Show enters and exits. Hide enters and exits.
| 00:05:08 | peterhil leaves the room. | |
| 00:06:38 | nari leaves the room. | |
| 00:09:09 | cored enters the room. | |
| 00:12:54 | tarcieri leaves the room. | |
| 00:13:03 | tarcieri enters the room. | |
| 00:16:30 | Maledictus leaves the room. | |
| 00:27:01 | scoopr leaves the room. | |
| 00:31:15 | lopex leaves the room. | |
| 00:35:29 | joearnold enters the room. | |
| 00:37:16 | benburkert leaves the room. | |
| 00:40:57 | benburkert enters the room. | |
| 00:41:58 | stepheneb enters the room. | |
| 00:55:00 | SoreGums leaves the room. | |
| 00:57:55 | jp_tix leaves the room. | |
| 00:58:21 | jptix enters the room. | |
| 00:58:39 | joearnold leaves the room. | |
| 01:09:07 | imajes enters the room. | |
| 01:14:26 | nari enters the room. | |
| 01:14:26 | macournoyer enters the room. | |
| 01:19:24 | benburkert leaves the room. | |
| 01:21:39 | benburkert enters the room. | |
| 01:26:48 | evan enters the room. | |
| 01:46:57 | seydar enters the room. | |
| 01:48:55 | seydar | salutations |
| 01:54:24 | rue | Evening, scoutster |
| 01:56:04 | jackdempsey leaves the room. | |
| 01:56:06 | seydar | don't remind me. I should be working on a merit badge right now |
| 01:56:15 | seydar | nah, i'll do it tuesday/wednesday |
| 01:56:22 | rue | Probably not an IRC merit badge |
| 01:56:22 | seydar | paper gets written tomorrow |
| 01:56:46 | seydar | unfortunately it's not. It's Citizenship in the Nation |
| 01:57:20 | rue | So, basically, it consists of standing around and maybe going to a baseball game? |
| 01:57:58 | seydar | i have to write about the constitution and talk about how America is so great and all |
| 02:03:48 | seydar | rue: how's AT&Ti going? |
| 02:03:53 | scoopr enters the room. | |
| 02:08:56 | scoopr leaves the room. | |
| 02:09:24 | scoopr enters the room. | |
| 02:13:51 | blowmage enters the room. | |
| 02:15:27 | joearnold enters the room. | |
| 02:18:32 | joearnold leaves the room. | |
| 02:26:09 | rue | seydar: Pretty swell, doing various different things |
| 02:28:52 | TheWordNerd enters the room. | |
| 02:39:08 | macournoyer leaves the room. | |
| 02:40:17 | seydar | rue: care to enlighten on those things? |
| 02:41:55 | macournoyer enters the room. | |
| 02:42:22 | seydar | expand* |
| 02:51:09 | macournoyer leaves the room. | |
| 02:58:21 | rue | seydar: Mainly stuff related to REST/web frameworks, a bit of Rubinius |
| 02:59:15 | evan | hey girls |
| 03:00:55 | rue | Well hello, sailor |
| 03:02:25 | evan | hows things? |
| 03:02:32 | evan | backlog was pretty quiet |
| 03:02:34 | drbrain | I'm about to get another RubyGems out the door! |
| 03:02:39 | drbrain | maybe by weeks-end |
| 03:02:43 | evan | cool! |
| 03:02:53 | evan | whats new in gemification land? |
| 03:03:05 | drbrain | sponsored by AT&T Interactive® |
| 03:03:11 | drbrain | lots of bug fixes |
| 03:03:15 | drbrain | development versions |
| 03:03:26 | drbrain | RSS feeds for gem repositories |
| 03:03:55 | evan | drbrain: has wycats ever explained the problem he had with requiring gem versions? |
| 03:04:02 | drbrain | yes |
| 03:04:03 | evan | he was tellin' me, sounded interesting. |
| 03:04:14 | evan | any thoughts on it? |
| 03:04:36 | drbrain | I know there's a patch floating around |
| 03:04:50 | evan | ah |
| 03:04:57 | drbrain | I don't remember if I incorporated it, it may not have had enough tests for my liking |
| 03:04:59 | evan | btw, if you want any help wading the patch queue, let me know. |
| 03:05:03 | drbrain | (aka, zero tests) |
| 03:05:12 | evan | since you're the only dev on rubygems atm it seems |
| 03:05:18 | drbrain | it's nearly empty |
| 03:05:24 | evan | ok, cool. |
| 03:05:33 | drbrain | Berger chewed through a bunch of patches in December/January |
| 03:05:40 | evan | ah, nice. |
| 03:05:48 | evan | I hung out with Luis a bunch this weekend |
| 03:05:49 | drbrain | I think he got bored |
| 03:05:50 | evan | he's awesome. |
| 03:05:58 | evan | well, feel free to add me |
| 03:06:01 | drbrain | oh, nice |
| 03:06:01 | evan | i'll help go through them |
| 03:06:32 | drbrain | right now, all the patches are awaiting cleanup/tests from the original authors |
| 03:07:04 | drbrain | bought my Japan tickets too! |
