke :
..github.com/parrot/parrot/issues/185, 188 (plan(0) not supported) by leto : github.com/parrot/parrot/issues/188, 189 (Build failure in alarm.c on Win32) by jnthn : github.com/parrot/parrot/issues/189, 190 (Debian package missing osutils.pbc) by Benabik :
..github.com/parrot/parrot/issues/190, 193 (load_bytecode on Windows) by Vidmich : github.com/parrot/parrot/issues/193, 199 (windows build broken) by coke : github.com/parrot/parrot/issues/199, 201 (Work-directory references in installed files) by he32 :
..github.com/parrot/parrot/issues/201, 202 (Rewrite tools/dev/nci_thunk_gen.pir to avoid dependency of pcre.) by bacek : github.com/parrot/parrot/issues/202, 203 (Opcode new_si is deprecated.) by bacek : github.com/parrot/parrot/issues/203 00:51
(parrot/parrot) Issues opened : 166 (Humbly suggesting improvements to workflow document) by gitster : github.com/parrot/parrot/issues/166, 169 (crow.pir should fill in digest information) by leto : github.com/parrot/parrot/issues/169, 170 (Release manager guide is too vague about
..how to update docs.parrot.org) by leto : github.com/parrot/parrot/issues/170, 173 (Script to cross-compile Parrot with qemu) by leto : github.com/parrot/parrot/issues/173, 174 (Fix and test .macro_local or remove it from our docs) by leto :
..github.com/parrot/parrot/issues/174, 175 (Sorting benchmarks) by leto : github.com/parrot/parrot/issues/175, 176 (GCC Compile Farm smokers are fragile) by leto : github.com/parrot/parrot/issues/176, 177 (Look at using the "restrict" keyword on pointers.) by petdance :
..github.com/parrot/parrot/issues/177, 182 (Segfault + seemingly crazy memory gobbling) by jnthn : github.com/parrot/parrot/issues/182, 183 (-t4 no longer gives sub-level output) by coke : github.com/parrot/parrot/issues/183, 185 (Memory corruption error) by coke :
..github.com/parrot/parrot/issues/185, 188 (plan(0) not supported) by leto : github.com/parrot/parrot/issues/188, 189 (Build failure in alarm.c on Win32) by jnthn : github.com/parrot/parrot/issues/189, 190 (Debian package missing osutils.pbc) by Benabik :
..github.com/parrot/parrot/issues/190, 193 (load_bytecode on Windows) by Vidmich : github.com/parrot/parrot/issues/193, 199 (windows build broken) by coke : github.com/parrot/parrot/issues/199, 201 (Work-directory references in installed files) by he32 :
..github.com/parrot/parrot/issues/201, 202 (Rewrite tools/dev/nci_thunk_gen.pir to avoid dependency of pcre.) by bacek : github.com/parrot/parrot/issues/202, 203 (Opcode new_si is deprecated.) by bacek : github.com/parrot/parrot/issues/203
(parrot/parrot) Issues opened : 166 (Humbly suggesting improvements to workflow document) by gitster : github.com/parrot/parrot/issues/166, 169 (crow.pir should fill in digest information) by leto : github.com/parrot/parrot/issues/169, 170 (Release manager guide is too vague about
..how to update docs.parrot.org) by leto : github.com/parrot/parrot/issues/170, 173 (Script to cross-compile Parrot with qemu) by leto : github.com/parrot/parrot/issues/173, 174 (Fix and test .macro_local or remove it from our docs) by leto :
..github.com/parrot/parrot/issues/174, 175 (Sorting benchmarks) by leto : github.com/parrot/parrot/issues/175, 176 (GCC Compile Farm smokers are fragile) by leto : github.com/parrot/parrot/issues/176, 177 (Look at using the "restrict" keyword on pointers.) by petdance :
00:52 aloha joined
aloha (parrot/parrot) Issues opened : 166 (Humbly suggesting improvements to workflow document) by gitster : github.com/parrot/parrot/issues/166, 169 (crow.pir should fill in digest information) by leto : github.com/parrot/parrot/issues/169, 170 (Release manager guide is too vague about 00:56
..how to update docs.parrot.org) by leto : github.com/parrot/parrot/issues/170, 173 (Script to cross-compile Parrot with qemu) by leto : github.com/parrot/parrot/issues/173, 174 (Fix and test .macro_local or remove it from our docs) by leto :
