Camelia, the Perl 6 bug

IRC log for #parrot, 2010-11-05

Parrot | source cross referenced

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

All times shown according to UTC.

Time Nick Message
00:01 cotto It's part of the parrot org on github now
00:01 cotto parrot/pir
00:01 cotto I'll fix that.
00:02 cotto fixed
00:02 cotto kid51++
00:14 kid51 is now known as kid51_at_dinner
00:15 dukeleto Code Review Needed: https://gist.github.com/663444
00:15 dukeleto that is the git filter-branch command that i am going to run
00:16 dukeleto if you have a problem with one of the name transformation rules, speak now, or forever deal with it.
00:16 * Tene reads.
00:16 dukeleto this was autogenerated from the CREDITS file
00:17 Tene Uh.... bad.
00:17 Tene Quoting fail.
00:17 dukeleto Tene: are you talking about githubs color rendering?
00:17 Tene You're trying to contain it in '', but you have: Curtis 'Ovid' Poe, Nuno 'smash' Carvalho, and Sean O'Rourke
00:18 dukeleto Tene: ah. I had fixed some quoting errors, but not all. I only took into account double quotes
00:18 Tene dukeleto: no, I'm talking about how you have: --env-filter '...''...''...'...'
00:18 dukeleto Tene: yes, i see now
00:18 dukeleto Tene: good thing I wrote a program to generate that :)
00:18 whiteknight I think my username should not be capitalized
00:18 Tene Also it's not sorted! :P
00:19 dukeleto whiteknight: that is a bug in the CREDITS file, but that is good to know. if your svn username is lowercased, it won't work
00:20 Tene dukeleto: can we have: "dukeleto") n="Stephen Weeks" ; m="tene@allalone.org" ;; ?
00:20 Tene ;)
00:20 whiteknight at least, my trac login is lowercased
00:20 dukeleto whiteknight: is your svn username all lowercase?
00:21 whiteknight I dont remember, i always use cached credentials
00:22 dukeleto whiteknight: it is lowercase
00:22 Tene looks to be all lowercase in 'svn log'
00:22 whiteknight yay! I'm not terribly mis-remembering things
00:23 cotto dukeleto, you can take the a.k.a. off my name
00:23 cotto probably the same for kid51
00:23 dukeleto this is no fun.
00:24 dukeleto the CREDITS file should be correct.
00:24 dukeleto there are a whole bunch of other a.k.a stuff
00:24 cotto give me a second then
00:25 dukeleto all these issues will be present in the current git svn clone i am making
00:25 dukeleto cotto: i can just remove some of them
00:25 dukeleto cotto: but i am sure somebody will complain
00:26 cotto I can commit the changes now if you like
00:26 dukeleto and I don't care. I want this repo converted.
00:26 dukeleto cotto: yes, please do that. But i will just modify my input file
00:26 cotto done
00:27 * cotto hopes we don't have a r49800
00:31 dukeleto https://gist.github.com/663458
00:32 dukeleto please look at that now. it is sorted, and has cotto's name fixed
00:32 * cotto will be able to show his gsoc pride tomorrow at $dayjob->new()
00:32 dukeleto Tene: ^^^
00:32 dukeleto cotto: nice :)
00:32 dukeleto is ash_ really a valid svn username?
00:33 cotto dukeleto, do you want to do the same for particle and infinoid?
00:34 cotto dukeleto, trac seems to like "ash" better: http://trac.parrot.org/parrot/changeset/49465
00:35 Tene dukeleto: single quotes don't work like that in sh.  Inside of sh single quotes, no escaping is handled at all, whatsoever.  the \ character is not treated specially.
00:36 Tene try for yourself: echo 'foo\'bar'
00:36 dukeleto ok, i fixed particle, jkeenan and infinoid's "a.k.a" business
00:36 dukeleto Tene: joy to the world
00:37 dukeleto Tene: i will just run a .sh from that command
00:37 Tene dukeleto: I'm sorry, I don't understand.
00:37 dukeleto Tene: but i may still run into the escaping problem
00:38 dukeleto Tene: i will run './some_junk.sh' for the --env-filter
00:38 dukeleto Tene: hopefully that will work?
00:39 Tene dukeleto: if you run it like that, your environment variable modifications will not be handled right.  You could instead do: source some_junk.sh
00:39 dukeleto Tene: thanks!
00:40 dukeleto Tene: but i can just put those env vars in the some_junk.sh
00:41 dukeleto there are still some other things to work on for the git migration, that can be done in parrallel
00:41 dukeleto whiteknight: you said you wanted to help?
