Da
Show enters and exits. Hide enters and exits.
| 00:00:06 | lopex leaves the room. | |
| 00:00:14 | chop3 enters the room. | |
| 00:07:09 | foysavas leaves the room. | |
| 00:07:23 | wycats leaves the room. | |
| 00:08:26 | stepheneb enters the room. | |
| 00:24:41 | stepheneb leaves the room. | |
| 00:24:46 | brainopia enters the room. | |
| 00:25:05 | nicksieger leaves the room. | |
| 00:29:17 | brennanc enters the room. | |
| 00:31:39 | imajes leaves the room. | |
| 00:33:09 | cypher23 leaves the room. | |
| 00:33:46 | antares enters the room. | |
| 00:34:03 | kw enters the room. | |
| 00:41:45 | jero5 leaves the room. | |
| 00:46:32 | brainopia leaves the room. | |
| 00:46:35 | fbuilesv enters the room. | |
| 00:47:22 | rue | Scary |
| 00:47:42 | rue | "We have had some conversations with people on the IRC channel #rubyonrails. Among those people, nobody has ever run Apache as non-root." |
| 00:48:07 | djwhitt | err... what? |
| 00:49:16 | djwhitt | don't most distros run it as www or apache? |
| 00:49:27 | rue | http://modrails.org/documentation/Security%20of%20user%20switching%20support.html |
| 00:49:33 | rue | One would think |
| 00:50:42 | drbrain | apache starts as root to bind port 80 |
| 00:50:48 | drbrain | then switches to nobody/www |
| 00:50:54 | drbrain | that's probably what they mean |
| 00:51:16 | kw leaves the room. | |
| 00:52:02 | djwhitt | ah, yeah, that would make more sense |
| 00:52:42 | hoopy | ruby enterprise edition? |
| 00:54:41 | drbrain | heh, rspec output changed |
| 00:55:10 | gnufied | ##c is one of the most *inspring* channels around |
| 00:55:13 | drbrain | in 0.7.mumble it was N specifications, in 1.mumble it's N examples |
| 00:56:04 | gnufied | is changing fork => fork and exec *everywhere* |
| 00:57:55 | moofbong_ enters the room. | |
| 00:58:02 | evan | gnufied: eh? |
| 00:59:12 | rue | Humm, interesting. Some of this is verbatim mod_scgi |
| 00:59:37 | rue | Bleh, enough |
| 00:59:48 | antares | drbrain: it happened in 0.8 AFAIR |
| 01:00:27 | gnufied | evan, ah, sorry a little unrelated. I was talking aloud. just fork leaks in ruby. as you mentioned in your presentation somewhere I remember. |
| 01:01:15 | evan | it doesn't leak |
| 01:01:21 | evan | it just uses a lot of memory. |
| 01:03:51 | moofbong leaves the room. | |
| 01:04:39 | drbrain | later, off to run an errand |
| 01:05:09 | evan | later. |
| 01:05:16 | kw enters the room. | |
| 01:06:23 | gnufied | evan, why? is it because entire parents resources are copied? |
| 01:06:42 | evan | yeah |
| 01:06:46 | evan | the way the GC works in 1.8 |
| 01:06:54 | evan | is that it goes through and sets a bit inside every object |
| 01:07:03 | evan | which causes the OS to copy all the pages for all the pages into the child |
| 01:07:42 | gnufied | so does an exec in child after fork fix that? |
| 01:08:07 | evan | could you pastie me some code that does that? |
| 01:08:51 | gnufied | evan, http://pastie.caboo.se/179460 |
| 01:09:20 | evan | well, fork/exec is the classic way to start a subprocess |
| 01:09:26 | brixen | heh, I've just been working on a similar thing today to parallelize and isolate spec runs |
| 01:10:04 | evan | btw, line 10 has about 6 too many parathises |
| 01:10:17 | evan | if pid = fork() |
| 01:10:36 | evan | or, more ruby-like |
| 01:10:43 | evan | pid = fork { ... } |
| 01:10:54 | djwhitt | brixen: parallelized spec runs? won't the output be confusing? |
| 01:10:56 | rue | ^^ |
| 01:11:16 | brixen | djwhitt: well, it has to be aggregated by the parent process ;) |
| 01:11:28 | gnufied | also in bar.rb: http://pastie.caboo.se/179461 |
| 01:12:10 | gnufied | evan, yeah sorry for that. but shouldn't exec free the copied resources ? |
| 01:12:12 | brixen | dgtized: and I haven't figured out a good way to do progress indicator like the marching dots |
| 01:12:32 | evan | gnufied: yeah |
| 01:12:34 | rue | Would that not work just fine? |
| 01:12:36 | brixen | er djwhitt ^^ |
| 01:12:50 | rue | If you just print dots, that is |
| 01:13:07 | djwhitt | brixen: well, as far as build bots are concerned that doesn't matter. I just thought it would be confusing if you got weird interlieved error messages |
| 01:13:15 | brixen | rue: actually, it seems to. print the . to stdout and separate the summary/bt info |
| 01:13:20 | brixen | rue: to aggregate later |
| 01:14:13 | Fobax | Have you seen this: http://izumi.plan99.net/blog/index.php/2008/01/14/making-ruby%E2%80%99s-garbage-collector-copy-on- write-friendly-part-7/ |
| 01:14:24 | Fobax | It cut MRI memory usage with fork to be MUCH better |
| 01:15:38 | Fobax | btw, evan, you may be interested to know that MRI with jemalloc instead of the default glib malloc is MUCH more memory efficient, though mildy slower |
| 01:15:49 | evan | oh, interesting. |
| 01:15:50 | Fobax | it may be worth making it an option in rubinius |
| 01:16:00 | evan | yes perhaps. |
| 01:16:19 | evan | though we don't make nearly as many calls out to malloc() as MRI |
| 01:16:26 | Fobax | that's good to hear |
| 01:16:38 | evan | mark/sweep still uses malloc() and free() directly, but thats going to change in time |
| 01:16:44 | evan | so that the mark/sweep can be mark/compact |
