Camelia, the Perl 6 bug

IRC log for #parrot, 2010-11-11

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:00 NotFound_b What's in that branch, then?
00:00 whiteknight chromatic: yeah, the timings on my box were screwey. It's a mem-constrained VM that was running while I was trying to install VisualStudio on the host system
00:00 whiteknight NotFound_b: that branch is old and I need to delete it
00:06 dalek roast: 91b1991 | KodiB++ | S12-construction/construction.t:
00:06 dalek roast: [construction.t] Added tests for an unreported bug fixed in Rakudo 037775eb.
00:06 dalek roast: review: https://github.com/perl6/roast/commit/9​1b19912f60b66acedae7f42a43c7d0b6173028f
00:06 dalek roast: b540d54 | KodiB++ | S (3 files):
00:06 dalek roast: Removed some Rakudo TODOs for passing tests.
00:06 dalek roast: review: https://github.com/perl6/roast/commit/b​540d54c526fefc694329acae1a88d6c963c8904
00:08 whiteknight I'm absolutely *thrilled* that the SET_FLAG()macro doesn't take an interp pointer as a parameter
00:08 whiteknight it just assumes that we have one, and that it's called "inter"
00:08 whiteknight "interp"
00:10 NotFound_b whiteknight: macros are fun, aren't they?
00:14 whiteknight they can be
00:14 whiteknight good macros are good
00:14 whiteknight bad macros are hell-evil
00:24 whiteknight bluescreen++
00:30 dngor left #parrot
00:30 dngor joined #parrot
00:31 janus left #parrot
00:50 theory joined #parrot
01:24 patspam joined #parrot
01:24 patspam left #parrot
01:25 NotFound_b left #parrot
01:49 dip left #parrot
01:52 whiteknight where is the PIRATE repo?
01:58 cotto parrot/pir
01:59 whiteknight okay, I wasn't sure if that was still it
02:00 whiteknight are we still calling it "PIRATE"?
02:01 cotto I am.
02:10 dngor_ joined #parrot
02:10 dngor left #parrot
02:14 whiteknight okay, then so will I
02:24 cotto arrrrrrr
02:26 whiteknight cotto: can you think of any instance where we would call Parrot_destroy, or Parrot_exit, but not call both?
02:27 whiteknight I mean, once Parrot_destroy is called, the various subsystems are dismantled and there is nothing else left to do except exit
02:27 whiteknight and you don't exit without destroying first, to free resources
02:28 whiteknight it looks like die_from_exception calls Parrot_exit without calling Parrot_destroy. That seems like a bug to me
02:30 cotto whiteknight, I don't know why someone wouldn't want to call both of those.
02:30 lidden left #parrot
02:31 whiteknight I'm thinking about having a single API function to call the both of them, since I can't think of a reason they should be two separate calls
02:31 whiteknight in the API, Parrot_exit doesn't "exit" the program, it only exits Parrot's context
02:32 whiteknight so it's safe to call that whenever without worrying about the embedding application being closed
02:32 cotto That's sensible.  You might want the embedding program to keep going after Parrot's done.
02:32 whiteknight exactly
02:33 rurban_ joined #parrot
02:35 rurban left #parrot
02:35 rurban_ is now known as rurban
02:37 bluescreen joined #parrot
02:43 cotto Destruction sounds like an implementation detail that shouldn't be exposed. +1 to what it sounds like you were going to do anyway.
02:48 whiteknight there are exit handlers that Parrot exposes and I think that's a good idea
02:49 whiteknight actually, I don't know if we want that to be visible to embedding applications
02:49 whiteknight because they already know when Parrot exits, because they do it
02:49 whiteknight maybe destruction should be an atexit handler
02:53 whiteknight cotto: what's your opinion on longopt_get? Is that functionality that libparrot should provide, or do we cut it off into the embedding application (parrot.exe)?
02:57 cotto How would it be used by an embedding application?
02:58 cotto It strikes me as something that's specific to a command-line usage of libparrot.
02:59 cotto Wouldn't an embedding application would deal with those options programmatically.
02:59 cotto ?
03:02 whiteknight well, at the core it's a mechanism to help with commandline arg parsing
03:02 whiteknight the data that makes it parrot.exe-specific can be set
03:03 whiteknight though it really has nothing to do with parrot at that level
03:03 whiteknight decision made: not part of the API
03:03 cotto right
03:03 atrodo i'd say longopt is something outside the scope of libparrot
03:03 cotto +1
03:03 whiteknight done and done
03:03 atrodo (too slow)
03:04 whiteknight and on that note, it's time for bed
03:04 whiteknight goodnight, gentlemen
03:04 cotto night
03:04 whiteknight left #parrot
03:25 cotto whiteknight++ for his blog post that mentions 6.0
03:26 plobsing joined #parrot
03:27 bluescreen left #parrot
03:27 cotto The world will be an interesting place when 6.0 comes out.
