Show enters and exits. Hide enters and exits.
| 00:21:08 | tarcieri | On Thu, Jun 11, 2009 at 1:11 AM, Yukihiro Matsumoto<matz@ruby-lang.org> wrote: |
| 00:21:08 | tarcieri | > It was me introduced a bug. I neglect to run the test this time. |
| 00:21:08 | tarcieri | > Unfortunately, 1.9 maintainers seemed to trust me, who they shouldn't |
| 00:21:09 | tarcieri | > trust. I hope we could learn something from this experience. |
| 00:21:20 | tarcieri | *facepalm* |
| 00:23:13 | tarcieri | lesson: don't trust the language creator to run tests when making security fixes on his own language? |
| 02:59:08 | manveru | tarcieri: don't trust nobody! |
| 17:41:25 | weepy | HELLO ... Hello ... hello ... h,,,, ... ,,,,, . . . . |
| 17:43:13 | absentia | waiting.. |
| 17:44:27 | weepy | - seems the logs on donttreadonme aren't updating .... I thought it seemed very quiet in the rubinius room ! |
| 17:52:43 | dbussink | weepy: they are updated afaik, it's just really quiet the last few days |
| 18:10:53 | brixen | weepy: evan and I are in SF for a couple days |
| 18:11:16 | brixen | weepy: feel free to chat us up :) |
| 18:18:24 | dbussink | brixen: doing productive pair programming or more like a conf thingy? |
| 18:32:09 | weepy | ah SF - cosy - I like that place |
| 18:32:53 | weepy | yes you're right -> I can see my message in the logs :-) |
| 18:33:34 | weepy | Im looking forward to a semi stable version I can try out! |
| 18:46:18 | brixen | dbussink: I'm getting the jit tut atm |
| 18:46:47 | brixen | weepy: you should try your code now and let us know what doesn't work |
| 18:51:49 | weepy | brixen: does it run rails ? |
| 18:52:17 | weepy | nearly all my code is fairly heavily entwined |
| 18:52:29 | weepy | (with rails) |
| 19:05:00 | brixen | weepy: it runs some rails code and we're fixing stuff |
| 19:05:17 | brixen | for example, we fail ~5-10% of the AR tests if you run them by hand |
| 19:07:02 | weepy | so close .. ! nice one |
| 19:07:44 | brixen | weepy: yeah, we just need to get more people trying stuff out |
| 19:08:06 | weepy | where are the more up to date installation instructions ? |
| 19:08:22 | brixen | well, you don't need to instaall |
| 19:08:25 | brixen | -a |
| 19:08:31 | weepy | instaaaaal |
| 19:08:39 | brixen | just clone, rake, and run bin/rbx |
| 19:08:43 | weepy | ok |
| 19:08:51 | brixen | or ln vm/vm to anywhere |
| 19:08:54 | weepy | it has llvm version in there ? |
| 19:09:22 | brixen | you can even move vm/vm somewhere and rename it, but don't change the name of the dev root dir |
| 19:09:43 | weepy | Receiving objects: 9% (11105/113991), 3.32 MiB | 324 KiB/s |
| 19:09:44 | dbussink | brixen: still have some stuff that gives me a "Tried to use object of type Object (42) as type Array (5)" error in extlib specs that could point at some issue i guess |
| 19:09:45 | weepy | cooool |
| 19:09:54 | dbussink | dunno if that is easy fixable though |
| 19:09:54 | brixen | if you get llvm from svn and check it out to vm/external_libs/llvm, rake will build it |
| 19:10:15 | brixen | dbussink: ok, point me to the command to run |
| 19:10:45 | weepy | you should put some decent instructions on the github README |
| 19:10:47 | dbussink | brixen: i'll see if i can make a simple repro |
| 19:11:01 | evan | dbussink: yeah, that extlib's LazyArray |
| 19:11:05 | evan | dbussink: I made it do that rather than crash |
| 19:11:27 | evan | LazyArray breaks the Object#dup protocol we use |
| 19:11:30 | brixen | weepy: you mean like "See doc/getting_started.txt. |
| 19:11:33 | evan | so we need to figure out the way to fix that. |
| 19:12:25 | weepy | briken: lol yes - make it *bold* |
| 19:12:48 | dbussink | evan: ah, do you consider it abusing? |
| 19:12:55 | dbussink | evan: or something rbx should properly support? |
| 19:13:04 | evan | we should support it |
| 19:13:04 | brixen | weepy: heh ok |
| 19:13:07 | evan | even though it's ugly. |
| 19:13:23 | boyscout | Style cleanup - 9298d53 - Evan Phoenix |
| 19:13:23 | boyscout | Use a different define name - de05b18 - Evan Phoenix |
| 19:13:23 | boyscout | Add bounds check to Tuple#at inline code - 915ad6e - Evan Phoenix |
| 19:13:23 | boyscout | Fix edge case for stopping LLVM's background thread - fbd7147 - Evan Phoenix |
| 19:13:23 | boyscout | Cleanup cruft - 5cc3a84 - Evan Phoenix |