| 01:17:02 | Fobax | I'm guessing that jemalloc is *slower*, but just much better at allocating small chunks of memory |
| 01:17:07 | evan | yeah |
| 01:17:14 | evan | i've read the heap analysis posts aboutit |
| 01:17:23 | evan | it works a lot harder to fit the allocations together |
| 01:17:46 | Fobax | I did some C profiling yesterday, and found that garbage collection is using between 35 and 50% of my cpu time in MRI |
| 01:18:05 | Fobax | I'm considering doing things like making jemalloc not thread safe, since that will give a 5% speed up on its own |
| 01:18:27 | evan | ah, true. |
| 01:19:17 | Fobax | how's rubinius going btw? How soon can I use it in production? |
| 01:20:12 | rue | Tomorrow!* |
| 01:20:15 | evan | we're aiming to have it running rails a bit by railsconf |
| 01:20:27 | rue | *DependingOnYourProduction.VoidWhereProhibited |
| 01:20:28 | Fobax | when is railsconf? |
| 01:20:40 | evan | end of may |
| 01:20:55 | evan | we took a small step backwards in terms of time because of my new VM work |
| 01:20:55 | Fobax | sweet |
| 01:21:02 | evan | which may or may not have been a good idea, time wise. |
| 01:21:06 | evan | but it's a good idea code wise |
| 01:21:18 | evan | *meh* we'll live. |
| 01:22:11 | chop3 leaves the room. | |
| 01:22:17 | webmat | evan: you said previously it would help make the code more DRY. Does it look like the VM's going to be smaller code-wise? |
| 01:22:35 | evan | yeah |
| 01:22:36 | evan | it does. |
| 01:22:38 | webmat | nice |
| 01:26:07 | cored enters the room. | |
| 01:29:12 | rue | webmat: The code is in the repo if you want to take a peek at it |
| 01:29:35 | d2dchat_ leaves the room. | |
| 01:30:01 | webmat | rue: yep, I saw that in my last pulls :-) Right now I'm busy with something else tho |
| 01:38:43 | headius leaves the room. | |
| 01:38:44 | headius_ enters the room. | |
| 01:39:41 | chop3 enters the room. | |
| 01:41:40 | webmat | is the 'gem' subcommand of rbx supposed to be working? the source index update seems very slow |
| 01:42:10 | rue | If it is slow, it means it is working! |
| 01:42:24 | rue | Someone is pushing 3000 objects onto the stack *cough* |
| 01:42:43 | evan | one man's "this is slow" is another man's "this is working!!" |
| 01:42:55 | webmat | Ok, I'll let him push for a while then :-) |
| 01:43:18 | webmat | evan: true, it's better than getting an exception :-) |
| 01:43:38 | rue | I need to monkeypatch that for us |
| 01:44:53 | srbaker enters the room. | |
| 01:46:33 | imajes enters the room. | |
| 01:54:33 | AndrewO enters the room. | |
| 01:55:58 | srbaker leaves the room. | |
| 01:59:04 | madsimian enters the room. | |
| 01:59:24 | webmat | I'm trying to install gems on an installed rbx... Does an uninstall delete the gem directory or does it stay there? |
| 01:59:45 | TheVoic1 enters the room. | |
| 02:00:10 | webmat | (scratch the 'trying', the index update just finished and it worked) |
| 02:00:36 | imajes leaves the room. | |
| 02:00:52 | imajes enters the room. | |
| 02:01:14 | evan | huzzah! |
| 02:01:48 | webmat | So will my gems survive the next uninstall/install? :-) |
| 02:02:03 | evan | yeah |
| 02:02:05 | webmat | of rbx I mean |
| 02:02:06 | imajes leaves the room. | |
| 02:02:09 | webmat | cool |
| 02:02:09 | evan | should |
| 02:02:14 | webmat | could |
| 02:02:22 | webmat | I'll check |
| 02:10:02 | imajes enters the room. | |
| 02:11:45 | TheVoice leaves the room. | |
| 02:17:09 | loincloth leaves the room. | |
| 02:18:20 | d2dchat enters the room. | |
| 02:43:30 | loincloth enters the room. | |
| 02:45:28 | ezmobius leaves the room. | |
| 02:45:42 | VVSiz_ enters the room. | |
| 02:56:06 | cored leaves the room. | |
| 02:56:27 | imajes leaves the room. | |
| 02:59:31 | JimRoepcke_ enters the room. | |
| 02:59:31 | JimRoepcke leaves the room. | |
| 03:02:45 | VVSiz leaves the room. | |
| 03:03:24 | stepheneb enters the room. | |
| 03:11:41 | ctennis leaves the room. | |
| 03:13:49 | jayWHY leaves the room. | |
| 03:17:03 | stepheneb_ enters the room. | |
| 03:18:05 | gnufied leaves the room. | |
| 03:19:04 | stepheneb leaves the room. | |
| 03:29:10 | rubuildius_ppc leaves the room. | |
| 03:33:38 | jtoy enters the room. | |
| 03:34:49 | chop3 leaves the room. | |
| 03:38:17 | mkescher enters the room. | |
| 03:48:17 | Maledictus leaves the room. | |
| 03:57:53 | loincloth leaves the room. | |
| 03:59:51 | d2dchat leaves the room. | |
| 04:09:21 | webmat | it'd be awesome if Rubinius had a wiki... any plans to that effect? |
| 04:09:27 | gnufied enters the room. | |
| 04:10:37 | webmat | Maybe even one based on Merb's wiki, but with the Rubinius web page look |
| 04:16:06 | kw leaves the room. | |
| 04:22:15 | seydar enters the room. | |
| 04:23:16 | seydar | howdy y'all |
| 04:25:19 | seydar leaves the room. | |
| 04:44:43 | crayz_ enters the room. | |
| 04:45:02 | mkescher leaves the room. | |
| 04:45:36 | crayz__ leaves the room. | |
| 04:46:45 | obvio leaves the room. | |
| 04:47:26 | rue | webmat: That is not a bad idea, actually |
| 04:47:58 | obvio enters the room. | |
