<< 26-09-2013 >>
00:08:54*MFlamer quit (Quit: Page closed)
00:48:14reactormonkAraq, found the seq stuff?
01:02:09*fowl quit (Quit: Leaving)
01:15:49*DAddYE quit (Remote host closed the connection)
01:17:02reactormonkwhy exactly does map(@[1,2,3], inc)
01:17:06reactormonk... fail in nimrod i?
01:17:22reactormonkstdin(1, 14) Error: instantiate 'inc' explicitely
01:31:06reactormonkanyway, I don't really get the instantiate 'echo' explicitely
01:31:29reactormonkoh, good old ()
01:32:51reactormonkis it just me or is $ in js a noop?
01:33:08reactormonksomehow the generated code doesn't look like it should
01:35:56*q66 quit (Quit: Leaving)
01:38:33reactormonk echo($a)
01:38:35reactormonktranslates to
01:38:45reactormonk rawEcho(a_34332);
01:38:47reactormonkbut should be
01:39:17reactormonk rawEcho(cstrToNimstr(a_34332));
02:39:43*brson quit (Ping timeout: 260 seconds)
02:41:27*brson joined #nimrod
03:03:03*brson quit (Ping timeout: 252 seconds)
03:05:01*brson joined #nimrod
03:19:14*DAddYE joined #nimrod
03:23:39*DAddYE quit (Ping timeout: 248 seconds)
03:27:52*MFlamer joined #nimrod
03:37:54reactormonkany way to do controlled effect invocation aka lazy aka FP?
03:43:42*capisce quit (Read error: Connection reset by peer)
03:49:36*capisce joined #nimrod
03:56:04*ltbarcly joined #nimrod
04:09:19*OrionPK quit (Quit: Leaving)
04:12:18MFlamerreactormonk: not sure what you mean. like with a monad?
04:21:04*DAddYE joined #nimrod
04:26:34*brson quit (Quit: leaving)
04:27:55*DAddYE quit (Ping timeout: 259 seconds)
04:49:30*dyu joined #nimrod
04:49:34*zhtx joined #nimrod
05:02:53*XAMPP quit (Quit: Drink all the Booze; Hack all the Things; Kill all the Humans;)
05:15:58Araqreactormonk: lazy == pass a function/closure to be evaluated
05:16:15AraqrawEcho should be correct
05:16:35Araqand no luck, didn't find the issue about $ for seq
05:23:52*DAddYE joined #nimrod
05:28:06*DAddYE quit (Ping timeout: 245 seconds)
05:32:27*MFlamer quit (Ping timeout: 250 seconds)
05:47:33*zhtx quit (Quit: Leaving.)
05:56:00*NewGuy joined #nimrod
05:56:31*ltbarcly quit (Quit: Computer has gone to sleep.)
05:56:58NewGuyAnybody here?
06:01:15*NewGuy quit (Quit: Page closed)
06:24:44*DAddYE joined #nimrod
06:31:40*DAddYE quit (Ping timeout: 264 seconds)
06:37:32*Associat0r joined #nimrod
06:37:32*Associat0r quit (Changing host)
06:37:32*Associat0r joined #nimrod
06:42:45*filwit joined #nimrod
06:43:28*filwit quit (Client Quit)
06:53:52*Araq_ joined #nimrod
07:20:24*DAddYE joined #nimrod
07:57:25*gsp joined #nimrod
09:10:08*DAddYE quit (Remote host closed the connection)
09:10:35*DAddYE joined #nimrod
09:14:46*DAddYE quit (Ping timeout: 246 seconds)
09:30:26*Araq_ quit (Read error: Connection timed out)
09:31:07*Araq_ joined #nimrod
09:41:20*shodan45_ joined #nimrod
09:41:48*shodan45 quit (Ping timeout: 245 seconds)
09:52:24*shodan45_ quit (Read error: Connection reset by peer)
09:55:35*shodan45_ joined #nimrod
10:03:26*EXetoC joined #nimrod
10:10:24*shodan45_ quit (Ping timeout: 252 seconds)
10:11:10*DAddYE joined #nimrod
10:12:06*shodan45_ joined #nimrod
10:16:04*ltbarcly joined #nimrod
10:16:23*shodan45_ quit (Ping timeout: 245 seconds)
10:16:25*ltbarcly quit (Client Quit)
10:17:26*DAddYE quit (Ping timeout: 240 seconds)
10:18:04*shodan45_ joined #nimrod