| 03:07:05 | rue | Nothing much going on. brixen is working on the profiler? |
| 03:07:09 | evan | drbrain: yay! |
| 03:07:16 | evan | drbrain: you'll have a blast |
| 03:07:20 | evan | i'm sad I can't be there |
| 03:07:29 | evan | patterson is an awesome travel partner, as you know. |
| 03:07:29 | drbrain | my gf Jamie is going too |
| 03:07:36 | drbrain | yeah |
| 03:07:37 | evan | I can't imagine the trouble the 3 of you will get into. |
| 03:07:39 | manveru | yahoo, dev versions! |
| 03:07:56 | evan | so, since I can't imagine it, I REQUIRE photo proof afterward! |
| 03:08:03 | evan | :D |
| 03:08:08 | drbrain | hehe |
| 03:08:10 | manveru | now i can do a release for every single commit :) |
| 03:08:17 | evan | manveru: success! |
| 03:08:19 | drbrain | manveru: indeed |
| 03:08:21 | rue | I fixed Method#call, and (I think) dynamic locals, but I have not cleaned up stuff. |
| 03:08:43 | evan | rue: did you commit those fixes? |
| 03:09:21 | rue | No |
| 03:09:28 | evan | ok |
| 03:09:32 | evan | i'm curious what was busted |
| 03:09:36 | rue | The former is trivial, the latter not so much |
| 03:10:17 | cored leaves the room. | |
| 03:10:58 | evan | dynamic locals == eval introduced locals |
| 03:11:13 | evan | ie, get_eval_local/set_eval_local |
| 03:11:17 | rue | evan: I think the reproduction for latter was something like `def foo; eval "a = 1"; [1].each {|i| eval "p a"}; end; foo` |
| 03:11:29 | evan | ah ha! |
| 03:11:32 | evan | i thought I fixed that... |
| 03:11:38 | evan | ah |
| 03:11:40 | evan | i see why |
| 03:11:40 | rue | Yeah. I fixed it by recursing up the vscopes |
| 03:11:45 | evan | i hardcoded it to just use @parent |
| 03:11:53 | evan | but it needs to walk up to the scope where @parent == nil |
| 03:11:55 | evan | ie, the top. |
| 03:11:57 | rue | But I figure there is a better way, maybe pushing a different frame |
| 03:12:07 | evan | push where? |
| 03:12:24 | evan | well, if it fixes stuff |
| 03:12:26 | evan | go ahead and push it |
| 03:12:32 | evan | we can also sort out a better way later |
| 03:12:37 | evan | i'm trying to be more chill about stuff |
| 03:12:45 | evan | like I was when the project started |
| 03:13:29 | seydar | it'd be funny if you showed up in IRC high everyday |
| 03:13:32 | evan | oh btw, any.io is AWESOME |
| 03:13:47 | evan | i love being able to use finger and DNS for twitter. |
| 03:14:01 | evan | seydar: i probably wouldn't get much work done |
| 03:14:09 | evan | my track record for working while high is pretty bad. |
| 03:14:12 | rue | push_variables, was it? I did not look through it but it pushes the of_sender rather than the one where the local is found or something, so there is double traversal |
| 03:14:21 | rue | I do not recall now |
| 03:14:36 | evan | ok |
| 03:14:42 | evan | no sweat |
| 03:14:51 | drbrain | even beer + programming does not necessarily work out well |
| 03:14:57 | evan | no |
| 03:14:59 | evan | rarely in my case |
| 03:15:03 | evan | i get sleepy |
| 03:15:09 | evan | and I get code like |
| 03:15:16 | evan | a = ob.runa.uhhueaheeaaooooooooooooooo |
| 03:15:23 | evan | as my head hits the keyboard |
| 03:16:36 | seydar | i once wrote a letter to a prof that finished with 'end' |
| 03:16:41 | seydar | but that wasn't because i was drunk |
| 03:16:44 | seydar | i was just tired |
| 03:16:57 | seydar | OMFG story time kids! |
| 03:17:11 | rue | tilman wanted to change the rbc format and I think libc rewired JIT |
| 03:17:30 | rue | Well, in potentia |
| 03:17:50 | seydar | so my friend and I were acting really drunk on the green, and we made eye contact with a cop. then we got in our car, pulled off, and his lights turned on and he pulled us over |
| 03:18:08 | rue | This is on a golf court? |
| 03:18:17 | seydar | oh. sorry you non new englanders |
| 03:18:26 | seydar | a green is a nice lil' field in the center of town |
| 03:18:28 | jackdempsey enters the room. | |
| 03:18:42 | rue | So essentially yes |
| 03:18:45 | rue | Go on |
| 03:18:46 | evan | rue: yeah, i saw libc's change |
| 03:18:54 | evan | jit needs a lot more work than he did |
| 03:18:59 | evan | because of the execution change |
| 03:19:02 | seydar | rue: that was my story. we got pulled over on the suspicion of drunk driving |
