| 00:01:01 | * | rapind quit (Ping timeout: 250 seconds) |
| 00:02:24 | BlaXpirit | how to turn seq[int] into array[5, int] ? |
| 00:02:51 | Siecje | You can't put procedures in a table? |
| 00:03:15 | def- | Siecje: the procs need to be procvars for that |
| 00:04:40 | BlaXpirit | .eval import tables; proc test() {.procvar.} = echo "hi"; var t = {"test": test}.toTable(); t["test"]() |
| 00:04:43 | Mimbus | BlaXpirit: hi |
| 00:04:59 | ekarlso | flaviu: https://zetta.io/en/ back up |
| 00:05:20 | * | Jehan_ quit (Quit: Leaving) |
| 00:05:22 | * | _303_ joined #nim |
| 00:07:59 | * | fowlmouth quit (Quit: Leaving) |
| 00:08:48 | BlaXpirit | I want to turn seq[int] into array[n, int] where n is known at compile time or make a function that can return array[n, int] when given an n known at compile time |
| 00:09:18 | def- | BlaXpirit: n: static[int]? |
| 00:09:33 | BlaXpirit | sounds nice |
| 00:10:28 | BlaXpirit | seems to work, thanks |
| 00:11:15 | * | TnTReX joined #nim |
| 00:12:24 | BlaXpirit | ended up with proc urandom*(size: (static[Natural]){lit}): array[size, uint8] |
| 00:12:51 | Siecje | def-: I'm getting indentation error (3, 26) https://dpaste.de/v57Z |
| 00:13:21 | BlaXpirit | Siecje, I think :type goes before {.procvar |
| 00:16:04 | * | ob_ joined #nim |
| 00:16:23 | Siecje | it is shown after a variable here http://nim-lang.org/manual.html#deprecated-pragma |
| 00:16:47 | BlaXpirit | Siecje, that's different :| |
| 00:17:02 | BlaXpirit | some people have raised concerns over this seeming inconsistency |
| 00:17:18 | BlaXpirit | seems logical to me though |
| 00:18:26 | Siecje | https://dpaste.de/4nAE#L6 Error: ordinal type expected |
| 00:18:35 | def- | {"/": index} |
| 00:18:57 | def- | that's the same as [("/", index)] btw |
| 00:19:29 | Siecje | Thanks |
| 00:20:22 | def- | you only need {.procvar.} if you want to use the proc as a variable from another module. inside the same module it's not necessary |
| 00:21:01 | * | Matthias247 quit (Read error: Connection reset by peer) |
| 00:21:03 | Araq | and btw somebody should start a .procvar discussion somewhere |
| 00:21:24 | Araq | I'm not sure the benefits outweigh the problems it causes for newbies |
| 00:21:36 | BlaXpirit | uhh which problems |
| 00:21:37 | def- | Araq: what's the alternative? |
| 00:21:46 | dom96 | Araq: hah. I was right! |
| 00:21:53 | Araq | def-: be sloppy about it |
| 00:22:11 | Araq | somebody adds a default parameter --> code breaks, so what |
| 00:22:17 | nimnoob123 | quick question: say i have a seq of X type objects and add Y, Z which are objects of X, if I called myseq[0].XObjectMethod() from Y or Z, it would use the X's version unless I ref of X object else it won't override X's definition? |
| 00:22:37 | BlaXpirit | nimnoob123, you cannot do it if it is not a ref |
| 00:22:43 | def- | Araq: and we just document that it doesn't work with default parameters? |
| 00:22:43 | Araq | def-: it's not like any other language cares about this problem |
| 00:22:48 | dom96 | We need to improve our functional programming features before 1.0. |
| 00:22:54 | nimnoob123 | was wondering what was going on with my code |
| 00:23:09 | BlaXpirit | nimnoob123, it will behave as you expect if you define *methods* not procs |
| 00:23:09 | def- | dom96: yes, if everything was procvar by default that would be great in this regard |
| 00:23:32 | flaviu | ahh, finally: http://i.imgur.com/E88OsOS.png |
| 00:23:44 | BlaXpirit | nimnoob123, basically, read on `method` |
| 00:23:47 | dom96 | flaviu: ? |
| 00:23:49 | flaviu | compare to http://nim-lang.org/assets/images/bg.jpg |
| 00:23:52 | BlaXpirit | o.o |
| 00:24:05 | def- | flaviu: looks so much better, never use jpeg for that |
| 00:24:21 | BlaXpirit | 380 KiB |
| 00:24:51 | flaviu | Don't worry, I'll get it down |
| 00:25:16 | BlaXpirit | but yeah, looks cool |
| 00:25:16 | dom96 | Araq: I still don't understand the problem. |
| 00:26:07 | def- | Araq: if this means `+` and everything else can be used as procvars by default, that would be great |
| 00:27:07 | Araq | def-: no, that doesn't mean '+' works as a procvar |
| 00:27:20 | def- | oh right, `+` is magic |
| 00:27:24 | Araq | '+' is a magic that ends up as a CPU instruction |
| 00:27:42 | Araq | we could generate the trivial wrapper for you |
| 00:27:52 | def- | would be cool for functional code |
| 00:27:58 | Araq | but I'm not sure that's the best idea for a systems programming language |
| 00:28:14 | nimnoob123 | so in order to make all that work the parent object needs to ref the RootObj as well, curious to why that is? (noob asking noob things ;]) |
| 00:28:29 | * | Jesin quit (Quit: Leaving) |
| 00:29:04 | nimnoob123 | guessing it's just a way of telling the compiler that the parent object is an abstract object? |
| 00:30:20 | * | Jesin joined #nim |
| 00:30:30 | def- | flaviu: can't get it smaller than this: http://ddnet.tw/E88OsOS.png |
| 00:31:03 | Araq | nimnoob123: I don't understand your questions |
| 00:31:35 | BlaXpirit | Araq, it's just about polymorphism and methods |
| 00:32:45 | Araq | right but I'm not sure "need to ref the RootObj as well" means the right thing |
| 00:33:21 | BlaXpirit | maybe RootObj here is not actually "the" RootObj, but a root object... who knows |
| 00:34:30 | nimnoob123 | it is the RootObj the built in type, the parent class has to ref the RootObj in order for any of what I said to work |
| 00:35:07 | BlaXpirit | uh yeah |
| 00:35:25 | nimnoob123 | the question was: is ref object of RootObj just a way of telling the compiler that it's an abstract class |
| 00:35:32 | dom96 | Just so we don't forget: https://github.com/Araq/Nim/issues/2172 :P |
| 00:35:50 | BlaXpirit | nimnoob123, maybe just read some generic topics on polymorphism |
| 00:36:11 | BlaXpirit | in C++ one would simply say that there can be no polymorphism without pointers |
| 00:36:14 | BlaXpirit | ref is like a pointer |
| 00:37:27 | BlaXpirit | dom96, uhh isn't that a big performance penalty? maybe someone should write about this there to prevent stupid questions |
| 00:38:32 | nimnoob123 | all I'm asking is why does the parent object need to use 'ref' http://pastebin.com/XhMBCUuu |
| 00:38:43 | * | Jesin quit (Ping timeout: 250 seconds) |
| 00:39:05 | Araq | ha that's not a "Language design: undesired property" by any means |
| 00:39:19 | Araq | it was a *desired* property for years! |
| 00:39:21 | dom96 | Araq: relabel it as appropriate then |
| 00:39:31 | dom96 | it's undesired now! |
| 00:39:40 | dom96 | it's been undesired by me for years |
| 00:39:50 | nimnoob123 | oops forgot to remove the count stuff |
| 00:40:08 | Araq | whatever |
| 00:40:16 | dom96 | nimnoob123: it doesn't if your child objects don't use it. |
| 00:41:00 | nimnoob123 | dom96, it won't show the output of the child objects it'll default to the parent |
| 00:41:22 | nimnoob123 | without ref |
| 00:42:30 | * | Demon_Fox joined #nim |
| 00:42:42 | nimnoob123 | http://pastebin.com/ZxtPqXQ7 outputs shutup twice |
| 00:42:58 | flaviu | def-: Actually: https://dl-2.kraken.io/cbf90b934916858f7589ee10f7b4d1c6/background.png |
| 00:43:33 | flaviu | 148K :D. It's lossy though, some of the gradients are of lower quality. |
| 00:43:44 | nimnoob123 | was just curious why the parent required the ref keyword, i understand why the child objects would need it |
| 00:43:45 | def- | flaviu: that looks very bad for me |
| 00:44:47 | def- | nimnoob123: because you put them into a seq the children would get cut off |
| 00:46:33 | nimnoob123 | which goes back to my original question is that just a way for the compiler to know that parentObj is an abstract class |
| 00:48:23 | * | lavender quit (Ping timeout: 240 seconds) |
| 00:49:37 | def- | i don't understand this question, sorry |
| 00:51:32 | nimnoob123 | alright, it doesn't really matter tbh |
| 00:54:16 | * | kjo joined #nim |
| 00:58:04 | * | rapind joined #nim |
| 00:59:58 | Siecje | I'm not sure why it expects 'FutureBase' and how can I figure out what is happening https://dpaste.de/mgEM#L14,21 |
| 01:02:39 | * | rapind quit (Ping timeout: 265 seconds) |
| 01:04:22 | dom96 | Siecje: Your proc has no return type |
| 01:05:03 | Siecje | dom96: If I put the type after I get invalid indentation |
| 01:05:18 | def- | which looks right, because cb should not have a return type to be used in serve() |
| 01:05:29 | * | rapind joined #nim |
| 01:05:31 | def- | Siecje: proc cb(req: Request): Future[string] {.async.} = |
| 01:06:35 | dom96 | indeed. What are you trying to do? |
| 01:06:56 | * | Trustable quit (Remote host closed the connection) |
| 01:08:54 | * | Trustable joined #nim |
| 01:09:10 | Siecje | dom96: Just play around. |
| 01:09:25 | dom96 | I think you want to replace the "test" in line 15 with routes[$req.url.path]() |
| 01:09:50 | Siecje | dom96: true, but it's not getting that far. |
| 01:09:54 | Siecje | https://dpaste.de/Y8Lx |
