Da
Show enters and exits. Hide enters and exits.
| 00:00:34 | Defiler | but time to watch 24 first :) |
| 00:05:27 | evan | mmm |
| 00:05:28 | evan | cool |
| 00:05:34 | evan | new variable system is working nicely. |
| 00:06:03 | brixen | nice |
| 00:10:47 | boyscout | Quit as early as possible if we're building on an unsupported platform - 04ba109 - Wilson Bilkovich |
| 00:18:52 | boyscout | CI: rubinius: 04ba109 successful: 3463 files, 13862 examples, 41517 expectations, 0 failures, 0 errors |
| 00:24:35 | Defiler | evan: http://github.com/evanphx/rubinius/issues/issue/332 |
| 00:24:52 | evan | ok, thanks. |
| 00:24:58 | evan | i'll put them in place now. |
| 00:25:17 | Defiler | cool |
| 00:25:26 | evan | done. |
| 00:52:30 | boyscout | Updated MSpec source to a11ca3cf. - ac4d7be - Brian Ford |
| 00:52:30 | boyscout | Updated CI specs to RubySpec 7e8b572e. - 31d008a - Brian Ford |
| 00:53:08 | brixen | well, planned on getting c-api specs moved over |
| 00:53:16 | brixen | instead just fixed messed up specs :/ |
| 00:54:46 | Defiler | what would it take to get the build working on 1.9? I might be willing to spend a minute on that sometime |
| 00:55:08 | brixen | various stuffs |
| 00:55:15 | brixen | making sure melbourne builds |
| 00:55:24 | brixen | accounting for #require changes |
| 00:55:33 | brixen | feel free to whack on it |
| 00:55:41 | Defiler | well there's some really immediate thing where gitlib can't be found |
| 00:55:59 | brixen | yeah, cus #require changed |
| 00:56:10 | brixen | could probably put . in $: |
| 00:56:18 | brixen | gotta catch a bus |
| 00:56:23 | brixen | back in a bit.. |
| 01:00:46 | boyscout | CI: rubinius: 31d008a successful: 3463 files, 13862 examples, 41516 expectations, 0 failures, 0 errors |
| 01:24:24 | Defiler | man, shell script condition syntax is so wack |
| 01:24:40 | Defiler | I'm trying to write a condition that is true when the output of a certain grep is blank |
| 01:24:48 | Defiler | and it's like "error: hate huuumaaaaaanns" |
| 01:25:12 | Defiler | error: put your hand in the box |
| 01:25:31 | explody | er, why not just check the return code from grep |
| 01:26:10 | explody | 0 if it hits, 1 if it doesn't |
| 01:26:31 | Defiler | oh, I didn't know it did that. handy |
| 01:26:57 | explody | if `grep -q thing /some/file`; then echo "yay"; else echo "nope"; fi |
| 01:27:20 | Defiler | aha, nice |
| 01:58:05 | matthewd | brixen: Thanks! |
| 01:59:54 | matthewd | brixen: I'll note that "called on a module" wasn't distinguishing from a call on an array, but from a call on a class |
| 02:03:56 | brixen | matthewd: ok, that makes sense |
| 02:04:08 | brixen | but if that's a distinction, where's the spec using Class.new? |
| 02:04:15 | brixen | I may have missed it |
| 02:05:36 | matthewd | I didn't add it; I'm drawing a distinction from the already-specified "normal" behaviour |
| 02:06:08 | brixen | could you point me to the file again? |
| 02:07:12 | matthewd | core/module/method_defined.spec |
| 02:09:16 | brixen | hmm |
| 02:09:40 | brixen | a module has no ancestors |
| 02:10:14 | brixen | are you saying contrasts with this first humunguous spec at the top? |
| 02:10:51 | brixen | I'm having trouble seeing the contrast |
| 02:11:01 | brixen | what was rbx doing special with Kernel? |
| 02:11:02 | matthewd | True |
| 02:11:14 | matthewd | It was searching Object (and thus Kernel) always |
| 02:11:23 | brixen | from a module? |
| 02:11:38 | matthewd | Which is correct (and newly specced) behaviour for alias, module_function, public, etc... but not method_defined? |
| 02:12:07 | brixen | ok |
| 02:12:33 | brixen | trying to think of how to describe this |
| 02:12:47 | brixen | the "instance method defined on Kernel" is kinda misleading |