| 03:19:05 | evan | but his is a good first step |
| 03:19:22 | rue | Well, yeah |
| 03:19:43 | rue | seydar: There are no riding leopards involved, are there? |
| 03:20:01 | seydar | rue: no. and what did you say to 'pikkulla'? i accidentally closed the window |
| 03:20:20 | rue | "Pikku <whatever>lla" or "pienella" |
| 03:21:22 | seydar | thanks! |
| 03:21:28 | seydar | pulls out his dictionary |
| 03:22:04 | rue | See, a really good story starts with something like "So this one time, we had eaten morning glory seeds and after getting over the paranoia of having poisoned ourselves, we wandered out to town. The only thing I remember from early on in the night is when we had joined a queue somewhere and I had lied down on my back because I was laughing so hard, but then it got interesting |
| 03:22:31 | seydar | WTF |
| 03:22:36 | seydar | that was the best story ever |
| 03:22:39 | seydar | it made zero sense |
| 03:22:57 | evan | exactly. |
| 03:22:59 | rue | There you go |
| 03:23:03 | evan | see, in finland |
| 03:23:09 | evan | they get 10x the exposure to cosmic rays |
| 03:23:11 | evan | as everyone else. |
| 03:23:34 | seydar | hahaha my parents think that's awesome |
| 03:23:38 | seydar | you're famous in my household |
| 03:23:40 | benburkert leaves the room. | |
| 03:23:42 | rue | Great |
| 03:24:00 | evan | rue: now you've got a family to cook for you on the east coast |
| 03:24:01 | seydar | you should really try to go to goruco |
| 03:24:38 | rue | evan: That might explain why my high school career planning document stated my dream profession as, and I quote, "opium-addled wandering minstrel" |
| 03:25:47 | seydar | rue: isn't it 4 in the morning in finlandia? are you like this at all times or just when you're tired? |
| 03:25:47 | rue | seydar: Yeah, but naawh. I forgot when it was again anyway, and I am already going to Spain :P |
| 03:26:02 | evan | your high school counsler sounds soo much cooler that mine |
| 03:26:02 | rue | It is 05:25 |
| 03:26:12 | seydar | oh hey, that's cool. ditch out on the second greatest conference of all time |
| 03:26:12 | evan | mine just told me that my GPA was too low to get into any good colleges. |
| 03:26:14 | evan | fuck that guy. |
| 03:26:20 | rue | Hah |
| 03:26:25 | seydar | i now have a lowered opinion of you |
| 03:26:35 | seydar | evan: we have this program called naviance, that basically says the same thing |
| 03:26:43 | rue | Damn, do I have to kick Superman's ass again? |
| 03:26:47 | seydar | except statistically in a way that's difficult to argue against |
| 03:26:52 | benburkert enters the room. | |
| 03:27:04 | seydar | /kick rue for your failure to go to NYC |
| 03:28:17 | rue | Well.. my wife will be in Ohio in 2 weeks.. it counts for, like, 1/3 visit right? |
| 03:29:51 | rue | I should probably go to bed in a bit, but I am watching "Terry Jones' Medieval Lives" |
| 03:29:55 | rue | (Part 5) |
| 03:30:17 | rue | I was thinking starting a formal polyphasic sleep rhythm today |
| 03:31:31 | seydar | wait |
| 03:31:34 | seydar | hold on |
| 03:31:46 | seydar | first: wife? |
| 03:32:13 | seydar | second: i tried polyphasic sleeping but my rents stopped me. we should bond over PPSing |
| 03:33:18 | rue | This basmatic curry rice is superb. Mm |
| 03:34:03 | rue | Yeah, I will try 3 + 3*20, which is somewhat close to my normal patterns |
| 03:34:32 | seydar | you know what's odd? that's exactly what i did, too |
| 03:34:39 | seydar | now explain the wife business |
| 03:34:45 | seydar | this i did not know |
| 03:36:33 | rue | Well, erm, when a man and a woman (or, occasionally, a man and a man or a woman and a woman, or in some cases any combination of one or more of the above) love eachother very much.. |
| 03:37:16 | seydar | does she speak finnish as well? |
| 03:37:26 | rue | No |
| 03:37:56 | seydar | 'tis a shame |
| 03:38:04 | seydar | but i'm sure she's perfect in every other way! |
| 03:38:08 | seydar | besides the not speaking finnish part |
| 03:39:25 | rue | Minor slight |
| 03:39:37 | krawek enters the room. | |
| 03:40:33 | seydar | does she deal well with the whole "rue lives in finland" bit? |
| 03:40:33 | seydar | sorry if that's too personal |
