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.
cotto fizzbuzz in ruby without ruby: experthuman.com/programming-with-nothing 00:11
00:35 NotFound joined
benabik cotto: I had to skip ahead to the end just to be sure he wasn't trying to develop the lamba calculus from scratch. :-D 01:07
cotto it does have that kind of feel 01:09
I don't even want to think about how floating point would work
benabik Hah. Well you can encode booleans and lists, so you can just use lists of booleans and IEEE 754. :-D 01:10
cotto ow
benabik I suppose you could use lists of Church numerals to be continued fractions. 01:12
01:16 whiteknight joined
benabik Really, with the untyped lambda calculus any solution is pretty painful. 01:18
whiteknight good evening, #parrot
benabik o/ whiteknight
whiteknight hello benabik
dalek itor: 31f731e | NotFound++ | src/Guitor.winxed:
fix Window.SetSelectionOwner
 01:26
whiteknight I want to try to get interp_api_cleanup mergable 01:29
01:37 preflex_ joined
whiteknight nqp builds with that branch and is passing tests so far. That's a good sign 01:42
dalek itor: 8db3d53 | NotFound++ | examples/pokedit.winxed:
keyboard select and copy in example pokedit
NotFound I'm near to the point of closing the circle: using a parrot based winxed written text editor to hack on parrot and winxed. 01:44
whiteknight that would be totally awesome
double awesome 01:45
NotFound And of course, to hack on the editor itself. 01:46
whiteknight of course
benabik NotFound: You could use compreg to evaluate snippets of code directly in the editor. :-D 01:51
NotFound benabik: yeah 01:56
benabik Something I do regularly while experimenting with code is have winxed open in Vim and run :!winxed % to check that it does what I want. 01:57
NotFound benabik: I wrote years ago a text mode text editor with perl embeded that allows that. Pokedit will be the next generation of such beast. 01:58
whiteknight it would make some of my Rosella work much easier 02:00
NotFound The problem is that more people will blame us about discriminating windows ;) 02:04
benabik There are X servers available for windows. 02:07
NotFound benabik: yeah, but I think most users won't even try to install any. 02:09
dalek rrot: 6f1762b | Whiteknight++ | / (77 files):
Merge branch 'master' of github.com:parrot/parrot into separator_fixes
 02:13
itor: 50c6995 | NotFound++ | examples/pokedit.winxed:
use the primary selection and buffer copy clipboard
 02:22
whiteknight t/compilers/pge/p5regex/p5rx.t .............. 523359/960 02:27
....lolwat?
I had to kill the test, it just kept going
ttbot Parrot 6f1762be MSWin32-x86-multi-thread make error tt.taptinder.org/cmdinfo/62230 02:31
whiteknight bleh 02:32
wtf, I thought I had that branch tested on windows
02:39 alvis joined
whiteknight I'll fix it tomorrow morning. Goodnight 02:58
dalek rrot/interp_api_cleanup: 0322da8 | jkeenan++ | / (2 files):
Update MANIFEST and .gitignore.
 03:20
rrot: b428cd5 | jkeenan++ | .gitignore:
Have git ignore some .o files generated during testing.
 03:31
03:34 Psyche^ joined 03:39 Psyche^ joined
dalek itor: d230497 | NotFound++ | examples/pokedit.winxed:
selection with mouse in example pokedit
 03:44
03:46 nbrown joined
ttbot Parrot b428cd5d MSWin32-x86-multi-thread make error tt.taptinder.org/cmdinfo/62270 03:46
03:51 davidfetter joined
dalek href="https://parrot.github.com:">parrot.github.com: 660d252 | (Alvis Yardley)++ | / (3652 files):
Add docs from v3.0.0 to v3.8.0.
 04:39
href="https://parrot.github.com:">parrot.github.com: b6c27f7 | (Alvis Yardley)++ | / (3652 files):
Merge branch 'ayardley/earlier_docs_versions'
benabik That's a big commit. 04:45
alvis++
dalek href="https://parrot.github.com:">parrot.github.com: af70c55 | (Alvis Yardley)++ | index.html:
Changed 'Earlier' to 'Previous' in the sub-heading 'Earlier Documentation Releases'.
 04:48
benabik alvis: You should add yourself to CREDITS so dalek gives you karma.