| 02:12:54 | brixen | that was an implementation error |
| 02:12:56 | matthewd | So I guess technically, this contrasts with that, more than a (arguably missing) "returns false for a method defined in some arbitrary non-ancestor class" |
| 02:13:07 | brixen | yeah |
| 02:15:07 | brixen | perhaps: it "does not search Object or Kernel when called on a module" ? |
| 02:20:37 | matthewd | k. Worthwhile adding a "# unlike alias and friends" comment, to clarify why it's not quite as pointless as it sounds? |
| 02:21:40 | brixen | certainly |
| 02:22:17 | brixen | when you've done everything you can to make it clear, adding a # comment is always an option |
| 02:22:45 | brixen | anyway, thanks for working on this! |
| 16:11:57 | brixen | morning folks |
| 16:12:17 | brixen | how about attacking some c-api specs... |
| 16:15:49 | evan | morning |
| 16:32:53 | toulmean | brixen: https://twitter.com/toulmean/status/14774213226 |
| 16:32:57 | toulmean | you're crazy man |
| 16:33:05 | brixen | toulmean: :D |
| 16:34:03 | brixen | toulmean: I won't disagree with you and many others wouldn't either :) |
| 16:34:34 | brixen | is also frying an egg while chatting... |
| 16:56:35 | toulmean | brixen: so you can load a simple .class file ? |
| 16:57:07 | brixen | toulmean: nah, I can parse a simple class file |
| 16:57:15 | toulmean | so just parse |
| 16:57:17 | brixen | which is almost loadingi it |
| 16:57:37 | toulmean | then the goal is to load it in memory and make it available somehow ? |
| 16:57:37 | brixen | I'll define "load" to mean I have it in a format that the translater can consume |
| 16:57:41 | brixen | it's not there yet |
| 16:57:44 | toulmean | ok |
| 16:57:49 | toulmean | impressive. |
| 16:57:50 | brixen | the goal is to run it, via translation |
| 16:58:03 | brixen | well, maybe some day :) |
| 16:58:08 | toulmean | so to run it: you have people give you the main method ? |
| 16:58:14 | brixen | it's just a hobby to get me more info about other vm's |
| 16:58:17 | toulmean | or you want to interface it jruby style ? |
| 16:58:24 | toulmean | ok |
| 16:58:27 | brixen | no, it would just run |
| 16:58:39 | brixen | in this case, just like you did java Hello with Hello.class |
| 16:58:50 | toulmean | what would be the difference with shelling out ? |
| 16:58:57 | toulmean | why all the pain :) ? |
| 16:58:58 | brixen | shelling out? |
| 16:59:07 | brixen | I don't follow |
| 16:59:19 | toulmean | %x[java Hello.class] |
| 16:59:39 | brixen | um, why would I do that? :P |
| 16:59:41 | toulmean | oh ok |
| 16:59:55 | toulmean | it's more for consumers like me |
| 16:59:56 | brixen | the goal is to integrate java running *in* rbx |
| 17:00:11 | brixen | so, I don't want to use the jvm |
| 17:00:16 | toulmean | yep. |
| 17:00:28 | brixen | heh |
| 17:00:34 | toulmean | can you call out methods or whatnot and get results |
| 17:00:42 | toulmean | and you know, integrate them back in ruby |
| 17:00:50 | brixen | that would be stage 2 |
| 17:00:53 | toulmean | oh well. I'm asking too much. |
| 17:01:00 | toulmean | that sounds like a nice alt to rjb |
| 17:01:01 | brixen | stage one is just to make rbx look like a jvm to some java code |
| 17:01:16 | toulmean | good enough for me. |
| 17:01:16 | Defiler | brixen: What if I wanted to become the maintainer of Net::HTTP |
| 17:01:23 | Defiler | how would I go about doing that |
| 17:01:24 | brixen | I think ultimately, it will be easier to do RI (ruby integration) on rbx than JI on jvm |
| 17:01:44 | toulmean | brixen: just beware that the .class format changed a few times in between java versions |
| 17:01:52 | brixen | Defiler: hmm, yugui set up a thing to ask to be maintainer of a lib |
| 17:02:01 | brixen | Defiler: this is in stdlib, yes? |
| 17:02:08 | Defiler | yes |