..github.com/parrot/parrot/issues/174, 175 (Sorting benchmarks) by leto : github.com/parrot/parrot/issues/175, 176 (GCC Compile Farm smokers are fragile) by leto : github.com/parrot/parrot/issues/176, 177 (Look at using the "restrict" keyword on pointers.) by petdance :
..github.com/parrot/parrot/issues/177, 182 (Segfault + seemingly crazy memory gobbling) by jnthn : github.com/parrot/parrot/issues/182, 183 (-t4 no longer gives sub-level output) by coke : github.com/parrot/parrot/issues/183, 185 (Memory corruption error) by coke :
..github.com/parrot/parrot/issues/185, 188 (plan(0) not supported) by leto : github.com/parrot/parrot/issues/188, 189 (Build failure in alarm.c on Win32) by jnthn : github.com/parrot/parrot/issues/189, 190 (Debian package missing osutils.pbc) by Benabik :
..github.com/parrot/parrot/issues/190, 193 (load_bytecode on Windows) by Vidmich : github.com/parrot/parrot/issues/193, 199 (windows build broken) by coke : github.com/parrot/parrot/issues/199, 201 (Work-directory references in installed files) by he32 :
..github.com/parrot/parrot/issues/201, 202 (Rewrite tools/dev/nci_thunk_gen.pir to avoid dependency of pcre.) by bacek : github.com/parrot/parrot/issues/202, 203 (Opcode new_si is deprecated.) by bacek : github.com/parrot/parrot/issues/203 00:57
01:04 aloha joined
aloha (parrot/parrot) Issues opened : 166 (Humbly suggesting improvements to workflow document) by gitster : github.com/parrot/parrot/issues/166, 169 (crow.pir should fill in digest information) by leto : github.com/parrot/parrot/issues/169, 170 (Release manager guide is too vague about 01:06
..how to update docs.parrot.org) by leto : github.com/parrot/parrot/issues/170, 173 (Script to cross-compile Parrot with qemu) by leto : github.com/parrot/parrot/issues/173, 174 (Fix and test .macro_local or remove it from our docs) by leto :
..github.com/parrot/parrot/issues/174, 175 (Sorting benchmarks) by leto : github.com/parrot/parrot/issues/175, 176 (GCC Compile Farm smokers are fragile) by leto : github.com/parrot/parrot/issues/176, 177 (Look at using the "restrict" keyword on pointers.) by petdance :
..github.com/parrot/parrot/issues/177, 182 (Segfault + seemingly crazy memory gobbling) by jnthn : github.com/parrot/parrot/issues/182, 183 (-t4 no longer gives sub-level output) by coke : github.com/parrot/parrot/issues/183, 185 (Memory corruption error) by coke :
01:07 aloha joined 01:09 aloha joined
aloha (parrot/parrot) Issues opened : 227 (Test. Please ignore.) by bacek : github.com/parrot/parrot/issues/227 01:10
(parrot/parrot) Issues closed : 227 (Test. Please ignore.) by bacek : github.com/parrot/parrot/issues/227
01:11 aloha joined
bacek_at_work cotto, ping 01:17
msg cotto, feel free to setup more github projects to watch. Using "!setgithubproject #channel user/projectname" syntax. 01:18
aloha OK. I'll deliver the message.
bacek_at_work !stgithubproject #parrot parrot/parrot.github.com 01:19
!setgithubproject #parrot parrot/parrot.github.com
aloha, !setgithubproject #parrot parrot/parrot.github.com
meh...
aloha (parrot/parrot.github.com) Issues opened : 1 (Automate main index.html creation) by leto : github.com/parrot/parrot.github.com/issues/1, 2 (Docs for each version of Parrot) by leto : github.com/parrot/parrot.github.com/issues/2, 4 (Add section to index.html on "Development 01:21
..Languages") by ayardley : github.com/parrot/parrot.github.com/issues/4
dalek itor: ab7f567 | NotFound++ | / (3 files):
action queue in Controller
 01:31
01:51 whiteknight joined
whiteknight good evening, #parrot 01:55
dalek CT: 35cf74b | Whiteknight++ | src/packfile/ (8 files):
Add in a few off-the-top-of-my-head files for a Packfile layer to add a prettier interface over the Parrot Packfile PMCs
 02:10
whiteknight msg benabik I just brain-dumped a few files into PACT. Check out src/packfile/* for the start.
aloha OK. I'll deliver the message.