| 04:48:57 | webmat | rue: wouldn't that rule? not only lower the barrier to contributing code / spec / code documentation but also lowering the barrier to wiki documentation ;-) |
| 04:53:56 | Packet_Monkey leaves the room. | |
| 04:54:40 | Packet_Monkey enters the room. | |
| 04:54:50 | Packet_Monkey leaves the room. | |
| 04:58:55 | rue | I suppose; The Lighthouse page system is not all that great |
| 04:59:05 | rue | Defiler: About? |
| 04:59:38 | webmat | Oh, is there a wiki on the rubinius lighthouse? |
| 05:02:32 | rue | Semi sorta. Think you need to be a member to be able to edit and the system sucks (not that we are restrictive with the edit rights) |
| 05:03:35 | rue | It is the "Pages" section |
| 05:04:22 | webmat | Ok, never noticed before :-) |
| 05:05:01 | webmat | I'll try to pimp it in a future Rubinius for the Layman... If I can actually get around to pushing those out. |
| 05:05:39 | rue | Hehe |
| 05:05:47 | rue | Started out good! |
| 05:06:07 | webmat | thanks, your comment was much appreciated :-) |
| 05:06:17 | webmat | I almost didn't publish it |
| 05:06:18 | rue | Actually we should really just go for it and set up a "proper" site |
| 05:06:26 | webmat | agrees |
| 05:06:36 | srbaker enters the room. | |
| 05:07:44 | webmat | rue: you work at EY? if not, get Evan to ask Jed about the Merb wiki |
| 05:07:54 | rue | Yeah, I do |
| 05:07:55 | webmat | how that's going for them |
| 05:08:13 | rue | Defiler is, as I understand, the web wrangler for us now |
| 05:08:31 | webmat | cool, then ask Jed directly? I'm not sure if he works for EY or he's just in the Merb community |
| 05:08:44 | webmat | ask exmobius maybe |
| 05:08:50 | webmat | ezmobius* |
| 05:09:22 | webmat | banana fingers on tiny Apple keyboard |
| 05:11:29 | rue | Heh, yeah, we can bother ez |
| 05:12:10 | srbaker leaves the room. | |
| 05:12:41 | jtoy leaves the room. | |
| 05:13:40 | trythil enters the room. | |
| 05:21:10 | be9 enters the room. | |
| 05:21:25 | webmat | So instead of actually writing part 2 of the Layman series, tonight I fiddled with my irbrc. And wrote 'irbrc for the runtime-tramp' :-) http://programblings.com/2008/04/11/irbrc-for-the-runtime-tramp/ |
| 05:21:44 | rubuildius_ppc enters the room. | |
| 05:22:15 | webmat | Contains a great way to safely require when your fiddle around with many runtimes that may not support everything you usually load in you irbrc |
| 05:27:29 | rue | Cool! Except for the not writing part 2 ;) |
| 05:27:43 | webmat | It's halfway done :-) |
| 05:28:04 | webmat | I'll try to push it out at the beginning of the week :-) |
| 05:28:40 | stepheneb_ leaves the room. | |
| 05:30:23 | gnufied leaves the room. | |
| 05:31:45 | rue | Mine is exceptionally bare: http://pastie.org/179534 |
| 05:35:37 | rubuildius_ppc | Eric Hodel: ee32e0b1d; 1990 files, 6481 examples, 22601 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/179537 |
| 05:36:03 | rue | That was slightly delayed |
| 05:42:29 | RyanTM leaves the room. | |
| 05:51:27 | KirinDav enters the room. | |
| 05:53:27 | jtoy enters the room. | |
| 05:55:52 | webmat | good night everyone :-) |
| 05:56:10 | rue | Nites! |
| 05:56:47 | nicksieger enters the room. | |
| 05:57:18 | webmat leaves the room. | |
| 06:08:37 | AndrewO leaves the room. | |
| 06:09:07 | wifelette leaves the room. | |
| 06:11:28 | KirinDav leaves the room. | |
| 06:17:38 | agile enters the room. | |
| 06:21:34 | brennanc leaves the room. | |
| 06:45:37 | mernen leaves the room. | |
| 06:47:17 | fbuilesv leaves the room. | |
| 07:07:03 | yaroslav enters the room. | |
| 07:12:09 | dysinger leaves the room. | |
| 07:24:47 | yaroslav leaves the room. | |
| 07:25:23 | yaroslav enters the room. | |
| 07:43:43 | benburkert leaves the room. | |
| 07:49:19 | benburkert enters the room. | |
| 07:54:13 | KirinDav enters the room. | |
| 08:11:13 | mentz enters the room. | |
| 08:17:23 | benburkert leaves the room. | |
| 08:18:50 | benburkert enters the room. | |
| 08:21:18 | hornbeck leaves the room. | |
| 08:31:09 | maduyb__ enters the room. | |
| 08:42:11 | reima enters the room. | |
| 08:42:20 | reima | Morning. |
| 08:43:12 | jtoy leaves the room. | |
| 08:57:10 | sk enters the room. | |
| 08:58:38 | rue | Morning |
| 08:58:43 | rue | And also night |
| 09:06:04 | sk | good afternoon |
| 09:14:02 | wycats enters the room. | |
| 09:16:14 | rue leaves the room. | |
| 09:17:49 | wycats leaves the room. | |
| 09:17:52 | sk leaves the room. | |
| 09:17:53 | rubuildius_ppc leaves the room. | |
| 09:17:53 | wubo leaves the room. | |
| 09:18:23 | wycats enters the room. | |
| 09:18:25 | sk enters the room. | |
| 09:18:25 | rubuildius_ppc enters the room. | |
| 09:18:25 | wubo enters the room. | |
| 09:18:28 | thehcdreamer enters the room. | |
| 09:20:47 | KirinDav leaves the room. | |
| 09:22:36 | yaroslav leaves the room. | |
| 09:25:41 | maduyb__ leaves the room. | |
| 09:26:29 | olabini enters the room. | |
| 09:37:29 | TheVoic1 leaves the room. | |
| 09:43:12 | wubo leaves the room. | |
| 09:43:12 | rubuildius_ppc leaves the room. | |
| 09:43:12 | sk leaves the room. | |
| 09:43:12 | wycats leaves the room. | |
| 09:43:30 | wycats enters the room. | |
| 09:43:30 | sk enters the room. | |
