Da
Show enters and exits. Hide enters and exits.
| 00:15:07 | boyscout | Add task :todos to do a simple grep for todo marks in kernel/. - 580e56e - Eero Saynatkari |
| 00:15:07 | boyscout | Added documentation (and some @todos) for ARGF. - 8e1d27b - Eero Saynatkari |
| 00:15:07 | boyscout | while true -> loop in argf.rb - 701854e - Eero Saynatkari |
| 00:18:09 | boyscout | CI: Build 701854e failed. http://ci.rubini.us/rubinius/builds/701854eccb0674214f58823c3a05bb7d0fd8b09a |
| 00:18:56 | evan | rue! |
| 00:18:58 | evan | you didn't run the tests! |
| 00:20:49 | boyscout | Revert "while true -> loop in argf.rb" - 2ee9610 - Evan Phoenix |
| 00:21:20 | rue | I did! :/ |
| 00:21:30 | evan | while true is better than loop anyway. |
| 00:22:22 | rue | N-uh! |
| 00:22:34 | rue | Although it seems to work better here...reviewing |
| 00:22:36 | evan | uhuh! |
| 00:22:50 | evan | loop do is probably the slowest way to loop ever. |
| 00:23:49 | boyscout | CI: Build 2ee9610 failed. http://ci.rubini.us/rubinius/builds/2ee9610723ac82a92f3f6ced3b36cff29b1b879e |
| 00:23:52 | rue | Oh! I see, I thought I did not include that change |
| 00:23:57 | evan | hey wtf. |
| 00:24:31 | rue | I gots it |
| 00:24:34 | evan | you changed read |
| 00:24:36 | evan | i see. |
| 00:24:42 | evan | in a documentation commit |
| 00:24:43 | evan | ! |
| 00:24:45 | evan | :) |
| 00:25:11 | rue | Yep, I thought I blocked it out but it snuck in |
| 00:25:25 | evan | crafty code. |
| 00:29:27 | tarcieri | holy crap |
| 00:29:33 | evan | tarcieri: HEY I HEAR REIA IS DEAD |
| 00:29:34 | evan | IS THAT TRUE |
| 00:29:37 | evan | :D |
| 00:29:38 | tarcieri | lol |
| 00:29:42 | boyscout | Oops. Re-enable nil check for ARGF#read, comments for future. - 9beb172 - Eero Saynatkari |
| 00:29:43 | tarcieri | uhh, no |
| 00:29:49 | evan | i'm kidding |
| 00:29:51 | evan | i read your post on it. |
| 00:29:52 | tarcieri | and holy crap, you made Ruby's parser spit out intelligable errors |
| 00:29:59 | evan | yeah! |
| 00:30:03 | evan | you love it, don't ya. |
| 00:30:04 | tarcieri | AMAZING! |
| 00:30:06 | tarcieri | yes |
| 00:30:12 | rue | Notes for future code archaeologists |
| 00:31:01 | evan | rue: when 1.0 comes out, we should get a leather bound notebook and write a bunch of rubinius notes in it |
| 00:31:11 | evan | then weather it and put in a library in venice. |
| 00:31:19 | tarcieri | heh |
| 00:31:19 | evan | so the future code archeoligists have a good time. |
| 00:31:52 | tarcieri | I always wondered why Ruby's parser gave such shitty errors, then I looked at parse.y and my eyes proceeded to bleed profusely |
| 00:32:16 | evan | well, syntax errors are really hard to report well |
| 00:32:17 | rue | tarcieri: Really hard to concentrate on error improvement with bleeding eyes, nodnod |
| 00:32:23 | tarcieri | heh |
| 00:32:30 | evan | back a few years ago |
| 00:32:36 | evan | when I was working on the backtrace format |
| 00:32:39 | tarcieri | evan: I'm still playing around with a PEG for Reia... really awesome to be able to give line and column of syntax errors |
| 00:32:41 | boyscout | CI: 9beb172 success. 3005 files, 11472 examples, 35615 expectations, 0 failures, 0 errors |
| 00:32:42 | evan | I did a survey of every language I could find |
| 00:32:54 | evan | to see all possible existing backtrace formats |
| 00:33:02 | evan | tarcieri: rubinius has that too |
| 00:33:06 | tarcieri | nice |
| 00:33:19 | evan | it's hard to communicate exactly what went wrong there though |
| 00:34:57 | tarcieri | yeah errors are hard, heh |
| 00:35:37 | tarcieri | starting with sane labels for all the terminals in your grammar probably helps a lot |
| 00:36:01 | tarcieri | unexpected kEND, expecting $end! you TARD |
| 00:37:30 | evan | yeah |
| 00:37:37 | evan | that one happens so often |