02:38 nbrown joined 03:12 benabik joined 03:17 schmooster joined 03:18 jsut joined 04:16 JimmyZ joined
dalek itor: a6a19d9 | NotFound++ | examples/pizarra.winxed:
use a few more binders in example pizarra
 05:06
05:26 rfw joined 05:40 sri joined 05:41 nine joined
dalek rrot: 25d36d2 | dukeleto++ | docs/binaries/parrot.pod:
[doc] Clean up parrot binary POD
 05:53
dukeleto !pr 05:54
aloha Open pull requests for parrot/parrot.github.com : 0 pull requests open ()
dukeleto bacek_at_work: where is the list of commands that aloha responds to? 05:56
bacek_at_work++ # github integration
06:39 JimmyZ joined
JimmyZ !help github 06:39
!help github 06:40
cotto ~~ 06:48
bacek ~~ 07:53
dukeleto, erm. I don't think that such list exists
!pr parrot/parrot
aloha Open pull requests for parrot/parrot.github.com : 0 pull requests open ()
bacek aloha, help github
aloha bacek: I don't know anything about 'github'.
bacek aloha, help github::announce 07:54
aloha bacek: Announce new/changed issues and pull requests, and, soon, pushes.
bacek aloha, help github::pullrequests
aloha bacek: Monitors outstanding pull requests on a GitHub project.
..
..Allows use of a !pr command to fetch the current count of open pull requests
..including a tally by user.
..setting the user_github_project setting in the bot's store).
bacek aloha, help karma
aloha bacek: Gives karma for or against a particular thing. Usage: <thing>++ # comment, <thing>-- # comment, karma <thing>, explain <thing>.
bacek aloha, list
aloha, help 07:56
aloha bacek: Ask me for help about: msg, convert, status, vars, karma, auth, chanop, seen, maths, github::announce, infobot, clock, translate, github::pullrequests, loader (say 'help <modulename>').
bacek dukeleto, yes, something like this
aloha, help clock 07:57
aloha bacek: No help for module 'Clock'. This is a bug.
bacek aloha, clock?
aloha bacek: LAX: Wed, 23:57 PST / CHI: Thu, 01:57 CST / NYC: Thu, 02:57 EST / UTC: Thu, 07:57 UTC / LON: Thu, 07:57 GMT / BER: Thu, 08:57 CET / TOK: Thu, 16:57 JST / SYD: Thu, 18:57 EST
08:18 mj41 joined 08:56 rfw joined
nine Good morning, #parrot 09:31
09:40 AzureStone joined 09:59 lucian joined
mls morning! 10:03
tadzik morning 10:06
dalek kudo/nom: d9890fb | moritz++ | t (2 files):
bump to a NQP revision which implements :i

also mark some tests as #icu that now need it		 10:07
11:12 autark joined 11:16 JaphMalik joined
dalek kudo/take-recont: 2453985 | moritz++ | src/core/Num.pm:
return integers from rounders
 11:16
kudo/take-recont: 8c38b99 | moritz++ | / (2 files):
use nqp::p6recont_ro in &take
mls nine: the Poiner_Array code looks ok to me 11:21
nine: I guess you run into trouble because you mark PMCs belonging to another interpreter, thus belonging to another GC
nine: each PMC is always linked into one pointer array, thus belongs to a GC 11:22
afk -> lunch 11:33
11:39 nbrown joined
nine mls: did you find a place where I mark a PMC belonging to another interp? Because I was quite careful not to do that. And why would deactivating the re-use of removed cells make my problem disappear? 12:07
12:11 AzureStone joined 12:17 lucian joined
dalek rrot/threads: f6ae12e | nine++ | / (61 files):
Merge branch 'master' into threads
 12:58
rrot/threads: 9ed2d22 | nine++ | / (2 files):
New function Parrot_thread_create_proxy
rrot/threads: 0cb1f00 | nine++ | src/ (3 files):
Remove obsolete check for parent_interpreter from Parrot_cx_init_scheduler

Every interp needs its own scheduler. So it makes no longer sense to not create a scheduler in a child interp. This now makes it in theory possible to run a completely independent interp in another thread by making clone_interpreter no longer strictly neccessary.
rrot/threads: 72907ae | nine++ | / (3 files):
Distribute scheduled tasks onto worker threads

The scheduler now tries to be a little smarter by finding the worker thread with the lowest task count and scheduling the new task onto this thread.
