Index

Show enters and exits. Hide enters and exits.

06:22:40parndthey, hpricot RSTRUCT_PTR -> http://github.com/evanphx/rubinius/issues/issue/210 -- is there any likelihood of this being supported any time soon?
06:25:25brixenparndt: I would say not likely
06:25:40brixenRStruct has an RBasic field in it
06:25:51brixenMRI is very different inside
06:25:58brixenany reason you can't use nokogiri?
06:26:17parndtnot really, just code that hooks onto hpricot, just interested that it doesn't work :-)
06:26:39brixenwell, it's not really reasonable to expect it to work on other than rbx
06:26:54brixenif you know anything about C coding, that is a horrible way to write software
06:27:06brixenmucking around in the insides of some other programs data
06:27:23brixener work on other than MRI, rather
06:28:01brixenwe try to support a lot of the MRI C-API
06:28:11brixenbut that is the *API*, not the internal data structures
06:28:20parndtwell I'll see if we can get rid of our reliance on it :-) thanks Brian
06:28:41brixencool
06:29:10brixenwould be interesting if someone could sanitize hpricot
06:29:18brixenmaybe I should look at that...
06:29:50parndtwell as I'm not a C coder I couldn't really help you with that.
06:30:14brixenyeah, there's a few people who would like it to work
06:33:02parndtwell it'd make our lives easier
06:33:18parndtbut I'm guessing it wouldn't be a straightforward undertaking
06:33:34brixenprobably not, no
06:33:43parndtcongrats on 1.0 btw
06:34:09brixenthanks!
06:39:56slavahi brixen
06:41:15brixenhi slava
06:42:32brixenparndt: it looks like it wouldn't be too hard to replace RSTRUCT usage with a function API
06:42:52brixenparndt: I'll play around with it later
06:42:57parndtbrixen: wow, awesome!
06:43:24brixenI'd have to patch hpricot of course
06:43:29parndtnaturally
06:43:30brixenis it maintained?
06:43:37parndtyeah
06:43:40brixenok
06:43:51parndthttp://github.com/hpricot/hpricot
06:45:07brixencool
06:45:17brixensleeps
06:45:24parndtnight :-)
06:45:30brixennight!
13:55:44huklhey there
13:57:57huklis Array implemented in ruby as well ? if so - where can i find its implementation in the source ?
13:58:38hukli'm currently on a slow network and cloning takes ages
13:59:26matthewdhukl: Technically yes... but it's a wrapper around a rather array-like VM-primitive, Tuple
13:59:48huklmatthewd, i see
14:00:10hukldoes that apply for strings and other basic types as well ?
14:00:36dbussinkhukl: string uses ByteArray as a primitive type
14:01:10hukli wondered last night how one would implement Array and String in ruby ;)
14:01:12matthewdhukl: kernel/{bootstrap,common}/{string,array}.rb
14:01:40huklcould you give me the path of the files where its defined ? would love to take a look at it
14:01:41huklah
14:01:43huklthanks
14:01:48huklslow internet sucks
14:01:49dbussinkhukl: well, you still need some more basic storage primitive structure
14:02:37huklcan i ask you some more basic questions ? i'm kind of curious how it all works
14:02:41dbussinkhukl: you can see a tuple as a storage space that allows for x pointers to objects
14:02:49dbussinkhukl: where x is variable sized
14:02:51matthewdhukl: As you'll see, while Tuple and ByteArray provide basic indexable storage, most of the methods that make Array and String feel like Array and String, are written in Ruby
14:02:57dbussinkso that can be used to implement ruby's array semantics
14:03:17hukli see
14:03:29dbussinkhukl: so array tracks the number of elements in use itself
14:03:49dbussinkand the capacity of the underlying tuple, growing and shrinking the tuple as needed
14:05:15dbussinkhukl: where growing is basically allocating a new tuple with a bigger size and copying the elements over
14:07:12huklokay - uhm another thing i'd love to know is whether all the stuff that is implemented in ruby is evalutated / jitted/ bytecoded the moment i execute ruby or start up irb ? or would i need to "recompile" rubinius if i'd change something on Array for example ?