00:41 Tene Hmm.  It's not clear to me at all how env-filter works.  Normally, any environment modifications only affect the environment of any process created by that process, never the parent process.
00:42 dukeleto Tene: it sources what is inside --env-filter for every git commit, and those magic variables change who the author name/email info
00:42 Tene Yeah, I thought so.
00:42 Tene In that case, invoking a new program will not work at all.  You need to source it.
00:42 dukeleto Tene: ok, will try that
00:43 Tene see: bash -c 'export FOO=omghai'; echo $FOO
00:45 whiteknight dukeleto: As much as I can, yes. I'm probably in bed within an hour
00:46 dukeleto whiteknight: feel free to port create_language.pl and mk_language_shell.pl to git before you go to bed
00:47 dukeleto Tene: i still have the problem of quotes inside of quotes
00:47 dukeleto Tene: i need to use variables to get around that?
00:48 Tene dukeleto: you won't if that's all in a separate file.
00:48 dukeleto Tene: i still have the problem.
00:48 Tene What you pasted is fine, except you need to remove the \s
00:48 Tene hm?
00:48 Tene post the proposed file contents
00:48 dukeleto Tene: i will try that
00:49 whiteknight dukeleto: I'll see what I can do. I'm no git expert, but I'll give it the old college try
00:49 whiteknight from what I remember about college, this is going to require some beer, apparentl
00:50 dukeleto "dukeleto") n="Jonathan "Duke" Leto" ; m="jonathan@leto.net" ;;
00:50 dukeleto Tene: that isn't valid
00:50 dukeleto Tene: it causes ./transform_parrot_authors.sh: 130: Syntax error: Unterminated quoted string
00:50 dukeleto whiteknight: there are both names with single and double quotes
00:50 dukeleto Tene: ^^^ was meant for you
00:50 * dukeleto can't be trusted to tab complete at this point
00:52 Tene Ah.
00:52 Tene dukeleto: can you post the entire thing?
00:52 dukeleto I think i've solved it
00:52 Tene you need to \" the "s, but leave the 's alone
00:54 whiteknight I'm pretty unfamiliar with this code. I doubt I'll be able to update it tonight
00:55 dukeleto whiteknight: no worries.
00:55 dukeleto whiteknight: isn't it fun to maintain two almost identical scripts?
00:56 dukeleto https://gist.github.com/663479
00:56 dukeleto Tene: ^^^ take a look at that. It seems to not have syntax errors, which is a start.
00:57 Tene dukeleto: the alternative is to use a slightly awkward quoted heredoc, if you don't want a separate file.
00:57 Tene Yeah, that looks fine to me, afaict.
00:59 dukeleto Rewrite e927c24745ffe50d968f9d175db0a35590231294 (1/40300)eval: 1: source: not found
00:59 dukeleto can't use 'source', since git uses 'sh'
00:59 Tene Ah.
00:59 dukeleto sh -c works
00:59 Tene in sh, 'source' is spelled '.'
01:00 Tene sh -c won't work.
01:00 dukeleto filter branch we go!!!
01:00 Tene that launches a separate process.
01:00 dukeleto Tene: hmm. it *seems* to be working.
01:00 Tene is it actually rewriting?
01:00 Tene If it is, git must be doing something clever
01:00 dukeleto Tene: can't tell until it is done.
01:01 dngor joined #parrot
01:01 Tene See, it will *run* just fine, but it won't affect the environment variables, unless git does something special.
01:01 dukeleto Tene: i am using . now
01:01 dukeleto Tene: i belive you. i killed it and i am now using . :)
01:01 dukeleto filter branch is doing a few dozen commits per second, and is much faster
01:02 dukeleto it is on commit 1000 already
01:02 * dukeleto crosses fingers
01:02 cotto great!
01:03 Tene FTR, the way to get arbitrary-delimiter quoting in sh is:
01:03 Tene FOO=$(cat <<'END'
01:03 Tene ...
01:03 Tene END
01:03 Tene )
01:03 Tene everything up to END is literally quoted, no interpolation or escaping or anything.
01:03 * dukeleto kills the old git svn clone process. Good riddance.
01:04 Tene the ''s around END are important.
01:04 sorear ln -s $( which parrot-nqp ) ~/bin/sh  # problem solved
01:04 cotto very good riddance
01:04 sorear (delete ~/bin/sh ASAP because a lot of other stuff will break)
01:06 dukeleto filter-branch seems to be more than 10% done already
01:17 kid51_at_dinner is now known as kid51
01:17 kid51 Should we be refraining from, ya know, doing actual work?
01:18 * cotto suspects not until dukeleto explicitly says so
01:19 cotto I think he can still incrementally update just fine
01:20 cotto Once we have all the pieces in place, I'm hoping to send something to parrot-dev, etc spelling out the process so that everyone knows when and where it'll be safe to commit.