:-D
alvis hello benabik.
benabik: Ok, I'll do that. thanks. 04:49
benabik alvis: Thanks for doing the work on the doc versions.
sorear karma Alvis Yardley
aloha Alvis Yardley has karma of 16.
alvis benabik: you're welcome. thanks for doing the work on PACT. i'm looking forward to using it.
benabik I came up with a cunning plan for PACT. Since I'm going for my PhD, I should still be able to apply for GSoC. Three months of paid time on PACT. :-D 04:50
alvis benabik: wow! that works. :) 04:51
benabik Assuming whiteknight++ doesn't run away with it while I'm doing my thesis.
alvis benabik: never know. he definitely has wheels! 04:52
ok, gtg. and thanks sorear on the karma report. i'll fix up the CREDITS. 'night all. 04:55
dalek rrot: 0775735 | alvis++ | CREDITS:
Update CREDITS
 05:03
05:08 jsut joined
ttbot Parrot 07757354 MSWin32-x86-multi-thread make error tt.taptinder.org/cmdinfo/62308 05:19
benabik ttbot is going to be very unhappy until Win32 is fixed from that merge. 05:21
6f1762b, which merged separator_fixes (for anyone curious)
05:25 rfw joined 06:06 alin joined 08:36 mj41 joined 08:42 alin joined 08:48 alin left 09:48 lucian joined 10:02 bacek joined 10:10 preflex_ joined 10:28 baest joined
dalek kudo/nom: 3504a85 | moritz++ | src/ (2 files):
throw a X::Attribute::Undeclared when encountering an undeclared attribute
 10:31
10:39 snearch joined 10:45 snearch joined 10:46 snearch joined 12:55 baest joined 13:01 baest_ joined 13:03 baest joined 13:13 whiteknight joined
whiteknight good morning, #parrot 13:13
bleh, I still can't do a normal build on windows, much less find the problem that's screwing up taptinder 14:04
My compiler is giving me crap that it can't find external symbol _environ. So I rip out all the code that uses that symbol (which I think is not necessary) and then miniparrot segfaults 14:05
miniparrot isn't using any of the environ stuff anyway, so it shouldn't be an issue
dalek rrot: 6f46971 | Whiteknight++ | / (3 files):
Fix the build on my particular flavor of windows (x86-64, win7, cl 16.0-x64)

Remove all the integer-keyed lookup logic from Env Pmc, since that behavior is not really used and isn't speced. pbc_dump doesn
  't appear to have the PMCNULL variable linked for some reason, but also doesn't seem to use it, so just dummy it up. Remove a setjmp call from task.pmc which appears to be using an uninitialized buffer and causing a segv here.		 14:16
whiteknight now we just have to wait and see if taptinder is going to like that or not 14:19
it built succesfully for me here and is running tests 14:20
I think part of my problem is that I'm using MSVC here only, and not using strawberry perl. I suspect the taptinder that was failing is using strawberry 14:24
dalek rrot: 98e3edc | Whiteknight++ | t/pmc/env.t:
Remove t/pmc/env.t tests that used the iterator and array-index operations of Env
 14:28
ttbot Parrot 6f469715 MSWin32-x86-multi-thread make error tt.taptinder.org/cmdinfo/62361 14:34
whiteknight ...nope 14:37
14:38 benabik joined
whiteknight benabik! Do you have access to a windows machine? 14:38
benabik whiteknight: yes
(Well, a virtual machine.) 14:39
whiteknight benabik: Can you try a build with master HEAD? It's building on my windows machine, but taptinder is still not happy
benabik whiteknight: Can do.
whiteknight it looks like that failing taptinder box is using msvc not strawberry like I assumed, so I won't bother installing strawberry right now 14:40
benabik Ah, well, I use Strawberry as well. I suppose I could clone it onto a school machine. I think they have MSVC. 14:41
whiteknight just try it with strawberry. 14:42
more data points is good
I *think* that taptinder machine belongs to Coke
dalek rrot: 9a21498 | Whiteknight++ | / (2 files):
Change WIN32 to _WIN32 to get the library path logic working correctly on my machine. I don't know which of the two (or something different entirely) should be used for real. Fix t/pmc/env.t
 14:45
benabik Build started. 14:46
ttbot Parrot 9a21498b i386-linux-thread-multi make error tt.taptinder.org/cmdinfo/62387
whiteknight FFFFUUUUUU
benabik At least it fails quickly? 14:47
whiteknight no, that last one was a stupid typo
benabik Just got the same error, actually. :-D
whiteknight here comes the stupid fix for the stupid typo 14:48
whiteknight--
dalek rrot: 001f65d | Whiteknight++ | src/library.c:
fix stupid typo
moritz just just be #else
moritz not #elif
whiteknight moritz: yeah, way ahead of you
ttbot Parrot 98e3edcc MSWin32-x86-multi-thread make error tt.taptinder.org/cmdinfo/62376
whiteknight it is way too early for this kind of crap
benabik Error parsing option config_lib.pir? 14:49
whiteknight yeah, I think the path separators on that machine aren't being read correctly
windows sometimes will let you use / instead of \\ in paths, but sometimes it will mis-parse the args and give weird errors 14:50
benabik src/call/context.c:682: failed assertion 'Parrot_pcc_get_regs_used(interp, ctx, REGNO_STR) > idx' 14:51
whiteknight ...lolwat?