| 01:10:08 | Siecje | Future[void] says expects future got string |
| 01:11:22 | dom96 | https://gist.github.com/dom96/630021f932c34153a57c |
| 01:14:28 | Siecje | dom96: why does eliminating result matter? |
| 01:14:59 | dom96 | Siecje: the callback for asynchttpserver shouldn't return anything |
| 01:15:32 | Siecje | it doesn't it calls something that does though |
| 01:15:50 | * | sillesta quit (Ping timeout: 252 seconds) |
| 01:17:00 | dom96 | 'result' is the implicit variable which is always returned in a proc |
| 01:17:40 | Siecje | oh really |
| 01:17:56 | * | qwr joined #nim |
| 01:18:06 | Siecje | That's different, and yeah that was the issue |
| 01:18:40 | dom96 | http://nim-lang.org/tut1.html#result-variable |
| 01:19:46 | * | BlaXpirit quit (Quit: Quit Konversation) |
| 01:20:49 | * | gokr_ quit (Quit: IRC for Sailfish 0.9) |
| 01:20:54 | Siecje | so you can't access result outside of the scope? |
| 01:21:10 | Siecje | result is always declared or only when it is used? |
| 01:21:36 | * | TnTReX left #nim ("Leaving") |
| 01:22:35 | flaviu | dom96: https://github.com/nim-lang/art look good? |
| 01:22:59 | dom96 | Why does it have html? |
| 01:23:56 | flaviu | dom96: Concept pages |
| 01:24:18 | dom96 | perhaps the repo should be called "concepts" then? |
| 01:24:29 | * | infinity0 quit (Ping timeout: 252 seconds) |
| 01:24:30 | * | infinity0_ joined #nim |
| 01:24:30 | * | infinity0_ quit (Changing host) |
| 01:24:30 | * | infinity0_ joined #nim |
| 01:24:33 | dom96 | or "assets" |
| 01:24:49 | flaviu | "assets" sounds fine. |
| 01:25:01 | dom96 | It would be nice to split the website from the Nim repo |
| 01:25:27 | dom96 | but Araq probably wouldn't want that. |
| 01:27:42 | flaviu | dom96: Please also replace the background on the website with https://raw.githubusercontent.com/nim-lang/assets/master/Art/background-optimized.png |
| 01:28:29 | dom96 | create a PR for that |
| 01:32:16 | def- | flaviu: are you sure we need them all? many of these files are just concepts and not actually used |
| 01:32:42 | def- | flaviu: would be nice to have just the important images available on the website (logo, logo + name) |
| 01:33:10 | flaviu | def-: You want me to throw away data? :O |
| 01:33:35 | flaviu | IMO it's best to keep all data. It doesn't cost anything to do so. |
| 01:34:04 | def- | flaviu: in case someone searches for the logo and logo+name would be nice to have a proper place for them |
| 01:34:27 | def- | flaviu: also: http://ddnet.tw/logo-optimized.svg |
| 01:35:22 | def- | i can't even find an svg anywhere with just logo+name |
| 01:35:44 | flaviu | def-: Not much point in optimizing that, it still has lots of hidden logo variants. |
| 01:35:56 | def- | flaviu: oh, right |
| 01:42:46 | * | Boscop_ is now known as Boscop |
| 01:43:30 | Boscop | nim seems to gain some traction |
| 01:43:56 | whitenoise | I introduced my coworkers to Nim today |
| 01:44:00 | Boscop | it's popping up on blogs etc. being compared to rust |
| 01:44:03 | whitenoise | they promptly installed it on their laptops |
| 01:44:21 | * | Sembei joined #nim |
| 01:44:23 | Boscop | i remember when this channel had like 10 ppl |
| 01:44:37 | Boscop | whitenoise: is it production ready? |
| 01:44:55 | whitenoise | Boscop: Nim? It depends on what you mean by "production ready" |
| 01:45:18 | * | Pisuke quit (Ping timeout: 250 seconds) |
| 01:45:45 | whitenoise | there are some bugs, but there are bugs in most everything as far as I am aware, even very mature languages. Nim's currently downfall, in my opinion, for being "production ready" is its immaturity. In another language, you could easily import a library for anything you set your mind to, whereas in Nim you would likely have to do quite a bit more coding to |
| 01:45:45 | whitenoise | get your desired effect because the library does not yet exist. |
| 01:45:57 | flaviu | dom96: https://github.com/Araq/Nim/pull/2173 |
| 01:46:00 | whitenoise | this in my mind would probably make a lot of corporate environments not favor it at the moment |
| 01:46:15 | whitenoise | but as far as a hobby language, I'm digging it. |
| 01:47:31 | * | lavender joined #nim |
| 01:47:34 | whitenoise | I also only have a few days of experience with it, so my opinion should not carry a ton of weight. |
| 01:50:25 | * | Jesin joined #nim |
| 01:54:11 | TEttinger | It does seem to have a lot of features that are more common in "mature languages", like the fine tuning over performance (Ruby didn't even get a JIT compiler for years!) |
| 01:54:50 | TEttinger | c2nim seems to make binding stuff a lot easier |
| 01:55:10 | TEttinger | so if a C lib exists, using from Nim won't be hard |
| 01:55:17 | TEttinger | (potentially) |
| 01:55:31 | flaviu | Well, creating an idiomatic wrapper will take some work. |
| 01:56:15 | flaviu | But that depends on the complexity of the library. |
| 01:56:24 | TEttinger | yep |
| 02:02:11 | * | qwr quit (Remote host closed the connection) |
| 02:03:25 | * | Trustable quit (Remote host closed the connection) |
| 02:06:14 | * | darkf joined #nim |
| 02:13:04 | * | chemist69_ joined #nim |
| 02:16:36 | * | chemist69 quit (Ping timeout: 264 seconds) |
| 02:17:45 | * | gsingh93 joined #nim |
| 02:19:56 | * | qwr joined #nim |
| 02:25:21 | * | sampwing quit (Ping timeout: 252 seconds) |
| 02:25:22 | * | phira quit (Ping timeout: 240 seconds) |
| 02:25:36 | * | xeizlif quit (Ping timeout: 250 seconds) |
| 02:26:53 | * | phira joined #nim |
| 02:27:18 | * | xeizlif joined #nim |
| 02:28:42 | whitenoise | is there a way to do a constructor in Nim? |
| 02:29:14 | whitenoise | i guess you would just write your own "new" method and then run it |
| 02:32:21 | flaviu | whitenoise: Nim doesn't have constructors (although see my proposal https://github.com/Araq/Nim/issues/1693). The convention is to have an initFoo if it's a value, and a newFoo if it's heap-allocated. |
| 02:43:48 | def- | I worked on the ideas page again, any feedback welcome: https://github.com/Araq/Nim/wiki/GSoC-2015-Ideas |
| 02:44:24 | flaviu | has it been submitted yet? |
| 02:44:29 | def- | still have 20 hours |
| 02:44:31 | flaviu | > 16 hours, 15 minutes remaining |
| 02:44:39 | flaviu | That's not much time. |
| 02:44:41 | def- | right, 16 hours |
| 02:44:57 | def- | that's why I'm working on it, obviously ;) |
| 02:45:19 | def- | The application is also finished I'd say: https://github.com/Araq/Nim/wiki/GSoC-2015-Organization-Application |
| 02:46:34 | def- | I even added 3 fun projects from my personal todo list |
| 02:47:02 | * | reem quit (Remote host closed the connection) |
| 02:56:57 | * | gokr quit (Ping timeout: 250 seconds) |
| 02:57:03 | * | fowl joined #nim |
| 03:01:18 | flaviu | def-: I improved some awkward phrasing: https://github.com/Araq/Nim/wiki/GSoC-2015-Organization-Application/_compare/4f5760fb335829bcd585e73e97df95131497280a...ac587f6ae55ff3d8a380d64867b1c36c575b5fbe |
| 03:01:43 | def- | Thanks |
| 03:07:16 | * | reem joined #nim |
| 03:09:09 | * | banister quit (Quit: My MacBook has gone to sleep. ZZZzzz…) |
| 03:13:20 | * | reem quit (Remote host closed the connection) |
| 03:14:47 | * | reem joined #nim |
| 03:17:43 | * | reem quit (Remote host closed the connection) |
| 03:18:53 | * | filwit joined #nim |
| 03:19:25 | * | nimnoob123_ joined #nim |
| 03:20:14 | nimnoob123_ | woo first nim contribution, simple, but useful :D sdl2.image didn't work on windows because it didn't set the lib at the top |
| 03:20:14 | filwit | flaviu: you can put the website work under WTFPL or another permissive license of your choosing (BSD, MIT, GPL) |
| 03:20:46 | def- | nimnoob123_: good job |
| 03:20:59 | nimnoob123_ | :D thanks def- |
| 03:23:21 | * | reem joined #nim |
| 03:24:42 | * | saml_ joined #nim |
| 03:24:45 | * | nande quit (Remote host closed the connection) |
| 03:25:04 | filwit | flaviu: i mostly find licensing issues petty blockades to real work (even though I understand they're important to get right), so as long as someone else doesn't try and claim responsibility for my original work, I'm fine with whatever you choose or is standard to the rest of Nim-lang stuff. |
| 03:26:16 | filwit | flaviu: on that note. Thank you for organizing this and putting it in a repo. |
| 03:26:57 | * | reem quit (Remote host closed the connection) |
| 03:28:09 | fowl | im trying to bootstrap but build.bat uses gcc |
| 03:28:10 | def- | filwit: do you have the logo+name as an svg somewhere? |
| 03:28:34 | * | reem joined #nim |
| 03:28:41 | filwit | def-: flaviu made a repo here: https://github.com/nim-lang/assets |
| 03:28:57 | def- | fowl: may be easier to bootstrap from an existing nim installation then |
| 03:29:06 | def- | filwit: is it in there? |
| 03:29:56 | filwit | def-: the logo? |
| 03:30:16 | def- | filwit: logo+name as on the top of the website |