01:23 dukeleto filter branch has done about half of history already
01:23 cotto 'bout time something went quickly
01:35 lidden left #parrot
01:38 whiteknight why didn't you use filter-branch initially?
01:38 cotto it meant writing code, while the import didn't
01:38 whiteknight ok
01:39 cotto if the import had worked properly, it would have been faster
01:49 kid51 dalek is completely offline, which means we're not getting our last svn commit messages
01:50 cotto correct
01:52 dukeleto Ref 'refs/heads/trunk' was rewritten
01:52 dukeleto real48m30.970s
01:52 dukeleto user39m21.810s
01:52 dukeleto sys13m36.990s
01:54 dukeleto https://gist.github.com/663533 <-- shortlog on the new repo
01:54 dukeleto still some bugs
01:54 dukeleto whiteknight: looks like you had both a lower case and an upper case name. Thanks for making life complicated.
01:55 kid51 left #parrot
01:58 cotto Mark Glines and Infinoid show up separately
02:10 plobsing joined #parrot
02:12 cotto moritz's middle initial should probably have a period
02:16 cotto Geoff Broadwell and kid51 have their nicks appended in parentheses
02:23 dngor_ joined #parrot
02:23 cotto "Infinoid <Infinoid@d31e2699-5ff4-0310-a27c-f18f2fbe73fe>" looks suspicious
02:24 dngor left #parrot
02:35 cotto dukeleto, do you have the patience for another iteration of fixes?
02:36 dngor_ is now known as dngor
02:47 whiteknight left #parrot
02:56 theory joined #parrot
03:02 dukeleto cotto: yeah. i took a break. I am at PDX Hackathon
03:03 particle1 joined #parrot
03:04 particle left #parrot
03:04 dukeleto aloha, msg whiteknight your GCI task are unusably ambiguous. If you can add links and follow the template, that would be awesome
03:04 aloha dukeleto: OK. I'll deliver the message.
03:06 atrodo So what exactly is the difference between the import and the filter-branch methods?
03:07 dukeleto aloha: git svn clone vs git filter-branch
03:08 atrodo so, instead of redo-ing a svn clone, you filtered your existing one?
03:10 dukeleto aloha: yes. git is much faster than git+svn
03:10 aloha dukeleto: Okay.
03:10 dukeleto atrodo: ^^^
03:10 * dukeleto can't seem to tab-complete today
03:11 atrodo okay.  that makes more sense
03:37 theory left #parrot
03:48 mikehh left #parrot
04:00 theory joined #parrot
04:09 theory left #parrot
04:16 * dukeleto just made a bunch of Parrot-related GCI tasks
04:21 cotto are you doing round 2 of filter-branch now?
04:23 dukeleto cotto: nope.
04:24 dukeleto cotto: i am enjoying myself at a bar, without that laptop :)
04:24 dukeleto cotto: should we email parrot-dev and ask people if they are happy with the names?
04:24 cotto dukeleto: have a moderate amount of fun then
04:24 dukeleto cotto: i don't want people complaining after it happens, but i don't know how much to care about it now
04:24 cotto sounds like a good idea
04:28 dukeleto cotto: sould i ask parrot-dev about the author map? or is it not worth it?
04:28 dukeleto cotto: also, what about converting mk_language_shell.pl and create_language.pl ? Does it matter if we do them after the git conversion?
04:29 cotto My concern is that they could get dropped on the floor and bitrot, but I also don't want them holding the move any further than it's already been held up.
04:29 cotto +1 to going ahead and fixing them post-move
04:32 dukeleto cotto: sounds good to me. they are the only tools left to convert, so they will happen soon after the conversion, if anybody cares about them
04:32 dukeleto cotto: what do you think about the "revision" key in parrot_config ?
04:32 dukeleto cotto: do we make it some huge number post-git-conversion, so old tools get the message that they don't have a new enough parrot?
04:33 cotto dukeleto: that makes some amount of sense.
04:34 cotto though after git o'clock "revision" doesn't have much meaning
04:35 cotto "mu" would be a fitting, if unhelpful, value
04:36 dukeleto cotto: converting those tools depends on certain parrot_config keys existing, so they are non-trivial to port
04:36 dukeleto cotto: think of every script that expects to ask parrot it's "revision"
04:36 dukeleto cotto: git should just give a huge number, that never existed in SVN, to signal "you are too old, bitches"
04:36 cotto Yes.  We can't just drop it from parrot_config all at once.