| 17:02:11 | toulmean | I can give you places where to pick up hard code to parse at some point. |
| 17:02:14 | brixen | toulmean: yes, that's why it's got a version # |
| 17:02:16 | Defiler | http://redmine.ruby-lang.org/wiki/8/Maintainers |
| 17:02:21 | toulmean | ok going to work |
| 17:02:23 | toulmean | ah ok |
| 17:02:23 | Defiler | shows no maintainer |
| 17:02:35 | brixen | Defiler: I think you can just request it then |
| 17:02:42 | brixen | I forget the details |
| 17:02:51 | Defiler | yugui_zzz: I volunteer to maintain net/http. I already have many patches. |
| 17:03:12 | evan | Defiler: email is much better. |
| 17:03:18 | brixen | heh "I beseech you, let me make good codez" |
| 17:03:28 | Defiler | yeah but I'm busy maintaining net/http hehe |
| 17:03:29 | brixen | Defiler: yeah, I'm thinking ruby-core email there |
| 17:03:33 | brixen | haha |
| 17:03:47 | Defiler | ok, cool. I'll send one out today when I get a chance |
| 17:03:58 | Defiler | If it's going to be in there, it should be hardcore |
| 17:04:01 | brixen | when you get a break from all the Net::HTTP fun |
| 17:04:59 | Defiler | I'm working on taking all the 'legacy' api features in it as well, and moving them to a separate module |
| 17:05:09 | Defiler | which is only loaded when you call one of those methods |
| 17:05:23 | Defiler | and legacy will include a ton of things that aren't deprecated yet |
| 17:05:37 | brixen | sounds cool |
| 17:05:50 | Defiler | and then I will stop it from allocating a hojillion procs every time it wants to do anything at all |
| 17:06:12 | brixen | but you might *need* those procs |
| 17:06:25 | Defiler | right now there is a core method that supports every feature that net/http exposes, ever, including getting notified every time it reads a chunk from the stream |
| 17:06:34 | brixen | haha |
| 17:06:37 | Defiler | and even the bare-bones non-block methods end up calling down into it |
| 17:06:59 | Defiler | I'm fine to retain compatibility with the API though; that is certainly the whole point of being in stdlib |
| 17:07:21 | evan | Defiler: yes, changing the hot loop from using timeout() would be nice. |
| 17:07:26 | Defiler | word |
| 17:07:54 | Defiler | Right now it is like a swarming barely-contained beehive of closures |
| 17:08:17 | Defiler | snake mating season in siberia |
| 17:08:26 | evan | hah |
| 17:35:53 | evan | whats up with matthewd's bug. |
| 17:36:01 | evan | it should be an easy fix |
| 17:36:06 | evan | i'd like to get it closed. |
| 17:38:26 | brixen | I think he's got the fix |
| 17:38:30 | brixen | and the specs |
| 17:38:39 | brixen | but he didn't get a chance to finish it up last night |
| 17:38:49 | brixen | we discussed the spec changes I suggested though |
| 17:39:08 | brixen | matthewd: ping |
| 18:03:40 | Defiler | brixen: around? |
| 18:04:14 | brixen | yes |
| 18:04:26 | Defiler | I'm always needing to tell some piece of code which directory something that is calling it is in |
| 18:04:52 | Defiler | so I have lots of base_dir = File.expand_path("../..", __FILE__) |
| 18:04:53 | brixen | hmm, sounds like you need __DIR__ |
| 18:04:59 | Defiler | and then passing that base_dir into something |
| 18:05:32 | Defiler | but, in this particular case, what I really want is a method I can put in this gem that things can call to get back a Pathname for where they are |
| 18:05:41 | Defiler | e.g. Helper.where_am_i? |
| 18:05:52 | brixen | hmm |
| 18:06:03 | Defiler | I can think of a bunch of ways to implement that, but I'm wondering if you have an opinion on the most portable way |
| 18:06:05 | brixen | why are these things always needing to know? |
| 18:06:16 | Defiler | usually because they are/involve daemons |
| 18:06:22 | Defiler | and need to know where to go to re-exec themselves |
| 18:06:34 | brixen | hm |