14:11:03matthewdhukl: The kernel is pre-parsed; if you play in kernel/, you need to 'rake build'
14:18:15huklok
14:18:21huklthank you!
14:19:36dbussinkhukl: the bytecode is generated when code is loaded
14:19:47dbussinkhukl: jit only happens on hot code that is executed a lot
14:42:28huklone more question ;) is there a debugger that is working with rubinius ?
14:42:40huklwould be the perfect match probably
14:49:44matthewdhukl: evan was working on resurrecting the rbx debugger last week or so
14:50:28matthewdI believe he had it doing some neat things... but I haven't played with it yet
14:51:57hukli attended euruko this weekend and the guy who is maintaining ruby-debug demoed a complete rewrite with tons of cool features
14:53:08huklhttp://wiki.github.com/rocky/rbdbgr/cool-things
14:57:57dbussinkhukl: evan got it working again, but there isn't really a lot of documentation already afaik
17:01:42elliotcmHowdy gents. I wrote a spec to cover a bug in Rubinius's webrat, but for some reason when I run the bin/mspec command in the docs to run the spec, it seems to be using the webrat code from my normal C ruby. Any thoughts?
17:03:20brixenelliotcm: not sure what you mean by "Rubinius' webrat"
17:03:31brixencould you gist me a git diff or something
17:03:38elliotcmYeah lemme grab the other laptop
17:03:51elliotcmwhoops, cucumber on the brain
17:04:12elliotcmnot webrat, webrick
17:04:17brixenahh ok
17:07:50brixenelliotcm_: this might be easier: what makes you think it's running C-ruby webrick?
17:08:02elliotcmThis does: http://pastie.org/985969
17:08:13elliotcmNote line 10
17:08:25brixenwell yeah :P
17:08:30brixenlook at your invocation
17:08:38brixen-tr means find "ruby" on your pach
17:08:40brixener path
17:09:03brixenso, what you probably want is bin/mspe spec/ruby/library/net/...
17:09:04elliotcmOh. I was straight up using the example out of the documentation.
17:09:21brixenwell, see bin/mspec -h
17:09:32brixenso, rbx bundles mspec
17:09:47brixenbin/mspec some/spec/file runs with rbx by default
17:09:59brixenbecause it reads the spec/default.mspec file, which sets the target
17:10:19brixenanyway, try bin/mspec library/net/http/http/start_spec
17:10:23elliotcmIs line 11 in "write_a_ruby_spec.txt" correct, then?
17:10:39elliotcmWhat's the purpose of doing it that way if it doesn't actually run in Rubinius
17:10:40elliotcm?
17:10:56brixenok, let's start at the beginning :)
17:11:05brixenthe specs tell us what *MRI* does
17:11:17brixenso, to write a spec, you must run it, and it must pass on MRI
17:11:25brixenonce you got that, then you can run the spec on rbx
17:11:32brixenand make sure rbx behaves the same as MRI
17:11:41elliotcmOh I seee
17:11:47brixenthat's the whole point of rubyspec, to make sure rbx behaves the same as MRI
17:12:02brixenso, this howto is just about writing the rubyspec
17:12:33brixenthe next step would be doc/howto/fix_a_failing_spec.txt
17:12:43brixenbut I need to fix up these docs a bit more still
17:12:46elliotcmThat makes a bunch more sense.
17:12:51brixenheh, cool
17:12:59elliotcmI was thinking of the specs in the more traditional standalone manner.
17:13:08brixensure
17:13:57brixendid you read doc/specs.txt?
17:14:16brixenif stuff isn't making sense in the docs, we really appreciate hearing about it
17:14:47elliotcmYeah I did
17:14:55elliotcmlemme read it again and see if I skimmed over anything important
17:14:58brixenok, did it make sense
17:14:59brixenok
17:15:54brixengonna grab a shower, bbiab..