04:37 dukeleto cotto: i can see dropping the "revision" key after 1 or 2 stable releases, after the conversion
04:37 cotto but for now we need a strategery
04:37 dukeleto cotto: but we need to have a new key, that git-aware tools will use
04:38 dukeleto cotto: there is already a "version" key, but we need a key that represents a "sha1"
04:38 dukeleto cotto: for the case of depending on a specific sha1
04:39 cotto dukeleto, I don't like depending on a specific sha1.  I'm still holding out on us figuring out a sane way to determine how recent the source of a build is.
04:40 cotto what's the use case for depending on a sha1?
04:41 Tene cotto: consider 'git describe'
04:44 dukeleto cotto: there are 2 use cases
04:44 dukeleto cotto: a HLL/external project can depend on a version number or a sha1
04:45 dukeleto cotto: nothing in between
04:45 dukeleto cotto: for instance, development versions will depend on sha1's
04:45 dukeleto cotto: a project can only depend on a sha1 if parrot.git is available
04:45 cotto dukeleto, the problem I see is that that approach will fail if the user doesn't have that exact version of parrot installed
04:46 dukeleto cotto: otherwise, only a version number
04:46 dukeleto cotto: there is a git command to see if one sha1 is contained by another
04:46 dukeleto cotto: so that problem won't occur
04:46 dukeleto cotto: there is a way to say "do you have something newer than this sha1?"
04:46 cotto ok.  Access to a git repo isn't onerous for a development version.
04:49 dukeleto right. if parrot_config tells us parrot lives in a directory that happens to be a git repo, the thing can depend on a sha1. Otherwise, not.
04:50 dukeleto cotto: make sense?
04:50 cotto yup
04:54 dukeleto cotto: i sent something to parrot-dev, asking people what usernames they want
04:54 dukeleto cotto: perhaps that was dumb, but i will at least give people a chance
04:57 cotto Sure.  Give it over the weekend and we can make the move early next week.
04:58 cotto or sooner if you're not feeling that patient
04:58 cotto dukeleto, is that the last of the migration bugs or is there more work that can be done in parallel?
05:00 dukeleto cotto: there are still lots of things on the migration wiki
05:00 dukeleto cotto: the create_language and mk_language_shell scripts are probably the most important to work on
05:01 contingencyplan left #parrot
05:01 cotto dukeleto: do you think it could happen early next week?  Later than that is getting a little close to the 2.10 release.
05:03 contingencyplan joined #parrot
05:08 dukeleto cotto:  hard to say. I will be on a work trip from nov 7-12th
05:08 dukeleto cotto: but i want to get this shite done.
05:10 cotto dukeleto, I very much feel the same way
05:11 dukeleto cotto: good to know :)
05:11 dukeleto cotto: this has been dragging on too long. We need to cut the cord
05:11 cotto +yes
05:13 cotto dukeleto, I'd rather have a slightly rocky transition with a few minor artifacts than have this drag on for much longer while we make it perfect.
05:14 cotto I really want to tell people "Sorry for the bumps but we're done now.  Get back to work."
05:14 cotto s/tell/be able to tell/
05:22 dukeleto cotto: take a look at http://trac.parrot.org/parro​t/wiki/GoogleCodeIn2010Tasks
05:22 dukeleto cotto: i added a bunch of smaller tasks
05:22 * cotto looks
05:22 dukeleto cotto: one is to research (compare+contrast) Dalvik and Parrot
05:22 dukeleto cotto: please add any others that you can think of
05:23 cotto I'd love a Perl 6 shirt.
05:23 cotto good idea there
05:24 dukeleto cotto: too bad almost nobody wants to help
05:25 dukeleto cotto: i made almost every task. I don't get why people don't want google to pay for students to get involved in their projects
05:25 cotto I'll get after myself to finish up that list of PMCs that would be good for GCI students.
05:26 dukeleto cotto: thanks. but simpler non-code tasks are even more valuable
05:27 mikehh joined #parrot
05:29 cotto dukeleto, it might just be because it's a new and unfamiliar program.  When I think of where Parrot could use help, my instinct is to think almost exclusively in terms of code.
05:29 dukeleto cotto: yep. But we can use so much help in non-code tasks. Parrot doesn't know how to market itself.
05:29 dukeleto cotto: Parrot doesn't know how to bring in new-comers.
05:29 dukeleto cotto: Parrot doesn't have friendly documentation for newbies.
05:31 cotto dukeleto, historically we've been spoiled by a fairly regular stream of newcomers.  I agree that we're not marketers by default and that we need to do more to attract new contributors of all kinds.
05:31 sorear once upon a time, dalek ran on feather, and I had the ability to debug it
05:31 cotto sorear, I logged in to feather to look at it until I realized that I have no idea about where dalek even lives.