| 18:07:09 | Defiler | but there are lots of cases; right now I've got a pile of daemon tasks that would like to load a config file in a relative way |
| 18:07:26 | Defiler | but need to make sure they know proper paths, so they can do it all after a pair of forks |
| 18:07:57 | Defiler | __DIR__ would be great though |
| 18:08:01 | Defiler | but I need to run on 1.8.7 |
| 18:08:15 | brixen | I was kidding about __DIR__ :) |
| 18:08:19 | brixen | but hmm |
| 18:08:45 | brixen | I try to have relative things all relate to one absolute thing |
| 18:08:51 | Defiler | I've got a helper in this library already for determining your actual working directory |
| 18:09:01 | Defiler | so do I; with a daemon that happens to always be / though heh |
| 18:09:10 | brixen | other than Dir.pwd? |
| 18:09:24 | Defiler | yeah, Dir.pwd? can be confused by symlinks |
| 18:09:31 | brixen | so with processes you can put an ENV var in |
| 18:09:46 | brixen | this is so sysadminny |
| 18:09:51 | brixen | where is evan? |
| 18:09:59 | brixen | is not a sysadminny :) |
| 18:10:00 | evan | here |
| 18:10:04 | evan | sup? |
| 18:10:09 | brixen | evan: you so silent |
| 18:10:12 | Defiler | https://gist.github.com/8538b47cb48fe6841526 |
| 18:10:17 | evan | sorry, playing with the agent |
| 18:10:20 | brixen | heh |
| 18:10:32 | evan | and packing up to drive up to glendale to have lunch with tenderlove |
| 18:10:34 | brixen | evan: dare you to name our agent Nikita :) |
| 18:10:39 | evan | hah |
| 18:10:49 | evan | -Xnikita.activate |
| 18:10:53 | brixen | hehe |
| 18:11:06 | brixen | "playing with our agent" has a whole new connotation |
| 18:11:08 | brixen | :) |
| 18:11:16 | evan | was there a question you had for me? |
| 18:11:20 | brixen | anyway, Defiler wants __DIR__ :) |
| 18:11:28 | evan | ah |
| 18:11:34 | brixen | j/k but he has a question about loading daemons relative |
| 18:11:49 | Defiler | It's not really a question, since I already do it a lot |
| 18:11:53 | brixen | read the last screen of discussion |
| 18:11:58 | evan | File.dirname(CompiledMethod.of_sender.scope.active_path) |
| 18:12:04 | Defiler | it's mostly for advice about the most portable way to figure out the directory my caller is implemented in |
| 18:12:15 | Defiler | and I guess it's caller sadly |
| 18:12:21 | evan | no no |
| 18:12:23 | evan | look up. |
| 18:12:27 | evan | for Rubinius |
| 18:12:29 | Defiler | Has to work on 1.8.7 |
| 18:12:32 | Defiler | but I could do that when on rbx, yeah |
| 18:12:33 | brixen | well, that's not portable |
| 18:12:39 | evan | caller is all you've got. |
| 18:12:41 | brixen | you can special case it |
| 18:12:42 | evan | on MRI |
| 18:13:16 | brixen | if defined?(THE_AWESOME_IS_THIS) File.dirname(CompiledMethod.of_sender...) else caller() end |
| 18:13:53 | Defiler | we could make special return values for magic defined? arguments in rbx |
| 18:14:12 | Defiler | path = defined?(caller.active_path) |
| 18:14:12 | Defiler | heh |
| 18:14:17 | brixen | heh |
| 18:14:36 | brixen | a whole language in a language |
| 18:14:45 | brixen | since defined? processes expressions |
| 18:15:05 | brixen | but doesn't evaluate their truth value |
| 18:15:18 | Defiler | erlang?(start() -> io:format("Hello!").) |
| 18:23:36 | cremes | brixen: learned my lesson (again) on writing specs; wrote my first spec for my 0mq bindings and discovered a bug :) |
| 18:23:43 | cremes | you are the master |
| 18:24:01 | brixen | haha |
| 18:24:15 | cremes | the *very first* one... <sigh> |
| 18:24:18 | brixen | well, I'm glad you found a bug, but I am no master |
| 18:24:24 | brixen | hah |
| 18:24:26 | cremes | and it was on code that just could NOT be wrong |
| 18:24:30 | cremes | but it was |
| 18:24:33 | brixen | you know what happens when you find the first one? |