10:21:17*DAddYE joined #nimrod
10:25:28*DAddYE quit (Ping timeout: 246 seconds)
10:36:25*q66 joined #nimrod
11:14:24*DAddYE joined #nimrod
11:16:09*Boscop joined #nimrod
11:19:57*Boscop is now known as M4chine
11:20:04*M4chine left #nimrod (#nimrod)
11:20:26*DAddYE quit (Ping timeout: 245 seconds)
11:38:52*gsp quit (Ping timeout: 264 seconds)
12:17:30*DAddYE joined #nimrod
12:23:16*q66_ joined #nimrod
12:23:50*DAddYE quit (Ping timeout: 240 seconds)
12:26:48*q66 quit (Ping timeout: 252 seconds)
12:42:58*Araq_ quit (Quit: ChatZilla 0.9.90.1 [Firefox 23.0.1/20130814063812])
13:03:42reactormonkAraq, well, it produces the wrong output, as a is a cstring
13:13:24*EXetoC quit (Ping timeout: 257 seconds)
13:20:38*DAddYE joined #nimrod
13:26:56*DAddYE quit (Ping timeout: 260 seconds)
13:34:16*EXetoC joined #nimrod
13:50:26*Araq_ joined #nimrod
13:57:27*Araq_ quit (Quit: ChatZilla 0.9.90.1 [Firefox 23.0.1/20130814063812])
14:09:15*zhtx joined #nimrod
14:11:20*gsp joined #nimrod
14:23:51*DAddYE joined #nimrod
14:29:43*ltbarcly joined #nimrod
14:29:49*ltbarcly quit (Client Quit)
14:30:24*q66_ is now known as q66
14:30:28*DAddYE quit (Ping timeout: 246 seconds)
14:56:54*filwit joined #nimrod
15:06:29dom96hello
15:09:33*shodan45_ quit (Quit: Konversation terminated!)
15:10:19*shodan45 joined #nimrod
15:16:35filwithi dom
15:17:38filwitdom96: so i thought of a better name for a Nimrod game engine: hymn
15:17:50filwitdom96: since nimrod is a king from scripture
15:18:15filwitdom96: http://reign-studios.net/philipwitte/hymn/
15:18:45filwitjust a mock up of a logo mostly
15:20:05dom96that looks sexy
15:20:51dom96not sure about the name though, feels a bit too religious, but *shrug*
15:22:25filwityeah, that's why i added the red, haha
15:22:36filwitwanted it to be a play on things
15:22:44filwitsay 'hymn' but looks devilish
15:22:45dom96hymn from hell heh?
15:22:50filwityeah, lol
15:25:03*zhtx quit (Quit: Leaving.)
15:26:53*DAddYE joined #nimrod
15:33:07*DAddYE quit (Ping timeout: 246 seconds)
15:37:04*gsp quit (Ping timeout: 264 seconds)
15:39:55*MFlamer joined #nimrod
15:48:45*gsp joined #nimrod
15:54:12*zhtx joined #nimrod
15:55:31Araqhttp://www.reddit.com/r/programming/comments/1n3su5/want_to_learn_ada_now_you_can_free_online_courses/
15:55:51AraqRust is mentioned, Nimrod of course not
15:55:59Araqso ... do something about it
15:58:46NimBotAraq/Nimrod master 4b179da Araq [+0 ±1 -0]: documentation improvements
15:58:46NimBotAraq/Nimrod master 2119c66 Araq [+5 ±8 -0]: the compiler can now deal with multiple modules of the same name
15:59:57p0nceRust hype is at an all-time high
16:00:59*gsp quit (Read error: Operation timed out)
16:01:28Araqand yet Rust copied no feature from Ada afaik and Nimrod copied every good feature of it *cough*
16:01:39Araqincluding subranges
16:06:40*zhtx quit (Quit: Leaving.)
16:08:52MFlamerAraq: I'm having some trouble locating the source of a bug in my table. Do you have any recommendations? How might --stacktrace:on help?
16:09:37Araqah the fun with threading bugs
16:09:59MFlamerYes, but this happens even with a single thread.
16:10:08Araqoh cool
16:10:10dom96Araq: You could do something about it too you know
16:10:23MFlamerUnder certain conditions it just quits
16:10:38Araqwithout a stack trace?