| 03:41:23 | evan | she lives with rue, in finland. |
| 03:42:06 | rue | ^^ |
| 03:43:39 | seydar | WHOA |
| 03:43:44 | seydar | oh right that makes sense |
| 03:43:50 | seydar | so is she learning finnish? |
| 03:44:02 | rue | She is fairly well-adjusted, considering she has to be around me continuously |
| 03:45:17 | rue | She is supposed to take a class in the summer. There is {un,}fortunately no need to speak Finnish |
| 03:45:40 | evan | helsinki is pretty fucking metropolitan |
| 03:45:53 | evan | plus, like swedishi, finish doesn't get ya far in the world |
| 03:46:02 | evan | so most of them there peoples know english too |
| 03:46:07 | evan | as rue has shown us |
| 03:46:28 | evan | rue: hopefully i'm not protraying that wrong |
| 03:47:08 | rue | Not really, considering together we make up about the population of Southern California |
| 03:47:25 | rue | Not quite, even. NYC, and that is counting Iceland too |
| 03:47:31 | seydar | it is my sincere hope that finland will become an economic powerhouse and everyone will do business in finnish |
| 03:48:06 | evan | i believe that was true at one point |
| 03:48:21 | evan | maybe it wasn't finnish, but it was something like that. |
| 03:48:31 | rue | Latin? :) |
| 03:48:45 | evan | hah |
| 03:48:49 | rue | The Swedes got around until the 20th century |
| 03:49:25 | evan | before they started teaching things other than swedish to the population? |
| 03:50:03 | evan | thats probably because they were trying to lay claim to north for so long |
| 03:50:03 | jackdempsey leaves the room. | |
| 03:51:23 | rue | Too busy fighting.. and there has never really been that much of a pressure to colonise conquered areas here in Europe to the degree of getting the folks to change their language |
| 03:51:35 | evan | yeah |
| 03:51:56 | evan | sorta like how we only needed timezones when people could communicate over long distances for the first time |
| 03:53:47 | rue | Well, the actual colonies no-one spoke a language that *anyone* in the occupying force understood, unlike continentally, so they imported the tongue too |
| 03:54:11 | evan | on the other hand |
| 03:54:20 | evan | thats handy for sending messages across battle lines |
| 03:56:36 | rue | Yeap |
| 03:57:23 | seydar | speaking of iceland, i have a book on icelandic if anyone's interested |
| 03:57:43 | evan | i wonder if whats why klingon is so hard to understand |
| 03:57:49 | rue | Funny how everyone forgot about Iceland |
| 03:58:05 | evan | who? |
| 03:58:07 | evan | whats iceland? |
| 03:58:28 | seydar | iceland would be cooler if they hadn't welcomed bobby fischer so warmly |
| 04:02:33 | rue | I thought they were endothermic |
| 04:13:08 | seydar | hahahaha |
| 04:13:26 | seydar | ok. good night children. rue you are the best story teller ever |
| 04:31:24 | imajes leaves the room. | |
| 04:42:42 | libc | evan, still here? |
| 04:42:49 | evan | yep |
| 04:42:52 | evan | saw your patch |
| 04:42:54 | evan | thats fine to commit |
| 04:42:57 | evan | but the jit needs a lot more work |
| 04:43:05 | libc | evan, http://pastie.org/438055 |
| 04:43:07 | evan | did you actually run it with your patch? is it working for you? |
| 04:43:19 | libc | evan, it's working… |
| 04:43:28 | libc | I changed CALLS_TIL_JIT to 0 and run spec |
| 04:44:07 | evan | ok, cool |
| 04:44:08 | evan | commit that |
| 04:44:12 | evan | i'm not sure I like the array one |
| 04:44:27 | libc | what should I fix in it ? |
| 04:44:35 | evan | i don't like your 'def cast_array' in helpers/generator.rb |
| 04:44:38 | evan | why are you testing @last? |
| 04:45:23 | evan | actually though |
| 04:45:27 | evan | i'm rereading now |
| 04:45:40 | libc | because compiler/generator does that |
| 04:46:01 | evan | it shouldn't |
| 04:46:18 | evan | or rather, where does it do that? |
| 04:46:29 | evan | wow. |
| 04:46:32 | evan | wtf is that.. |
| 04:46:36 | evan | did I do that? |
| 04:46:37 | evan | hm. |
| 04:46:49 | evan | this is probably my fault |
| 04:46:50 | evan | one sec. |
| 04:46:58 | stepheneb leaves the room. | |
| 04:47:18 | evan | yep |
| 04:47:22 | evan | totally my fault |
| 04:47:24 | evan | commit that! |
| 04:47:46 | evan | i'm bad and I don't think there is a spec for that peephole opt |