| 18:24:40 | cremes | tip of the ice berg? |
| 18:24:51 | brixen | it's an inductive set |
| 18:25:02 | cremes | given k, k + 1 is buggy too? |
| 18:25:11 | brixen | where the first one invokes the succ function |
| 18:25:11 | brixen | heh |
| 18:25:11 | brixen | pretty much |
| 18:25:18 | brixen | and it recurses |
| 18:25:25 | brixen | nasty beasts :) |
| 18:25:32 | cremes | argh... well, at least i won't keep causing regressions as i muck around |
| 18:25:43 | brixen | indeed |
| 18:26:35 | cremes | btw, do you know of a way to test that a finalizer gets called on an object when it gets gc'ed? |
| 18:31:04 | Defiler | brixen: OK, so, imagine you had, say, 8 to 10 things that had this same need for boilerplate: https://gist.github.com/ea07b7cda6d890201ea4 |
| 18:31:20 | Defiler | what would you factor out and where would you put it? heh |
| 18:35:28 | brixen | I would be inclined to make an Environment class or some such |
| 18:35:37 | brixen | that is created and passed to these guys |
| 18:35:46 | Defiler | yeah |
| 18:35:46 | brixen | or that there is a way for them to invoke |
| 18:36:00 | Defiler | task.environment maybe |
| 18:36:04 | Defiler | or .env |
| 18:36:04 | brixen | yeah |
| 18:36:07 | brixen | sure |
| 18:36:11 | Defiler | and that hands back an instance of something happy slappy |
| 18:36:17 | brixen | yep |
| 18:36:31 | Defiler | and I guess that has to use 'caller' to answer one of those questions |
| 18:36:35 | Defiler | oh well |
| 18:36:49 | brixen | sometimes there is no other way |
| 18:36:52 | brixen | hangs head |
| 18:37:04 | Defiler | at least rvm lets me run the test suite for it under a bunch of rubies |
| 18:37:07 | Defiler | handy |
| 18:37:39 | brixen | yeah |
| 18:37:46 | brixen | rvm rocks |
| 18:38:25 | brixen | mostly because the name reminds me of "Ruby virtual machine" |
| 18:40:11 | maharg | it'd be nice if jruby and rubinius had a uniform interface for getting at that kind of info |
| 18:40:20 | maharg | so that eventually mri has something to adopt |
| 18:40:34 | brixen | that would be interesting, yeah |
| 18:41:04 | brixen | jruby could make CompiledMethods :) |
| 18:41:50 | wayneeseguin | w00t! |
| 18:41:55 | wayneeseguin | -ish :-p |
| 18:42:01 | brixen | :D |
| 18:42:10 | brixen | wayneeseguin: j/k, rvm is awesome |
| 18:48:13 | wayneeseguin | lol |
| 18:48:17 | wayneeseguin | Trying. |
| 18:48:22 | wayneeseguin | Not an easy project. |
| 18:49:20 | brixen | no kidding |
| 19:04:26 | brixen | russian_vodka: what's you favorite? |
| 19:04:30 | brixen | er your |
| 19:04:35 | brixen | damn engrish |
| 19:33:01 | russian_vodka | brixen, "Russian Standart" |
| 19:33:34 | russian_vodka | brixen, http://www.sostav.ru/articles/rus/2008/07.06/news/images/90.jpg |
| 19:34:54 | brixen | russian_vodka: cool, I'll have to see if I can get that here somewhere |
| 19:35:17 | russian_vodka | brixen, I can send one to you ) |
| 19:35:52 | brixen | heh |
| 19:36:28 | brixen | russian_vodka: I should just visit Russia and have the real thing there :) |
| 19:36:54 | russian_vodka | oh. |
| 19:37:29 | russian_vodka | Yes, right. Take evan an come here, Moscow :) |
| 19:37:37 | russian_vodka | *an => and |
| 19:37:48 | brixen | that would be fun! :) |
| 19:38:30 | russian_vodka | girls, vodka & rock'n'roll :D |
| 19:41:06 | brixen | haha |
| 19:41:35 | brixen | russian_vodka: evan is married... I, on the other hand, am not :) |
| 19:45:44 | russian_vodka | brixen, girls for you and vodka & rock'b'roll for evan |
| 19:46:54 | brixen | mmm, interesting possibility |
| 19:47:14 | russian_vodka | :D |
| 19:47:29 | russian_vodka | brixen, ok. vodka for both ;) |
| 19:47:37 | brixen | heh |
| 20:42:01 | PierreY | Hi, I cannot build rubinius for a while, I got tons of errors while compiling grammar.y/cpp (Ubuntu 10.04) |