03:28 cotto (how much Parrot contributes to that interestingness depends on us)
03:28 * atrodo feels inspired
03:54 chromatic 6.0?
03:55 cotto http://whiteknight.github.com/201​0/11/04/parrot_of_the_future.html
03:56 dngor_ is now known as dngor
04:45 jsut joined #parrot
04:50 jsut_ left #parrot
05:02 chromatic left #parrot
05:27 theory left #parrot
06:09 PerlJam left #parrot
06:23 chromatic joined #parrot
06:30 dukeleto 'ello
06:33 dukeleto 41692 messages since the first one, 5.83 years ago, for an average of 1.22 hours between messages <-- http://cia.vc/stats/project/parrot
06:33 dukeleto they have a bug in how they calculate the age of our repo, or our repo is wonky. Who wants to bet?
07:17 rkazak joined #parrot
07:21 bacek aloha, humans
07:22 cotto allison, bacek
07:23 * cotto fails at autocomplete
07:23 cotto aloha, bacek
07:23 bacek cotto, wrong bot^W! :)
07:23 dukeleto how do i make a new config step run?
07:23 dukeleto i have it written
07:23 dukeleto but it doesn't get run
07:24 rkazak make file?
07:25 dukeleto rkazak: good guess. the thing that generates the thing that generates the makefile is more like it :)
07:25 cotto dukeleto, perhaps lib/Parrot/Configure/Step/List.pm has something to do with it?
07:26 dukeleto are we getting branch commits from dalek yet?
07:26 rkazak so what's failing the first, second or third?
07:27 dukeleto auto::sha1 -          Determine Parrot's sha1...f34b7f8b7d5413ae2a05ea3c55c0b0d19199dc49
07:27 dukeleto \o/
07:28 cotto we're modern now
07:28 dukeleto who is against me adding a new config step, without tests, so that moritz can do his thang, and then we add tests ?
07:28 dukeleto i can write a few basic tests
07:28 dukeleto but they won't be extensive
07:29 dukeleto i guess i can just leave them in a branch
07:29 dukeleto and rakudo people can test the branch, if it works for them, we can merge it
07:29 moritz +1
07:29 dukeleto moritz: i just added a 'sha1' parrot_config key, and i will have a git_describe key in a few more minutes
07:29 moritz \o/
07:30 bacek backslash oh slash
07:30 moritz sounds like a song title :-)
07:36 chromatic left #parrot
07:38 PerlJam joined #parrot
07:43 dukeleto Code review for my new auto::sha1 probe? https://github.com/parrot/parrot/c​ompare/master...sha1_config_system
07:43 dukeleto if y'all like that, i can make a auto::git_describe probe in the same exact way
07:43 dukeleto i am not sure if we need to be caching the git_describe probe as well
08:00 whiteknight joined #parrot
08:03 sorear everyone who can is now encouraged to migrate to push notifications
08:04 sorear they're much faster, use less resources, don't require a commit bit, and don't require you to list branches manually
08:04 dukeleto sorear++
08:04 dukeleto sorear: how do we do that?
08:04 sorear the post-commit URL to use is http://host04.appflux.net:5​000/dalek?t=freenode,perl6
08:04 fperrad joined #parrot
08:04 sorear dukeleto: go to your github repository page, click 'Admin', click 'Service Hooks', click 'Post-Receive URLs', enter that, click 'Update Settings'
08:06 * cotto does so now
08:06 dukeleto cotto: did it alread
08:07 dukeleto already, even
08:07 cotto orly
08:07 dukeleto sorear: that links is for #perl6 ?
08:07 dukeleto sorear: we would have ?t=perl,parrot ?
08:07 cotto dukeleto, do you show two urls?
08:07 dukeleto sorear: or is that URL you gave correct?
08:07 dukeleto cotto: yes
08:08 cotto ok.  I seem to have successfully added it too.  Maybe github dtrt and decided that one was enough.
08:08 dukeleto magnet,parrot ?
08:08 cotto one instance of a url, that is
08:08 JimmyZ joined #parrot
08:10 dalek parrot/git_describe: 2e10036 | dukeleto++ | .gitignore:
08:10 dalek parrot/git_describe: [config] Add cached git describe string to .gitignore
08:10 dalek parrot/git_describe: review: https://github.comhttps://github.com/parrot/parro​t/commit/2e10036d7d1df09b70c8c368c819b25e4bbf946b
08:10 dukeleto nice!
08:10 dukeleto sorear++
08:11 dukeleto sorear: your link is a bit wonky, though
08:11 dukeleto sorear: looks like you don't need to prefix it with https://github.com
08:11 cotto sorear++
08:12 cotto dukeleto, when would be an effective time to pester you about getting the parrot-commits script ready for use?