| 04:48:45 | libc | evan, ok, the 3rd thing I wanted to talk about is object_size… |
| 04:48:49 | evan | ok |
| 04:48:53 | evan | go for it |
| 04:50:14 | libc | evan, it's slow, I did this thing: http://pastie.org/438059 … Make virtual call only for Tuple, VariableScope and so on |
| 04:50:45 | libc | evan, it's twice as fast, but still slow… maybe return size in objectheader? |
| 04:50:55 | evan | how slow? |
| 04:50:59 | evan | how did you measure it? |
| 04:51:02 | libc | shark |
| 04:51:21 | libc | 6% before, 2.9 after |
| 04:51:25 | evan | ok |
| 04:51:27 | libc | it's with JIT |
| 04:51:28 | evan | let me read your patch |
| 04:51:35 | evan | what about without JIT? |
| 04:51:37 | evan | should be the same |
| 04:51:49 | libc | without JIT run_interpreter at the top ^_^ |
| 04:52:06 | libc | because it does all work… with JIT gc at the top |
| 04:52:35 | evan | i'm confused. |
| 04:52:42 | evan | i don't know why the JIT matters at all for this patch |
| 04:53:23 | libc | evan, with JIT you have got a lot of function calls with small percentage, without you have got 1 function call, it's the only thing… |
| 04:53:41 | evan | your patch is for object_size() |
| 04:53:47 | libc | yes |
| 04:53:52 | libc | gc does next_object |
| 04:53:54 | evan | which is completely unrelated to whether the JIT or interpreter are used |
| 04:53:55 | libc | which uses object_size |
| 04:54:05 | evan | let me finish reading the patch |
| 04:54:30 | libc | it has the other optimization, (about promted_insert), ignore that for now ^_^ |
| 04:58:22 | evan | ok, all read. |
| 04:58:25 | evan | looks good |
| 04:58:33 | evan | you've moved the type sizes into a static table |
| 04:58:47 | evan | so we don't have to go from type_id => TypeInfo object => object_size |
| 04:58:51 | libc | yes, for inlining and elliminate virtual call |
| 04:58:58 | evan | sure, that makes sense |
| 04:59:10 | evan | you might as well leave instance_size on TypeInfo anyway |
| 04:59:15 | evan | no problem having it 2 places |
| 04:59:31 | evan | maybe make the data member instance_size_ |
| 04:59:38 | evan | and have instance_size() just return it |
| 04:59:39 | libc | but I still have virtual call for tuples, variable_scopes etc |
| 05:00:03 | evan | yeah |
| 05:00:07 | evan | you can't get around that |
| 05:00:11 | evan | if it's not a virtual call |
| 05:00:13 | evan | it would be a switch |
| 05:00:19 | evan | which would have the same performance probably |
| 05:00:25 | evan | we could play with that as a next step |
| 05:00:28 | evan | after you commit this |
| 05:01:56 | evan | otherwise, looks fine |
| 05:02:18 | evan | put some comments about the new field_extract.rb code that checks TypeInfo |
| 05:02:58 | evan | also, you can use use inline instead of __inline__ |
| 05:03:39 | libc | __inline__ is a hack for now… It only works because this method is not used anywhere but baker.cpp |
| 05:04:02 | evan | right |
| 05:04:10 | evan | thats fine |
| 05:04:37 | evan | so, my comments make sense? |
| 05:05:19 | libc | I guess so |
| 05:06:23 | qbproger leaves the room. | |
| 05:07:09 | evan | you're clearly better at STL that me |
| 05:07:16 | evan | looking at how you rearranged promoted |
| 05:07:18 | imajes enters the room. | |
| 05:07:23 | evan | to avoid having to reallocate it |
| 05:08:20 | libc | vector is a wrapped array, you can do things like *(&a[0] +3) and get to a[3] |
| 05:08:41 | libc | it's based on that assumption (which is true if compiler follows C++ standard) |
| 05:09:32 | evan | cool |
| 05:09:41 | evan | if it works, it's fine by me! |
| 05:16:45 | joachimm enters the room. | |
| 05:39:35 | ezmob leaves the room. | |
| 05:44:48 | joachimm leaves the room. | |
| 06:15:13 | asap18 leaves the room. | |
| 06:37:08 | benburkert leaves the room. | |
| 07:10:56 | krawek leaves the room. | |
| 07:11:03 | scoopr_ enters the room. | |
| 07:16:00 | blowmage leaves the room. | |
| 07:20:00 | scoopr leaves the room. | |
| 07:29:37 | binary42 leaves the room. | |
| 08:11:06 | joachimm enters the room. | |
| 08:42:39 | headius leaves the room. | |
| 08:47:47 | headius enters the room. | |
| 08:56:58 | joachimm leaves the room. | |
| 08:59:01 | headius leaves the room. | |