The downside is that scheduling a task needs to allocate new PMCs in the thread interp. To not run into concurrency problems in the GC, the scheduler locks a mutex while running a task, so the scheduling may have to wait until the current task is preempted.
mls nine: I don't know. I'd add a "if (!Parrot_gc_pool_is_owned(interp, self->pmc_allocator, item)) abort();" line to gc_gms_mark_pmc_header(), so that it aborts if it tries to mark a PMC belonging to another pool 13:19
(this will slow down parrot quite a bit, so it's only for debugging purposes) 13:21
13:35 whiteknight joined, AzureSto_ joined
whiteknight good morning, #parrot 13:38
msg nine I saw your commit to distribute scheduled tasks to worker threads. Very nice. I wonder if we should add a second parameter to schedule op to say "schedule on current thread only" versus "schedule wherever you want" versus "schedule on the exact thread specified 13:40
aloha OK. I'll deliver the message.
whiteknight msg nine like a new schedule_p_p op, where the second parameter is a thread handle (and if null, is the current_thread) 13:41
aloha OK. I'll deliver the message.
dalek kudo/take-recont: a7eed04 | moritz++ | t/spectest.data:
run perl.t
 14:13
14:30 autark joined 14:41 alvis joined
PerlJam whiteknight: part of moritz's role with the Perl 6 Advent Calendar is to drum up contributors. You might want to do the same :-) 14:52
whiteknight That's the plan 15:06
it's about 50% "Here's cool stuff" and 50% "if you would like to volunteer... 15:07
PerlJam I can volunteer ideas of what I'd like to see, but it's doubtful that I'd be able to contribute given that I haven't really kept up with parrot 15:08
whiteknight yeah, ideas are definitely welcome 15:10
PerlJam Where should I put them? (we use the perl6-mu repo over in perl6land) 15:11
whiteknight just msg them to me
dalek itor: 237e728 | NotFound++ | src/Guitor (2 files):
sync the event queue before processing queued actions
 15:12
itor: 9279603 | NotFound++ | examples/pizarra.winxed:
fix handler of quit button
itor: 664c1ed | NotFound++ | examples/pizarra.winxed:
method naming more consistent
 15:17
itor: bc71a0f | NotFound++ | src/Guitor.winxed:
Pixmap class
 15:30
nine mls: thank you very much! That's exactly the line I need to make debugging this stuff so much easier. 15:31
whiteknight: I've been thinking about the same thing. The way it is now, schedule_immediate always schedules on the current thread while the schedule op never schedules anything on the main thread. I'd guess that this will work out quite well but giving more options to the HLL can only help 15:33
whiteknight: users could help greatly for such questions :) 15:34
15:36 Psyche^ joined
whiteknight yes 15:38
nine: schedule_immediate is a new op? 15:42
nine whiteknight: no, at least not yet. For now it's accessible through interp."schedule"(task) 15:45
whiteknight okay, that works too 15:47
nine But one should not even use interp."schedule". interp."schedule_proxied"(task, proxy_pmc) is what should be used for scheduling updates of another task's data
whiteknight so you can use that with a handle to any interp?
nine schedule_proxied schedules the given task on the interp that the proxy_pmc is proxying to. Had to do it this way because having a reference to another interp in a register just screwed up GC 15:48
15:49 Aleksandar joined
whiteknight okay 15:49
can you have a proxy for the foreign interp that doesn't screw up GC?
15:50 olaf joined
nine I really should have written down the different ways to make this work I tried. I think I tried using a proxy for the other interp but IIRC somewhere between find_method and get_namespace I always ran into trouble 15:52
whiteknight find_method I expect to be a big hassle, personally 15:53
if the method is just a Sub reference, the bytecode is immutable and there shouldn't be a problem
but if the method is a closure with hidden state that's being adjusted across threads, that's a problem 15:54
15:54 snearch joined
whiteknight or a coroutine with a living context 15:54
15:54 snearch joined
nine I guess simply adding an op would have been much easier than trying to get a method call to work 15:55
whiteknight I almost want to add an is_thread_safe flag on the PMC structure that would indicate objects that needed to be proxies and those that do not
method calls and certain vtable accesses are just going to require a lot more thought
nine Well for now I have a system that at least works. I'm certain that it can be improved in a lot of places :) But at least it should prove the concept and allows experiments further up the stack 15:58
whiteknight nine: Whenever you think you've hit a good stopping point, I can come in and start cleaning and stuff 15:59
nine I hate that I had to take a lock while a thread is executing a task. But as long as scheduling has to create new PMCs I just see no other way.