16:10:55MFlameryes
16:11:06Araqtry --debuginfo and run it under gdb
16:11:16MFlamerok
16:12:32Araqdom96: if I do everything myself, I might as well start implementing only the features that *I* need
16:14:15*gsp joined #nimrod
16:15:31Araqbbl
16:16:39MFlamerI'm not asking anyone to the work for me, just steer me in the right direction
16:17:09MFlamerI give gdb a shot. Never used it before
16:27:51*ltbarcly joined #nimrod
16:28:30*ltbarcly quit (Client Quit)
16:34:23dyuAraq: its probably not the feature, its the backers
16:35:49dyugolang has the least features, (less is more mantra by their core devs) ... but their community is growing larger by the day
16:36:45dyuit also has to do that it is already 1.0x, hence usable in production (don't quote me on that)
16:51:19*DAddYE joined #nimrod
16:53:54filwitNimrod is still the language "you never knew you wanted". Google and Mozilla actively promote Go & Rust (not sure about rust actually), and so more people are aware of their respective benefits
16:54:46filwitNimrod doesn't (yet) have many people promoting it in different circles, so it's no wounder it doesn't get mentioned much
16:55:05filwiti think Araq's talk should help Nimrod see more limelight
16:56:20dyuThe problem with nimrod is that the design is good, but the T/P prefix stuff is ugly :-) A small like that turns of newbies
16:56:41dyuA small detail*
16:56:58filwiti think a lot of people, like I almost did, will pass up Nimrod based on aesthetics alone (i don't ever care about Basic-style languages), unless they can see some presentation on what it's real power is
16:57:09filwityes dyu, i'm with you there
16:57:17filwitthat was my first, and still biggest complaint
16:57:21dyurust is good, but their compiler right now is friggin slow
16:57:41dyufilwit: yea me too, but I got used to it the more I keep on looking at the source
16:57:59filwitthe T/P thing feels so.. C-style, or 1980s
16:58:04dyuthe problem with golang? c-ffi overhead :-) https://github.com/dyu/ffi-overhead#results
16:58:27filwiti haven't used either go nor rust much
16:58:50filwitrust looks interesting, but i hate the syntax
16:58:52p0nceI tried Rust and was underwhelmed
16:59:28dyup0nce: couldn't try rust because it takes hours to bootstrap
16:59:38p0nceI wanted something like template<int> stuff
16:59:53p0ncebut you have to go the macro way from the start then
17:01:35*brson joined #nimrod
17:01:42dyuhere's a rust guy
17:01:45dyu:-)
17:03:07dom96oh cool. Nice benchmarks.
17:03:14dom96Didn't know go had an ffi overhead
17:03:41p0nceJNI-style
17:03:50dyuI knew it had an overhead, but didn't know it was that bad
17:03:56dom96I was going to try Rust too, but after 40 minutes of it bootstrapping I just gave up.
17:04:10dyui thought go would be faster than java (because everybody says jni is slooow)
17:04:17dyuin c ffi
17:04:23filwitNimrod is slower than C by 5 milliseconds in that test... that is a unforgivable sin
17:05:07*fowl joined #nimrod
17:05:07filwiti will now complain to Araq for the next two weeks until he dedicates all his time to optimizing the compiler
17:05:10dom96I bet the time between the C and Nimrod version fluctuates up and down a lot.
17:05:18dyudom96: indeed
17:05:26dyusometimes nimrod is faster
17:05:44dyuprobably some cpu warmup thingy
17:05:47dyudunno really
17:05:52filwitit's great to see how close they are though
17:06:13p0ncewe have a program at work to compare two version of our executable, basically perform a Welch test http://en.wikipedia.org/wiki/Welch%27s_t_test
17:06:16dom96I would do multiple runs and do an average.
17:06:33filwitof course, it's a very basic test
17:06:37p0ncebut yeah average of 10 run is usually sufficient
17:07:12filwitand you can't really benchmark that...
17:07:24filwitcause the compilers should optimize that loop away i think.
17:07:37filwitwait nevermind
17:07:48p0nceimpressive how Go manage to do consistently worse than anything that exist
17:07:48dyuI've run it multiple times, the difference is 1% ... e.g. The results are as stable as they get
17:10:24dyuwell, except for scripting languages
17:10:52dom96Any ideas why Go's overhead is so large?