| 09:00:52 | SoreGums enters the room. | |
| 09:02:20 | SoreGums leaves the room. | |
| 09:02:33 | SoreGums enters the room. | |
| 09:44:11 | botanicus enters the room. | |
| 09:45:54 | ennui enters the room. | |
| 10:01:40 | imajes_ enters the room. | |
| 10:03:16 | imajes leaves the room. | |
| 10:09:06 | lstoll enters the room. | |
| 10:11:29 | gnufied enters the room. | |
| 10:27:22 | VVSiz enters the room. | |
| 10:59:34 | naeu enters the room. | |
| 11:07:29 | nari leaves the room. | |
| 11:16:29 | jero5_ enters the room. | |
| 11:18:22 | Maledictus enters the room. | |
| 11:21:46 | mvr enters the room. | |
| 11:28:29 | jero5 leaves the room. | |
| 11:35:23 | gnufied leaves the room. | |
| 11:52:47 | mediogre enters the room. | |
| 12:01:47 | gnufied enters the room. | |
| 12:16:30 | chris2 enters the room. | |
| 12:44:56 | gnufied leaves the room. | |
| 12:59:42 | mvr leaves the room. | |
| 13:06:13 | imajes_ leaves the room. | |
| 13:09:30 | mediogre leaves the room. | |
| 13:16:35 | lstoll leaves the room. | |
| 13:22:56 | lstoll enters the room. | |
| 13:30:52 | lstoll leaves the room. | |
| 13:33:52 | asap18 enters the room. | |
| 13:57:56 | gavinstark enters the room. | |
| 14:19:49 | asap18 leaves the room. | |
| 14:37:01 | mutle leaves the room. | |
| 14:38:14 | mutle enters the room. | |
| 14:40:36 | gnufied enters the room. | |
| 14:47:09 | yipstar enters the room. | |
| 14:49:41 | joearnold enters the room. | |
| 14:50:05 | macournoyer enters the room. | |
| 14:56:14 | rue | libc: vector is specified to have that behaviour regardless of whether it is backed by an array, so far as I understand |
| 14:57:29 | rue | libc: Should use C++ casts rather than C style.. other than that, looks nice |
| 14:58:33 | stepheneb enters the room. | |
| 15:09:45 | botanicus leaves the room. | |
| 15:12:09 | outerim enters the room. | |
| 15:14:25 | therealadam enters the room. | |
| 15:21:35 | wmoxam enters the room. | |
| 16:01:31 | enebo enters the room. | |
| 16:15:40 | stepheneb leaves the room. | |
| 16:24:06 | jashmenn enters the room. | |
| 16:25:30 | antares_ enters the room. | |
| 16:27:04 | binary42 enters the room. | |
| 16:41:11 | joearnold leaves the room. | |
| 16:41:13 | blowmage enters the room. | |
| 17:11:23 | lopex enters the room. | |
| 17:13:31 | ennui leaves the room. | |
| 17:14:51 | joearnold enters the room. | |
| 17:16:58 | naeu leaves the room. | |
| 17:17:16 | headius enters the room. | |
| 17:18:32 | libc | rue, Are you talking about “(int)LastObjectType” ? |
| 17:34:36 | radarek enters the room. | |
| 17:37:56 | tilman | evan: you around? |
| 17:40:08 | stepheneb enters the room. | |
| 17:50:13 | joearnol_ enters the room. | |
| 17:50:32 | ezmob enters the room. | |
| 17:50:47 | nicksieger enters the room. | |
| 17:51:34 | joearnold leaves the room. | |
| 17:52:25 | brixen | tilman: my 2 cents on the rbc format.. |
| 17:52:45 | tilman | shoot |
| 17:52:48 | brixen | using rba's will give much better compression than ad hoc tweaks to the format |
| 17:52:56 | brixen | if size is an issue |
| 17:53:19 | brixen | for speed, the unmarshal code could probably be changed to read into a buffer and iterate with a pointer |
| 17:53:41 | tilman | shouldn't trivial size-related cleanups be done before going to compression? |
| 17:53:48 | brixen | typical installations shouldn't need to worry about size of runtime/ |
| 17:54:07 | brixen | not imo if it complicates the format |
| 17:54:24 | brixen | it's pretty simple and the compression will do better than your hand tweaking I think |
| 17:54:44 | brixen | kind of the old asm programmer complaining about a compiler optimizer |
| 17:54:45 | brixen | ;) |
| 17:55:11 | brixen | anyway, that's my thought |
| 17:55:16 | brixen | evan will decide |
| 17:55:23 | tilman | calling this tweaking is n't even fair imo |
| 17:55:27 | brixen | btw, almost have the profiler working |
| 17:55:39 | brixen | well, what would you call it? |
| 17:55:41 | tilman | the current format is really naive |
| 17:55:46 | brixen | and simple |
| 17:55:49 | tilman | yes |
| 17:55:56 | brixen | we used to have a complicated one |
| 17:56:05 | brixen | evan made it simpler |
| 17:56:16 | tilman | aha, so simplicity was a design goal |