| 03:30:19 | filwit | def-: i thought it was.. but the logo under Art/logo.svg isn't exactly correct now that I take a look |
| 03:30:31 | * | reem_ joined #nim |
| 03:31:02 | filwit | def-: either way, it's in that zip i gave you the other day |
| 03:31:20 | fowl | def-, how do i do that? |
| 03:31:30 | def- | filwit: it would be really nice to have an svg of logo+name. the 200px logo from the website isn't good enough if you want to print something about nim for example |
| 03:31:42 | def- | filwit: pretty sure it's not, i went through each file i think |
| 03:32:42 | filwit | def-: okay let me see why that wasn't included. I think it's just because I mostly finalized the rendering of that inside the website svg |
| 03:33:22 | def- | filwit: ah, i could extract it from the website svg |
| 03:33:27 | * | reem quit (Ping timeout: 265 seconds) |
| 03:33:32 | Varriount | Hi filwit! |
| 03:34:10 | filwit | def-: I can separate it out and give you a copy. I would rather people not be trying to pull the website apart layer-by-layer to try and get a usable copy of the logo. |
| 03:34:19 | filwit | Varriount: hey man |
| 03:34:43 | def- | filwit: exactly, we should put it into the assets-repo so it's easy to reuse |
| 03:35:08 | filwit | def-: but i'm not going to do that tonight. I'm just dropping by because flaviu sent me an issue on github and I thought I'd just talk to him here instead (but doesn't look like he's around ATM) |
| 03:35:49 | filwit | def-: I have a couple of updated to the new website support page to get to tomorrow, so I'll do it then. |
| 03:35:55 | filwit | updates* |
| 03:36:30 | filwit | def-: well... shit it only takes a minute.. let me just do it |
| 03:38:54 | flaviu | I'm here. |
| 03:39:21 | flaviu | It might not be a good idea to licence the art under a permissive licence. |
| 03:39:54 | flaviu | What if someone makes an unrelated project with that logo? |
| 03:40:51 | fowl | flaviu, the only honorable thing to do at that point is seppuku |
| 03:41:39 | filwit | flaviu: what do you suggest? |
| 03:41:52 | filwit | i like fowl's idea >:} |
| 03:42:11 | flaviu | https://wiki.archlinux.org/index.php/DeveloperWiki:TrademarkPolicy |
| 03:42:42 | fowl | Varriount, can you help me bootstrap on windows? |
| 03:42:58 | filwit | flaviu: looks good to me. run with that. |
| 03:43:44 | def- | fowl: you use the binary nim version and compile koch with it and basically follow the instructions to build from source |
| 03:44:17 | def- | actually just "nim -d:release c koch" and "./koch boot -d:release" |
| 03:44:37 | fowl | def-, missing a module "compiler_koch" |
| 03:45:14 | def- | huh, i don't understand |
| 03:45:32 | fowl | well generally old nimrods cant compile the newest koch |
| 03:45:37 | def- | oh, right |
| 03:45:54 | def- | that would mean you have to go through a few commits at a time, terrible |
| 03:46:19 | def- | just install mingw and bootstrap the regular way? |
| 03:46:57 | fowl | windows development |
| 03:47:04 | * | fowl shoots himself in the head instead |
| 03:48:03 | fowl | i'll look for cl.exe and link.exe and try to fix build.bat |
| 03:48:03 | ldlework | :P |
| 03:48:24 | def- | fowl: that may not work if there is gcc-specific stuff in the csources |
| 03:50:00 | filwit | def-, flaviu: NimLogo - http://www.reign-studios.net/philipwitte/nim/artwork/NimLogo.zip |
| 03:50:58 | filwit | that's got the BG color or the website in it, but will export with a transparent background. So it's designed to sit on anything dark. I'll make a reverse color version later for light backgrounds. |
| 03:51:03 | def- | filwit: thanks a lot |
| 03:51:11 | filwit | np |
| 03:51:25 | def- | strangely the svg doesn't show in firefox for me, but works in inkscape |
| 03:51:29 | fowl | def-, well i've built urho3d with visual studio, i'm sure it wont work with nim if i build nim with gcc |
| 03:51:39 | filwit | def-, flaviu: shit don't download that... didn't save before making the zip... :\ |
| 03:51:54 | def- | fowl: after building nim with gcc you can use that version of nim to build it with visual studio |
| 03:52:32 | filwit | def-, flaviu: okay, can download now (same link) |
| 03:52:51 | fowl | def-, oh.. ok |
| 03:52:54 | def- | filwit: still the same |
| 03:53:06 | filwit | def-: yes the blur operations are Inkscape specific filters |
| 03:53:44 | def- | ok, should probably make a plain svg version for people who want to embed it somewhere |
| 03:53:47 | filwit | def-: they're used for the shine marks and stuff (at least I think they're inkscape specific, i know inkscape has some special stuff but am not sure if blur is one of them) |
| 03:53:58 | filwit | def-: good point |
| 03:54:12 | def- | (still good to have the inkscape one for editing) |
| 03:54:31 | * | reem_ quit (Remote host closed the connection) |
| 03:55:05 | * | reem joined #nim |
| 03:55:22 | filwit | well I understand not everyone is going to know how to use Inkscape, so it's good to have a basic version.. but honestly for the best look you should just export from Inkscape at whatever resolution you need and use that (maybe that should be written into the official branding rules or something) |
| 03:55:53 | flaviu | Well, I couldn't be bothered to adapt Arch's page, I did https://github.com/nim-lang/assets/tree/master/Art instead. |
| 03:57:17 | flaviu | It can be rewritten in legalese whenever Araq talks to his lawyer (or one of 3DICC's lawyers?) |
| 03:57:43 | * | reem quit (Remote host closed the connection) |
| 03:57:58 | fowl | on windows system.int is 4 bytes? |
| 03:58:30 | def- | fowl: on all 32bit systems. if you use the x86 version, yes |
| 03:58:38 | flaviu | fowl: Int is always word size (C's size_t) |
| 03:58:46 | TEttinger | presumably because most exes on windows can run as 32-bit |
| 03:58:54 | filwit | okay, i updated the zip with a Logo_simple (no effects) version |
| 03:59:04 | fowl | hhh |
| 03:59:27 | fowl | when was i supposed to specify i wanted 64bit |
| 03:59:39 | TEttinger | you might not |
| 03:59:49 | def- | fowl: when downloading the binary here: http://nim-lang.org/download.html |
| 04:00:03 | fowl | i didnt use any binary |
| 04:00:12 | def- | filwit: maybe my firefox is just broken, doesn't work for me |
| 04:00:13 | filwit | flaviu: looks okay to me, but I'm far from legally qualified to judge :) |
| 04:00:16 | fowl | gcc then msvc |
| 04:00:30 | def- | ah, then you have 32bit gcc maybe |
| 04:00:31 | filwit | def-: hmm... i haven't tried in a browser yet, one sec |
| 04:00:40 | TEttinger | is your gcc 32-bit? msvc comes in a 32 and 64 bit version |
| 04:00:57 | def- | filwit: and there is a build64.bat? |
| 04:01:08 | def- | fowl, not filwit |
| 04:02:09 | def- | filwit: but if the effects can't be represented in regular svg, it doesn't matter anyway |
| 04:02:25 | * | fowl left #nim ("Leaving") |
| 04:02:44 | * | reem joined #nim |
| 04:02:46 | * | reem quit (Remote host closed the connection) |
| 04:03:22 | filwit | def-: well i removed the shines.. it's still the basic logo without some of the jazz, but it's okay. Unfortunately even saving as "plain svg" doesn't seem to make it work in the browsers.. i'm trying to figure out what's causing that. |
| 04:04:13 | * | reem joined #nim |
| 04:04:32 | * | reem quit (Remote host closed the connection) |
| 04:05:40 | flaviu | filwit: Tell people to use the png? |
| 04:06:26 | flaviu | Anyway, it's late my time. See you all tomorrow. |
| 04:06:50 | filwit | later |
| 04:07:35 | filwit | def-: yeah I'm not sure what's messing up the SVG output for browsers.. that's really strange (and something I need to figure out) |
| 04:08:22 | filwit | def-: but honestly I would rather people be using the logo with proper shines anyways, so I think we should just make a couple of different resolution PNGs for people to grab instead |
| 04:08:33 | def- | filwit: yes |
| 04:08:48 | * | reem joined #nim |
| 04:10:23 | * | rapind quit (Remote host closed the connection) |
| 04:11:30 | filwit | def-: okay, updated with 3 sizes which should work for everyone |
| 04:12:07 | flaviu | filwit: fyi, you can commit directly to the repo. |
| 04:13:12 | * | rapind joined #nim |
| 04:13:34 | * | ob_ quit (Ping timeout: 265 seconds) |
| 04:13:43 | filwit | flaviu: good point.. |
| 04:13:55 | * | reem quit (Remote host closed the connection) |
| 04:14:07 | filwit | flaviu: i'll do that, but I need to take off now. See you guys tomorrow |
| 04:14:17 | * | filwit quit (Quit: Leaving) |
| 04:14:17 | def- | thanks again and bye, filwit |
| 04:17:11 | * | reem_ joined #nim |
| 04:18:41 | * | reem__ joined #nim |
| 04:20:55 | * | fowl joined #nim |
| 04:20:58 | * | reem joined #nim |
| 04:21:47 | * | reem_ quit (Ping timeout: 265 seconds) |
| 04:23:14 | * | reem__ quit (Ping timeout: 265 seconds) |
| 04:25:39 | * | reem quit (Ping timeout: 265 seconds) |
| 04:31:08 | * | reem joined #nim |
| 04:31:16 | * | reem quit (Remote host closed the connection) |