17:10:56dyumostly ruby/python/node people have defected to go
17:11:21dyuits the gc locking/barrier
17:12:12dyujava has barriers too on jni calls, but I'm not sure what extra things go is doing that makes it that slow
17:17:15filwitGoogle's Go has built-in library c-ffi logging to report to the NSA..
17:17:25filwitobviously
17:20:44dyufilwit: are you still going with macros/templates to sort-of have the syntax you aimed for?
17:23:54*dyu thinks they mess up the stack trace info
17:25:56filwitdyu: what do you mean exactly?
17:26:16filwitdyu: yes, i have some plans to use Nimrod for a project I'm working on
17:27:17filwitdyu: but that will be in a couple of months. I'm currently playing with Nimrod's macro's to see what's capable and kicking around the idea of making a small game engine with it if I'm happy enough with what i can come up with
17:27:19*gsp quit (Ping timeout: 260 seconds)
17:27:45dyuhttps://gist.github.com/PhilipWitte/6648742 <-- you're gonna implement that with macros right?
17:28:32filwityes, actually someone already mostly did. but i'm making something similar, yes
17:29:08filwitprototype design: http://reign-studios.net/philipwitte/hymn/
17:29:09filwit:)
17:29:25filwitbtw, that syntax is wrong in a few places
17:29:36dyucool
17:29:44dyubut doesn't the stack trace get messed up?
17:29:52dyuwhen you get an exception
17:29:59filwithmmm... idk actually
17:30:00dyuyou won't be able to pin point the exact line
17:30:35fowlno
17:30:41filwitnimrod may have features to allow me to pass that information down the line
17:30:41filwiti'll have to ask Araq more about that sort of thing later
17:30:42dyuoh ok
17:31:09dyufowl: care to elaborate :-)
17:32:03filwitfowl: no to me or to dyu?
17:32:04fowlno
17:32:07filwitlol
17:32:17fowldyu, it just works, idk how
17:32:26filwitmagic
17:32:30dyucool
17:32:38fowlthe line/column info is from where the source is when its read, not where it ends up
17:32:50filwitthat {.magic.} pragma you see in the compiler internals.. that's real magic
17:34:28dyufilwit: who's andrew
17:39:32*gsp joined #nimrod
17:40:07*q66_ joined #nimrod
17:40:08*q66 quit (Read error: Connection reset by peer)
17:40:26*q66_ is now known as q66
17:44:28filwitdyu: my brother
17:44:40filwitdyu: buy why are you asking? where are you reading that?
17:45:17dyuhttp://www.reign-studios.net/
17:45:33dyufrom your prototype design
17:46:06filwitah, right
17:46:48filwityes, that's our new Unity plugins site
17:51:01reactormonkno curry, right?
17:51:10reactormonkAraq, problem found. wrong type annotation in the import.
17:51:40EXetoCmmmh, curry
17:52:39filwityeah i'm hungry
17:52:48filwittime for lunch i think
17:59:51*XAMPP joined #nimrod
18:01:56reactormonkhttp://pastebin.com/JzSHjmn2 anyone an idea why the result is always false?
18:05:03reactormonk(a == currentActivity) is
18:05:09reactormonk(a_34346 < workspace.currentActivity)
18:05:11reactormonkWTF.
18:05:29Araqlol. typo in the codegen?
18:05:41reactormonklikely.
18:06:07reactormonk ["", "", "($1 == $2)", "($1 == $2)"], # EqCString
18:06:08reactormonklooks good
18:06:21Araqwell you should do that:
18:06:30AraqmEqCString: [ ... ]
18:06:38Araqthis way the compiler checks the order
18:07:05reactormonkwhere?
18:07:15Araqin the line you just quoted?
18:07:28Araqbut you need to do that for the whole array
18:07:59reactormonk:-(
18:08:12AraqMFlamer: depending on your bug, you can also try ENDB and in particular its watchpoint feature. It's quite good.
18:10:48reactormonkwhere is ["", "", "($1 == $2)", "($1 == $2)"], # EqCString
18:10:49reactormonkused?
18:11:33Araqin line 484
18:11:53reactormonkwhich file?
18:11:59Araqjsgen.nim
18:12:32reactormonk arithAux(p, n, r, op, jsOps | luaOps)
18:12:36reactormonk484
18:12:59AraqjsOps is the name of the array
18:13:11reactormonkoh.