| 20:53:39 | brixen | PierreY: are you build with ruby 1.8? |
| 20:53:50 | PierreY | yes |
| 20:53:56 | brixen | PierreY: and did you follow getting_started.txt? |
| 20:53:58 | PierreY | build fails with ruby 1.9 |
| 20:54:21 | PierreY | arg, things have changed into getting_started.txt ? |
| 20:54:23 | brixen | could you gist me your output? |
| 20:54:43 | PierreY | I'm cloning th repository to restart from scratch |
| 20:54:44 | brixen | things haven't changed much, but I've clarified the steps in that file, yes |
| 20:54:51 | PierreY | right now |
| 20:55:00 | brixen | make sure you follow the steps for ubuntu |
| 20:55:10 | PierreY | and the problem is that my terminal buffer is too small to contain all error lines |
| 20:55:14 | brixen | and then gist me your terminal output |
| 20:55:19 | brixen | you can increase the buffer |
| 20:55:38 | brixen | are you running the gui or no? |
| 20:55:49 | PierreY | sure, I'm using gnome-terminal |
| 20:56:34 | brixen | ok, there's a setting... :) |
| 20:56:42 | brixen | I was just there yesterday, what was it... |
| 20:57:25 | PierreY | mmm preferences / scrolling / history lines ? |
| 20:57:32 | brixen | yeah |
| 20:57:55 | brixen | you found it faster than I could boot vbox |
| 20:58:13 | PierreY | vbox ? |
| 20:58:19 | brixen | looks like gnome beats windows... I'd still be trying to show the icons in control panel |
| 20:58:45 | brixen | virtualbox |
| 20:58:51 | brixen | www.virtualbox.org/ |
| 20:59:00 | PierreY | sure, I know that, but why do you have to use vbox ? |
| 20:59:10 | brixen | because I use os x |
| 20:59:19 | PierreY | oh, sorry :-) |
| 20:59:20 | brixen | so I don't have no use ubuntu :) |
| 20:59:23 | brixen | heh |
| 21:00:19 | PierreY | (still cloning... my bandwidth has decreased here down to 64kb... what a pity) |
| 21:00:27 | brixen | :( |
| 21:00:46 | brixen | I should snapshot this 9.10 and distupgrade to 10.04 |
| 21:00:50 | PierreY | my provider fault |
| 21:01:34 | PierreY | yeah 10.04 is close to osx now, the windo buttons are on the left side of the title bar ;-) |
| 21:03:08 | brixen | heh |
| 21:03:13 | brixen | the small things :) |
| 21:08:50 | PierreY | do you require bison or bison++ on ubuntu ? |
| 21:10:27 | PierreY | exit |
| 21:10:34 | PierreY | oups |
| 21:12:16 | brixen | yes |
| 21:12:59 | brixen | PierreY: if you discover things that are required but not listed in getting started, let us know |
| 21:15:32 | PierreY | ok... checked ruby : 1.8.7, dependencies : ok... building, now |
| 21:15:36 | brixen | vbox can I have my computer back now, kthx, and dump that openoffice shit in the bin... |
| 21:15:44 | brixen | PierreY: ok |
| 21:18:18 | PierreY | lubuntu is lighter than ubuntu and don't carry openoffice by default |
| 21:18:47 | BrianRice-work | wonders whether ubuntu will run out of prefixes... |
| 21:19:05 | PierreY | :-) |
| 21:19:46 | PierreY | after one letter prefixes, they will use two letter prefixes... be sure |
| 21:20:03 | PierreY | brixen: build failed ... |
| 21:21:50 | brixen | PierreY: I've opened my remote-viewing guide book but the image is very fuzzy |
| 21:21:55 | PierreY | https://gist.github.com/2fab5fc6cb3e65f810b1 |
| 21:21:57 | brixen | are you wearing a pink garter? |
| 21:22:03 | brixen | ah ok |
| 21:22:53 | brixen | PierreY: install ruby-dev |
| 21:22:58 | brixen | that should be on the list.... |
| 21:23:08 | brixen | or did you already? |
| 21:23:34 | brixen | PierreY: could you rake --trace and paste me everything? |
| 21:24:50 | PierreY | sorry ruby-dev was not installed |
| 21:24:54 | PierreY | shame on me |
| 21:24:56 | brixen | ok |
| 21:25:08 | evan | ok, back. |
| 21:25:17 | PierreY | build continues now |
| 21:25:25 | brixen | PierreY: excellent :) |