| 04:37:06 | * | jholland quit (Quit: Connection closed for inactivity) |
| 04:43:02 | * | reem joined #nim |
| 04:43:17 | * | reem quit (Remote host closed the connection) |
| 05:01:19 | * | gsingh93 quit (Ping timeout: 250 seconds) |
| 05:08:33 | * | reem joined #nim |
| 05:08:33 | * | reem quit (Remote host closed the connection) |
| 05:13:08 | * | lavender quit (Read error: Connection reset by peer) |
| 05:13:56 | * | reem joined #nim |
| 05:16:24 | * | gsingh93 joined #nim |
| 05:18:54 | * | reem quit (Remote host closed the connection) |
| 05:34:07 | * | milosn quit (Read error: Connection reset by peer) |
| 05:38:09 | * | rapind quit (Remote host closed the connection) |
| 05:39:30 | * | milosn joined #nim |
| 05:40:11 | * | rapind joined #nim |
| 05:41:59 | * | dhasenan quit (Remote host closed the connection) |
| 05:44:08 | * | saml_ quit (Ping timeout: 246 seconds) |
| 05:47:04 | * | Siecje left #nim (#nim) |
| 05:47:40 | * | rapind quit (Remote host closed the connection) |
| 05:54:05 | * | rapind joined #nim |
| 05:54:32 | * | Jehan_ joined #nim |
| 05:54:33 | * | dhasenan joined #nim |
| 05:54:49 | * | Jehan_ quit (Client Quit) |
| 05:57:50 | * | dhasenan quit (Remote host closed the connection) |
| 06:33:13 | * | darkf_ joined #nim |
| 06:33:28 | * | darkf quit (Ping timeout: 250 seconds) |
| 06:33:29 | * | darkf_ quit (Changing host) |
| 06:33:30 | * | darkf_ joined #nim |
| 06:34:52 | * | rapind quit (Remote host closed the connection) |
| 06:35:27 | * | darkf_ is now known as darkf |
| 06:50:14 | * | fizzbooze joined #nim |
| 06:52:49 | * | Varriount|Mobile joined #nim |
| 07:02:40 | * | darkf_ joined #nim |
| 07:06:23 | * | darkf quit (Ping timeout: 250 seconds) |
| 07:08:49 | * | brson quit (Quit: leaving) |
| 07:10:21 | * | darkf_ is now known as darkf |
| 07:17:18 | * | kjo1 joined #nim |
| 07:17:18 | * | kjo quit (Read error: Connection reset by peer) |
| 07:18:38 | * | kjo joined #nim |
| 07:18:38 | * | kjo1 quit (Read error: Connection reset by peer) |
| 07:20:00 | * | kjo1 joined #nim |
| 07:20:00 | * | kjo quit (Read error: Connection reset by peer) |
| 07:26:04 | * | kjo joined #nim |
| 07:26:04 | * | kjo1 quit (Read error: Connection reset by peer) |
| 07:27:27 | * | BlaXpirit joined #nim |
| 07:29:31 | * | kjo1 joined #nim |
| 07:29:32 | * | kjo quit (Read error: Connection reset by peer) |
| 07:31:04 | * | Varriount|Mobile quit (Quit: AndroIRC - Android IRC Client ( http://www.androirc.com )) |
| 07:34:01 | * | kjo joined #nim |
| 07:34:01 | * | kjo1 quit (Read error: Connection reset by peer) |
| 07:35:46 | * | rapind joined #nim |
| 07:36:40 | * | kjo1 joined #nim |
| 07:36:41 | * | kjo quit (Read error: Connection reset by peer) |
| 07:38:26 | * | chemist69_ quit (Quit: WeeChat 1.1.1) |
| 07:38:55 | * | reem joined #nim |
| 07:39:48 | * | kjo joined #nim |
| 07:39:48 | * | kjo1 quit (Read error: Connection reset by peer) |
| 07:40:01 | * | rapind quit (Ping timeout: 244 seconds) |
| 07:50:00 | * | gsingh93 quit (Quit: WeeChat 1.1.1) |
| 07:50:19 | * | gsingh93 joined #nim |
| 07:53:12 | * | kjo quit (Read error: Connection reset by peer) |
| 07:53:31 | * | kjo joined #nim |
| 07:53:41 | * | gsingh93 quit (Client Quit) |
| 07:54:01 | * | gsingh93 joined #nim |
| 07:59:45 | * | gsingh93 quit (Ping timeout: 252 seconds) |
| 08:01:53 | * | reem quit (Remote host closed the connection) |
| 08:06:08 | * | fizzbooze quit (Ping timeout: 252 seconds) |
| 08:10:19 | * | gokr joined #nim |
| 08:23:17 | * | dumdum joined #nim |
| 08:28:56 | * | nimnoob123_ quit (Quit: Page closed) |
| 08:29:02 | * | nimnoob123 quit (Quit: Page closed) |
| 08:36:48 | * | sillesta joined #nim |
| 08:40:30 | * | reem joined #nim |
| 08:54:06 | BlaXpirit | in unittest, is the suite name used anywhere at all? |
| 08:54:11 | * | TEttinger quit (Ping timeout: 246 seconds) |
| 08:54:20 | BlaXpirit | it's not written even when a test fails |
| 09:03:00 | * | Outlander joined #nim |
| 09:04:54 | * | Demon_Fox quit (Quit: Leaving) |
| 09:05:13 | * | Gonzih quit (Quit: WeeChat 1.0.1) |
| 09:06:25 | * | chemist69 joined #nim |
| 09:14:30 | * | BlaXpirit quit (Quit: Quit Konversation) |
| 09:20:19 | * | bjz joined #nim |
| 09:21:04 | * | bjz quit (Client Quit) |
| 09:21:46 | * | bjz joined #nim |
| 09:22:11 | * | chemist69 quit (Quit: WeeChat 1.1.1) |
| 09:28:20 | * | chemist69 joined #nim |
| 09:37:35 | * | rapind joined #nim |
| 09:39:23 | * | bjz quit (Ping timeout: 240 seconds) |
| 09:41:57 | * | rapind quit (Ping timeout: 244 seconds) |
| 09:44:35 | * | Outlander quit (Ping timeout: 246 seconds) |
| 09:52:21 | * | brson joined #nim |
| 09:55:32 | * | novist quit (Quit: ZNC - http://znc.in) |
| 09:57:24 | * | novist joined #nim |
| 10:05:27 | * | bjz joined #nim |
| 10:06:22 | * | infinity0_ quit (Ping timeout: 240 seconds) |
| 10:11:33 | * | infinity0 joined #nim |
| 10:15:09 | * | teapottime joined #nim |
| 10:21:04 | * | MajorWork joined #nim |
| 10:26:59 | * | MajorWork quit (Changing host) |
| 10:26:59 | * | MajorWork joined #nim |
| 10:29:49 | * | brson quit (Ping timeout: 264 seconds) |
| 10:31:51 | * | CrazyOwl joined #nim |
| 10:31:56 | CrazyOwl | hello guys |
| 10:34:44 | * | kuzy000_ joined #nim |
| 10:35:03 | * | jm116_ joined #nim |
| 10:35:22 | * | jm116_ quit (Client Quit) |
| 10:46:01 | * | Trustable joined #nim |
| 10:53:43 | * | reem quit (Remote host closed the connection) |
| 10:55:46 | * | Outlander joined #nim |
| 10:59:16 | Varriount | Hi CrazyOwl |
| 10:59:32 | Varriount | Araq: Is a sequence *castable* to an openarray? |
| 10:59:44 | Varriount | Or does some conversion need to be done? |
| 11:17:17 | * | rapind joined #nim |
| 11:21:25 | * | rapind quit (Ping timeout: 252 seconds) |
| 11:21:48 | * | Trustable quit (Remote host closed the connection) |
| 11:25:24 | * | novist quit (Quit: ZNC - http://znc.in) |
| 11:27:48 | * | novist joined #nim |
| 11:32:27 | * | novist quit (Quit: ZNC - http://znc.in) |
| 11:36:06 | * | endou________ joined #nim |
| 11:43:43 | * | endou_______ quit (*.net *.split) |
| 11:51:49 | * | banister joined #nim |
| 11:51:54 | * | banister quit (Max SendQ exceeded) |
| 11:52:45 | * | banister joined #nim |
| 11:57:38 | Araq | Varriount: no, it's not |
| 12:03:37 | * | akiradeveloper joined #nim |
| 12:18:51 | * | akiradeveloper quit (Remote host closed the connection) |
| 12:35:12 | * | pafmaf joined #nim |
| 12:35:48 | gokr | Varriount: AFAIK openarray can only be used for parameters, and it basically means "arrays or seq". I may be wrong. Its a bit of an oddity, but you would use it always unless you for some reason can't accept one or the other I presume. |
| 12:36:20 | * | antrix joined #nim |
| 12:45:18 | * | antrix quit (Quit: wtf) |
| 12:45:55 | * | qfire joined #nim |
| 12:51:55 | * | rapind joined #nim |
| 13:10:26 | * | kjo quit (Quit: Leaving.) |
| 13:10:40 | * | davidhq joined #nim |
| 13:20:30 | * | darkf quit (Quit: Leaving) |
| 13:28:53 | * | johnsoft quit (Ping timeout: 240 seconds) |
| 13:30:42 | * | johnsoft joined #nim |
| 13:44:32 | * | Jehan_ joined #nim |
| 13:47:08 | Jehan_ | Araq: I think the proper solution to the shallowCopy issues would be to make the second argument of shallowCopy a var parameter, too (and possibly introduce an unsafeShallowCopy with the current shallowCopy semantics where that is really needed). |
| 13:53:17 | Araq | Jehan_: aye, good idea, but breaks code |
| 13:54:19 | * | reem joined #nim |
| 13:55:04 | Jehan_ | Araq: Yup, I know. But I think it's ultimately not only the best, but the correct solution. So I think it should at the very least be something to think about. |
| 13:55:45 | Araq | dunno, 'shallowCopy' is declared unsafe already iirc |
| 13:56:49 | Araq | we can also always do something like --storeLiteralsInTheHeap:on|off |
| 13:57:35 | Araq | but then these need to become thread local, so hrm |
| 13:57:53 | * | rapind quit (Remote host closed the connection) |
| 13:57:56 | * | akiradeveloper joined #nim |
| 13:58:45 | * | reem quit (Ping timeout: 246 seconds) |
| 13:59:39 | * | emilsp joined #nim |
| 14:00:35 | * | CrazyOwl quit (Quit: Leaving) |
| 14:08:04 | * | NhanH quit (Ping timeout: 265 seconds) |
| 14:08:41 | * | akiradeveloper quit () |
| 14:10:00 | * | CARAM__ quit (Ping timeout: 265 seconds) |
| 14:11:28 | Jehan_ | Araq: Yeah, but the problem is that often you do need shallow copying for performance (e.g. it's what's causing almost all of the overhead with parsing the wikipedia dump), so a safe version would be really nice. |
| 14:12:38 | Araq | that strings copy is really unfortunate in retrospect but it's the one area where I picked consistency and ease of use over performance |
| 14:14:03 | Araq | plus it solves the "constant strings can be passed by ref without more type modifiers" problem |
| 14:15:46 | Araq | but looking at tables.nim you're right, this would work |
| 14:17:19 | * | Outlander quit (Ping timeout: 245 seconds) |