05:31 sorear but after the last feather crash, diakopter moved dalek to a completely private VPS and I can't touch it
05:32 dukeleto sorear: perhaps you just need to ask him for access to fix it?
05:33 sorear yeah
05:33 sorear just did so
05:34 dukeleto sorear++ # thanks for caring
05:34 cotto sorear++ indeed
05:35 cotto and if you can get access, maybe you can share the love and/or figure out why it's not posting wiki/ticket changes
05:36 dukeleto sorear: fixing dalek's wiki-fu would be awesome
05:36 * dukeleto attempts to leave the bar and get home
05:39 mikehh dukeleto: all my commits to svn have been based on my nick here, in fact I use the nick on github, how will that effect my commits to git - you have my full name there
05:43 cotto mikehh, if you tell github that the email address listed in the commit metadata is yours, it'll be linked to you.
05:47 cotto github lets you add as many extra email addresses as you need from the user control panel
06:04 mikehh cotto: the email address is the current one I use
06:04 cotto then you're set
07:58 jan left #parrot
08:18 contingencyplan left #parrot
08:21 fperrad joined #parrot
10:37 smash joined #parrot
10:37 smash hello everyone
12:27 whiteknight joined #parrot
12:28 whiteknight left #parrot
12:29 whiteknight joined #parrot
12:32 whiteknight msg dukeleto I would like all my commits under the "whiteknight" username, with that email address, thanks.
12:32 aloha OK. I'll deliver the message.
12:34 whiteknight good morning, #parrot
12:39 mikehh hi whiteknight
12:39 whiteknight hello mikehh, how are you today?
12:40 mikehh and yes, I think I also want commits under mikehh
12:40 mikehh whiteknight: passable, passable
12:43 mikehh whiteknight: how's the new job working out, other than the commute bit?
12:54 * mikehh got to change platforms - brb
12:54 mikehh left #parrot
13:01 jsut_ left #parrot
13:01 jsut joined #parrot
13:02 whiteknight pretty well, actually. The commute isn't bad at all really
13:02 whiteknight it's about an hour, but I can do computer work on the train, so it's a lot less stressful than driving
13:04 mikehh joined #parrot
13:08 bluescreen joined #parrot
13:15 darbelo joined #parrot
13:23 whiteknight msg dukeleto I fixed those two tasks in GCI. I have more to add, but am short on time right now
13:23 aloha OK. I'll deliver the message.
13:58 whiteknight We really should do more to make Parrot i18n compatible. Even if we selected a language at compile time, that would be far better than we have now
13:59 whiteknight Actually, I don't think that would be necessary. If we created hashes of error messages in PIR, Parrot could load a different msg hash at runtime depending on CLI flags
13:59 whiteknight I suspect we wouldn't want type, attribute, and method names to change based on language, but we could probably do something to support that at compile-time if we absolutely wanted to
14:00 darbelo If you want to do do it at compile time you can probably hijack some of the CONST_STRING() infrastructure without too much problem.
14:02 whiteknight True. Instead of calling CONST_STRING with an actual C string argument, we call it with a slug. The build looks up the slugs in a file of error messages based on current language
14:02 whiteknight CONST_STRING returns a Parrot STRING, a new CONST_LITERAL would simply place the literal text into the C code, like a preprocessor step
14:03 whiteknight anyway, my point is: if we had an i18n infrastructure in place, we could come up with a million translation projects for GCI students
14:04 whiteknight Plus, if system messages were stored in a hash at runtime, HLLs and user programs could trivially override their text
14:06 rurban joined #parrot
14:06 Andy joined #parrot
14:07 bluescreen left #parrot
14:16 rurban left #parrot
14:16 PerlJam joined #parrot
14:16 PerlJam left #parrot
14:17 PerlJam joined #parrot
14:22 bluescreen joined #parrot
14:46 cotto ~~
14:51 patspam joined #parrot
14:52 bluescreen left #parrot
14:54 tadzik joined #parrot
14:57 tadzik o/
15:03 bluescreen joined #parrot
15:05 dmalcolm joined #parrot
15:25 mtk joined #parrot
15:29 whiteknight You know what I would really really like? Perl6 syntax highlighters included in some major packages
15:29 whiteknight the code highlighting engines in MEdit and GEdit should definitely have it.  pygments too. Geshi obviously
15:29 whiteknight Geshi is used in Trac and MediaWiki, I believe
15:29 moritz whiteknight: vim7 comes with perl6.vim, iirc
15:30 atrodo Didn't vim ship it in this last version?