| 21:25:53 | PierreY | whay don't configure scripts check for required files ? |
| 21:26:00 | PierreY | from ruby-dev ? |
| 21:26:36 | brixen | not yet |
| 21:26:42 | PierreY | running 706 tests.... OK! ;-) |
| 21:26:49 | brixen | but that's on the list kinda |
| 21:26:54 | brixen | same with checking for readline.h |
| 21:27:35 | brixen | PierreY: you will probably see one spec failure about getting '0.0.0.0' instead of '127.0.0.0' |
| 21:27:51 | brixen | s/probably/may/ |
| 21:28:05 | brixen | PierreY: if you get other failures, please gist them |
| 21:28:39 | PierreY | I love the rubinius specs lines of dots |
| 21:28:48 | brixen | heh |
| 21:29:00 | brixen | PierreY: try bin/mspec ci -fa :) |
| 21:32:33 | PierreY | got that awful "F" in my lines of dots... due to 0.0.0.0 being equal to 127.0.0.1 |
| 21:32:44 | brixen | yeah spurious |
| 21:35:27 | PierreY | trying mspec ci -fa :-) I prefer dots |
| 21:35:37 | brixen | heh |
| 21:36:35 | PierreY | mspec without options gave me tons of errors and failures |
| 21:36:59 | brixen | yes, you have to use ci |
| 21:37:09 | brixen | tons =~ 500 or so |
| 21:37:29 | brixen | you can see what is failing with bin/mspec tag --list-all |
| 21:37:49 | brixen | er bin/mspec tag --list-all :files |
| 21:38:13 | PierreY | not 500, mspec itself seems to have failed before the end |
| 21:38:45 | PierreY | last line is "Backtrace:" with no backtrace |
| 21:38:55 | brixen | PierreY: hmm, so perhaps a segfault |
| 21:39:03 | brixen | PierreY: I'll upgrade to 10.04 and run it later |
| 21:39:18 | PierreY | so... how do I install gems ;-) |
| 21:39:38 | brixen | bin/rbx gem install foo |
| 21:40:02 | brixen | once you do, if it installs a script (like spec), bin/rbx -S spec to run it |
| 21:40:52 | brixen | PierreY: gotta run, bbl... but evan can probably help you out |
| 21:41:00 | evan | sure |
| 21:42:23 | PierreY | god, I'm installing rails 3.0 (rbx gem install rails --pre ?) in rubinius... |
| 21:42:49 | evan | cool |
| 21:43:30 | PierreY | somebody already tried to run a rails application with rbx in production ? |
| 21:44:26 | evan | in production, not that I know of. |
| 21:45:17 | PierreY | so... to run rails, I have to run bin/rbx -S rails ? |
| 21:50:56 | PierreY | INFO WEBrick 1.3.1 / INFO ruby 1.8.7 (2010-05-14)... does that mean that WEBrick is powered by MRI and not Rubinius ? |
| 21:52:24 | PierreY | no... there is no ruby process running, but rbx yes, maybe a bug in WEBrick |
| 21:53:02 | maharg | I bet you that's a hard coded string in the MRI libs for WEBrick and it just wasn't changed in the rbx import |
| 21:55:21 | evan | thats not a bug, thats just how WEBrick reports it |
| 21:55:35 | evan | we report ourself as 1.8.7 since some code checks for that |
| 21:55:38 | evan | and WEBrick is reporting |
| 21:56:23 | evan | that. |
| 21:56:39 | maharg | yeah just looked. That makes sense |
| 22:01:05 | PierreY | thank you for your help. I have to go. Congratulations for the 1.0 milestone ;-) |
| 22:28:54 | toulmean | brixen: evan: got a question for ruby core specs |
| 22:29:07 | toulmean | File.mtime(file) returns a time object |
| 22:29:25 | toulmean | in Buildr, some specs fail as it seems mtime compares to the ms |
| 22:29:42 | toulmean | is there a spec for this ? Is there interest over standardizing over mtime comparison ? |
| 22:30:11 | toulmean | the funny thing is: File.mtime(a).to_s == File.mtime(b).to_s if they have been created in the same second. |
| 22:30:27 | toulmean | our spec is a big bogus so we stopped doing equality over mtimes. It's a bad idea overall. |
| 22:30:41 | toulmean | just curious if there should be a spec over this. |
| 22:31:14 | toulmean | also - it might be OS dependent. |
| 22:35:38 | toulmean | ok discussing it with headius on jruby |