| 17:56:50 | brixen | have you built rba's to check the compression? |
| 17:56:59 | brixen | curious what it would be |
| 17:57:08 | tilman | what is rba these days? |
| 17:57:15 | tilman | gzip'ed ar-chive? |
| 17:57:18 | tilman | or still zip? |
| 17:57:19 | brixen | yeah |
| 17:57:27 | brixen | mm, I think gzip |
| 17:57:30 | brixen | I'd have to check |
| 17:57:49 | brixen | man, fscking neighbors moved in with cats that screamm all day |
| 17:58:01 | brixen | I'm thinking I should get a dog :) |
| 18:02:16 | blowmage_ enters the room. | |
| 18:02:56 | macournoyer leaves the room. | |
| 18:12:07 | rue | libc: next_object() |
| 18:13:05 | libc | oh, ok, I'll change |
| 18:13:08 | jero5_ leaves the room. | |
| 18:13:54 | rue | rbas would only affect the runtime size, though, they still need to be uncompressed so it does not have any runtime benefit |
| 18:15:11 | rue | libc: Just one of my pet issues :) The other one is const correctness but that is a losing battle right now :P |
| 18:16:16 | macournoyer enters the room. | |
| 18:17:32 | jero5 enters the room. | |
| 18:18:01 | blowmage leaves the room. | |
| 18:30:58 | krawek enters the room. | |
| 18:34:51 | koei enters the room. | |
| 18:37:38 | koei | hi, what is the rubinius' jit strategy? using llvm's or write its own one? |
| 18:38:21 | libc | it had code for both, now both are broken |
| 18:38:59 | antares | koei: so apparently, strategy is to get it working again first ;) |
| 18:39:30 | koei | :). get it working again on which? |
| 18:39:34 | brixen | koei: we're still exploring strategies, but likely we will have both a simple JIT and a more complex one |
| 18:39:45 | brixen | the simple jit is being worked on now |
| 18:40:24 | antares | brixen: have two to eventually implement/switch to more complex, or have jit pluggable? |
| 18:40:51 | brixen | I suppose you could have it be pluggable |
| 18:40:59 | therealadam leaves the room. | |
| 18:41:10 | brixen | the idea is to have a simple one that is fast but doesn't do many opts |
| 18:41:23 | brixen | and a more complex one that can run in another thread and generate better code |
| 18:41:29 | brixen | for hot sections |
| 18:41:43 | rue | The idea with the simple JIT (and I think it has worked) was that it would allow better to figure out how the rest of the codebase fits with generating machine code |
| 18:42:27 | antares | gotcha |
| 18:42:28 | antares | thanks |
| 18:43:18 | therealadam enters the room. | |
| 18:46:21 | koei leaves the room. | |
| 18:47:14 | oda enters the room. | |
| 18:47:14 | boyscout | Call to_ary on splats - c9502b9 - Eugene Pimenov |
| 18:47:14 | boyscout | Rewire JIT - 51a2393 - Eugene Pimenov |
| 18:47:14 | boyscout | Baker GC optimizations. - 9735a95 - Eugene Pimenov |
| 18:53:11 | joearnold enters the room. | |
| 18:53:43 | joearnol_ leaves the room. | |
| 18:56:33 | boyscout | CI: 9735a95 success. 1438 files, 7105 examples, 23402 expectations, 0 failures, 0 errors |
| 18:56:44 | joearnol_ enters the room. | |
| 18:58:04 | joearnold leaves the room. | |
| 18:58:24 | botanicus enters the room. | |
| 19:01:44 | antares_ leaves the room. | |
| 19:08:16 | joachimm enters the room. | |
| 19:08:56 | mediogre enters the room. | |
| 19:11:06 | botanicus leaves the room. | |
| 19:19:07 | ennui enters the room. | |
| 19:29:43 | gnufied leaves the room. | |
| 19:30:40 | libc leaves the room. | |
| 19:39:31 | libc enters the room. | |
| 19:44:27 | joachimm_ enters the room. | |
| 19:46:15 | joearnold enters the room. | |
| 19:47:42 | joearnol_ leaves the room. | |
| 19:52:05 | imajes enters the room. | |
| 19:57:51 | mediogre leaves the room. | |
| 19:59:23 | yipstar leaves the room. | |
| 19:59:35 | chris2 leaves the room. | |
| 20:00:03 | joearnold leaves the room. | |
| 20:00:27 | joearnold enters the room. | |
| 20:00:52 | joachimm leaves the room. | |
| 20:11:34 | mdalessio enters the room. | |
| 20:14:32 | oda leaves the room. | |
| 20:28:00 | stepheneb_ enters the room. | |
| 20:33:44 | botanicus enters the room. | |
| 20:35:09 | joachimm enters the room. | |
| 20:35:41 | botanicus leaves the room. | |
| 20:35:46 | botanicus enters the room. | |
| 20:41:43 | stepheneb leaves the room. | |