| 19:13:44 | dbussink | evan: what's the cause, the initialize_copy stuff? |
| 19:13:49 | weepy | im confused about llvm - do i need edge version or no . doesn't mention it in docs |
| 19:14:01 | evan | dbussink: no. |
| 19:14:01 | brixen | weepy: you don't need it at all |
| 19:14:06 | weepy | ok kool |
| 19:14:13 | evan | dbussink: our Object#dup turns around and calls Object#copy_from |
| 19:14:23 | evan | LazyArray, because it uses instance_methods to setup it's proxy methods |
| 19:14:29 | evan | sets up a proxy for copy_from |
| 19:14:33 | brixen | weepy: but if you want to use the jit, you need to fetch llvm from svn because had a patch committed that we depend on |
| 19:14:37 | evan | which ends up breaking stuff. |
| 19:14:40 | brixen | weepy: I'll update the docs |
| 19:15:00 | dbussink | evan: ah, well, i can exclude copy_from from it |
| 19:15:16 | evan | no no |
| 19:15:21 | evan | we need to fix it |
| 19:15:28 | evan | not LazyArray |
| 19:15:47 | dbussink | ok, just trying to be helpful ; |
| 19:15:49 | dbussink | ;) |
| 19:15:52 | evan | yep yep. |
| 19:15:57 | evan | i'll fix it later today hopefully. |
| 19:16:00 | weepy | brixen: cool - im runnning - if you want more people trying out, you probably need a few more blog posts or sommit asking for such - there's never very much info on the blogosphere |
| 19:16:25 | brixen | weepy: definitely, we're planning some blog posts |
| 19:16:30 | brixen | weepy: you can post too :) |
| 19:16:56 | weepy | yip defo - |
| 19:20:18 | boyscout | CI: 5cc3a84 success. 2709 files, 10678 examples, 33478 expectations, 0 failures, 0 errors |
| 19:24:32 | weepy | Woop: |
| 19:24:34 | weepy | Finished in 164.138114 seconds |
| 19:24:35 | weepy | 2709 files, 10677 examples, 33476 expectations, 0 failures, 0 errors |
| 19:24:41 | weepy | Woop => Finished in 164.138114 seconds |
| 19:24:43 | weepy | 2709 files, 10677 examples, 33476 expectations, 0 failures, 0 errors |
| 19:28:53 | brixen | woop woop |
| 19:43:22 | boyscout | Optimize (out) splat creation - 4773e97 - Evan Phoenix |
| 19:50:15 | boyscout | CI: 4773e97 success. 2709 files, 10678 examples, 33478 expectations, 0 failures, 0 errors |
| 19:53:37 | weepy | ok so i did rake |
| 19:53:43 | weepy | then checked out llvm |
| 19:53:53 | weepy | now rake seems to run the tests |
| 19:54:04 | weepy | how to get it to compile llvm ? |
| 19:55:51 | weepy | trying: rake clean; rake build |
| 19:57:53 | brixen | did you put llvm in vm/external_libs/llvm? |
| 19:58:05 | brixen | then RBX_LLVM=1 rake |
| 20:21:26 | tarcieri | well |
| 20:21:34 | tarcieri | I'm going to go hang out with the Pythonistas again today |
| 20:21:43 | tarcieri | including one of the PyPy dudes working on their JIT |
| 20:21:52 | tarcieri | he's gonna demo it |
| 20:22:04 | tarcieri | so if there's source code infoz I can snag, I'll let you guys know |
| 20:22:07 | tarcieri | in case you want to peep it |
| 20:22:26 | tarcieri | christ that place is bizzaro world though |
| 20:22:29 | tarcieri | we don't use git, we use hg |
| 20:22:38 | tarcieri | we don't use cruisecontrol or integrity, we use buildbot |
| 20:22:59 | tarcieri | we don't use rcov, we use figleaf |
| 20:23:28 | evan | hehe |
| 20:25:12 | tarcieri | the whole concept of what they do is pretty cool though |
| 20:25:17 | tarcieri | everyone hooks their laptop up to a projector |
| 20:25:25 | tarcieri | so you can see what people are hacking on |
| 20:25:28 | tarcieri | and how |
| 20:25:57 | tarcieri | the PyPy dude had some crazy debugger tool he wrote in Pygame |
| 20:26:07 | tarcieri | for spelunking through stack traces |
| 20:26:11 | evan | hah |
| 20:26:17 | evan | yeah, i've seen something like that they had |
| 20:30:54 | tarcieri | for what it is, Erlang's visual debugger is pretty badass |
| 20:31:00 | tarcieri | but it's written using Tk |
| 20:31:05 | tarcieri | or something equally as ugly looking |
| 20:31:14 | tarcieri | Erlang JUST got bindings to wx |
| 21:24:31 | dgtized | evan: do we have any statistics on the standard sizes of method tables? |
| 21:30:42 | evan | dgtized: nope, but you could add some code to print them out |
| 22:21:48 | dgtized | evan: any suggestions on where in the program lifetime it would make sense to count? |