18:15:59MFlamerAraq: ok, will do
18:16:22AraqI can also have a look at your code if you want to
18:19:54reactormonkAraq, I still don't get your mEqCString statement
18:20:26reactormonkahh
18:20:31reactormonknow I do.
18:22:27MFlamerAraq: If you were talking to me, I invited you to the repo https://bitbucket.org/mflamer/nimutils/src/1dfa1523ab04e421c9e4403554376b2c0ce4a606/LockFreeHash.nim?at=master
18:23:08AraqMFlamer: yeah; looking at your code right now
18:23:37MFlamerthe code will build and the test will run as expected, if you change the "numTests" from 1100 to 1200 and run, it'll freeze
18:24:18MFlamerThe code may be a little messy....not quite ready for review yet
18:26:02reactormonkI copy-pasted the magic together and ran echo(repr(jsOps[mEqCString]))
18:26:12reactormonk... test.nim(164, 15) Error: index out of bounds
18:26:25Araqthe code is ugly ;-)
18:27:20Araq1. you should use 'ptr array[0..10_000_000, TEntry]' instead of PEntry for 'data'
18:27:46MFlamerok
18:28:04Araq2. your 0x000 stuff assumes 32 bit ?
18:28:10reactormonkfull source: http://sprunge.us/ZCSC
18:28:29MFlamerfor now, yes
18:28:42MFlamerI'll make it portable...
18:28:49Araqreactormonk: try without 'repr'
18:29:56reactormonkbetter.
18:30:35MFlamerI also stash a couple flags in the low bits, assuming alignment and endianness
18:30:52reactormonkecho(jsOps[mEqCString][2])
18:30:58reactormonkstill gives me an OOB
18:34:47AraqMFlamer: have you triedA ATOMIC_FULL or however it's called instaed of ATOMIC_RELAXED?
18:35:06Araq'relaxed' sounds fishy
18:35:27*io2 joined #nimrod
18:36:14Araqio2: http://nimrod-code.org/talk01/slides.html#%281%29
18:36:51io2Araq: hi, thanks!
18:36:55io2that's the full talk?
18:36:57io2awesome
18:37:17Araqthat are the slides that I used, yes
18:37:19dyuAraq: finally
18:37:22MFlamerIn some places, I am going for synchronization with ATOMIC_SEQ_CST(full barrier). I assumed it shouldnt matter for a single threaded problem...but I could try full barrier everywhere and see what happens
18:37:34MFlamercool, on the slides
18:39:36io2tweeted
18:39:38io2and g+ in now
18:39:46dyuTaintedString
18:41:18io2ok, done
18:41:32io2Araq: looking forward to more material :)
18:41:41Araqgah ...
18:41:44Araqoh well
18:44:12AraqMFlamer: he, true, it shouldn't matter for single threaded code
18:45:05reactormonkAraq, left ccgexprs, right jsgen http://pastebin.com/LdCqmGzA
18:45:07reactormonkgo figure
18:46:46Araqreactormonk: ah nice; and that's why the indexes need to be named
18:47:56Araqin my defense ... when I wrote that the language didn't support that feature :P
18:50:55MFlamerAraq: I have to bail for a meeting, I'll check the logs later for any comments. Thanks for your time
18:51:09fowlcool slides thanks Araq
18:51:21AraqMFlamer: alright
18:51:32*MFlamer quit (Quit: Page closed)
18:52:08*dyu quit (Quit: Leaving)
18:52:34Araqfowl: I'm not sure how comprehensible the slides are without my explanations
18:52:45Araqbut then you're the wrong guy to ask I guess ;-)
18:55:54filwitnice slides, reading..
19:01:21io2the slides are excellent, full of code samplees
19:01:29io2samples*
19:01:59io2Araq: video?
19:02:01*MFlamer joined #nimrod
19:02:10io2is there someplace where you have it perchance?
19:05:11reactormonkAraq, delete the equntracedref?
19:08:00Araqio2: dunno when the video is ready
19:08:12Araqmy bet: still a month
19:11:45io2Araq: ok, thanks
19:11:54io2I beat dom96 into tweeting this :P
19:11:55io2he he
19:14:11Araqbbl
19:19:40reactormonkAraq, yeah, you have the untraced ref in both lua and js ops, but not in the original one
19:20:17reactormonkwait, TMagic defines mEqUntracedRef
19:26:42fowlugh still no docs for basic2d/3d
19:26:54filwitbasic2d/3d?