17:17:21elliotcmYeah I read it all but it didn't click at the time. Either that's me skimming (I was trying to get up to speed with the whole thing) or a lack of clarity on what RubySpec actually is, or a combination of both.
17:18:20elliotcmFor example, this line:
17:18:21elliotcmThe specs at ./spec/ruby are a copy of RubySpec at a particular revision.
17:19:35elliotcmdoesn't actually tell me what RubySpec is or is about if you don't know already. With RSpec and MSpec floating around I think I just read this and filled RubySpec away as another thing I don't know about and will find out about later
17:19:57elliotcm*filed
17:21:15elliotcmOn a brighter note, I'm particularly pleased at the enthusiasm you have for encouraging contribution
17:21:34elliotcmand the effort you've gone to to make it less than painful
17:24:11elliotcmI think laying it out as clearly in the docs as you did for me just now would help, but I couldn't guarantee it because there's no specific order to the documentation (by the nature of it)
17:24:20elliotcmI didn't even notice the TOC until way later
17:24:57elliotcmEven then the TOC is alphabetical and not a recommended reading order
17:36:14elliotcmOh by the way, when asked which non-MRI implementation Matz felt had most potential, Rubinius got the vote.
17:38:37brixenawesome!
17:38:51brixenso, in the specs.txt I point you to the rubyspec.org website
17:38:54brixenthat has better docs
17:39:01brixenabout rubyspec per se
17:39:23brixenbut I'll make another pass over /doc/* and see what I can do
17:39:56brixenI've been trying to write fewer and fewer words so it's more likely the actual words will be read rather than skimmed :)
17:40:37elliotcmYeah, but the catch there is that a fresh face looking at this stuff has docs and docs queued up to be read and a website is going to go on the back burner
17:42:26elliotcmNonetheless, I'm going to reset my branch and start again
17:42:33elliotcmthanks for the clarity
17:49:37toulmeanbrixen: evan: arton accepted the changes for rjb, he mentioned it passed with MRI
18:02:12brixentoulmean: awesome
18:02:32toulmeanbrixen: your work paid off.
18:02:43brixenelliotcm: I think what I should do is add a sinatra app to rbx that would serve up the docs as a local site
18:02:47toulmeanNow I'm thinking... I want to have a CI build of Buildr over rubinius.
18:02:57toulmeandid you release 1.0.1 ?
18:02:59brixenelliotcm: that way I can use hyperlinks to make it easier to browse
18:03:12brixentoulmean: not yet, but maybe this week
18:03:29brixenI'll be in and out this week because I'm going to a local conference
18:03:36toulmeanbrixen: I can base off trunk for now, but as soon as 1.0.1 is out I'll use it
18:03:41toulmeanbrixen: lucky you
18:03:54brixenthen railsconf the following week, but I think evan will probably release 1.0.1 before RC
18:04:09toulmeanbrixen: are you paid to work on rubinius full time ? Or for more stuff ?
18:04:17toulmeanjust wondering the economics
18:04:19elliotcmbrixen: Sounds good. let me know if you want a hand
18:04:22brixenrubinius and rubyspec
18:04:32brixenelliotcm: ok, will do
18:04:41brixenelliotcm: I appreciate the feedback on the docs
18:05:01brixenelliotcm: it's a tough task, no one wants to do it, but really necessary :)
18:05:23toulmeanok
18:05:50dbussinkbrixen: are both you and evan heading to railsconf?
18:06:04brixendbussink: yes! :)
18:06:06elliotcmNo problem, thanks for writing them.
18:06:22brixendbussink: I think we're planning on tackling you, you know, just 'cause :)
18:06:43dbussinkbrixen: tackling me? how did i deserve that? :P
18:06:52brixendbussink: :)
18:07:18brixendbussink: nah, I'm glad you're going, it's been a long time since RC pdx
18:07:35dbussinkbrixen: true that yeah
18:08:38brixentoulmean: so, re rjb, did you get buildr specs running?
18:08:56toulmeanbrixen: still not. Would do it today.