08:13 dalek dalek-poller: 51c6b9b | sorear++ | push.psgi:
08:13 dalek dalek-poller: Fix link generation for push mode JimmyZ++
08:13 dalek dalek-poller: review: https://github.comhttps://github.co​m/sorear/dalek-poller/commit/51c6b9​bce3513b62723affe979df889c3d879813
08:14 cotto Mmmm.  Metabrokenness.
08:16 JimmyZ left #parrot
08:18 whiteknight left #parrot
08:18 rkazak left #parrot
08:19 dalek dalek-poller: 320e165 | sorear++ | push.psgi:
08:19 dalek dalek-poller: Try a better multichannel way (2)
08:19 dalek dalek-poller: review: https://github.com/sorear/dalek-poller/comm​it/320e165a71e5861c31cce4a2441cff628b6b7473
08:19 dalek dalek-poller: ebba952 | sorear++ | push.psgi:
08:19 dalek dalek-poller: Try a better multichannel way
08:19 dalek dalek-poller: review: https://github.com/sorear/dalek-poller/comm​it/ebba952b0e7b0dad8b7fcadc55d3af16a074d765
08:19 dalek dalek-poller: 51c6b9b | sorear++ | push.psgi:
08:19 dalek dalek-poller: Fix link generation for push mode JimmyZ++
08:19 dalek dalek-poller: review: https://github.com/sorear/dalek-poller/comm​it/51c6b9bce3513b62723affe979df889c3d879813
08:19 dalek parrot/git_describe: b9f40d5 | dukeleto++ | t/configure/063-git_describe_to_cache.t:
08:19 dalek parrot/git_describe: [config] Make the git_describe cache tests pass
08:19 dalek parrot/git_describe: review: https://github.com/parrot/parrot/commit/​b9f40d587f854419850ca055efb54cda0ea18380
08:19 dukeleto cotto: when i am awake, probably tomorrow
08:19 * dukeleto is currently zombie_duke
08:20 dalek dalek-poller: a74c570 | sorear++ | push.psgi:
08:20 dalek dalek-poller: Try a better multichannel way (3)
08:20 dalek dalek-poller: review: https://github.com/sorear/dalek-poller/comm​it/a74c5707cf9a97b517e3404e883bc7f741929713
08:20 * cotto is currently going_to_sleep_cotto
08:21 sorear hey, that looked like a success 'in the wild'
08:29 dalek parrot/generational_gc: a6f582a | bacek++ | src/gc/gc_ms2.c:
08:29 dalek parrot/generational_gc: Add description of new-to-be-implemented GenGC algorithm.
08:29 dalek parrot/generational_gc: review: https://github.com/parrot/parrot/commit/​a6f582ae903e2cc8503f3da5c24d53e303411b5f
08:29 bacek Yay! sorear++!
08:30 bacek Anyway, I would like some review of  https://github.com/parrot/parrot/commit/​a6f582ae903e2cc8503f3da5c24d53e303411b5f
08:30 bacek I probably missed some important bits
08:34 cosimo left #parrot
08:34 sorear https://github.com/perl6/mu/bl​ob/master/misc/dalek-push.txt
08:42 cosimo joined #parrot
08:49 bacek sorear, are small rants about dalek-poller code style accepted? Or better to fork-and-patch it? :)
09:00 sorear bacek: dalek-poller is a legacy code base that I spent the last few days cleaning up... I think it used to be Perl 4
09:00 sorear if you want to rant, go ahead
09:00 bacek sorear, hollei schitt!
09:00 sorear you will get sympathy from me, not anger
09:01 bacek https://github.com/bacek/dalek-poller
09:02 bacek last few commits
09:02 bacek I didn't test it (yet). Still installing Plack on my box.
09:02 sorear plack is only needed for push mode.
09:02 bacek At least it compiles.
09:04 bacek sorear, ok. I sent pull request.
09:07 bacek afk # kind of dinner
09:14 GeJ bacek: what is the "Roots" phase?
09:14 bacek GeJ, collecting of roots objects.
09:14 * sorear sleep
09:14 bacek basically - tracing C stack and marking interp
09:15 bacek (In my sample - just arbitrary choice of objects)
09:27 dalek parrot/generational_gc: 86871e5 | bacek++ | src/gc/gc_ms2.c:
09:27 dalek parrot/generational_gc: Add one more step to GC.
09:27 dalek parrot/generational_gc: review: https://github.com/parrot/parrot/commit/​86871e50abf467d84a5815288d44dd44b6f2df84
09:27 bacek sorear++ # push notifications are bloody fast!
10:30 masonkramer left #parrot
10:33 rurban_ joined #parrot
10:33 masonkramer joined #parrot
10:35 rurban left #parrot
10:35 rurban_ is now known as rurban
10:43 contingencyplan left #parrot
11:05 whiteknight joined #parrot
11:07 bacek left #parrot
11:19 bacek joined #parrot
11:42 dip joined #parrot
12:07 bluescreen joined #parrot
12:15 whiteknight left #parrot
12:18 lidden joined #parrot
12:28 bluescreen left #parrot
12:55 mikehh left #parrot
13:15 bluescreen joined #parrot
13:24 seanstickle joined #parrot
13:30 whiteknight joined #parrot
13:33 whiteknight good morning, #parrot
13:34 tadzik good afternoon whiteknight
13:59 whiteknight hello tadzik, how are you today?