19:27:40fowlyea
19:27:47filwiti mean.. what is it?
19:27:55filwitsome nim module i've never heard of?
19:28:02fowlyea
19:28:12fowlfor vectors/points
19:28:19fowlits new
19:28:42filwitk i'll look it up
19:28:57fowlthey're in lib/pure
19:29:02NimBotAraq/Nimrod master 2c48c36 Simon Hafner [+0 ±1 -0]: deleted EqProc so the magic aligns again
19:29:03Araqfowl: just read the sources then; nimbuild on x86 produces the new docs but it's broken
19:29:28Araqer excuse me, reactormonk?
19:29:32reactormonkoops.
19:29:55reactormonkneed to set my git remotes correctly.
19:30:05Araqyes
19:30:14Araqand fyi the build broke
19:30:26reactormonkcan you repair it to the previous state?
19:30:34Araqjoin #nimbuild to see what you did
19:31:11Araqgah, I'm not good at git, you revert please
19:31:18reactormonkrevert or delete?
19:31:27Araqrevert
19:32:16NimBotAraq/Nimrod master 0b10cd8 Simon Hafner [+0 ±1 -0]: Revert "deleted EqProc so the magic aligns again" - wrong remote.... 2 more lines
19:32:31dom96io2: how dare you!
19:32:40io2:P
19:32:50dom96heh, just kidding. It's cooler for me to retweet stuff about Nimrod from time to time anyways.
19:33:47io2I should be doing something with nimrod on my own eventually, been playing with it for some time now and I need a rather good excuse time - wise as well to put it to work on something production - worthy.
19:34:40dom96As soon as I am done with my University applications (and GTA 5) I should have some more time for Nimrod.
19:34:51dom96My github contributions graph looks pretty bad currently.
19:35:02fowli want to play GTA5 so bad
19:35:54fowlhey so i cant do anything with github because i forgot the password i put on my rsa thing.. is that NSA backdoor public yet
19:37:36Araqso ... it's fixing the corruption or merging the new VM into the compiler ...
19:37:41Araqhmm difficult decision
19:39:33reactormonkAraq, yup, fixed it. How do I write a test for it?
19:40:16Araqwe have js tests
19:40:21dom96Araq: Both. At the same time.
19:40:26reactormonkand nodejs running them?
19:40:47dom96Araq: No forum thread with the slides link?
19:40:50Araqwell the tester doesn't do that yet
19:41:00Araqdom96: feel free
19:41:28dom96I'm too busy
19:42:13Araqhas anybody seen Riddick 3 btw?
19:42:26reactormonkhow do I generate a cstring btw?
19:42:34reactormonkwell, get one. literla, whatever.
19:43:24fowlAraq, started watching it but i was drunk and passed out before half way through it
19:43:36Araqlol
19:43:44Araqno spoilers!
19:46:06reactormonkwhat's the cstring literal?
19:46:07fowlok
19:46:11fowlit looked good
19:46:36Araqconst c = cstring("abc") # should work
19:46:53reactormonkkk
19:47:56reactormonktest.nim(1, 20) Error: internal error: evalOp(mEqCString)
19:47:59reactormonkuh
19:49:16reactormonkhttp://sprunge.us/LjBY
19:52:55*io2 quit ()
19:57:54fowlhi is this channel about Green Day's album "Nimrod"? whats you guy's favorite song off the album??
19:58:11Araqthat album sucks
19:59:41reactormonk of mEqStr: result = newIntNodeT(ord(getStr(a) == getStr(b)), n)
19:59:47reactormonkhm...
20:00:55reactormonkadded of mEqCString: result = newIntNodeT(ord(getStr(a) == getStr(b)), n)
20:01:18reactormonknot sure if it's correct
20:06:48Araqfor evals.nim?
20:07:01Araq*semfold.nim ?
20:08:00Araqreactormonk: IMHO constant folding shouldn't support == for cstrings as that comparison only compares addresses
20:08:09Araqwhich are unknown at compile time
20:09:51reactormonkoh
20:10:37*fowl quit (Read error: Connection reset by peer)
20:12:10reactormonkAraq, kk
20:28:53*fowl joined #nimrod
20:34:45Araqgood night
20:40:59*Associat0r quit (Read error: Connection reset by peer)
20:42:06*Associat0r joined #nimrod
20:42:06*Associat0r quit (Changing host)
20:42:06*Associat0r joined #nimrod
20:58:13*DAddYE quit (Remote host closed the connection)
20:59:14*DAddYE joined #nimrod
21:17:12*MFlamer quit (Quit: Page closed)
21:20:12*EXetoC quit (Quit: WeeChat 0.4.1)
21:23:40*NewGuy joined #nimrod
21:24:12NewGuyHowdy guys, how are you doin?