| 14:18:33 | Araq | Jehan_: so do you know the cause of the slowdown of the wikipedia parsing? |
| 14:18:33 | Jehan_ | Araq: Hmm, I don't have a fundamental problem with copying being the default for mutable strings and seqs, because it protects you against some really nasty and surprising aliasing problems. It's just that (IMO) there needs to be a safe way to avoid the overhead. I've mostly been using `shallow` in such situations, but there are cases where that's not the best solution. |
| 14:19:11 | Jehan_ | Araq: Yeah, I've run it through a profiler. It's pretty much all due to string copies. |
| 14:19:21 | Araq | but why? |
| 14:19:22 | * | pafmaf quit (Quit: This computer has gone to sleep) |
| 14:19:37 | Araq | why the slowdown after 600MB? |
| 14:19:38 | Jehan_ | parsexml.nim mostly, as far as I can tell. |
| 14:20:02 | Araq | I skimmed the code and it uses streaming quite well |
| 14:20:17 | Araq | the string copies should simply slow it down all the time |
| 14:20:20 | Jehan_ | Oh, that. I haven't dug into that, but my guess is simply that data structures grow to certain size. There doesn't seem to be a heap explosion. |
| 14:20:24 | Araq | and not after 600MB of output |
| 14:20:44 | Jehan_ | Yeah, but the slowdown depends on the sizes of strings being copied. |
| 14:21:06 | Araq | so it's just that the data is unfortunate after 600MB? |
| 14:21:23 | Jehan_ | If there's a GC problem, so far I haven't seen it show up in the profiler. |
| 14:22:06 | Jehan_ | I'm not 100% certain, because it takes quite a while to get there, especially since I was profiling with -d:release --opt:none to prevent inlining. |
| 14:22:24 | * | NhanH joined #nim |
| 14:22:40 | Jehan_ | It may be either different data or accumulated data or reduced cache locality or something like that. |
| 14:24:45 | * | CARAM__ joined #nim |
| 14:25:40 | Araq | I suspected it's some IO related problem |
| 14:26:19 | Araq | maybe some buffering failure or fragmentation of the heap or even on the hd |
| 14:28:17 | Araq | btw Delphi uses copy on write for strings which I still consider the best design out there but introduces implicit destructors as you need precise refcount values for this to work, so I didn't implement it |
| 14:46:04 | Araq | Jehan_: anyway, you convinced me, please create an issue for that or a PR |
| 14:48:56 | * | banister quit (Quit: My MacBook has gone to sleep. ZZZzzz…) |
| 14:49:46 | teapottime | sorry to interject, but doesn't Delphi also imposes a lock every time you update a string? Is this the price for precise refcounting or just an implementation decision for them? |
| 14:54:19 | teapottime | I'm guessing Nim won't have that problem so much because of the thread-local GC tho |
| 14:56:38 | * | BlaXpirit joined #nim |
| 15:00:11 | * | Trustable joined #nim |
| 15:01:56 | def- | Araq: the wikidump problem seems to be much easier, i'm currently debugging |
| 15:04:35 | def- | (or the part i can reproduce at least, it never comes to a halt for me) |
| 15:06:08 | * | pafmaf joined #nim |
| 15:10:40 | * | pregressive joined #nim |
| 15:11:07 | * | Siecje joined #nim |
| 15:12:04 | Araq | teapottime: it uses atomic increments and decrements which is faster than locks |
| 15:12:24 | Araq | it's also not entirely correct, but in practice it works well enough |
| 15:15:00 | def- | Araq: I call readBuffer a few billion times and suddenly it takes 30 times as long |
| 15:15:41 | Araq | nice, readBuffer is a than layer over C's IO |
| 15:15:45 | Araq | *thin |
| 15:15:52 | def- | i think this causes the slowdown |
| 15:16:28 | * | kjo joined #nim |
| 15:20:01 | Siecje | Does nim have decorators? |
| 15:21:24 | saml | what's corators? |
| 15:22:30 | Araq | Siecje: yes, you can do that proc p() {.m.} = ... |
| 15:23:01 | Araq | and then 'p' is passed to the macro m which can rewrite the proc definition |
| 15:23:39 | Jehan_ | Araq: Will do that later tonight or this weekend, once I have the time. |
| 15:24:52 | Siecje | Araq: How would you change the result of one of the parameters? |
| 15:26:46 | * | pregressive quit (Remote host closed the connection) |
| 15:33:11 | * | rapind joined #nim |
| 15:37:01 | Araq | Siecje: via a macro? |
| 15:37:33 | * | rapind quit (Remote host closed the connection) |
| 15:40:10 | def- | Araq: my bad, seems to be something with the GC and strings after all |
| 15:40:28 | Araq | lol |
| 15:40:36 | Araq | pity |
| 15:42:13 | * | davidhq quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…) |
| 15:45:54 | * | rapind joined #nim |
| 15:48:25 | * | pregress_ joined #nim |
| 15:48:46 | * | pregress_ quit (Remote host closed the connection) |
| 15:49:05 | * | pregressive joined #nim |
| 15:50:32 | * | rapind quit (Ping timeout: 265 seconds) |
| 15:52:29 | * | kapil___ quit (Quit: Connection closed for inactivity) |
| 15:57:45 | * | NhanH quit (Remote host closed the connection) |
| 15:57:45 | * | endou________ quit (Remote host closed the connection) |
| 15:57:46 | * | CARAM__ quit (Remote host closed the connection) |
| 15:59:02 | * | erisco joined #nim |
| 15:59:16 | erisco | is it possible to use Nim without the garbage collector? |
| 16:02:27 | Araq | erisco: yes. |
| 16:08:37 | * | gsingh93 joined #nim |
| 16:14:14 | * | rapind joined #nim |
| 16:14:26 | * | NhanH joined #nim |
| 16:18:45 | emilsp | is it possible to use ssl/tls with asynchttpserver library ? |
| 16:20:24 | * | rapind quit (Remote host closed the connection) |
| 16:21:31 | * | endou________ joined #nim |
| 16:22:36 | * | Sembei quit (Max SendQ exceeded) |
| 16:23:49 | * | Sembei joined #nim |
| 16:24:17 | erisco | Araq, thanks |
| 16:24:21 | * | erisco left #nim ("Leaving") |
| 16:27:18 | * | CARAM__ joined #nim |
| 16:31:04 | dom96 | emilsp: Not currently. |
| 16:31:34 | dom96 | emilsp: You can put nginx in front of it (which you should be doing anyway) and set up ssl with nginx thoguh. |
| 16:31:36 | dom96 | *though |
| 16:33:32 | emilsp | dom96, ok, I will have to look into that. |
| 16:36:00 | * | dybt joined #nim |
| 16:40:15 | * | mpthrapp joined #nim |
| 16:45:22 | dybt | Hi. In https://github.com/flaviut/nim-by-example/blob/master/content/opengl/triangle.md theres a line with `glGetAttribLocation(shaderProgram, "position").gluint` and I'm not sure what exactly the .gluint is doing. The function returns a GLint, which is defined as `type GLint* = int32`. I see that theres a dot operator, but I don't see any overloads in https://github.com/nim-lang/opengl . Is it some kind of type conversion? |
| 16:46:40 | sillesta | pretty sure it's a type cast |
| 16:46:57 | sillesta | and the following methods takes a GLuint i think, so conversion is needed |
| 16:47:02 | dybt | is that a normal syntax for it? |
| 16:47:23 | dybt | I couldn't see anything like that in the manual |
| 16:47:49 | sillesta | 2.gluint and gluint(2) is the same |
| 16:47:59 | dybt | ooh. Of course |
| 16:48:06 | sillesta | appending .gluint is more painless than wrapping with () i'd say |
| 16:48:22 | dybt | yeah. that makes sense |
| 16:49:05 | dybt | Thanks. I feel silly now for not getting that. haha |
| 16:49:17 | sillesta | i got a bit confused as well :D |
| 16:49:35 | sillesta | opengl is a bit weird about return types vs. parameters as well |
| 16:49:59 | sillesta | return signed int to check for errors with -1, pass as unsigned for some reason |
| 16:50:41 | * | rapind joined #nim |
| 16:51:00 | * | sampwing joined #nim |
| 16:52:51 | dybt | Yeah. I always thought just reserving a location to use as an error value would make more sense |
| 16:53:59 | dybt | although I guess having to do the conversion, you're less likely to forget that it could be an error value and to check for it |
| 16:56:01 | * | _303_ quit (Quit: WeeChat 1.1.1) |
| 16:56:31 | * | rapind quit (Remote host closed the connection) |
| 16:57:17 | * | filwit joined #nim |
| 16:59:47 | filwit | Araq: https://github.com/Araq/Nim/pull/2181 |
| 17:00:02 | * | pregressive quit (Remote host closed the connection) |
| 17:00:39 | filwit | Araq: i need to take off right now, so I can either make you some screenshots when I get back, or you can dup+pull that to test it out yourself. |
| 17:00:45 | * | pregressive joined #nim |
| 17:01:11 | * | pregressive quit (Remote host closed the connection) |
| 17:01:26 | * | pregressive joined #nim |
| 17:01:27 | filwit | either way, bbl |
| 17:01:30 | * | filwit quit (Client Quit) |
| 17:07:34 | * | teapottime quit (Ping timeout: 246 seconds) |
| 17:07:47 | * | rapind joined #nim |
| 17:14:30 | * | rapind quit (Remote host closed the connection) |
| 17:16:34 | * | emilsp quit (Read error: Connection reset by peer) |
| 17:21:00 | * | ChrisMAN quit (Read error: Connection reset by peer) |
| 17:21:41 | * | dybt quit (Quit: Leaving) |
| 17:26:02 | * | Varriount|Mobile joined #nim |
| 17:26:30 | * | MajorWork quit (Quit: Leaving) |