whiteknight I never thought the system would be lock-free, only that we wouldn't *need* to expose locking to the user 16:00
dalek itor: 4c406da | NotFound++ | / (3 files):
fix and improve Pixmap
nine whiteknight: well the lock should not have any measurable performance inpact on a thread's execution. It's just scheduling a new task on the thread that may be blocked for quite a long time making my little test program just much less impressive :) 16:01
whiteknight Can we add something like a "queue of tasks which need to be scheduled"? 16:05
a separate queue in the scheduler that expects asynchronous inputs 16:06
then all we need to do is lock that queue when reading or writing to it, and otherwise don't lock
we are running out of tasks for GCI 16:16
We have 6 tasks left until we replenish on the 16th
and at the moment, we only have 6 more tasks scheduled to be published on the 16th
aloha coverage? 16:20
aloha whiteknight: coverage is cv.perl6.cz or tapir2.ro.vutbr.cz/cover/cover-results/
NotFound whiteknight: task suggestion: rewrite tools/release/crow.pir and runtime/parrot/library/Crow.pir and improve its robustness and diagnostic messages 16:21
whiteknight NotFound: rewrite it in winxed? 16:22
NotFound In winxed or in nqp.
(of course I'll like winxed better) 16:23
dukeleto ~~ 16:26
there are some simple TT's about improving features on crow.pir as well 16:27
perhaps converting it to winxed will make those scripts more enjoyable to maintain
NotFound dukeleto: That's the intention.
dukeleto the VLC people are just creating 50 tickets like "write an assembly function"
a bit of an abuse of the system, but still seemingly allowed 16:28
if necessary, we can create 3-5 tickets about the same thing for very large tasks
I think that is very reasonable, and will defend it to Carol if somebody has the audacity to accuse of cheating again.
NotFound It's not a bad idea after all, a lot if tiny tasks may mean more students taking contact with the project. 16:29
dukeleto manifest__t/manifest/02-regenerate_file.t is failing 16:33
in fulltest
anybody seen that?
whiteknight nope
dalek itor: 6c2b308 | NotFound++ | examples/edit.winxed:
use a generic non fixed width font and allow to specify its size
 16:39
whiteknight dukeleto: We had a few refactoring tasks in the last group that I might like to see a second attempt at, just to see what other students come up with 16:43
dukeleto whiteknight: go for it 16:46
whiteknight: nothing in the rules against competing refactorings
whiteknight I'm still finding a few gems in the coverage results, but they aren't as obvious as "bring file X up to Y coverage" 16:47
some bits of dead code, especially in IMCC, could use the scalpel
compilers/imcc/optimizer.c is especially loaded with untested and possibly dead code 16:48
dalek itor: 2769e75 | NotFound++ | examples/pizarra.winxed:
rename 'quit' as 'close' in example pizarra
 16:54
whiteknight bleh, that whole file needs to be killed
dukeleto +1 deleting IMCC files 16:55
16:58 dmalcolm joined
dalek kudo/nom: a2f3f46 | jonathan++ | t/spectest.data:
S13-overloading/metaoperators.t passes entirely, so turn it on.
 17:06
itor: 1a692ba | NotFound++ | src/Guitor.winxed:
fix menu width
 17:23
17:32 japhb joined 17:33 benabik joined
benabik o/ #parrot 17:33
17:42 fperrad joined 17:51 aaanderse joined
dalek itor: 23dba7a | NotFound++ | src/Guitor (2 files):
improve selection events
 18:00
Aleksandar Can someone tell me how to get exponentation in winxed whiteknight tell me with $loadlib "trans_ops" and then ${ pow out, in, n } but i get error saying expected expresion near , 18:05
18:09 japhb joined
Aleksandar i find my mistake 18:10
NotFound No need to pirop, there is a pow builtin. 18:11
And exp, too.