| 09:43:30 | rubuildius_ppc enters the room. | |
| 09:43:30 | wubo enters the room. | |
| 09:47:18 | cypher23 enters the room. | |
| 10:05:12 | benburkert leaves the room. | |
| 10:11:38 | trythil leaves the room. | |
| 10:30:36 | GMFlash leaves the room. | |
| 10:35:52 | smparke2 enters the room. | |
| 10:37:55 | wubo leaves the room. | |
| 10:37:55 | rubuildius_ppc leaves the room. | |
| 10:37:55 | sk leaves the room. | |
| 10:37:55 | wycats leaves the room. | |
| 10:38:37 | wycats enters the room. | |
| 10:38:37 | sk enters the room. | |
| 10:38:37 | rubuildius_ppc enters the room. | |
| 10:38:37 | wubo enters the room. | |
| 10:39:10 | smparke1 leaves the room. | |
| 10:39:12 | sk leaves the room. | |
| 10:46:23 | qwert666 enters the room. | |
| 10:54:28 | reima leaves the room. | |
| 11:04:19 | yaroslav enters the room. | |
| 11:11:35 | codebrulee enters the room. | |
| 11:11:37 | smparke1 enters the room. | |
| 11:12:09 | smparke2 leaves the room. | |
| 11:30:25 | yaroslav leaves the room. | |
| 11:35:30 | yugui enters the room. | |
| 11:37:55 | chris2 enters the room. | |
| 11:38:26 | lopex enters the room. | |
| 11:39:28 | joachimm enters the room. | |
| 11:45:28 | context leaves the room. | |
| 11:45:32 | context enters the room. | |
| 12:02:48 | chris2 leaves the room. | |
| 12:05:04 | Skip enters the room. | |
| 12:12:10 | Maledictus enters the room. | |
| 12:12:45 | codebrulee leaves the room. | |
| 12:27:26 | stepheneb enters the room. | |
| 12:28:18 | yaroslav enters the room. | |
| 12:32:00 | stepheneb_ enters the room. | |
| 12:36:39 | stepheneb__ enters the room. | |
| 12:40:39 | ctennis enters the room. | |
| 12:44:39 | stepheneb leaves the room. | |
| 12:48:36 | radarek enters the room. | |
| 12:48:56 | stepheneb_ leaves the room. | |
| 12:49:08 | loincloth enters the room. | |
| 12:53:11 | yaroslav leaves the room. | |
| 12:58:09 | chris2 enters the room. | |
| 13:05:18 | gnufied enters the room. | |
| 13:11:20 | mentz_ enters the room. | |
| 13:14:33 | chris2 | anyone know gregor schmidt? |
| 13:14:50 | chris2 | oooh |
| 13:14:54 | chris2 | i actually do :P |
| 13:15:00 | mentz__ enters the room. | |
| 13:17:53 | perdix enters the room. | |
| 13:25:52 | _mk_ enters the room. | |
| 13:26:05 | _mk_ leaves the room. | |
| 13:29:02 | mentz leaves the room. | |
| 13:29:17 | mentz enters the room. | |
| 13:32:38 | mentz_ leaves the room. | |
| 13:33:09 | mentz__ leaves the room. | |
| 13:48:39 | brainopia enters the room. | |
| 13:53:40 | olabini leaves the room. | |
| 13:59:26 | brainopia leaves the room. | |
| 14:00:02 | jero5 enters the room. | |
| 14:04:51 | febeling enters the room. | |
| 14:06:09 | imajes enters the room. | |
| 14:14:11 | stepheneb enters the room. | |
| 14:14:28 | stepheneb__ leaves the room. | |
| 14:14:55 | stepheneb_ enters the room. | |
| 14:26:12 | gnufied leaves the room. | |
| 14:33:36 | boyscout | 1 commit by Charles Oliver Nutter |
| 14:33:37 | boyscout | * Added instance_eval spec for defining methods under immediates.; e8fd8e6 |
| 14:35:09 | Defiler | haha I had no idea you could do that |
| 14:35:10 | Defiler | 1.instance_eval { def foo; end } |
| 14:35:14 | loop enters the room. | |
| 14:35:15 | Defiler | horrible |
| 14:36:11 | fbuilesv enters the room. | |
| 14:36:25 | stepheneb leaves the room. | |
| 14:37:06 | dbussink | well, it can't be done on 1 ;) |
| 14:37:59 | Defiler | Oh, right.. it is a spec for what can't be done |
| 14:38:03 | Defiler | *phew* |
| 14:41:22 | rubuildius_amd64 | Charles Oliver Nutter: e8fd8e696; 1990 files, 6478 examples, 22572 expectations, 0 failures, 0 errors; http://rafb.net/p/eP7DLP42.html |
| 14:41:51 | dbussink | Defiler: hmm, rubinius does allow it :P |
| 14:43:40 | dbussink | instance_eval doesn't work properly in the first place though |
| 14:44:08 | Defiler | Really? |
| 14:44:19 | dbussink | it defines it on all objects of that class |
| 14:44:33 | dbussink | not only the single instance |
| 14:45:05 | febeling leaves the room. | |
| 14:46:33 | rubuildius_ppc | Charles Oliver Nutter: e8fd8e696; 1990 files, 6481 examples, 22601 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/179654 |
| 14:50:37 | Defiler | Oh, you mean the specific case of a 'def' inside the evaled code? |
| 14:50:55 | dbussink | well, i just tested instance_eval in mri and rubinius |
| 14:52:31 | chris2_ enters the room. | |
| 14:54:37 | chris2 leaves the room. | |
| 14:59:00 | yugui leaves the room. | |
| 15:11:23 | hornbeck enters the room. | |
| 15:20:32 | stepheneb_ leaves the room. | |
| 15:22:16 | smparkes leaves the room. | |
| 15:28:13 | febeling enters the room. | |
| 15:37:50 | chris2 leaves the room. | |
| 15:42:06 | chris2 enters the room. | |
| 15:42:18 | boyscout | 1 commit by Charles Oliver Nutter |
| 15:42:19 | boyscout | * A few more instance_eval specs, for non-immediate numerics and cvars.; 5883dd7 |
| 15:43:03 | headius | instance_eval is supposed to execute against a singleton class |
| 15:43:26 | headius | in cases where you aren't allowed a singleton class, MRI then executes it against nil and has checks for that on method definition, etc |
| 15:45:08 | Defiler | Wow. |
| 15:45:34 | headius | edge case, to be sure |