18:09:58brixentoulmean: ok
20:02:22dbussinki can hear the silence
20:04:54somebodyoO
20:11:36Defileroooohhhhhhh
20:12:15dbussinksssshhhh
20:12:26dbussinkDefiler: going to railsconf?
20:14:11Defilernope
20:14:22dbussinkah, too bad
20:14:24Defilertoo much money these days
20:14:44DefilerAlso, I'm moving right now :)
20:14:54Defilerbut I may try to sneak down by train just to hang out for a night
20:15:02dbussinkah, well, speaking makes a difference money wise, so that's good
20:15:10dbussinkgoing to nyc for a few days after the conf :)
20:15:35DefilerCool
20:16:06dbussinkyeah, been there for like one and a half day few years ago, way too short
20:16:32DefilerI can probably hand you my membership cards at various places and you can go to town
20:17:55dbussinkDefiler: like clubs etc.?
20:18:07dbussinkperhaps if there's a must see / go somewhere :)
20:18:10Defilermuseums, botanic gardens, movie halls, etc, heh
20:18:25Defilerand I actually have an unused citypass somebody left when they stayed here
20:18:36Defilerwhich is good for those boat tours around manhattan, etc
20:19:05Defilerso if you want to get all tourist-y let m eknow
20:19:21dbussinkwe'll probably go touristy yeah :)
20:19:47dbussinkliving somewhere in manhattan? or way too expensive?
20:27:17DefilerI'm not a huge manhattan fan
20:27:22DefilerI live in brooklyn
20:29:21dbussinkah, probably better for proper living, mainly going there for the touristy things ;)
20:29:33dbussinkbut want to wander around a bit in some other place too :)
20:31:04Defilerthe brooklyn botanic garden is awesome
20:31:26Defilerand there are good museums, theatres, restaurants, parks, etc, around here
20:31:57Defilermanhattan is the go-to place for standing on top of tall things and looking around though
20:34:26dbussinkDefiler: i might be bugging you for restaurant tips
20:34:32Defilerplease do
20:34:35DefilerI have Opinions.
20:35:35Defilerlike that you need to end up where these guys are parked http://vendr.tv/video/wafels-and-dinges/
20:35:48Defilerhttp://www.wafelsanddinges.com/location.html
20:37:56dbussinkDefiler: hehe, it travels around i see yeah :)
20:38:07Defilerit roams the night winds
20:38:19Defilerbringing tidings of waffels covered in insane things
20:38:21dbussinkbut belgian isn't that special for me :P
20:38:28DefilerYeah I know but it's good :)
20:39:00Defilerbut there are some more straight-facedly tasty options, for sure.
20:39:53dbussinkwhat's like a must go place then?
20:40:16Defilerrecently I've been in love with this place called 'Buttermilk Channel' in Carroll Gardens
20:40:23Defilerso damn good
20:40:58Defilerfranny's pizza, ippudo ramen, vinegar hill house, hrm..
20:41:05Defileryeah there's a whole list
20:41:21Defileral di la in brooklyn is awesome
20:41:33dbussinkDefiler: do they actually have buttermilk? :)
20:41:39Defileryeah
20:41:42dbussinkcool :)
20:41:49dbussinki can never get that stuff abroad
20:41:56dbussinkone of the first things i take when back home :P
20:41:59Defilerand lobster pot pie and dandelion-leaf salad and oh god
20:43:40dbussinki'll put them on the list :p
20:44:39Defilersuperb selection of oysters if you are a fan of that
20:44:44Defilerone of the few places I will eat them
20:47:15dbussinki tried oysters once, not my kind of thing
20:47:46dbussinkwhat i often find the hardest is getting decent breakfast though, dinner is not that problematic usually
20:49:30Defilerlittle owl is good, gramercy tavern, rose water
20:49:52DefilerI like dizzy's as well though it is fairly out of the way unless you are visiting prospect park
20:59:39dbussinkDefiler: thnx for the tips already, going to get some sleep first :)
21:01:24Defilercool