21:28:06NewGuyWell seems like nobodies here, so I'll just leave the question for whoever comes by:
21:29:18NewGuyI'm implementing a form of the Nulstein game architecture (http://software.intel.com/en-us/articles/do-it-yourself-game-task-scheduling/).
21:29:23Araqhi NewGuy
21:30:23NewGuyI've already done it in C++, and porting to Nimrod was a fun task for me. But, even though I talked to Araq on the forum about it, I wanted to clarify real quick:
21:31:19NewGuyThe system sends function pointers to the worker threads, which run the code in the procedures. These procedures read from one copy of the game data, and write to another copy (and only to the object's own data)
21:31:44fowlhi
21:31:54NewGuyAraq: Howdy, my good sir.
21:32:00NewGuyfowl: I missed you.
21:32:44AraqI'm sleeping already, better be quick with your question ;-)
21:33:41fowlNewGuy, im around
21:33:42NewGuySo if I went about doing this, what are the implications with memory? I know that Nimrod is thread-local GC (deferred reference counting with cyclic collection, right?), so would this all have to be manually managed or would the GC still work?
21:34:10Araqthat's a tough question in general
21:34:18Araqsending proc vars is just fine
21:34:29Araqand sending data via channels is too
21:34:31NewGuyAraq: Yeah, sorry. I understand that.
21:34:49Araqaccessing globals is fine unless it's GC'ed memory then you have to be careful
21:35:06NewGuyYeah, from our last conversation I know passing proc vars are fine, as long as they aren't clojures that implicity share their environment
21:35:13Araqyup
21:35:32Araqyou repeat the question, I repeat the answer :P
21:36:11NewGuyBut is it safe to modify the data from other threads?
21:36:19Araqto go forward you should pastebin some example code
21:36:32NewGuyYeah, let me do that real quick. Should have done it from the get-go
21:36:57NewGuyAraq: You at all fluent in C++?
21:37:09Araqfluent enough I think
21:48:27Araqtick tick tick
21:48:34NewGuyI'm so sorry, don't stay up on my accord.
21:48:44Araqyour time is running out
21:48:45NewGuyIt's taking a bit to write it all out and such.
21:49:14Araqok no problem, just ask the question on the forum then
21:49:20NewGuyWill do
21:49:26NewGuySorry for this, have a great night!
21:49:30Araqalright, good night
21:49:44*NewGuy quit (Quit: Page closed)
21:58:54*OrionPK joined #nimrod
22:06:13*Amrykid quit (Excess Flood)
22:06:42*Amrykid joined #nimrod
22:34:26filwitno chance of a nimrod Kate color theme out there?
22:34:45fowlfor aporia/
22:35:21filwitno, i just realized that would sound backwards
22:35:21filwiti mean, a scheme, for KDE Kate
22:35:54fowlah
22:35:56fowlprobably not
22:36:04filwitit's starting to hurt my eyes, lol, i should probably just build Aporia
22:36:12fowlunless you can import a vim scheme
22:36:20filwitbut i really like Kate, and I'd rather have things under one editor
22:36:23fowlor gtksourceview language
22:36:25fowldefinitoin
22:36:32filwitactually, i think you can import a vim scheme in Kate
22:36:41filwitnot .lang files though
22:37:14filwitnah, nevermind, only kate specific ones
22:37:25filwitguess i'll have to make one..
22:38:53*gsp quit (Ping timeout: 245 seconds)
22:41:23fowlfilwit, forgot my password to my public key how do i get past this
22:41:26fowlshould i just make a new one
22:41:31fowl(RSA)
22:41:54filwituh... idk
22:42:25filwitnot sure you meant to ask me, but I don't know what password to what public key you are speaking of
22:42:40fowlyea no kidding you dont know the password
22:42:56filwiti am so confused...
22:43:02fowlnevermind
22:44:46fowlthe question was, what would you do in this situation
22:44:51fowlnot asking you what my password is.. lol
23:02:47filwitah i see
23:03:09filwityeah i guess it was kinda obvious now that i re-read it