Aleksandar i have forgot to write pow 18:12
NotFound Aleksandar: use the builtin and winxed take care of the loadlib when needed. 18:13
nine whiteknight: having a separate queue might be possible and sounds like a great idea. I guess it would work as long as the actual scheduling later cannot possibly cause any garbage collection in the main thread. 18:19
benabik whiteknight: ping
nine mls++ # finding still shared globals by the minute thanks to your suggestion! 18:20
whiteknight benabik: pong 18:25
benabik Ack! Now I have to break up the long aloha msg. :-D
whiteknight: First off, the packfile stuff looks pretty good. :-) API might need tweaking, but actual usage will bear that out best. 18:28
whiteknight benabik: thanks! It was a quick brain-dump and I didn't want to work on it too much more till you looked at it 18:29
benabik whiteknight: Can we reliably separate operands and results? I know there are in/out markers for op arguments, but is there a guarantee that out comes before in? And what happens with inout?
whiteknight obvious there is a lot of functionality missing :)
that's a good question, I don't know
benabik We might just want to treat all of them as just arguments, and let the order and usage determine which is which. 18:30
whiteknight that's fine too
probably easier that way anyway
benabik Easier is good. :-D
whiteknight okay if you like the general shape of things, I can start beating it all up into shape or you're welcome to poke at it yourself 18:31
benabik The artificial "one result" thing caused by the tree structure is something I'm unhappy with in POST. It tends to be true, but what happens when it's not? Blah.
whiteknight once I get something buildable, I'll start adding some build infrastructure
benabik I'm unlikely to poke it in the next week.
whiteknight okay, I don't know what my schedule looks like 18:32
benabik Maybe I'll turn this little pile of notes into comments. :-/
It's a race! Whoever has free time first works on code. :-D
whiteknight according to me: schedule empty. According to my wife: schedule FULL AND EVERYBODY PANIC
benabik Sounds about right.
whiteknight holidays drive her completely insane
benabik I stay sane mostly by ignoring them. I have the excuse of PhD applications this year that my wife can't force me to pay attention. :-D 18:33
(Downside is that I really do have to do work for these apps.
whiteknight man, I really need to get moving on my own PhDness
benabik I don't have any plans post-MS other than PhD, so I kinda need to apply. 18:34
Otherwise come fall I have no job. :-(
atrodo jobs are overrated 18:35
benabik atrodo: But home and food are not.
atrodo benabik> One word: coupons
benabik atrodo: I dunno if my mortgage company takes coupons, but I'll have to check. ;-) 18:36
atrodo because according to TLC, i can feed a family of 100 with $3.29
cotto ~~ 18:45
dalek CT: 3b91fbf | benabik++ | src/packfile/ (7 files):
Add packfile commentary

I prefixed all my commentary with BCG: so they can be easily searched for and killed when the time comes.		 19:25
benabik whiteknight: I pushed some commentary. Just decided to do it inline for simplicity. 19:26
dalek itor: 2c1484b | NotFound++ | / (3 files):
initial support for property changes and send events, SelectionRequest events
itor: 13e7579 | NotFound++ | examples/text.winxed:
copy to clipboard option in example text
Aleksandar is there a way in winxed how a function can call itself without using its name ? 19:27
NotFound Aleksandar: in general no. What are you trying to do? 19:29
Aleksandar fucntion that returns recursive function
benabik Does var f = function() { f() }; work? 19:31
(Obviously it diverges, but beyond that.)
NotFound It should.
cotto benabik++ 19:32
istr winxed having some magic constant that referred to the name of the current function 19:33
Aleksandar hmm it compiled i will test it now
NotFound The only pitfall I see is that it forces the local function to be a closure even if no more lexicals are used.
cotto: It has, but the name is not very useful other than for debug messages. 19:34
cotto ok 19:35
NotFound Maybe I must add a magic constant that refers to the current function by its subid. 19:36
Aleksandar ok it worked thanks a lot :) 19:38
whiteknight if we need true anonymous recursion we're going to want to use a Y combinator
luckily for everybody involved, I happen to have an implementation of one that automatically memoizes results...
20:48 zby_home joined 20:52 lucian joined 21:10 GodFather joined 21:13 GodFather_ joined
dalek itor: 766e443 | NotFound++ | / (2 files):
selection and copy to clipboard in EditBox
 21:48
23:25 whiteknight joined
whiteknight good evening, #parrot 23:32
benabik o/ whiteknight 23:33
23:34 davidfetter joined
dalek sella/gh-pages: 01aca9d | Sasho++ | winxed/examples.md:
Examples added
 23:35
sella/gh-pages: 0802b1f | Whiteknight++ | winxed/examples.md:
Merge pull request #36 from Alekssasho/gh-pages

Google Code In Task Expression-less Switch Documentation
whiteknight hello benabik 23:39

Powered by Cro and the Raku® Programming Language.

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

Thank you!