| 15:54:22 | rubuildius_ppc | Charles Oliver Nutter: 5883dd78a; 1990 files, 6481 examples, 22601 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/179663 |
| 15:57:20 | rubuildius_amd64 | Charles Oliver Nutter: 5883dd78a; 1990 files, 6478 examples, 22572 expectations, 0 failures, 0 errors; http://rafb.net/p/3QCoN144.html |
| 15:59:46 | dysinger enters the room. | |
| 16:02:25 | mentz_ enters the room. | |
| 16:08:01 | mentz__ enters the room. | |
| 16:08:44 | mentz__ leaves the room. | |
| 16:09:32 | smparkes enters the room. | |
| 16:19:41 | KirinDav enters the room. | |
| 16:20:11 | mentz leaves the room. | |
| 16:23:45 | jtoy enters the room. | |
| 16:24:39 | GMFlash enters the room. | |
| 16:25:49 | mentz_ leaves the room. | |
| 16:32:04 | febeling | why should one not install currently? thats what the page on installation in lighthouse says... it does not give a reason thouigh |
| 16:32:56 | dbussink | it used to have some problems |
| 16:33:02 | dbussink | dunno the state now |
| 16:40:36 | fbuilesv | febeling: If you plan to be developing the process to reinstall each time you pull something from the repo can be a bit boring with changes happening every day |
| 16:41:40 | RyanTM enters the room. | |
| 16:49:48 | fbuilesv leaves the room. | |
| 16:51:02 | fbuilesv enters the room. | |
| 16:56:30 | ttmrichter leaves the room. | |
| 16:59:13 | RyanTM leaves the room. | |
| 17:00:08 | Defiler | febeling: There have, at various points, been problems related to which copy of a library rubinius loads.. the installed one or the one from the checkout |
| 17:00:24 | Defiler | but installation should work fine |
| 17:01:00 | hangon enters the room. | |
| 17:01:24 | RyanTM enters the room. | |
| 17:02:14 | benburkert enters the room. | |
| 17:03:49 | yaroslav enters the room. | |
| 17:05:16 | d2dchat enters the room. | |
| 17:10:02 | jtoy leaves the room. | |
| 17:16:26 | anteaya enters the room. | |
| 17:17:10 | joachimm_ enters the room. | |
| 17:20:05 | jayWHY enters the room. | |
| 17:30:00 | joachimm leaves the room. | |
| 17:32:32 | dbussink | hmm, is it ok to do a sleep inside a spec? |
| 17:33:02 | sk__ enters the room. | |
| 17:33:18 | sk__ leaves the room. | |
| 17:38:53 | dbussink | i don't like like it, but don't know another way to spec this |
| 17:39:01 | headius | spec what |
| 17:39:10 | dbussink | File::Stat#<=> |
| 17:39:26 | dbussink | it uses mtime to determine the result |
| 17:39:53 | dbussink | now i create a file, wait a short period and then stat the file |
| 17:40:39 | dbussink | i can make assumptions like "it's not likely that / is changed since the new file", but that doesn't feel very good either |
| 17:41:03 | headius | mmm |
| 17:41:23 | headius | well it's hard to spec things with temporal restrictions without things like sleep |
| 17:44:45 | yaroslav_ enters the room. | |
| 17:44:53 | perdiy enters the room. | |
| 17:45:08 | perdix leaves the room. | |
| 17:47:16 | febeling_ enters the room. | |
| 17:49:57 | brixen | dbussink: preferable to use a very small increment on the sleep, but we do have specs that use it |
| 17:50:25 | therealadam enters the room. | |
| 17:53:57 | febeling leaves the room. | |
| 17:54:13 | yaroslav_ leaves the room. | |
| 17:54:33 | yaroslav_ enters the room. | |
| 17:55:33 | yaroslav leaves the room. | |
| 17:56:32 | yaroslav_ leaves the room. | |
| 17:56:45 | yaroslav enters the room. | |
| 17:58:03 | KirinDav leaves the room. | |
| 17:58:37 | jnicklas enters the room. | |
| 18:02:17 | joachimm enters the room. | |
| 18:05:43 | hangon leaves the room. | |
| 18:06:21 | hangon enters the room. | |
| 18:12:02 | rubuildius_ppc leaves the room. | |
| 18:15:00 | dbussink | brixen: well, mtime's have a resolution of 1 second |
| 18:17:53 | joachimm_ leaves the room. | |
| 18:18:31 | dbussink | brixen: hmm, and sleep also doesn't guarantee that the time() clock returns a timestamp that is a second later |
| 18:27:14 | trythil enters the room. | |
| 18:43:00 | gnufied enters the room. | |
| 18:46:11 | imajes leaves the room. | |
| 18:46:37 | imajes enters the room. | |
| 18:50:07 | jnicklas leaves the room. | |
| 18:51:09 | jnicklas enters the room. | |
| 18:56:22 | _mk_ enters the room. | |
| 18:56:36 | scoopr | oh, mod_rails is out |
| 18:59:19 | jayWHY leaves the room. | |
| 19:00:09 | jayWHY enters the room. | |
| 19:06:37 | imajes leaves the room. | |
| 19:08:54 | yaroslav leaves the room. | |
| 19:28:33 | hangon leaves the room. | |
| 19:30:28 | d2dchat leaves the room. | |
| 19:31:03 | rby enters the room. | |
| 19:33:37 | loincloth leaves the room. | |
| 19:34:43 | _mk_ leaves the room. | |
| 19:38:03 | chris2 leaves the room. | |
| 19:41:25 | be9 leaves the room. | |
| 19:42:22 | guessmethod enters the room. | |
| 19:42:28 | guessmethod leaves the room. | |
| 19:47:30 | jayWHY leaves the room. | |
| 19:55:11 | jayWHY enters the room. | |
| 19:55:42 | jnicklas leaves the room. | |
| 20:28:25 | hangon enters the room. | |
| 20:30:30 | RyanTM leaves the room. | |
| 20:31:05 | zimbatm_ enters the room. | |
| 20:31:14 | RyanTM enters the room. | |
| 20:41:54 | dbussink | anyone in? |
| 20:46:26 | scoopr | just barely hanging in |