14:00 tadzik whiteknight: cool. Fixing qw// in Rakudo
14:05 whiteknight oh, very nice
14:07 dalek tracwiki: v3 | plobsing++ | PlobsingTaskList
14:07 dalek tracwiki: add new tasks / remove completed tasks
14:07 dalek tracwiki: http://trac.parrot.org/parrot/wiki/Plob​singTaskList?version=3&amp;action=diff
14:14 tadzik Still waiting for cotto or somebody to introduce me to Lorito
14:15 hudnix left #parrot
14:15 whiteknight tadzik: What do you want to know about Lorito?
14:15 whiteknight plobsing++
14:16 tadzik whiteknight: that's what I chose to focus on in my Parrot involvement. What is currently being done, what is to be done, what can I help with
14:16 hudnix joined #parrot
14:17 whiteknight ah, gotcha
14:17 plobsing left #parrot
14:17 moritz tadzik: you know about http://trac.parrot.org/parrot/wiki/LoritoRoadmap ?
14:17 whiteknight I haven't been keeping up with all the design details that cotto and others have been working out
14:18 tadzik moritz: yep
14:19 hudnix left #parrot
14:22 whiteknight I am really looking forward to Lorito, but I don't think I will be able to play a big role in its development and implementation initially
14:24 atrodo tadzik> Unless you want to checkout my prototype and play with it
14:24 atrodo tadzik> https://github.com/atrodo/lorito
14:25 tadzik atrodo: thanks, I'll look around
14:25 whiteknight internally, I think Parrot should mostly stay the same. We will still have a runloop for interpreted code that reads in the Lorito bytecode and executes it. What will change is the compiler front-end, and probably the storage method of opdefinitions in memory
14:25 tadzik atrodo: is there a TODO-list somewhere?
14:25 whiteknight we wont need all the complicated op_info_t garbage that current PASM opcodes need to carry around
14:26 atrodo tadzik> Unfortunately, only in my head
14:26 whiteknight once we have a lorito compiler front-end for testing, we can port IMCC to output the new bytecode instead of it's current output format, and we can start updating other compilers to output it as well
14:26 whiteknight and once we have lorito bytecode being imported into Parrot, we can do things with it internally like on-the-fly optimizations, JIT, etc
14:27 contingencyplan joined #parrot
14:27 atrodo whiteknight> Actually, i think targeting pirate and nqp-rx first is a better route
14:28 whiteknight what do you mean "better route"?
14:28 whiteknight better than what?
14:28 atrodo better than porting imcc first
14:28 whiteknight oh, that's fine then
14:29 whiteknight yeah, I don't care which order other compilers are updated in, and personally I would rather IMCC die in a fire than be updated to continue working with The New Parrot (c)
14:29 atrodo whiteknight> for some reason, I already knew that
14:33 hudnix joined #parrot
14:41 whiteknight atrodo: I don't know how you would have known about my absolute hatred of IMCC. I keep my feelings on the matter secret
14:47 atrodo whiteknight> I have a seventh sense about these things
14:53 whiteknight blah. Parrot build fails on Win64
14:53 whiteknight "Unknown PMC type to thaw 0"
14:54 dukeleto 'ello
14:54 tadzik ~~
14:56 whiteknight maybe I configured wrong
15:00 whiteknight ...no, my second attempt failed even worse
15:01 whiteknight attempt to access code outside bytecode segment
15:02 whiteknight ...and the third attempt was worse still. Miniparrot hung
15:03 * dukeleto tests gsoc_nci
15:04 Andy HOLY SHIT
15:04 Andy I just deleted parrot/parrot
15:04 Andy I meant to delete petdance/parrot.
15:04 Andy I'm sorry.
15:05 Andy dukeleto: Ping?
15:05 Andy whiteknight: ping?
15:06 moritz bacek_at_work had the last commit, so he surely has the complete repo
15:06 Andy Right.
15:07 Andy But I don't know how you created it in the first place.
15:07 Andy so didn't want to go trying
15:07 atrodo You may also be able to contact github to see if they can do anything
15:07 Andy The big warnings seem pretty clear that you're SOL.
15:08 * moritz would like to be able to drop organization privs, and regain them explicitly
15:09 dukeleto Andy: pong
15:09 dukeleto Andy: what did you do?
15:09 Andy Clearly, I do not need rights to drop the entire repo just to be able to commit.
15:09 dukeleto Andy: seriously?
15:09 Andy Yes, seirously.
15:09 Andy parrot/parrot, bang, gong.