benabik Running .\\parrot.exe pbc_to_exe.pbc pbc_to_exe.pbc
whiteknight and that's after a realclean and a git pull?
benabik After a git pull. It had just compiled pbc_to_exe... 14:52
I ran configure, maybe I forgot to clean?
whiteknight I don't know, but that's a very weird error to see 14:53
I see those kinds of errors most often when trying to run an old bytecode file with a newer parrot
benabik Yeah. Maybe I hadn't cleaned and one of the libraries was old. Starting over. 14:56
whiteknight taptinder hasn't crapped on my last commit yet, so that's a good sign 14:57
all this path logic and library logic really needs a massive rewrite 15:00
Almost the entire src/library.c file needs the boot 15:01
I suspect Parrot_init_library_paths can be rewritten in winxed and added to frontend/parrot2/prt0.winxed 15:04
same with Parrot_lib_updte_paths_from_config_hash
all these path-handling routines should probably just be moved to src/platforms/* 15:06
benabik whiteknight: Build complete, running tests 15:07
whiteknight okay, that's a good sign
and taptinder is still silent about my last commit too
moritz just tried to build rakudo on latest parrot, and got 15:09
===SORRY!===
get_iter() not implemented in class 'Env'
whiteknight moritz: damnit. I didn't think Rakudo was using that get_iter on Env
moritz whiteknight: oh, you removed it?
whiteknight moritz: yeah, it was breaking the windows build
damnitdamnitdamnit 15:10
moritz now it breaks the rakudo build :(
whiteknight okay, let me look for a replacement
okay, I think I see what the thing was doing 15:15
it is returning a list of all names of environment variables
ttbot Parrot 001f65d0 MSWin32-x86-multi-thread make error tt.taptinder.org/cmdinfo/62427 15:16
benabik whiteknight: And there's the failure. Just took longer. :-(
May want to just revert the branch (and attempted fixes) until we can figure out how to fix MSVC. 15:18
whiteknight: Test failures on Strawberry - t/pmc/eval.t 9, t/dynpmc/os.t 13, t/library/cgi_query_hash (planned 7, ran 1), t/library/nciutils.t 2-5 15:22
I seem to recall this testing cleanly, but that may have been a very old version.
Hm. Yesterday my parrot was 3.10.0-91-g2b3bb7e, and I'm pretty sure it tested cleanly. 15:23
benabik <3s `git describe --tags @{yesterday}` 15:30
dalek itor: 4bec695 | NotFound++ | src/Guitor.winxed:
minor fixes and more checks in drawable handling
 15:48
itor: ad334e4 | NotFound++ | examples/pokedit.winxed:
scroll with mouse wheel and improvements in selection handling
 15:52
benabik Wow. The git mailing list taught me today that Mail.app doesn't like utf-16be messages. O.o 15:57
whiteknight: fullest on master on my Win7/Strawberry box: smolder.parrot.org/app/projects/rep...ails/24898 16:03
*fulltest
whiteknight bleh, what's the incantation to undo that merge?
sorear git revert maybe? 16:04
benabik git revert <merge> should DRTR, I think.
moritz aye
benabik Hm. May need to specify a parent with -m# 16:05
whiteknight that seems to have removed the merge commit, but not the subsequent fix commits 16:06
I basically want to roll back several commits
benabik git-revert can take multiple commits, or uhm... 16:07
sorear benabik: DTRT?
[you mean]
benabik Do The Right Thing
Yes.