| 00:37:41 | evan | and it's a terrible message. |
| 00:37:48 | evan | thats why I fixed it up so specially. |
| 00:37:57 | tarcieri | yeah awesome |
| 00:50:09 | boyscout | appease the compiler gods with missing includes for uintptr_t - 21874f9 - Charles Comstock |
| 00:53:11 | boyscout | CI: 21874f9 success. 3005 files, 11472 examples, 35615 expectations, 0 failures, 0 errors |
| 00:56:15 | rue | I guess that does it for now, sleepy. |
| 07:52:03 | dbussink | morning |
| 09:19:16 | rue | Hola |
| 09:24:09 | ujihisa | ¡Hola! |
| 14:03:02 | manveru | well, seems like my build fails because of a missing require? |
| 14:03:39 | manveru | rakelib/kernel.rake tries to use Digest::SHA1 but only requires digest |
| 14:07:55 | manveru | hm, no, that's not it |
| 14:09:23 | manveru | ah, it still used my system ruby... |
| 14:09:40 | manveru | for reference: library not found for class Digest::SHA1 -- digest/sha1 |
| 14:09:49 | manveru | means you're running 1.9... |
| 14:13:48 | manveru | and next one |
| 14:13:54 | manveru | /home/manveru/.rvm/ruby-1.8.7-p174/bin/ruby -Ilib lib/bin/compile.rb -V -T default lib |
| 14:13:55 | manveru | ./lib/etc.rb:4: uninitialized constant Etc::FFI (NameError) |
| 14:17:13 | dbussink | manveru: what's your rubyopt? |
| 14:17:48 | manveru | oh |
| 14:17:52 | dbussink | manveru: the last issue shows up if you have for example lib as a default load path, so system ruby then picks up rubinius stuff and fails |
| 14:17:59 | manveru | -rubygems |
| 14:18:07 | manveru | thanks, works without |
| 14:18:45 | manveru | yay, it built |
| 14:19:09 | manveru | now let's try it from scratch and tell wayneeseguin to unset RUBYOPT for the install |
| 14:31:27 | rue | Wee |
| 14:31:42 | rue | Should probably add RUBYOPT to the installation docs |
| 14:38:06 | wayneeseguin | manveru: it's supposed to unset rubyopt :/ |
| 14:38:11 | wayneeseguin | See ruby-installer line 128 |
| 17:23:18 | hooopy | marco!! |
| 17:27:38 | brixen | polo? |
| 17:32:13 | evan | polo! |
| 17:37:55 | evan | so, I broke exception handling |
| 17:37:57 | evan | and ran the specs |
| 17:38:07 | evan | and seems with broken exception handling |
| 17:38:19 | evan | there is a fork bomb that is triggered |
| 17:38:26 | evan | i've ground my machine to a halt a few times. |
| 17:38:55 | brixen | hmm |
| 17:47:55 | rue | Sweet |
| 17:48:11 | rue | Everyone saw http://ruby-std.netlab.jp yes? |
| 17:48:26 | brixen | I saw, yes |
| 17:48:34 | brixen | we should probably tweet it up |
| 17:48:40 | rue | I am only about 100 pages into it :P |
| 17:51:53 | evan | i've gotten paper copies of previous drafts. |
| 17:52:17 | evan | wow, they decided to actually spec methods. |
| 17:52:26 | evan | they hadn't done that in any other drafts. |
| 17:55:31 | evan | HAHA! |
| 17:55:36 | evan | they've spec |
| 17:55:45 | evan | lambda { |(a,b)| } as |
| 17:55:49 | evan | implementation dependent |
| 17:55:54 | evan | we can just delete the code to support it! |
| 17:55:55 | evan | :D |
| 17:56:56 | brixen | hehe |
| 17:57:06 | brixen | wow, I didn't see the method specs before |
| 17:57:32 | brixen | this is a lot of work |
| 17:57:58 | Zoxc | likes how bindings is not present :) |
| 17:58:17 | evan | we should figure out how to get rubyspec attached to this |
| 17:58:28 | evan | perhaps we should go along the path we discussed before |
| 17:58:41 | evan | basically, tag our specs that conform to sections in this doc |
| 17:58:49 | brixen | yeah |
| 17:59:02 | brixen | we should probably not unless these #s are staying constant |
| 17:59:07 | evan | tag spec/frozen/core/kernel/p.rb officialspec:15.3.1.1.12 |
| 17:59:24 | brixen | s/unless/until/ |
| 17:59:30 | evan | yeah, we should ask yuigi about that |
| 18:00:28 | Zoxc | too bad I don't care about 1.8 :( |