| 20:46:40 | dbussink | hehe |
| 20:46:56 | dbussink | been testing sleep in rubinius, doesn't really sleep as much as it should |
| 20:47:42 | djwhitt | insomnia? |
| 20:48:13 | dbussink | suffers from severe sleep depravation :P |
| 20:48:24 | dbussink | http://pastie.caboo.se/179725 |
| 20:48:29 | dbussink | that's a one second sleep |
| 20:50:42 | jnicklas enters the room. | |
| 20:50:46 | hangon leaves the room. | |
| 20:58:53 | hangon enters the room. | |
| 21:05:47 | wycats leaves the room. | |
| 21:10:39 | enebo enters the room. | |
| 21:23:35 | rue enters the room. | |
| 21:23:38 | evan | dbussink: thats libev suffering |
| 21:24:00 | dbussink | evan: that's where i ended up to yeah |
| 21:24:06 | dbussink | too |
| 21:24:37 | dbussink | but half a second is quite a lot |
| 21:24:54 | evan | quite |
| 21:25:02 | evan | we should probably figure out why it did that. |
| 21:25:21 | evan | i'm debugging the new event/libev C++ code now |
| 21:25:26 | evan | soo much nicer. |
| 21:25:43 | enebo_ enters the room. | |
| 21:25:44 | dbussink | nice :) |
| 21:25:49 | dbussink | making good progress? |
| 21:26:47 | evan | yeah |
| 21:27:02 | evan | this experiment is taking longer than i originally thought it would |
| 21:27:05 | evan | but thats ok |
| 21:27:16 | evan | because everyone seems to be happy with the advances it's giving us |
| 21:27:35 | ezmobius enters the room. | |
| 21:28:00 | tarcieri | I had trouble with libev's timer watchers |
| 21:28:12 | tarcieri | there's a bit in the libev documentation on it |
| 21:28:26 | tarcieri | but what it suggests to do to correct it is backward |
| 21:31:50 | evan | hrm. |
| 21:32:03 | evan | oh crap. |
| 21:32:12 | evan | timer is in cpu time, isn't it |
| 21:32:14 | evan | not wallclock time. |
| 21:32:44 | evan | ie, doing a sleep() to make it advance doesn't make it advance at all |
| 21:33:20 | tarcieri | I don't get exactly how it keeps track of time internally |
| 21:33:46 | tarcieri | but you can calibrate it against the wall time |
| 21:34:07 | dbussink | why do i always get the feeling i'm digging up nasty things? :P |
| 21:34:33 | evan | hm. strangly, it appears that ev_now() is only updated after the loop runs |
| 21:34:35 | tarcieri | timeout = interval + ev_time() - ev_now(loop_data->ev_loop); |
| 21:34:41 | tarcieri | yeah exactly |
| 21:34:49 | evan | ok, there we go. |
| 21:34:59 | evan | now ev_now() is showing that 2 seconds went by. |
| 21:35:04 | evan | but why didn't it fire my timer event... |
| 21:35:24 | rue | Morning, gents |
| 21:35:32 | fbuilesv | hola |
| 21:35:35 | boyscout | 2 commits by Matthias Reitinger |
| 21:35:35 | boyscout | * Fixed incorrect check for object equality in BigDecimal#nonzero? spec; 00547bc |
| 21:35:35 | boyscout | * Eliminated use of to_s to check for NaN in BigDecimal specs; e840379 |
| 21:35:41 | evan | allo |
| 21:36:40 | evan | i'm doing some debugging before heading to the beach |
| 21:37:08 | tarcieri | evan: what was ev_now() when you initialized the timer? |
| 21:37:15 | dbussink | beach? /me is jealous |
| 21:37:22 | tarcieri | particularly compared to the wall clock |
| 21:37:22 | evan | umm. let me check. |
| 21:37:46 | evan | gets out a giant flavor flav clock to perform proper debugging with |
| 21:37:54 | tarcieri | heh |
| 21:38:06 | enebo leaves the room. | |
| 21:38:54 | tarcieri | evan: I had timers getting like up to a second off for no apparent reason until I calibrated the values I initialized the timers with against ev_time() |
| 21:39:09 | fbuilesv | read clock as cock and mixing that with Flavor Flav was not nice at all :( |
| 21:39:20 | tarcieri | haha |
| 21:39:28 | evan | wtf. |
| 21:39:53 | evan | hah |
| 21:40:17 | evan | ok, well, libev is processing my timer event |
| 21:40:26 | evan | because the at field is advancing |
| 21:40:53 | tarcieri | just at the wrong time? |
| 21:41:52 | evan | i checked the at field before the loop ran, and it was set for 2 seconds in the future |
| 21:42:00 | evan | i checked after a sleep and after the loop ran again |
| 21:42:05 | evan | the at field has moved forward |
| 21:42:13 | evan | and the time it was originally has past |
| 21:42:50 | tarcieri | weird |
| 21:43:29 | dgtized leaves the room. | |
| 21:43:37 | evan | ok, i should get read for the beach. |
| 21:43:41 | evan | libev can wait. |
| 21:43:50 | evan | the sands of time call. |
| 21:43:56 | dbussink | how warm is it there now? |
| 21:44:10 | evan | today? i've heard it's supposed to be 92 |
| 21:44:18 | evan | doesn't feel that hot though |
| 21:44:20 | evan | probably 80 though |
| 21:44:31 | dbussink | so no swimming? |
| 21:44:36 | dbussink | or being a real man :P |
| 21:44:36 | evan | in the ocean! |
| 21:44:49 | evan | the heat here in southern california is a dry heat |
| 21:44:53 | evan | it's not humid at all |
| 21:45:03 | evan | so the need for continuely spending hot days in the pool isn't as high |
| 21:45:22 | dbussink | if it's over 80 here, you can't really do anything anymore |
| 21:45:41 | evan | i'm sure it's humid there though in the summer right? |
| 21:46:34 | dbussink | yeah, when i was in california it was really nice and not that humid indeed |