I apparently shouldn't be allowed near anything important with a keyboard today… Shame I work in a sysadmin office.
whiteknight okay, if I keep touching this I'm just going to break it worse 16:13
...I think it's reverted? 16:19
moritz hasn't seen a push 16:21
whiteknight the offending commits appear to not be visible on github anymore
I don't think dalek reports things that aren't new commits 16:22
benabik Rewound master?
moritz did you git push --force?
whiteknight basically, yes
benabik I still see the merge, but I don't see the ones from earlier today. 16:23
whiteknight okay, well that's a start
so why the hell didn't the merge revert?
or, why doesn't it look reverted?
benabik I've reverted the merge locally, can push if needed. 16:25
moritz if you want to continue the push --force route, you could
git reset --hard f43f011d15
git push --force
# but I'd prefer to simply revert the commit
benabik prefers not to alter time and more than needed.
whiteknight okay, how do I know which number to use for the -m option in git revert? 16:27
benabik It's -m2
moritz whiteknight: the merge commit has two hashes
benabik It's the mainline, counting from 1. 16:28
moritz if you do git log $hash for each hash, you see which branch they refer to
and you can see that the first is the branch, and the second is master
benabik Simple version: if it's "merge branch into master" use -m1 if it's "merge master into branch" use -m2 :-D
Checking the hashes is more reliable. :-D
dalek rrot: 1a3b2f2 | Whiteknight++ | / (4 files):
Revert "Merge branch 'master' of github.com:parrot/parrot into separator_fixes"

This reverts commit 6f1762be1aa6c3b617696e4bc4a954a239ad45f4, reversing changes made to f43f011d1534c3660bd8fdb161323dcdebfb5009.		 16:29
whiteknight that seems to do it, I think 16:32
benabik With minor amounts of time travel. :-/
But yes.
whiteknight whatevs 16:33
I was really hoping that branch would just merge in cleanly. Nothing is ever easy
windows--
And this environment crap has been preventing me from building here, for whatever reason, in weeks 16:34
benabik If I get spare time today, I'll see about getting Parrot to build on the school computers. They have MSVC but I don't know about perl.
whiteknight that would be a big help, if you can pull it off
benabik: do they have matlab?
benabik Alternatively, I could install MSVC on my machine. :-/ 16:35
whiteknight matlab comes bundled with perl, last I checked, though it might not be in your path
benabik We used to have it, not sure anymore. 16:36
whiteknight okay
benabik Looks like `perl` is 5.12 in C:\\Perl64\\ with a patch labeled "ActivePerl", so I guess Activestate. 16:39
whiteknight yeah
benabik Might need to figure out how to locally install a couple modules. But other than that should be good to go. 16:40
dalek rrot/win32_64_fixes: b6c8207 | Whiteknight++ | src/pmc/env.pmc:
Fix a preprocessor directive in Env PMC, and add in some TODO notes about places we need to fix for windows (environ global is deprecated there)
 16:47
benabik Shouldn't -DWIN32 be added by Configure? 16:48
whiteknight I don't know, but it isn't 16:49
the compiler defines _WIN32
benabik We seem to use _WIN32 and WIN32 somewhat randomly. :-/
whiteknight WIN32 seems to never be defined 16:50
so anywhere that WIN32 is used, it's probably an error
unless other windows are different from mine
benabik It's in config/init/hints/mswin32.pm
It's a Parrot #define, not a compiler #define. 16:51
dalek itor: 0e0f052 | NotFound++ | examples/pokedit.winxed:
horizontal scroll
 16:53
17:00 cogno joined
dalek itor: 4fdd495 | NotFound++ | src/Guitor.winxed:
fix DrawPoint
 17:04
17:19 cogno joined 17:44 cogno joined 18:14 cogno joined
dukeleto ~~ 19:06
dalek rrot/ayardley/documentation_revisions: ea26f42 | alvis++ | docs/ (2 files):
Updated and relocated 'TODO_revisions' to 'docs/revisions'.
 19:07
rrot/ayardley/documentation_revisions: 4ef8e79 | alvis++ | docs/revisions/docs/manuals/specification/ (14 files):
Renamed 'docs/revisions/docs/manuals/reference' to 'docs/revisions/docs/manuals/specification'.
rrot/ayardley/documentation_revisions: 1190fa8 | alvis++ | docs/revisions/docs/manuals/reference/.gitignore:
Removed 'reference' directory because going with 'specification' directory instead.
rrot/ayardley/documentation_revisions: 5912255 | alvis++ | docs/revisions/docs/manuals/specification/ (9 files):
Added remaining chapter placeholders to the VM Specification and Reference Implementation.
rrot/ayardley/documentation_revisions: f8389e5 | alvis++ | docs/revisions/docs/manuals/specification/appa_parrot_intermediate_programming_language.pod:
Remove (rename) 'appa_parrot_intermediate_programming_language.pod'
rrot/ayardley/documentation_revisions: 7923fd1 | alvis++ | docs/revisions/ (6 files):
Added an informational note file.