| 17:26:38 | * | Matthias247 joined #nim |
| 17:29:29 | * | Jehan_ quit (Quit: Leaving) |
| 17:29:35 | * | rapind joined #nim |
| 17:38:49 | * | Jehan_ joined #nim |
| 17:40:36 | * | pregressive quit (Remote host closed the connection) |
| 17:47:19 | * | emilsp joined #nim |
| 17:47:38 | Araq | Jehan_: aha so sysstr line 80 is guilty? very interesting |
| 17:48:10 | Jehan_ | Essentially, yes. |
| 17:48:16 | Araq | and also very easy to fix :-) |
| 17:49:07 | Jehan_ | I'm also thinking that it would be nice if it were possible to avoid the bzero() in this case (though that's not easy). |
| 17:49:36 | * | sepisoad joined #nim |
| 17:50:41 | sepisoad | how to use Compile pragma with a file that is nested inside another directory: {.compile: "c_src/file.c".} |
| 17:51:36 | sepisoad | nim seems to fails at compiling a c source file that is not in the current directory or i'm wrong |
| 17:53:25 | Araq | Jehan_: doesn't look too hard, but requires a separate entry point into the GC |
| 17:53:53 | Jehan_ | Araq: Yeah, that's what I meant by "not easy". :) |
| 17:57:50 | Araq | hrm I wonder why I picked the capacity and not the size for copying |
| 17:58:47 | * | rapind quit (Remote host closed the connection) |
| 17:59:01 | Araq | ah yeah I thought "capacity is only 1.5 times as big as length, so it cannot hurt and it gives us a bit room to grow" |
| 17:59:24 | Araq | that's of course completely wrong when you consider setLen |
| 18:00:07 | * | Siecje left #nim (#nim) |
| 18:03:06 | Araq | sepisoad: pretty sure that works, but sometimes you need to use -f to force recompilation of that file |
| 18:04:24 | sepisoad | how to use -f |
| 18:05:21 | Araq | nim c -f foo.nim |
| 18:07:15 | * | rapind joined #nim |
| 18:11:09 | sepisoad | oops, my bad, i was totally wrong |
| 18:11:54 | * | rapind quit (Remote host closed the connection) |
| 18:13:41 | * | d3m1gd joined #nim |
| 18:15:46 | * | Matthias247 quit (Read error: Connection reset by peer) |
| 18:17:39 | * | rapind joined #nim |
| 18:17:47 | elasticdog | stupid question...when updating an existing Nim install (from Git), do you still need the csources, or can you just compile koch and then do `koch boot -d:release`? |
| 18:19:42 | dom96 | elasticdog: That's not stupid at all. The answer is unfortunately: it depends. |
| 18:20:45 | dom96 | elasticdog: Most of the time `koch boot` should work. You may need to occasionally rebuild from the csources. |
| 18:21:20 | elasticdog | dom96: so the "safest" way to update is to `git pull` and then just run the build.sh each time? |
| 18:21:32 | elasticdog | when would the csources be necessary? |
| 18:22:01 | dom96 | elasticdog: I wouldn't call it safest. Nothing bad will happen to your Nim compiler if you try to `koch boot` and the bootstrap process fails. |
| 18:22:36 | dom96 | The csources are necessary when your compiler can no longer build the latest compiler due to a change which breaks bootstrapping. |
| 18:22:46 | dom96 | This happens rarely though. |
| 18:23:48 | dom96 | You should be simply git pulling and executing koch boot. If koch boot fails then you can grab the latest C sources and bootstrap from them. |
| 18:24:16 | elasticdog | gotcha...I was thinking the csources were essentially obsolete now that the Nim compiler exists, but only used for the initial bootstrapping of a fresh install |
| 18:24:30 | elasticdog | but I see that the csources are also getting updates...thanks for the details |
| 18:24:35 | dom96 | That is pretty much the case. |
| 18:24:37 | * | pregressive joined #nim |
| 18:25:06 | dom96 | If the compiler's code uses features that were just implemented in the compiler then C sources need to be rebuilt. |
| 18:25:56 | * | pregressive quit (Client Quit) |
| 18:26:18 | * | kjo quit (Quit: Leaving.) |
| 18:26:55 | * | pregressive joined #nim |
| 18:27:44 | * | kjo joined #nim |
| 18:34:24 | * | ChrisMAN joined #nim |
| 18:34:28 | * | ChrisMAN_ joined #nim |
| 18:34:55 | * | ChrisMAN_ quit (Client Quit) |
| 18:35:27 | * | dumdum quit (Ping timeout: 256 seconds) |
| 18:45:50 | * | nimnoob123 joined #nim |
| 18:46:37 | sepisoad | i found a weird bug in .compile pragma, when i defined a c source file with ".cc" extension the linker failed to find symbols, but after i renamed my c source file extention to ".c" it worked |
| 18:47:12 | Araq | sepisoad: I think that's a feature ... |
| 18:47:19 | Araq | a misfeature |
| 18:47:29 | sepisoad | how come? |
| 18:47:52 | Araq | let me see |
| 18:48:16 | dom96 | Osproc.startProcess broken on Windows? https://gist.github.com/dom96/e4609fbd573c13585dc0 |
| 18:49:39 | Araq | sepisoad: no, I misremembered, but a 'cc' extension tells GCC it's a C++ file |
| 18:49:40 | dom96 | oh. I see the problem. |
| 18:49:45 | dom96 | That error could be improved. |
| 18:50:54 | * | ehaliewicz joined #nim |
| 18:50:56 | Araq | sepisoad: so that definitely explains why you get linker errors as it compiled with the c++ compiler |
| 18:51:09 | Araq | that's not a bug in the .compile pragma though |
| 18:51:55 | nimnoob123 | can anyone recommend a nim project on github to learn from which isn't a binding of a library |
| 18:52:22 | def- | nimnoob123: learn what from? |
| 18:52:34 | sepisoad | an the c++ name mangling makes finding symbols impossible, right? |
| 18:53:05 | BlaXpirit | def-, just link your article :p |
| 18:53:12 | def- | BlaXpirit: haha, no! |
| 18:53:13 | nimnoob123 | things I don't understand, easier to read code and see real examples of how something works in action |
| 18:54:00 | nimnoob123 | and im wondering if there are any projects on nim worth checking out that could be a good learning experience from reading the source |
| 18:54:28 | reactormonk | Araq, what data type is there to hold binary data? |
| 18:54:38 | BlaXpirit | reactormonk, string? |
| 18:54:48 | reactormonk | BlaXpirit, that's null-terminated, isn't it? |
| 18:55:05 | * | brson joined #nim |
| 18:55:12 | Araq | no, it's not really |
| 18:55:13 | def- | nimnoob123: there is a lot of code in the Nim repo itself, and you can see most external nim programs here: http://nim-lang.org/lib.html#nimble (I haven't read the code of much outside the standard library honestly) |
| 18:55:20 | BlaXpirit | length is determined by a number, not by zero byte |
| 18:55:31 | BlaXpirit | you can store binary data in string, but not in cstring |
| 18:55:32 | Araq | reactormonk: use 'string', the trailing zero byte is for C interop |
| 18:55:41 | BlaXpirit | there is always array[char] ........ |
| 18:55:42 | def- | nimnoob123: i also have an article where I develop a small project: http://howistart.org/posts/nim/1 |
| 18:55:45 | BlaXpirit | or uint8 |
| 18:55:55 | reactormonk | Araq, apparently items relies on the zero? |
| 18:56:10 | * | NimBot joined #nim |
| 18:56:22 | * | fizzbooze joined #nim |
| 18:56:25 | BlaXpirit | reactormonk, apparently not |
| 18:56:30 | Araq | reactormonk: no, that was just a bug for repr because that's included before the compiler knows about the items for string |
| 18:56:48 | Araq | so it picked the items for 'cstring' instead. nasty bug |
| 18:56:54 | Araq | but not too hard to find |
| 18:56:58 | reactormonk | Araq, ok, thanks for clarifing that. |
| 18:57:27 | * | mpthrapp quit (Remote host closed the connection) |
| 18:58:32 | * | pregressive quit (Remote host closed the connection) |
| 19:00:26 | * | pregressive joined #nim |
| 19:00:35 | * | pregressive quit (Remote host closed the connection) |
| 19:01:04 | * | pregressive joined #nim |
| 19:01:17 | * | pregressive quit (Remote host closed the connection) |
| 19:01:47 | * | pregressive joined #nim |
| 19:03:24 | d3m1gd | is it possible inside a macro to tell if function is defined or not, and define a stub if not defined? Or maybe someone has a better way of top-down development |
| 19:04:10 | BlaXpirit | d3m1gd, compiles |
| 19:04:13 | def- | d3m1gd: when declared(myFunction)? |
| 19:04:29 | d3m1gd | thanks! |
| 19:04:41 | Araq | that doesn't work as you think it does, def- |
| 19:04:41 | BlaXpirit | well that's even better |
| 19:04:44 | BlaXpirit | oh. |
| 19:04:53 | def- | Araq: too bad, why not? |
| 19:04:56 | Araq | but the macro can *emit* this declared check |
| 19:05:11 | Araq | def-: because it gets resolved when the macro is compiled |
| 19:05:20 | Araq | def-: not when it's expanded |
| 19:06:02 | def- | that's what i expected, I guess I should have written a few more words |
| 19:06:17 | BlaXpirit | in any case, `compiles` works for me https://github.com/BlaXpirit/nim-random/blob/c200e2036e77a54b0d1dc5ffcc6da8893172ea32/src/random/common.nim#L41 |
| 19:06:48 | Araq | BlaXpirit: since you are using it, I should mark it .experimental. (just kidding) |
| 19:07:56 | * | rapind quit (Remote host closed the connection) |
| 19:08:39 | d3m1gd | sooo, will it be possible to wrap a block in a macro and traverse ast to find non-defined functions and autodefine them, or is it something crazy? |