| 21:46:57 | evan | thats one awesome thing about the west coast of north america |
| 21:47:10 | evan | the climate is not humid at all |
| 21:47:29 | dbussink | but it's still cold here right now |
| 21:47:41 | dbussink | around 10 celcius |
| 21:47:56 | dbussink | 53.6 F according to my calculator |
| 21:48:04 | evan | thats not that cold. |
| 21:48:16 | dbussink | no, but usually it's a bit warmer this time of year |
| 21:48:16 | evan | it gets down to that here in the evenings |
| 21:48:23 | evan | true |
| 21:48:28 | evan | what will the tulips do!! |
| 21:48:29 | evan | :) |
| 21:48:31 | evan | ok, off to the beach. |
| 21:48:34 | evan | later gents |
| 21:48:36 | dbussink | have fun! |
| 21:50:07 | rue | Be back in a bit, I am going to solve this Leyton mystery once and for all |
| 21:50:19 | dbussink | leyton mystery? |
| 21:50:37 | djwhitt | nintendo ds game |
| 21:51:32 | dbussink | ah, didn't know that |
| 21:51:43 | dbussink | the new mario kart for the wii is cool though |
| 21:51:47 | dbussink | works pretty well |
| 22:00:47 | radarek leaves the room. | |
| 22:07:25 | anteaya leaves the room. | |
| 22:10:43 | hangon_ enters the room. | |
| 22:11:09 | hangon leaves the room. | |
| 22:17:02 | zenspider leaves the room. | |
| 22:22:11 | fbuilesv | ci bots died? |
| 22:22:50 | djwhitt | mine's up |
| 22:22:55 | djwhitt | looks like ppc is down though |
| 22:23:45 | fbuilesv | djwhitt: pushed some spec changes like 20 mins ago, no answer S: |
| 22:24:36 | djwhitt | fbuilesv: hmm... mine's been running specs for about 40 minutes |
| 22:25:28 | fbuilesv | djwhitt: I own at killing bots :-) |
| 22:25:39 | djwhitt | hehe |
| 22:25:40 | djwhitt | indeed |
| 22:25:44 | djwhitt | let me check the log... |
| 22:26:44 | djwhitt | no errors in the logs |
| 22:27:06 | djwhitt | I'll kill and restart and see what happens |
| 22:27:23 | rubuildius_amd64 | Matthias Reitinger: 00547bc56; bin/ci failed! http://rafb.net/p/IBrp1Y69.html |
| 22:27:33 | enebo_ leaves the room. | |
| 22:27:34 | djwhitt | ignore that |
| 22:27:51 | djwhitt | I suppose I should have stoped the bot and then killed the spec run |
| 22:27:52 | djwhitt | oh well |
| 22:28:21 | rubuildius_amd64 leaves the room. | |
| 22:28:43 | rubuildius_amd64 enters the room. | |
| 22:28:51 | rubuildius_amd64 leaves the room. | |
| 22:29:04 | rubuildius_amd64 enters the room. | |
| 22:29:09 | _mk_ enters the room. | |
| 22:31:23 | anteaya enters the room. | |
| 22:32:13 | febeling_ leaves the room. | |
| 22:37:38 | imajes enters the room. | |
| 22:41:37 | djwhitt | fbuilesv: spec run after last restart is still going. I suspect it's the same problem we noticed before on 64 bit. I'm testing on 32 bit right now to confirm |
| 22:41:44 | boyscout | 1 commit by Federico Builes |
| 22:41:45 | boyscout | * More specs for REXML::Element; 7638548 |
| 22:42:22 | fbuilesv | djwhitt: I just tested on Mac and Ubuntu again to make sure on 32 bits and it worked just fine. And I didn't know you were rerunning the changes, should'nt have pushed those last spec changes |
| 22:43:36 | rubuildius_amd64 leaves the room. | |
| 22:43:53 | djwhitt | fbuilesv: no biggy. I'm taking the build bot down for now |
| 22:49:19 | djwhitt | cremes, cremes_: if you're arround it looks like rubuildius_ppc is down |
| 22:51:40 | rubuildius_ppc enters the room. | |
| 22:52:02 | cremes | djwhitt: just restarted it |
| 22:52:43 | djwhitt | cremes_: cool, I've temprorarily taken my bot down because of extremely long spec runs (was at 40 min and still going) |
| 22:53:25 | djwhitt | fbuilesv: 32bit run came out fine too |
| 22:53:34 | cremes | djwhitt: okay; i'm going to need some of your auto-restart lovin' so i don't have to watch this 'bot so closely |
| 22:54:12 | fbuilesv | cremes: was the 40 mins spec run running the BigDecimal specs changes or did it happen before? |
| 22:54:27 | djwhitt | cremes_: I'll get the latest code tested this weekend. then you can switch to that |
| 22:54:46 | djwhitt | fbuilesv: you intend that last message for me? |
| 22:55:08 | fbuilesv | djwhitt: oh, nvm, thought cremes had the same issue, |
| 22:55:37 | djwhitt | no, I don't think so. I think it's amd64 only |
| 22:56:06 | djwhitt | probably intel too, but I have no way of testing that |
| 22:57:10 | djwhitt | fbuilesv: to answer your question, run times had been starting to creep up again before your commit |
| 22:57:58 | fbuilesv | djwhitt: I see. Let me see if I can get my slice to run this |
| 22:58:15 | djwhitt | fbuilesv: slice? as in slicehost? |
| 22:58:38 | fbuilesv | djwhitt: yes, it's a big one for the company so it _should_ not die |
| 22:58:59 | ttmrichter enters the room. | |
| 23:03:33 | wmoxam enters the room. | |
| 23:04:16 | rubuildius_ppc | Federico Builes: 763854840; 1995 files, 6495 examples, 22615 expectations, 0 failures, 0 errors; http://pastie.caboo.se/paste/179785 |
| 23:12:31 | _mk_ leaves the room. | |
| 23:16:13 | die_gabel enters the room. | |
| 23:19:37 | tokengeek enters the room. | |
| 23:22:47 | hangon_ leaves the room. | |
| 23:23:10 | rby leaves the room. | |
| 23:23:31 | hangon enters the room. | |
| 23:24:35 | hornbeck leaves the room. | |