15:09 Andy gone
15:09 dukeleto Andy: wow, andy.
15:10 * dukeleto fixes this
15:10 Andy I know we Just have to recreate the repo and push back to it.
15:10 dukeleto Things will change.
15:11 Andy Surely there's some middle ground between "commit" and "obliterate repository"
15:11 dukeleto Andy: this is not cool. You killed all of our permissions
15:11 Andy Again, big apologies.  I meant to blow away my fork.
15:11 Andy dukeleto: I am under no delusions that it is cool.
15:12 dukeleto Andy: i expect beer and cookies for fixing this. This is not a good time for me to fix this problem.
15:13 dukeleto And now we will have strict permissions.
15:13 Andy I think that's sensible.
15:13 Andy If it wasn't me now, it would have been someone else in the future.
15:14 Andy If I hadn't told us all, would we have even had any way of figuring out what happened?
15:14 moritz what happened, yes
15:14 moritz who did it, probably not
15:14 dukeleto If anybody pushed code in the last 8 hours, I don't have that in my local repo, so y'all will have to push those branches again.
15:15 dukeleto Andy: yes, i can see that the repo was deleted. To see who did it, i would just look in my github feeds
15:15 Andy I don't see it in my feed.
15:15 Andy on me.
15:15 whiteknight okay, I just set it so the developers team only has Push&Pull access, not adminstrative rights on parrot/parrot
15:16 moritz +1
15:16 Andy Thanks.
15:16 dukeleto whiteknight: why wasn't it that alreada?
15:16 dukeleto whiteknight: did someone change it to that? I didn't set it up like that.
15:16 whiteknight dukeleto: I don't know who set up the teams in the first place. I was only give metacommitter privs two days ago
15:17 dukeleto I've mostly fixed parrot.git. Some branches or peices of branches might not be there.
15:17 dukeleto Since i only have parrot.git from ~10hrs ago
15:17 dukeleto IF YOU HAVE PUSHED CODE IN THE LAST 12 HOURS, PLEASE PUSH AGAIN.
15:17 Andy I will announce to parrot-dev.
15:18 dukeleto Andy: thanks
15:18 dukeleto Andy: it is mostly fixed.
15:18 Andy Thank you.
15:18 Andy Again, big apologies.
15:18 dukeleto did anybody make new branches in the last 12 hours?
15:18 whiteknight Andy: it's no big deal. Like you said, it would have happened eventually
15:19 whiteknight I added in the email alerts to parrot-commits again. I don't know what post-receive urls we were using so I didn't add those back
15:20 dukeleto which is why dalek is not working
15:20 moritz so let's just pull parrot stuff again
15:20 moritz for dalek
15:21 moritz oh, it's still in there
15:21 moritz so dalek should still report commits, albeit slower
15:22 Andy maybe put something in /topic?
15:23 * dukeleto is done putting out fires for now
15:23 Andy Anything I should add to this mail?  http://nopaste.snit.ch/25524
15:23 whiteknight dukeleto++
15:23 whiteknight Andy: looks good to me
15:24 whiteknight I normally do a fresh pull every morning. I didn't today. Damnit
15:25 Andy oK, sending.
15:25 whiteknight Andy++
15:26 Andy ok send
15:26 Andy sent
15:26 Andy I double checked my screen about nine times before delting petdance/parrot just now, which was my original intent. :-)
15:26 seanstickle left #parrot
15:26 whiteknight Actually, I have a clone I made this morning when I got in, but I used git:// instead of git@ to clone it
15:27 whiteknight let me get my ssh key together, and I should be able to update it
15:27 Andy God bless this brave new world of dvcs
15:27 moritz whiteknight: you can just edit .git/config to use the git@ URL
15:32 atrodo https://github.com/parrot/p​arrot/tree/generational_gc < Looks like it works?
15:32 whiteknight moritz: yeah, I already did that. I had to set up an ssh key first
15:32 whiteknight I'm trying to push with my clone, but it's not doing anything
15:33 moritz then you probably don't have anything newer than github
15:33 whiteknight yeah, apparently
15:34 Andy also posted this http://support.github.com/discussions/si​te/2177-request-make-delete-this-reposit​ory-button-include-the-repository-name
15:34 nopaste left #parrot
15:34 TonyC left #parrot
15:35 moritz Andy++
15:35 Andy Down with them!  Down with them!
15:35 whiteknight my push would have updated all branches too, right?
15:35 Andy And when I say "them" I mean "vague pronouns"
15:36 moritz whiteknight: by default 'git push' only pushes the current branch
15:36 whiteknight okay, how do I make it do all branches?
15:36 moritz no idea :(
15:37 whiteknight where is branch information stored, /refs/heads/?
15:38 Andy HEY!  My Admin button on parrot/parrot is missing!
15:38 Andy :-)
15:38 Andy I'm gonna make a gratuitous commit/push to make sure that still works.