15:30 moritz and can be used through Text::VimColor
15:30 whiteknight okay, that's Vim. We need more
15:31 whiteknight I think Geshi should definitely have perl 6 support. pygments is what's used on github, where rakudo lives
15:31 moritz problem is, Perl 6 isn't easy to parse
15:31 atrodo Vim is pretty much all I need
15:31 moritz maye I should experiment with it nonetheless
15:31 mtk left #parrot
15:32 whiteknight it probably doesn't need to be perfect. separating constants, variables, symbols, and braces would be a great start
15:32 rurban joined #parrot
15:32 mtk joined #parrot
15:38 darbelo Whatever vim is doing, the others can probably do too.
15:40 cotto Unfortunately Perl 6 syntax highlighting makes slow computers less responsive.
15:41 cotto where "slow computers" means anything <2gHz
15:42 whiteknight For most cases, I don't think we need super-complicated parsers.
15:42 whiteknight you only need to differentiate between a few basic types of tokens
15:43 whiteknight highlighting keywords and braces could be done extremely easily. sigiled variables too
15:43 cotto Sure.  Even something minimal that did a few things correctly most of the time would be nice.
15:44 whiteknight example code on Wikipedia wouldn't need to be fancy, for instance.
15:45 whiteknight Part of me would also like PIR syntax highlighters in a variety of places, but if  that language dies tragically as I am hoping it would, that would be a wasted effort
15:46 whiteknight I suspect winxed could be highlighted using javascript highlighting rules
15:47 darbelo Probably. The syntax is basically the same.
15:47 darbelo It's pretty much just PIR with javascript syntax.
15:48 darbelo And a few less misfeatures.
15:49 mtk left #parrot
16:04 patspam left #parrot
16:05 patspam joined #parrot
16:11 tadzik after looking at Google CodeIn tasks, I could translate some stuff to Polish, not even as a special task
16:16 mikehh tadzik: we probably need you to mentor some GCI student for them to do it
16:20 tadzik cool
16:21 lidden joined #parrot
16:22 mikehh left #parrot
16:25 jsut_ joined #parrot
16:30 jsut left #parrot
16:32 rurban left #parrot
16:35 whiteknight cotto: I've got i18n on the brain today. I'm going to try to put together a draft design I think
16:36 whiteknight if you have any ideas, let me know. I'll send you a draft when I have it ready
16:36 dukeleto 'ello
16:37 whiteknight hello dukeleto++
16:37 dukeleto whiteknight: good localtime()
16:37 dukeleto whiteknight: thank you very much for improving the GCI tasks
16:38 dukeleto i split up some of the translation tasks into many small tasks
16:38 dukeleto added a "research" task to compare and contrast Dalvik and Parrot
16:38 dukeleto added some "make a t-shirt" tasks for Parrot and Perl 6
16:39 whiteknight research to compare/contrast with Neko might be nice too
16:40 dukeleto whiteknight: i was gonna add that right now :)
16:40 dukeleto whiteknight: are you reading my mind?
16:40 whiteknight some kind of a parrot census thing might also be nice: There are plenty of projects on Parrot that we don't know about
16:41 whiteknight I'm just brain-storming now
16:41 dukeleto whiteknight: something like the yearly Git survey
16:41 dukeleto whiteknight: please continue brainstorming :)
16:41 atrodo I think that's a great idea
16:45 dukeleto whiteknight: i just added the "compare neko+parrot" task
16:47 whiteknight in terms of the Product Management team, if we could have somebody put us together a list of projects depending on Parrot, their current status, and developer contact info, that would be invaluable
16:47 whiteknight we do have a list on trac somewhere, though it's ugly
16:52 dukeleto Task : Create a short video/screencast about how Perl 5, Perl 6 and Parrot VM are related
16:52 * dukeleto just created that
16:53 dukeleto whiteknight: that sounds like a good idea
16:58 theory joined #parrot
16:59 tadzik left #parrot
17:02 tadzik joined #parrot
17:05 dukeleto http://google-opensource.blogspot.com/2010/1​1/announcing-accepted-organizations-for.html <--- We got accepted to Google Code-In!
17:05 atrodo Yay!
17:06 * dukeleto needs to write a blog post for parrot.org and blog.perlfoundation.org
17:07 jan joined #parrot
17:09 mikehh joined #parrot
17:09 whiteknight dukeleto++
17:10 whiteknight I promise to get several more tasks into the list this weekend
17:10 dukeleto i didn't realize it, but only 20 orgs got into Code-In. We should be very proud of ourselves.
17:10 dukeleto whiteknight++
17:11 jnthn Wow.