rrot/ayardley/documentation_revisions: 127afd4 | alvis++ | docs/revisions/TODO_revisions:
Updates to revision effort's TODO
rrot/ayardley/documentation_revisions: c456225 | alvis++ | docs/revisions/docs/mans/ (25 files):
Renamed directory from 'manuals' to 'mans'.
rrot/ayardley/documentation_revisions: 726de08 | alvis++ | docs/revisions/docs/mans/ (23 files):
Renamed directories 'developer' to 'dev' and 'specification' to 'vmspec'.
rrot/ayardley/documentation_revisions: dd0e8ba | alvis++ | docs/revisions/docs/mans/vmspec/.gitignore:
Add (mostly empty) files which layout out the chapter organization of the 'PVM Specification and Implementation Reference' document.
rrot/ayardley/documentation_revisions: 0129e09 | alvis++ | docs/revisions/docs/mans/vmspec/ (4 files):
Reordered the appendicies.
rrot/ayardley/documentation_revisions: 0a43668 | alvis++ | docs/revisions/docs/mans/vmspec/pr (2 files):
Updates to 'proposal.txt' and 'preface.pod'.
rrot/ayardley/documentation_revisions: 42f5086 | alvis++ | docs/revisions/docs/mans/ (23 files):
Removed old directories 'developer' and 'specification'.
rrot/ayardley/documentation_revisions: 33e61c6 | alvis++ | docs/revisions/docs/manuals/ (27 files):
Removed old 'manuals' directory'.
benabik What happened to heuristic merges?
nine Wow...I've spent so much time in gdb that I recognize PMCNULL instantly by it's address... 19:09
benabik :-D 19:10
cotto ~~ 19:15
dalek itor: f5c6580 | NotFound++ | examples/pokedit.winxed:
status line and fixes in cursor position
 19:17
19:19 snearch joined
cotto #ps in 8 19:21
dalek itor: 0301085 | NotFound++ | examples/pokedit.winxed:
fix status line drawing
 19:22
dukeleto benabik: those may have been broken when aloha got updated recently 19:35
benabik dukeleto: Updating aloha broke dalek? 19:36
dalek itor: 758e328 | NotFound++ | examples/pokedit.winxed:
update status on mouse
 19:43
itor: 4f44e09 | NotFound++ | examples/pokedit.winxed:
update status after paste
 19:47
cotto johnkary.net/git-1-7-8-changes-for-...developer/ 19:49
dalek rrot/threads: 8aa61aa | nine++ | / (5 files):
Proxy namespaces and classes as well

With this change, threads have read only access to their parent's namespaces and classes. Though this avoids any problems with cloning namespaces and allows task tests to use the imported ok function and thread tests to pass, it's still not perfect.
Classes can be instantiated on a thread if they have been instantiated before in the main thread.		 20:00
20:01 bluescreen joined
nine I fear, Parrot_Class_instantiate will need some refactoring so I can run the first time instatiation stuff on the main thread and create the object in the child thread 20:03
benabik You can do it! 20:05
nine I'll see what I can do tomorrow :) Time for bed now 20:10
20:14 dmalcolm joined 20:48 cosimo joined
dalek itor: 2e06c10 | NotFound++ | examples/pokedit.winxed:
alert window and save menu item
 21:13
21:16 mj41 joined 21:17 Yuki`N joined 22:21 schmooster joined
dalek kudo/nom: ad9e72c | jnthn++ | src/Perl6/Actions.pm:
Fix when, so $/ will be properly set.
 22:27
Yuki`N msg dukeleto 2nd round of tasks comes out soon...do you have anything in mind?
aloha OK. I'll deliver the message.
dalek itor: e13c0bc | NotFound++ | examples/pokedit.winxed:
cut menu option and ctrl-x
 22:33
22:39 davidfetter joined
dalek itor: 2acc193 | NotFound++ | examples/pokedit.winxed:
pos and line starting at 1 in status line
 22:50
kudo/nom: 32dc1ed | jnthn++ | src/core/Failure.pm:
Fix double-stringification of Failure causing death bug; fix suggested by mls++.
 22:56
kudo/nom: 483b3fc | jnthn++ | docs/ROADMAP:
Take a crack at getting us a ROADMAP that isn't over a year out of date. Patches welcome - this sure won't be perfect.
 23:18
kudo/nom: 81a1031 | jnthn++ | docs/ROADMAP:
Attempt a clarification (tadzik++).
 23:26
kudo/nom: 6fd6552 | (Geoffrey Broadwell)++ | docs/ROADMAP:
Add four of my projects to the ROADMAP
 23:43

Powered by Cro and the Raku® Programming Language.

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

Thank you!