15:39 moritz whiteknight: seems like
15:39 nopaste joined #parrot
15:39 moritz so   for i in `ls .git/refs/heads/`; do git push $i origin/$i; done  might work
15:39 moritz if you always named your local branches like the remote
15:41 whiteknight well, git push --all pushes all refs in /refs/head/
15:41 * moritz should read docs instead of writing shell scripts :-)
15:44 Andy OK https://github.com/parrot/parrot/commit/​2d7a93202c4158ad6e9a4fd1db0022ae2427c409 got up there ok, so non-admins can push.
15:47 dalek parrot: 2d7a932 | petdance++ | README:
15:47 dalek parrot: remove an extra dash
15:47 dalek parrot: review: https://github.com/parrot/parrot/commit/​2d7a93202c4158ad6e9a4fd1db0022ae2427c409
15:51 stilgar joined #parrot
15:54 Andy left #parrot
16:01 whiteknight ha. You can't help but laugh about that kind of stuff.
16:02 whiteknight whatever doesn't kill us will make us stronger
16:02 moritz except when it breaks us.
16:06 whiteknight right, except that
16:06 mikehh joined #parrot
16:06 whiteknight but this didn't break us
16:08 theory joined #parrot
16:33 NotFound Looks like I'm not the only one not well used to git and github ;)
16:34 NotFound Did I need to clone again, or can keep using my current tree?
16:34 whiteknight you should be good with your current tree. Try a git pull to see if you get anything new
16:34 moritz you can keep your tree
16:35 whiteknight SVN repos have a Guid, and creating a new repo will have a different Guid even if the contents are copied
16:35 whiteknight but in Git, I don't think there is anything like a single repo identifier
16:36 atrodo there isn't.  everything is identified by the sha1 of it's content
16:36 whiteknight that's what I thought
16:37 NotFound About the thing of the revision number to be checked by languages et al... will not be easier to use tags.
16:37 NotFound ?
16:37 NotFound Tag the revision you want to use, and check it.
16:37 moritz it requires all parrot users to commit to parrot
16:37 moritz not practical
16:39 NotFound Then tell people to use only already tagged revisions.
16:40 moritz that's not practical either
16:40 NotFound I don't think any one that is not working close with the parrot team should use any arbitrary point in the repo.
16:41 moritz well, we often want to encourage rakudo users to try rakudo with newer parrot revisions, in order to check of regressions
16:41 moritz previously we simply did this with echo $new_revision > build/PARROT_REVISION
16:41 moritz and Configure.pl --gen-parrot did The Right Thing[tm]
16:42 moritz and once git-describe output is available from parrot_config, I'll fix rakudo's Configure.pl so that it works again
16:43 NotFound Fast-forwarded master to 2d7a93202c4158ad6e9a4fd1db0022ae2427c409
16:43 NotFound Looks like it's working,
16:44 moritz yep, that's the newest commit in master
16:45 * moritz -> norsk course
16:46 dmalcolm joined #parrot
16:47 NotFound You chenged the Developers group to make its members unable to kill the repo again. That means that I, as member of Owners still can do it, sin't it?
16:48 whiteknight NotFound: yessir. Your killing power is still at 100%
16:48 * NotFound resist temptation....
16:48 cotto at least you know just how much damage you can do
16:48 whiteknight the golden rule of github: Delete not, lest ye be deleted also
16:49 whiteknight I think that was in the bible
16:49 whiteknight the little-known 11th-15th commandments all dealt with DVCS
16:50 cotto If I was sleepy this morning, that ended when I saw andy's message.
16:50 whiteknight ha
16:50 whiteknight luckily we've got dukeleto on our side. He destroys problems with an axe
16:51 whiteknight although we are going to need to make sure he gets some rest eventually
16:51 NotFound Well, now we know a way to make all parroteers awake and alert.
16:52 cotto If you do that on a regular basis, I might be taking a vacation to Spain. ;)
16:52 cotto worst alarm clock ever
16:53 NotFound cotto: you're welcome. The Pope is gone, now Santiago is walkable again.
16:53 whiteknight I've always wanted to go to spain
16:54 plobsing joined #parrot
16:55 NotFound whiteknight: You're welcome too.
16:57 NotFound This is my place: http://maps.google.es/maps?hl=en&amp;i​e=UTF8&amp;ll=42.872083,-8.535932&amp;​spn=0.00061,0.001195&amp;t=h&amp;z=20
16:59 whiteknight oh fun
17:04 NotFound Who is the maintainer of http://tapir2.ro.vutbr.cz/cover/ ?
17:05 NotFound Is a nice service, will be nice to keep updating it.
17:06 Andy joined #parrot
17:10 whiteknight that is a good question. It is a very nice service
17:12 dukeleto NotFound: mj41 is the maintainer
17:12 dukeleto seen mj41?
17:12 aloha mj41 was last seen in #perl6 33 days 22 hours ago joining the channel.