| 19:09:21 | BlaXpirit | I don't see a need for a macro |
| 19:09:23 | Araq | that's something you really should do differently, I think |
| 19:09:36 | d3m1gd | i am open for suggestions =) |
| 19:09:42 | BlaXpirit | methods or generics work wonders |
| 19:11:39 | d3m1gd | ok, that is true. And i wanted to learn some macro fu on real life example.. |
| 19:12:20 | Araq | d3m1gd: you can read & parse the module at compile-time, modify it and write back the result. in theory. |
| 19:12:32 | Araq | dunno if anybody ever tested this |
| 19:13:38 | d3m1gd | Araq: like: myMacro: import myModule? |
| 19:14:15 | Araq | no, like using staticRead and writeFile in a macro |
| 19:15:45 | d3m1gd | Araq: staticRead, omg, i am amazed by your creation more by every day |
| 19:16:34 | d3m1gd | Araq: and actually i would like to help, if there is something easy enough, which could also help learning the language |
| 19:16:56 | * | rapind joined #nim |
| 19:17:00 | Araq | learn the language and then come back and look at the bugs marked with "easy" or "stdlib" |
| 19:17:25 | Araq | I think we got rid of the very hard "easy" issues by now |
| 19:17:52 | d3m1gd | will do |
| 19:23:20 | * | Outlander joined #nim |
| 19:26:36 | * | sepisoad quit (Quit: Leaving) |
| 19:28:00 | * | pafmaf quit (Quit: This computer has gone to sleep) |
| 19:35:25 | d3m1gd | has some question about anon proces from futures. tiny case is here: https://gist.github.com/d3m1gd/e8d280b53ac3d00697ff |
| 19:35:31 | d3m1gd | it does not compile |
| 19:36:34 | def- | d3m1gd: applyTo10((a: int) => echo(a)) |
| 19:37:02 | def- | but i made an issue about this a few hours ago: https://github.com/Araq/Nim/issues/2179 |
| 19:37:13 | d3m1gd | def-: nope |
| 19:37:20 | d3m1gd | def-: still does not |
| 19:37:29 | def- | d3m1gd: works for me |
| 19:37:37 | def- | d3m1gd: on current devel branch of Nim |
| 19:38:12 | def- | you're right, it's broken in 0.10.2 |
| 19:39:19 | Araq | it's broken for me too ... and I'm on 0.10.3 beyond devel |
| 19:41:34 | Araq | oh wait, you added the :int |
| 19:41:50 | Araq | yeah that works |
| 19:46:35 | * | Jesin quit (Remote host closed the connection) |
| 19:50:24 | * | sepisoad joined #nim |
| 19:53:32 | * | rapind quit (Remote host closed the connection) |
| 19:54:01 | sepisoad | do we have to call dealloc() on a cstring, like when we pass a cstring to a c function and the function does a heap allocation and ... |
| 19:55:01 | * | renesac joined #nim |
| 19:55:49 | Araq | sepisoad: if you don't do the alloc() you MUST not call dealloc() either |
| 19:56:02 | Araq | it's not a wrapper for C's free()! |
| 19:56:48 | sepisoad | but the c function will allocate some memory and won't free it up, how that memory is handled |
| 19:57:17 | sepisoad | in c i would free that memory manually when i'm done with its value |
| 19:57:52 | Araq | that's sloppy design, in Nim you wrap the free() and then do the same of course |
| 19:58:27 | d3m1gd | Araq: so adding int then. Does it not destroy reason for lightweight syntax for anonymous proces? Like they can figure it out from calling function definition |
| 19:58:35 | Araq | but the C lib won't work reliably on Windows since there is not a single "C memory manager" but multiple depending on which runtime you link against |
| 19:59:42 | Araq | d3m1gd: so it's in development. which part of 0.10.x didn't you understand? |
| 20:07:03 | * | kjo quit (Quit: Leaving.) |
| 20:13:00 | * | sepisoad quit (Ping timeout: 264 seconds) |
| 20:16:01 | * | sampwing quit (Ping timeout: 252 seconds) |
| 20:16:19 | * | sepisoad joined #nim |
| 20:16:24 | * | sepisoad quit (Read error: Connection reset by peer) |
| 20:21:36 | * | Boscop_ joined #nim |
| 20:22:54 | * | Jesin joined #nim |
| 20:24:55 | * | jsudlow quit (Remote host closed the connection) |
| 20:25:17 | * | Boscop quit (Ping timeout: 245 seconds) |
| 20:26:58 | * | panzone joined #nim |
| 20:30:46 | * | kjo joined #nim |
| 20:32:53 | * | Boscop_ quit (Read error: Connection reset by peer) |
| 20:33:32 | * | Varriount|Mobile quit (Quit: AndroIRC - Android IRC Client ( http://www.androirc.com )) |
| 20:35:43 | * | rapind joined #nim |
| 20:41:56 | * | jholland joined #nim |
| 20:42:42 | * | nimnoob123 quit (Quit: Page closed) |
| 20:56:57 | d3m1gd | Araq: sorry for asking, just wanted to clarify this one out, because it was working and got broken. I am ok with it if you know about the issue |
| 20:57:53 | Araq | d3m1gd: afaik it never worked |
| 20:57:58 | Araq | did it work? |
| 20:58:50 | d3m1gd | Araq: it did, if return type is specified on calling proc's argument. I think. Will do a quick test |
| 21:00:40 | d3m1gd | Araq: works on 0.10.2: https://gist.github.com/d3m1gd/e8d280b53ac3d00697ff |
| 21:03:58 | * | BlaXpirit quit (Quit: Quit Konversation) |
| 21:04:21 | * | BlaXpirit joined #nim |
| 21:11:02 | * | Boscop joined #nim |
| 21:16:29 | Araq | d3m1gd: hrm ok |
| 21:19:36 | * | Demon_Fox joined #nim |
| 21:19:50 | * | dumdum joined #nim |
| 21:23:40 | * | reem joined #nim |
| 21:36:34 | * | kjo quit (Quit: Leaving.) |
| 21:36:52 | * | panzone quit (Quit: Leaving...) |
| 21:40:24 | * | flaviu quit (Remote host closed the connection) |
| 21:40:39 | * | flaviu joined #nim |
| 21:44:13 | * | infinity0 quit (Ping timeout: 255 seconds) |
| 21:46:28 | * | infinity0 joined #nim |
| 21:48:42 | * | sampwing joined #nim |
| 21:52:06 | * | nande joined #nim |
| 21:55:37 | * | Mat4 joined #nim |
| 21:55:43 | Mat4 | hi all |
| 22:00:57 | Araq | servus Mat4 |
| 22:02:53 | Varriount | Araq: Is there a calling convention imposed on iterators? |
| 22:03:11 | Varriount | Actually, nevermind, I'll just look that up myself. |
| 22:03:50 | * | chemist69 quit (Quit: WeeChat 1.1.1) |
| 22:04:20 | Araq | Varriount: thank you for fixing the build status buttons :-) |
| 22:04:30 | Mat4 | hi Araq |
| 22:06:25 | * | reem quit (Remote host closed the connection) |
| 22:07:01 | * | reem joined #nim |
| 22:07:14 | Varriount | Araq: I was thinking that iterators could be directly sent values after they've started, by taking advantage of the fact that cdecl is a caller-cleanup convention |
| 22:08:43 | Varriount | But since iterators don't have a fixed calling convention, it wont work... |
| 22:10:46 | * | reem quit (Remote host closed the connection) |
| 22:12:24 | * | reem joined #nim |
| 22:12:37 | * | sampwing quit (Ping timeout: 252 seconds) |
| 22:13:07 | * | reem quit (Read error: Connection reset by peer) |
| 22:13:33 | * | reem joined #nim |
| 22:14:02 | Araq | Varriount: I see what you mean but I don't see why thats necessary |
| 22:16:48 | * | Jehan_ quit (Quit: Leaving) |
| 22:17:00 | * | TEttinger joined #nim |
| 22:19:29 | * | sps joined #nim |
| 22:19:53 | * | sps is now known as Guest91095 |
| 22:20:25 | * | Guest91095 quit (Client Quit) |
| 22:22:25 | * | reem quit (Remote host closed the connection) |
| 22:22:41 | * | dumdum quit (Ping timeout: 256 seconds) |
| 22:25:00 | * | enquora joined #nim |
| 22:25:33 | Araq | Varriount: we need your string assignment optimization in the compiler. what was the reason it never worked? |
| 22:26:36 | * | reem joined #nim |
| 22:28:20 | * | kjo joined #nim |
| 22:30:11 | * | saml quit (Quit: Leaving) |
| 22:31:32 | * | filwit joined #nim |
| 22:34:37 | Araq | filwit: thanks for the website update |
| 22:34:48 | filwit | Araq: i see you uploaded it, so assume you didn't have any question or comments about the site changes |
| 22:35:00 | filwit | np, that went faster than last time :) |
| 22:35:33 | filwit | i'll hopefully have the forum PR this weekend |
| 22:35:37 | flaviu | filwit: fyi, there are better ways to merge upstream into the current branch |
| 22:35:52 | filwit | the tabs being out-of-order with the website really bug me now :| |
| 22:36:00 | filwit | flaviu: eh? |
| 22:36:23 | flaviu | https://gist.github.com/flaviut/0bfeca0cf9bc5fe53a51 |
| 22:36:38 | flaviu | see "If a PR grows out of " |
| 22:37:32 | filwit | flaviu: i don't know enough about git to understand what that's doing exactly, I'll need to look things up |
| 22:37:54 | filwit | flaviu: i was under the impressions that rebasing was only for stripping unneeded cruft from a large repo.. |
| 22:37:56 | ldlework | filwit: Pro Git is potentially the best piece of technical writing I've read |
| 22:38:05 | ldlework | http://git-scm.com/book/en/v2 |
| 22:38:14 | flaviu | It basiclly replays all your commits on top of the given reference (commit hash, branch, etc). |
| 22:38:16 | filwit | ldlework: thanks for the link, i'll take a look |
| 22:38:22 | * | rapind quit () |
| 22:39:06 | filwit | flaviu: can you explain what the downside of doing it the way I have been? |
| 22:39:36 | flaviu | It's not aesthetically pleasing |
| 22:40:12 | flaviu | Not a big deal really, but it often happens out of ignorance, not conscious choice. |