| 18:00:48 | evan | this covers 1.9 too. |
| 18:00:50 | evan | it has to. |
| 18:01:11 | brixen | yeah, 1.9 is mostly a superset of 1.8 |
| 18:01:20 | brixen | the vast majority of behavior is the same |
| 18:01:34 | brixen | it's not a precise superset though |
| 18:01:36 | rue | evan: Conformant program may not rely on implementation-specific behaviour ;) |
| 18:01:46 | evan | huzzah! |
| 18:02:01 | evan | that means that |(a,b)| has been officially deprecated. |
| 18:02:12 | rue | Provisionally :P |
| 18:02:21 | evan | ah yes, provisionally |
| 18:02:35 | Zoxc | |(a,b)| is invalid :D |
| 18:03:01 | evan | Zoxc: I mean that as block args |
| 18:04:06 | rue | I do not like using tags for this purpose |
| 18:04:52 | Zoxc | why does that work in ruby anyway? (and how does it work) |
| 18:05:45 | evan | it works because matz added support for it in masgn |
| 18:05:49 | evan | and block args reuses masgn |
| 18:06:02 | evan | it unpacks a array's elements |
| 18:06:20 | evan | if i have |
| 18:06:28 | evan | foo { |a,(b,c)| ... } |
| 18:06:32 | evan | and foo does |
| 18:06:36 | evan | yield 1, [2,3] |
| 18:06:37 | rue | Similar to `a, (b, c), d = [1, [2, 3], 4]` |
| 18:06:41 | evan | a == 1, b == 2, c == 3 |
| 18:10:10 | evan | huh, the 3rd argument to backtrace has been deprecated too |
| 18:10:13 | evan | er. |
| 18:10:15 | evan | to raise |
| 18:10:18 | evan | the backtrace argument |
| 18:10:35 | rue | Or forgotten? |
| 18:10:38 | Zoxc | no backtrace faking! |
| 18:11:04 | evan | *shrug* |
| 18:11:14 | evan | says that arg count > 2 is impl. dependent |
| 18:11:28 | evan | I've overloaded our 3rd argument to support chained exceptions |
| 18:11:45 | rue | Hm, might be worth a query |
| 18:11:49 | evan | raise MoreGenericError, "something generic happened", more_specific |
| 18:11:59 | evan | where more_specific is another Exception object |
| 18:12:05 | Zoxc | can you use custom backtrace objects? |
| 18:12:14 | evan | in rbx? |
| 18:12:17 | evan | on in MRI? |
| 18:12:21 | evan | s/on/or/ |
| 18:12:53 | Zoxc | in ISO-RUBY |
| 18:13:00 | evan | looks like no |
| 18:13:38 | Zoxc | can rbx print arguments in backtraces? |
| 18:14:05 | evan | we've discussed it |
| 18:14:08 | evan | but not currently, no. |
| 18:14:21 | evan | adding support for that isn't too hard. |
| 18:15:10 | Zoxc | quite handy since sometimes arguments are passed in *args, which means you need to look through I million method to find out how the method you're looking at should be used =P |
| 18:16:23 | evan | yep. |
| 18:16:26 | evan | it's not too hard to do |
| 18:16:35 | evan | need to just make room for them in a Location object |
| 18:16:46 | evan | and have the VM add them when it generates the Location object array |
| 18:31:00 | Zoxc | wonders if ISO/IEC 646 is compitable with UTF8 |
| 18:32:03 | Zoxc | oh and why does an syntax error in Kernel#eval print the messages terminate ruby? |
| 18:35:17 | evan | eh? |
| 18:35:33 | evan | i'm not aware of what you speak |
| 18:41:34 | rue | An uncaught SyntaxError, you mean? |
| 18:43:03 | rue | Same as any other uncaught error, and SyntaxError not being < StandardError means you need to look for it or ScriptError specifically |
| 18:47:46 | evan | yeah, Kernel#eval raises SyntaxError |
| 18:47:52 | evan | you need to rescue it |
| 18:47:56 | evan | if you don't want it to bubble up. |
| 18:49:13 | dbussink | Zoxc: that iso spec specifies the standard 7 bit latin chars |
| 18:49:41 | dbussink | Zoxc: basically the lower ascii characters, which are utf-8 compatible |
| 18:49:44 | Zoxc | it's a successor of ASCII implying changes |
| 18:50:40 | dbussink | Zoxc: "The final 1991 version of the code is identical to ASCII" |
| 18:50:58 | dbussink | originally they replaced the $ sign with an international currency symbol, but that was reverted |