17:12 dukeleto :(
17:20 whiteknight blah
17:20 cotto_work ~~
17:21 whiteknight shit. /me really needs to start writing up more GCI tasks
17:22 whiteknight This whole "I'm a lazy sack and I'll do it later" thing is really not working for me
17:22 * dukeleto has a baseball bat and some cookies, if that will motivate whiteknight
17:23 whiteknight you know the old saying: you catch more tuits with cookies than with baseball bats
17:25 whiteknight writing tasks should be so much easier considering the sorry state of some of our documentation
17:26 NotFound Do bats play baseball?
17:26 whiteknight NotFound: by sonar
17:29 * dukeleto imagines a flying mammal hitting a grand slam
17:33 M_o_C joined #parrot
17:50 dukeleto dalek should work again
17:50 dukeleto i fixed the post-receive URL
17:58 theory left #parrot
17:59 theory joined #parrot
18:01 chromatic joined #parrot
18:04 dukeleto A DINGO ATE YOUR HYPHEN!
18:05 cotto_work I'm glad it didn't go for the colon.
18:05 dukeleto cotto_work: ba dump ching
18:08 janus joined #parrot
18:24 cotto_work http://andrewducker.livejournal.com/2233017.html <- That gives me an idea.
18:25 atrodo Millions of programmers would rejoin in song on that day
18:27 whiteknight what would people say about having a C++ wrapper around the new Parrot API?
18:28 NotFound Reinventing the jvm and the java applets?
18:29 NotFound whiteknight: I've thinked about a C++ api several times, and concluded that the C api is not mature enough.
18:30 whiteknight NotFound: true, but I am in the middle of maturing the C API
18:30 atrodo cotto_work> it's on hacker news too http://news.ycombinator.com/item?id=1893686
18:31 whiteknight I have a  C++ coder friend who is looking for something to do to keep his skills sharp
18:31 rurban_ joined #parrot
18:35 rurban left #parrot
18:35 plobsing left #parrot
18:35 rurban_ is now known as rurban
18:37 cotto_work atrodo: that's where I saw it.
18:38 atrodo Between that and the GameBoy emulator in javascript article i saw last week, i have half a notion to reimplement my lorito prototype in js
18:39 whiteknight if we had a working implementation of JS on Parrot, you could do it in that
18:39 cotto_work If you have the tuits, go for it.
18:40 NotFound atrodo: do it in winxed and then we'll be able to run parrot on parrot on parrot....
18:40 atrodo turtles all the way down
18:40 whiteknight except the turtles compile each other into themselves
18:41 NotFound For extra points, make it in the intersetction of javascript and winxed.
18:42 whiteknight since Winxed is pretty close to javascript, maybe that should be the bootstrapping language used to write the javascript compiler
18:43 whiteknight seems fitting to me
18:43 NotFound Not so close for anything other than short examples.
18:43 whiteknight well, it is closer than writing a JS compiler in NQP
18:43 NotFound A lot.
18:43 whiteknight and it's far better than PIR or NQP
18:45 NotFound For people more used to the C family that to perl, at least.
18:45 NotFound than
18:49 NotFound If someone wants a relatively simple project: write a javascript parser in winxed, able to emit warnings about undeclared globals and things like that.
18:50 whiteknight put that on the BigProjects page
18:52 whiteknight We should add Winxed to smolder
18:53 NotFound Note that winxed is (C) me, not parrot.
18:54 NotFound Just in case someone cares.
18:55 whiteknight ah, that is a good point
18:56 whiteknight people can still use it. It's copyleft, right?
18:56 NotFound Yeah, AL
18:58 atrodo I would think that having parrot based languages smoldered is a good thing
18:58 simon_ "smoldered"?
18:58 NotFound atrodo: I agree, but maybe someone objects to using pafo resources for non owned projects.
19:01 whiteknight rakudo is in the smolder list, and they aren't owned by Parrot
19:02 theory left #parrot
19:02 NotFound I don't have any problem, I mention it just in case.
19:03 whiteknight if you're interested in adding it to smolder, I'm fine with it too
19:04 NotFound The test suite is too short, I'm not sure if it will pay the effort.
19:04 jan left #parrot
19:06 M_o_C left #parrot
19:07 dukeleto NotFound: i think people are talking about spectest
19:07 dukeleto NotFound: or you are trolling us ;)
19:07 whiteknight TROLLS EVERYWHERE
19:08 NotFound spectest? Someone wrote a winxed spec? ;)
19:08 whiteknight NotFound: yeah, and you're *way* off it
19:09 dukeleto EMAXTROLLS
19:10 whiteknight We are now trolling at maximum efficiency!