17:11 whiteknight I don't see many perl-related tasks in that list
17:11 jnthn Congrats, and nice work to everyone here who's been pushing the effort. :-)
17:11 atrodo Wow indeed
17:11 whiteknight I assume we're going to get more, and that Rakudo people will be putting some tasks in?
17:12 jnthn Is it open to having some Perl 6 tasks added?
17:12 jnthn OK. :-)
17:12 jnthn What are the dates people can submit?
17:12 jnthn I can probably think of and add some that I'd be happy to mentor.
17:13 whiteknight jnthn: We've been adding tasks to the list already. You can add them at any time if you want
17:13 whiteknight http://trac.parrot.org/parro​t/wiki/GoogleCodeIn2010Tasks
17:13 jnthn whiteknight: OK.
17:14 jnthn whiteknight: No, I meant the dates there will be students able to take them. (more)
17:14 dukeleto jnthn: i've been adding Perl 6 tasks
17:14 jnthn $dayjob is kinda busy at the moment so I'd want to be able to give them proper attention.
17:14 dukeleto whiteknight: there are perl tasks. Keep scrolling
17:14 whiteknight ok
17:15 dukeleto jnthn: we can add tasks at any time
17:18 whiteknight What would it take to embed a new script handler into FireFox?
17:18 whiteknight I know it  can be done in IE using ActiveX
17:18 whiteknight at least, for older versions of IE. I don't know about IE8 or IE9
17:25 atrodo whiteknight> I've wondered that myself
17:26 whiteknight I want to start a project to embed Parrot in C#. And then I want to use that in another project to allow using Parrot bytecode files in windows services and WCF web services
17:26 whiteknight somewhere along the line, if I could get it running in a browser, I would love to do that as well
17:27 dukeleto whiteknight: sounds interesting
17:27 whiteknight the C# wrapper should be pretty thin, and will follow in parallel with the embedding API work, since function mappings will typically be 1:1
17:29 jnthn whiteknight: fwiw, I did manage to at least create a Parrot interpreter from C# once.
17:30 jnthn Using the externals stuff.
17:30 whiteknight jnthn, yeah I did it once too and wrote a blog post about it
17:30 whiteknight that's why I know it should be simple for me to do it simultaneously with other API work
17:30 jnthn Oh, I'm not sure a blug.
17:30 jnthn *I
17:30 whiteknight ...I didn't understand that response at all
17:30 whiteknight :)
17:31 jnthn I didn't write a blog post on it. :-)
17:31 jnthn .oO( Must save ablaut abuse for #perl6 )
17:41 * whiteknight is envious that the mono people are using MediaWiki
17:41 whiteknight and all we get is trac
17:42 dukeleto Let's only migrate one thing at a time :)
17:43 whiteknight no, I don't have any hopes that we will migrate away from trac at this point
17:49 contingencyplan joined #parrot
18:04 whiteknight part of me wonders why we chose Trac from the universe of all ticket trackers
18:04 whiteknight Oooh, a new theme for our trac might be a nice project
18:13 darbelo There's a universe of ticket trackers? I'd be hard pressed to name more than a handful.
18:15 darbelo Heck, I don't think I've used anything other than Trac and Bugzilla for extended periods.
18:21 whiteknight http://en.wikipedia.org/wiki/Comp​arison_of_issue_tracking_systems
18:22 atrodo Ya, there's a lot of them.  Not too many good ones
18:22 whiteknight I've used FogBugz, Bugzilla, RT, Gemini, JIRA, OTRS, TFS, not to mention things like sourceforge, googlecode, and github that we probably can't make use of on our own
18:27 darbelo I don't think we can use the various hosted or non-OS solutions. And if you take the timeframe of the migration into account. There's not much left on the list.
18:29 darbelo Trac doesn't strike me as a horrible too horrible a choice, really.
18:33 whiteknight I find it's wiki in particular to be extremely poor
18:33 whiteknight maybe I'm just a wiki snob
18:34 whiteknight the ticket tracker portion of it is decent enough
18:34 darbelo It's only poor if you compare to a real wiki :)
18:34 whiteknight acceptability though obscurity? Sounds like a microsoft strategy
18:35 darbelo It's more of a 'acceptability through reduced expectations' thing.
18:37 darbelo I don't expect Trac's wiki to be comparable to say MediaWiki.
18:38 darbelo Trac is a bug tracker that just happens to bundle a smallish wiki for your convenience.
18:42 darbelo Think SQLite vs PostgreSQL. You expect different things from each.
18:44 whiteknight I would like a lot more from our wiki software, eventually. I look at what Mono has, or Mozilla (both MediaWiki) and they use them to great effect hosting documentation and things
18:47 darbelo True, a better wiki could help. If it leads to better documentation and newbie-friendliness I'd even argue that losing bug tracker integration isn't a high price to pay.