| 19:01:42 | evan | the official spec does finally describe the difference between #clone and #dup nicely |
| 19:02:31 | rue | Neat, I can co-opt it if it is better than the description I wrote |
| 19:19:24 | cremes | evan: working with 'main' author to produce a small code example showing id2ref breakage |
| 19:19:34 | evan | ok |
| 19:19:37 | evan | tell ara I said hi. |
| 19:19:40 | cremes | his code doesn't call id2ref anywhere; might be a bug with Module.module_eval |
| 19:19:46 | cremes | will do |
| 19:19:50 | evan | it's in fattr |
| 19:19:57 | evan | is where the call to _id2ref is. |
| 19:20:05 | evan | ara also wrote fattr |
| 19:20:13 | cremes | oh... |
| 19:21:00 | evan | http://github.com/ahoward/fattr/blob/master/lib/fattr.rb#L128 |
| 19:21:32 | cremes | texting this info to ara now; good find |
| 19:22:43 | evan | cool. |
| 19:42:02 | dbussink | evan: still deep in fixing the nasty jit issue? |
| 19:42:09 | evan | yep |
| 19:42:21 | evan | trying to debug why my change destablizes things. |
| 19:42:24 | dbussink | evan: or is there light at the end of the tunnel ? |
| 19:43:20 | evan | there always is |
| 19:43:25 | evan | sometimes the light is just around a couple of bends |
| 19:43:28 | evan | and I can't see the bends yet |
| 19:43:28 | evan | :D |
| 19:43:45 | evan | it's just software afterall. |
| 19:44:41 | dbussink | hehe, i know the drill, hit some annoying issues today as well |
| 19:45:09 | slava | hi evan |
| 19:47:25 | evan | hi slava. |
| 19:48:29 | slava | i've been packing and moving for the last week or so, time to do some work again |
| 19:49:08 | evan | where ya moving? |
| 19:49:32 | slava | i'm in canada staying with parents for two weeks, then going to NZ. time for a new adventure |
| 19:50:56 | evan | fun! |
| 19:51:03 | evan | well, off to lunch. |
| 19:51:34 | slava | minneapolis was a pretty nice place to live |
| 19:58:14 | dwaite | afternoon |
| 21:20:55 | rue | Hi hi |
| 22:32:19 | dwaite | quiet day |
| 22:33:47 | rue | Or are we just talking on some secret channel? |
| 22:59:41 | evan | nope, just quiet. |
| 23:32:07 | evan | so, we're going to need a way to refer to the "IPA Ruby Standardization WG Draft" |
| 23:32:19 | evan | and we're going to need to push on the wording for it hard |
| 23:32:23 | evan | it's only day 1 |
| 23:32:30 | evan | and people are already confusing it with rubyspec. |
| 23:33:04 | Zoxc | calls it iso-rubyspec |
| 23:35:49 | evan | not a bad name |
| 23:35:52 | evan | but still confusing |
| 23:36:03 | evan | but it conflates that they're related |
| 23:36:31 | Zoxc | in-search-if-a-ruby-spec :) |
| 23:36:40 | Zoxc | of* :/ |
| 23:36:45 | evan | i think |
| 23:36:49 | evan | ruby standardization |
| 23:36:51 | evan | is clearer. |
| 23:36:54 | evan | but thats what it is. |
| 23:39:52 | brianmario | anyone run into a pegged CPU when trying a rake gems:install from a rails 2.3.4 app? |
| 23:40:14 | brianmario | starts out pegged at 180%+ then slowly drops down to ~100% and stays that way |
| 23:40:32 | brianmario | memory usage grows ~500k every few seconds indefinitely |
| 23:40:42 | brianmario | or at least, for 30min before I killed it |
| 23:44:23 | evan | hm |
| 23:44:24 | evan | nope |
| 23:44:36 | brianmario | trying to find the hot spot |
| 23:44:39 | evan | but we've gotten reports, and i've seen, some bad news cases with gem install |
| 23:44:41 | evan | ok! |
| 23:44:52 | evan | it's possible it's a bug in Socket |
| 23:51:26 | brianmario | not sure if I'm tracking it down with the right tool, but Shark has __memcpy as the 6th item from the top |
| 23:51:58 | brianmario | running a Time Profile, for about 2min on the hung rbx proc |
| 23:53:33 | evan | hm. ok. |
| 23:53:37 | evan | that might just be the GC |