| 22:40:14 | filwit | flaviu: in what way? does it create a branch or something? basically all i do is merger with upstream, push to my fork, then create a PR |
| 22:40:27 | filwit | merge* |
| 22:42:26 | filwit | flaviu: perhaps I forgot to pull upstream before making the PR this time (was going fast) |
| 22:44:01 | flaviu | filwit: 1 sec, I'm creating a picture. |
| 22:48:32 | flaviu | filwit: http://i.imgur.com/Rx6D4Md.png |
| 22:49:10 | * | Mat4 is now known as Mat4-coding |
| 22:49:26 | filwit | flaviu: okay I'll take a look, but let's discuss this privately or on nim-offtopic |
| 22:49:26 | * | ehaliewicz quit (Remote host closed the connection) |
| 22:55:49 | Varriount | Meep |
| 22:56:30 | Varriount | Araq: I don't know. It kept crashing. |
| 22:56:34 | * | sampwing joined #nim |
| 22:56:46 | Varriount | Araq: I don't even know if I have it anymore. |
| 22:57:03 | Araq | Varriount: no worries |
| 22:58:23 | Varriount | Araq: One thing about the optimization is that there needs to be a decrease size threshold. |
| 23:00:07 | * | saml joined #nim |
| 23:01:01 | * | sampwing quit (Ping timeout: 252 seconds) |
| 23:02:31 | * | willwillson joined #nim |
| 23:02:50 | * | ob_ joined #nim |
| 23:03:38 | saml | how can I make this parallel? https://github.com/saml/helloweb/blob/master/hello.nim |
| 23:04:05 | saml | nim hello world web server is faster than node.js but much slower than java |
| 23:04:15 | saml | cause java uses all cores by default |
| 23:07:43 | Varriount | saml: The async framework doesn't have support for multithreading yet. |
| 23:07:52 | saml | ah thanks |
| 23:07:55 | * | saml quit (Quit: Leaving) |
| 23:08:04 | Varriount | ._. |
| 23:08:14 | * | Boscop_ joined #nim |
| 23:08:16 | ldlework | Varriount: did you want to say something else? |
| 23:08:39 | Varriount | ldlework: Well, I was going to say that even if Nim is slower, it probably uses less memory than Java |
| 23:08:45 | ldlework | Also non-blocking IO is a completely different concept than concurrency |
| 23:08:56 | Varriount | ldlework: Yes, I know that. |
| 23:09:07 | ldlework | Oh okay. |
| 23:09:20 | ldlework | Do we have any sort of coroutine support yet? |
| 23:09:36 | BlaXpirit | iterators..... |
| 23:09:37 | ldlework | I mean couldn't he have just used threads? |
| 23:09:39 | federico3 | shouldn't it run behind a webserver that spawns one process per core? |
| 23:10:00 | ldlework | federico3: for languages and platforms that can't do concurrency 'in-house' sure |
| 23:10:14 | Varriount | ldlework: We don't have built in coroutines yet either. |
| 23:10:19 | * | Demon_Fox quit (Quit: Leaving) |
| 23:10:20 | ldlework | Varriount: gotcha |
| 23:10:39 | BlaXpirit | how are iterators not coroutines? |
| 23:10:44 | Varriount | ldlework: Though, what do you mean by coroutines? |
| 23:10:58 | federico3 | ldlework: you might want to restart the workers from time to time (e.g. when deploying new code) without disruption |
| 23:11:20 | * | Boscop quit (Ping timeout: 246 seconds) |
| 23:11:36 | Varriount | Or wrap the servers in an https proxy |
| 23:11:46 | BlaXpirit | Nim has iterators almost on par with Python's, and people say Python has coroutines... |
| 23:12:07 | ldlework | federico3: sure, but you can use multiple hosts behind balancing for that |
| 23:12:38 | * | banister joined #nim |
| 23:12:45 | * | banister quit (Max SendQ exceeded) |
| 23:12:50 | ldlework | I guess I didn't mean coroutines, since those are cooperative and involve one thread |
| 23:13:04 | * | banister joined #nim |
| 23:13:12 | federico3 | ldlework: of course, but sometimes you want to do both |
| 23:13:29 | ldlework | federico3: most clustering and failover tooling works with high-availability products |
| 23:13:37 | ldlework | not, say nginx or apache... |
| 23:13:56 | ldlework | So I wouldn't recommend going multi-process for the mere justification of continuous deployment |
| 23:14:51 | federico3 | well, I've seen more than 2 layers of load-balancing (in a very broad sense) in place, but I get what you mean |
| 23:15:25 | flaviu | def-: Did you get the GSOC app in? |
| 23:16:00 | flaviu | Also, is anyone here a potential student? |
| 23:16:00 | Varriount | ldlework: Coroutines can decide where execution has continued after they've yielded control. |
| 23:16:07 | ldlework | wow goroutines are preemptive!? |
| 23:16:30 | ldlework | Varriount: right, they're cooperative and generall single process. It was the wrong word to use. |
| 23:17:06 | ldlework | I can't really think of a language that has natural first-class parellel tasking |
| 23:17:18 | ldlework | that doesn't involve making a thread and passing a callable to it, or something |
| 23:17:27 | gokr | Didn't Fortress do lots of that? |
| 23:17:34 | ldlework | never even heard of it |
| 23:17:58 | Varriount | flaviu: I am a potential student |
| 23:18:02 | gokr | ldlework: http://en.wikipedia.org/wiki/Fortress_%28programming_language%29 |
| 23:18:11 | gokr | Guy Steele's baby |
| 23:18:20 | federico3 | greenlets and asyncio like in Python? |
| 23:18:26 | gokr | Loops were by default parallell in Fortress. |
| 23:18:34 | Varriount | Woah. |
| 23:19:12 | ldlework | federico3: both greenlets and asyncio are both cooperative concurency, not parelellism |
| 23:19:14 | gokr | It was a very interesting language, but ceased development in 2012 |
| 23:19:32 | federico3 | oh, *true* parallelism |
| 23:19:33 | ldlework | oh greenlets are parellel |
| 23:19:36 | * | nande quit (Ping timeout: 264 seconds) |
| 23:19:44 | ldlework | wait no |
| 23:19:47 | federico3 | ldlework: nope |
| 23:19:50 | ldlework | tasklets are, but not greenlets |
| 23:19:56 | ldlework | jesus this nomenclature is fucking horrid |
| 23:20:20 | Varriount | ldlework: Welcome to computer science. Please take a ticket and sit in the waiting area. |
| 23:20:26 | ldlework | heh |
| 23:21:01 | ldlework | Anyway, we should Nim tasklets. That's be neat. |
| 23:21:30 | Varriount | Python's generators became coroutines with the inclusion of 'yield from' |
| 23:21:45 | flaviu | ldlework: do I hear you volunteering to implement them? :P |
| 23:21:50 | ldlework | haha |
| 23:22:00 | ldlework | Varriount: indeed. pretty awesome imo. |
| 23:22:15 | * | nande joined #nim |
| 23:22:20 | ldlework | flaviu: I left out the verb |
| 23:22:24 | ldlework | perhaps I meant 'dream about' |
| 23:22:36 | federico3 | ldlework: oh and your ticket number is 33 and you've been allocated seat n. |
| 23:22:41 | federico3 | Segmentation fault |
| 23:22:59 | * | ldlework goes to CS purgatory and waits for the great GC in the sky. |
| 23:23:38 | Varriount | ldlework: I guess coroutines could be implemented in Nim by having the compiler place trampolining mechanisms wherever an iterator returns |
| 23:23:57 | * | BlaXpirit quit (Quit: Quit Konversation) |
| 23:24:36 | ldlework | But they still wouldn't be parellel and wouldn't help that guy get more performance |
| 23:24:43 | Varriount | *shrug* |
| 23:24:50 | * | reem quit (Remote host closed the connection) |
| 23:25:01 | * | sillesta quit (Ping timeout: 250 seconds) |
| 23:25:20 | Varriount | ldlework: Think of a way to incorporate threading into the async framework. That's the key, |
| 23:25:27 | ldlework | Varriount: yeah |
| 23:25:32 | TEttinger | ldlework: I think APL or J could be implemented to run array operations in parallel by default? there's quite a lot of abstraction between J verbs and the hardware |
| 23:25:46 | ldlework | TEttinger: no idea what APL or J means |
| 23:25:51 | TEttinger | languages |
| 23:25:55 | ldlework | oh |
| 23:26:07 | TEttinger | in referense to <ldlework> I can't really think of a language that has natural first-class parellel tasking |
| 23:26:10 | Varriount | ldlework: Perhaps dispatch operations that are GC-safe could be automatically run in a separate thread? |
| 23:26:15 | TEttinger | they don't currently though |
| 23:26:23 | * | reem joined #nim |
| 23:26:30 | renesac | filwit: about the forum layout, once you have the reply box open you can't see the topic's page numbers/links anymore |
| 23:26:33 | ldlework | Yeah I'm sure they're ARE languages that have first class parellelism but I just couldn't think of any |
| 23:26:52 | renesac | or is it a problem for dom96 ? |
| 23:27:40 | filwit | renesac: i'll see about fixing it, thanks |
| 23:29:21 | * | sampwing joined #nim |
| 23:30:51 | * | nande_ joined #nim |
| 23:31:29 | * | nande quit (Ping timeout: 245 seconds) |
| 23:40:58 | * | pregressive quit (Read error: Connection reset by peer) |
| 23:41:57 | * | kuzy000_ quit (Ping timeout: 245 seconds) |
| 23:45:55 | def- | flaviu: yes, application is in |
| 23:47:41 | * | brson quit (Quit: leaving) |
| 23:50:28 | * | reem quit (Remote host closed the connection) |
| 23:52:18 | * | reem joined #nim |
| 23:52:20 | * | reem quit (Remote host closed the connection) |
| 23:59:37 | * | pregressive joined #nim |