| 23:24:38 | fbuilesv | djwhitt: the specs run complaining a lot about the File specs but they all seem to finish in around 60-70 secs |
| 23:25:19 | cypher23 leaves the room. | |
| 23:25:28 | djwhitt | fbuilesv: what size slice is that? |
| 23:25:38 | fbuilesv | djwhitt: 1024 |
| 23:25:45 | djwhitt | really... interesting |
| 23:25:58 | fbuilesv | what exactly? |
| 23:26:17 | djwhitt | the machine I'm running on has 2048 MB RAM |
| 23:27:07 | fbuilesv | djwhitt: although this is 1024 it has some apache, mysql, postfix and some usual stuff running |
| 23:27:29 | fbuilesv | 326MB free to be exact |
| 23:27:30 | djwhitt | fbuilesv: what distro? |
| 23:27:40 | fbuilesv | djwhitt: Ubuntu 7.10 |
| 23:28:04 | djwhitt | fbuilesv: it's an AMD machine right? |
| 23:28:30 | fbuilesv | djwhitt: I have no idea, uname -p returns unknown but I do know it's running 64 bit |
| 23:28:43 | fbuilesv | any suggestion on how to get the exact processor if it's running under zen? |
| 23:28:55 | djwhitt | fbuilesv: uname -a will probably tell you |
| 23:29:16 | fbuilesv | djwhitt: Linux santiago 2.6.18-xen #1 SMP Tue Feb 12 06:40:50 UTC 2008 x86_64 GNU/Linux |
| 23:29:17 | djwhitt | oh wait... uname -p should too |
| 23:29:31 | djwhitt | I wonder if it's Intel |
| 23:29:41 | fbuilesv | djwhitt: let me try to find that out |
| 23:30:01 | djwhitt | uname -a on my slice gives me this: Linux marvin 2.6.16.29-xen #1 SMP Sun Sep 30 04:00:13 UTC 2007 x86_64 Dual-Core AMD Opteron(tm) Processor 2212 HE AuthenticAMD GNU/Linux |
| 23:31:13 | fbuilesv | djwhitt: Slicehost too? |
| 23:31:19 | djwhitt | fbuilesv: yep |
| 23:31:23 | fbuilesv | weird, let me try to ask them |
| 23:31:36 | trythil leaves the room. | |
| 23:32:04 | fbuilesv | djwhitt: Dual-Core AMD Opteron(tm) Processor 2212 |
| 23:32:52 | djwhitt | interesting let me try it on my slice and see what happens (bot is actually on a differen machine) |
| 23:33:12 | fbuilesv | djwhitt: taking a shower, brb. |
| 23:39:05 | wmoxam leaves the room. | |
| 23:39:21 | seydar enters the room. | |
| 23:40:02 | seydar | ok, concurrency model question coming up |
| 23:40:21 | seydar | http://users.encs.concordia.ca/~grogono/Erasmus/E01.pdf someone tell me why not |
| 23:40:38 | seydar | basically, it lets the objects themselves take care of them being concurrent |
| 23:41:59 | jlindley enters the room. | |
| 23:43:47 | jnicklas leaves the room. | |
| 23:44:07 | seydar | although i probably misinterpreted it |
| 23:44:35 | Maledictus leaves the room. | |
| 23:46:51 | tarcieri | sounds an awful lot like the Actor model |
| 23:49:09 | seydar | yea |
| 23:49:10 | seydar | a |
| 23:49:20 | seydar | just now finished the article and read some comments on it |
| 23:49:33 | seydar | they all agree its pretty much the actor model. i think i posted the wrong link |
| 23:49:49 | seydar | because there is SOMETHING out there that lets objects concurencize themselves |
| 23:50:13 | srbaker enters the room. | |
| 23:50:47 | tarcieri | the gen_server pattern in Erlang is effectively a concurrent object |
| 23:51:10 | seydar | not familiar with that one. can you explain it more? |
| 23:51:12 | tarcieri | one which supports the same call semantics as normal objects, but can also multiplex asynchronous message reception |
| 23:51:32 | tarcieri | it's like an object that can behave synchronously or asynchronously at the same time |
| 23:51:37 | anteaya leaves the room. | |
| 23:52:27 | seydar | so isn't it deciding whether a method is defined as "asynchronous" or "synchronous"? |
| 23:53:07 | fbuilesv | djwhitt: how did it go? |
| 23:53:09 | tarcieri | the API exposed from a gen_server can be completely synchronous to the end-user |
| 23:53:26 | tarcieri | with concurrency occuring from handling incoming I/O events |
| 23:53:32 | djwhitt | fbuilesv: been running for about 4 min now |
| 23:53:48 | fbuilesv | djwhitt: the specs alone? |
| 23:53:53 | djwhitt | fbuilesv: yep |
| 23:53:57 | fbuilesv | weird |
| 23:54:02 | djwhitt | fbuilesv: were you running specs that were already compiled? |
| 23:54:17 | imajes leaves the room. | |
| 23:54:20 | fbuilesv | djwhitt: rake build && bin/mspec ci |
| 23:54:29 | djwhitt | fbuilesv: ah, hmm... |
| 23:55:07 | djwhitt | fbuilesv: it might just be swapping though. my slice only has 512MB RAM |
| 23:55:14 | djwhitt | fbuilesv: I'll let it run for a while |
| 23:55:17 | seydar | tarcieri: cool. and what would it take for threads to be concurrent in rubinius? |
| 23:55:21 | EugZol enters the room. | |
| 23:55:47 | fbuilesv | djwhitt: I ran on my personal 512MB and it did around 70s but it complained about File too |
| 23:55:56 | fbuilesv | it wasn't running anything besides nginx and few mongrels tho |
| 23:57:07 | djwhitt | I am running Gentoo... |
| 23:57:31 | hangon leaves the room. | |
| 23:57:32 | djwhitt | could be a distro thing. though it works fine on my 32 bit Gentoo system |
| 23:58:57 | seydar | btw everyone, the latest pull seems to kill sooooooooooo many things on Tiger/PPC |
| 23:59:29 | anteaya enters the room. | |
| 23:59:50 | fbuilesv | seydar: kill as in failures or as in specs that don't finish? |
| 23:59:58 | seydar | both |