| 20:47:39 | twbray enters the room. | |
| 20:48:36 | joachimm_ leaves the room. | |
| 20:50:32 | devinus enters the room. | |
| 20:54:59 | blowmage leaves the room. | |
| 20:55:29 | blowmage enters the room. | |
| 21:01:32 | joearnold leaves the room. | |
| 21:02:20 | joearnold enters the room. | |
| 21:02:44 | Maledictus leaves the room. | |
| 21:19:06 | imajes leaves the room. | |
| 21:21:23 | benny leaves the room. | |
| 21:39:50 | yipstar enters the room. | |
| 22:11:27 | radarek leaves the room. | |
| 22:15:15 | scoopr leaves the room. | |
| 22:16:26 | imajes enters the room. | |
| 22:19:41 | scoopr enters the room. | |
| 22:20:49 | max enters the room. | |
| 22:21:50 | devinus leaves the room. | |
| 22:27:33 | mutle_ enters the room. | |
| 22:29:16 | outerim leaves the room. | |
| 22:30:39 | macournoyer leaves the room. | |
| 22:32:01 | malumalu | I guess it is not intended to get one error when building with rake? |
| 22:35:02 | outerim enters the room. | |
| 22:38:17 | ennui leaves the room. | |
| 22:40:22 | scoopr leaves the room. | |
| 22:40:28 | scoopr enters the room. | |
| 22:44:19 | mutle leaves the room. | |
| 22:44:47 | rudebwoy_ leaves the room. | |
| 22:48:29 | gavinstark leaves the room. | |
| 22:50:17 | wmoxam leaves the room. | |
| 22:50:59 | joachimm leaves the room. | |
| 22:52:08 | rudebwoy enters the room. | |
| 22:53:08 | rue | malumalu: No, should not get errors. Are you using `rake` or `rake build`? |
| 22:54:55 | malumalu | just rake |
| 22:55:58 | malumalu | btw: not an build error, one test fails with an error |
| 22:58:23 | brixen | malumalu: what platform are you on? |
| 22:59:10 | malumalu | linux |
| 22:59:16 | malumalu | ubuntu 8.10 |
| 22:59:23 | brixen | ok, what test is it? |
| 22:59:48 | brixen | btw, 32 or 64bit? |
| 23:00:02 | malumalu | /spec/frozen/core/process/setrlimit_spec.rb:77 |
| 23:00:06 | malumalu | 32 |
| 23:00:11 | brixen | ah, a spec |
| 23:00:21 | brixen | I though it was a vm test |
| 23:00:25 | malumalu | eh, yeah, sorry |
| 23:00:27 | brixen | that spec is odd |
| 23:00:54 | brixen | anything out-of-ordinary with your system? |
| 23:01:06 | malumalu | I don't think so |
| 23:01:12 | brixen | I have 8.10 on a virtualbox vm, checking if I get that failure |
| 23:01:24 | malumalu | ok |
| 23:03:11 | brixen | that value is pulled in via the FFI const generator |
| 23:03:17 | brixen | hm |
| 23:03:57 | gramos enters the room. | |
| 23:05:34 | lstoll enters the room. | |
| 23:06:29 | qbproger enters the room. | |
| 23:10:08 | stepheneb_ leaves the room. | |
| 23:13:25 | therealadam_ enters the room. | |
| 23:13:25 | therealadam leaves the room. | |
| 23:14:18 | nari enters the room. | |
| 23:16:18 | brixen | malumalu: ok, finally built |
| 23:16:24 | brixen | I don't get a failure |
| 23:16:26 | malumalu | I think it doesn't help to know that max=0 and lim=140741783322623, does it? |
| 23:16:28 | malumalu | mh |
| 23:16:38 | brixen | could you pastie the output |
| 23:17:36 | brixen | also, could you pastie runtime/platform.conf |
| 23:21:06 | malumalu | http://gist.github.com/90978 |
| 23:27:00 | asap18 enters the room. | |
| 23:28:52 | malumalu | brixen: doesn't help much, I guess? |
| 23:28:55 | brixen | malumalu: hrm, I have the same values in platform.conf |
| 23:29:22 | malumalu | hm |
| 23:29:25 | brixen | malumalu: just a vanilla 8.10 install? |
| 23:29:57 | malumalu | no, but I didn't mess around with any system files ^^ |
| 23:30:11 | brixen | hm |
| 23:32:52 | gramos leaves the room. | |
| 23:32:54 | brixen | malumalu: I get 140741783322623 for both max and lim in that spec |
| 23:33:06 | brixen | you could try poking a bit to see why you get max=0 |
| 23:33:47 | brixen | er, wait, I get 140737488388096 rather |
| 23:35:11 | malumalu | brixen: hm, I'll have a look |
| 23:37:45 | malumalu | brixen: you have the same value for rbx.platform.process.RLIMIT_MEMLOCK? |
| 23:38:27 | brixen | checking.. |
| 23:38:46 | malumalu | mine is 8 |
| 23:39:00 | chris__ enters the room. | |
| 23:39:15 | brixen | yep |
| 23:39:28 | malumalu | hm |
| 23:40:29 | brixen | yeah, just rechecked platform.conf, all the rbx.platform.process are the same as you gist |
| 23:40:42 | malumalu | strange |
| 23:42:16 | therealadam leaves the room. |