19:10 * dukeleto puts the troll shields at maximum power
19:12 theory joined #parrot
19:50 bluescreen left #parrot
19:59 bluescreen joined #parrot
20:16 plobsing joined #parrot
21:01 stilgar left #parrot
21:22 dalek dalek-poller: 95a85a7 | sorear++ | push.psgi:
21:22 dalek dalek-poller: Fix handling of defined-or with arrays
21:22 dalek dalek-poller:
21:22 dalek dalek-poller: // puts the left side into scalar context, so @{foo} // () is wrong.
21:22 dalek dalek-poller: Changed it to @{ foo // [] }.
21:22 dalek dalek-poller: review: https://github.com/sorear/dalek-poller/comm​it/95a85a764421a0888370205a82030eab667c623e
21:22 dalek dalek-poller: efec941 | bacek++ | push.psgi:
21:22 dalek dalek-poller: Rename IGNORE into OK and always return it.
21:22 dalek dalek-poller: review: https://github.com/sorear/dalek-poller/comm​it/efec9417c98525225ab56d2e2c31d711a78c5e05
21:22 dalek dalek-poller: 95e4263 | bacek++ | push.psgi:
21:22 dalek dalek-poller: Reindent code.
21:22 dalek dalek-poller: review: https://github.com/sorear/dalek-poller/comm​it/95e4263c333f2234715d0223656e146d2efc07da
21:22 dalek dalek-poller: fd5ea6b | bacek++ | push.psgi:
21:22 dalek dalek-poller: Simplify (and speed-up) creating of @files list.
21:22 dalek dalek-poller: review: https://github.com/sorear/dalek-poller/comm​it/fd5ea6bbd11759aa5b6cea0d641ad77505aacade
21:22 dalek dalek-poller: 92fab84 | bacek++ | push.psgi:
21:22 dalek dalek-poller: Use space after C<if>
21:22 dalek dalek-poller: review: https://github.com/sorear/dalek-poller/comm​it/92fab84657f4b9bd685f055d9573278a5a377520
21:22 dalek dalek-poller: ac797df | bacek++ | push.psgi:
21:22 dalek dalek-poller: Use constant to return constant response.
21:22 dalek dalek-poller: review: https://github.com/sorear/dalek-poller/comm​it/ac797df05b3cf0987f52c99d6935e6ef1338562f
21:23 plobsing left #parrot
21:23 bacek aloha, humans
21:24 cotto_work aloha, bacek
21:24 dalek dalek-poller: ee0aebc | sorear++ | push.psgi:
21:24 dalek dalek-poller: Fix backward message output
21:24 dalek dalek-poller:
21:24 dalek dalek-poller: GitHub-- for wrong documentation. :/
21:24 dalek dalek-poller: review: https://github.com/sorear/dalek-poller/comm​it/ee0aebc3f65be7dc395a5409c59d73bd3df2dc84
21:24 bacek sorear, I think $commit->{modified} should be handled in same way as "added" and "removed". For consistency sake (and if we only add files without modifications)
21:24 bacek aloha, cotto_work
21:25 sorear Sounds good
21:31 whiteknight left #parrot
21:35 cotto_work I didn't realize that github's web interface could be used to make changes.  That could be really nice for small changes.
21:36 atrodo I saw that too.  That's really nice
21:37 chromatic It's very nice for fixing typos in POD.
21:38 cotto_work It's great for implementing new features too.
21:38 atrodo github is my new code editor
21:38 atrodo just need a CI server pulling on a post-recieve hook to test my changes and report back to me via email
21:38 atrodo much easier than my workflow now
21:49 plobsing joined #parrot
22:22 bluescreen left #parrot
22:26 dalek winxed: r673 | NotFound++ | trunk/winxedst1.winxed:
22:26 dalek winxed: rearrange a bit the generated code for for loops avoiding one goto in stage 1
22:26 dalek winxed: review: http://code.google.com/p/w​inxed/source/detail?r=673
22:29 masonkramer_ joined #parrot
22:29 masonkramer left #parrot
22:29 masonkramer_ is now known as masonkramer
22:30 Andy left #parrot
22:31 bacek_at_work ~~
22:35 fperrad left #parrot
22:36 Kulag left #parrot
22:37 Kulag joined #parrot
22:37 bluescreen joined #parrot
22:38 TonyC joined #parrot
22:41 whiteknight joined #parrot
22:42 Kulag left #parrot
22:44 Kulag joined #parrot
23:04 cotto_work ``
23:18 seanstickle joined #parrot
23:51 bluescreen whiteknight: PARROT_API should be a define to... ? an alias of PARROT_EXPORT ?
23:52 whiteknight bluescreen: I have no idea. Probably same as PARROT_EXPORT for now, to enable shared library exporting
23:52 bluescreen i figured that, I'm trying to compile it
23:53 whiteknight oh, have you had to make many fixes?
23:53 whiteknight I haven't gotten a chance to play with it tonight
23:54 plobsing left #parrot
23:55 bluescreen not yet, that's first failure
23:58 whiteknight ok

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

Parrot | source cross referenced