18:48 whiteknight We don't use the integration to great effect now anyway
18:48 darbelo Also true.
18:49 darbelo But I don't have the time to really help on any of this, so I'm hesitant to advocate more work for other people :)
18:51 whiteknight I'm not advocating either, just idly complaining
18:58 smash left #parrot
19:19 mikehh whiteknight: just looked at the wiki comparison, Trac seems to feature well
19:21 mikehh whiteknight: I tend to edit elsewhere then move it in
19:21 mikehh whiteknight: with approptiate markup in the wiki
19:23 dukeleto It's All Text, the firefox plugin, is a life saver.
19:23 mikehh not necessarily the best method, but I have used some bug-tracking systems that are truely horrible
19:23 dukeleto You click on a button on the text input field, and it pops up your $EDITOR, then replaces the content when you save+exit the editor
19:23 dukeleto I couldn't have written the GCI task page without it
19:25 whiteknight I have never heard of that extension before
19:25 mikehh :-}
19:25 whiteknight mikehh: in general terms, the Trac wiki really is lacking, even if it does outperform some other issue tracking software
19:26 dukeleto whiteknight: try it. it will make you a happier person.
19:26 whiteknight dukeleto: any happier, and I will burst
19:26 darbelo Unless you use another browser :)
19:26 dukeleto https://addons.mozilla.org​/en-US/firefox/addon/4125/
19:27 * dukeleto does not shed a tear for those that use other browsers ;)
19:27 dukeleto I tried Chrome for about 3 days. Then I wanted a browser that worked.
19:27 * dukeleto is sure chrome has gotten better, but is too attached to the bazillion awesome FF extensions
19:28 darbelo Heh. I'm giving Camino a spin on the work laptop, it's not awful.
19:29 darbelo But then, it's basically firefox pretending to be a OSX app.
19:32 darbelo But I don't really demand too much out of a browser. Mostly, I'm fine with anything that browses :)
19:34 sorear elinks has the F4 key for this :)
19:37 dukeleto i am concerned about getting no resonse about my email about the git conversion to parrot-dev
19:37 dukeleto i guess no one cares about how their names show up
19:39 darbelo My name shows up fine. FWIW.
19:40 PerlPilot joined #parrot
19:41 dukeleto darbelo++ # making my life easy
19:42 darbelo There seems to be trouble for a few names though. paul_the_greek's for example.
19:43 mikehh dukeleto: re: the list, I was wondering what happens if I commit as mikehh
19:43 PerlJam left #parrot
19:43 dukeleto darbelo: yeah. I saw that.
19:43 dukeleto mikehh: git has author names and emails. mikehh is a svn username
19:44 mikehh dukeleto: it is also my username on github
19:44 dukeleto mikehh: look at one of your git commits to see what your author name and email are
19:45 whiteknight dukeleto: I sent you an msg about it this morning
19:45 whiteknight dukeleto: you have the link to that gist handt?
19:45 whiteknight handy?
19:46 dukeleto whiteknight: i got your message, you are good :)
19:46 whiteknight w00t
20:04 whiteknight dukeleto: when are you pushing the finished masterpiece to github?
20:07 whiteknight I'm getting ansty in my pantsies
20:07 bluescreen left #parrot
20:18 mtk joined #parrot
20:20 plobsing left #parrot
20:22 bluescreen joined #parrot
20:25 whiteknight left #parrot
20:27 x3nU_ left #parrot
20:28 darbelo left #parrot
20:30 mtk left #parrot
20:41 mtk joined #parrot
20:41 cotto_work joined #parrot
20:42 atrodo cotto_work lives!
20:42 cotto_work so it would appear
20:50 mikehh All tests PASS (pre/post-config, make corevm/make coretest, smoke (#946) fulltest) at r49786 - Ubuntu 10.10 amd64 (g++-4.5 with --optimize)
20:53 mtk left #parrot
20:54 bluescreen left #parrot
21:26 cotto_work left #parrot
21:38 cotto_work joined #parrot
21:45 Themeruta joined #parrot
21:45 Themeruta is now known as NotFound_gf
21:45 NotFound_gf hi
21:49 sorear hi
22:04 patspam left #parrot
22:06 patspam joined #parrot
22:12 fperrad left #parrot
22:23 dngor_ joined #parrot
22:23 dngor left #parrot
22:24 dngor_ is now known as dngor
22:27 dngor left #parrot
22:47 bacek left #parrot
23:02 bacek joined #parrot
23:06 cotto_work left #parrot
23:23 rurban joined #parrot

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

Parrot | source cross referenced