<< 20-05-2017 >>
00:07:19*libman joined #nim
00:15:29*SusWombat joined #nim
00:15:35SusWombatHey everyone
00:18:03libmanhi SusWombat
00:23:56FromGitter<zacharycarter> o/
00:32:32*couven92 quit (Quit: Client Disconnecting)
00:38:36*bjz joined #nim
00:44:02libmanNew compiler from AMD is (like the one from Intel) sometimes faster than Clang or GCC. This is good for Nim compared to languages married to a single compiler backend.
00:44:12libmanhttp://www.phoronix.com/scan.php?page=article&item=amd-ryzen-aocc
00:44:54*chemist69 joined #nim
00:54:41*yglukhov quit (Remote host closed the connection)
00:57:41*bjz_ joined #nim
00:58:46*bjz quit (Ping timeout: 246 seconds)
01:00:52*Ven joined #nim
01:01:12*Ven quit (Client Quit)
01:01:32FromGitter<zacharycarter> @stisa : http://imgur.com/a/6uG2A your webgl example one produced with frag
01:01:56FromGitter<zacharycarter> working on writing a webgl sprite batch now
01:03:21*yglukhov joined #nim
01:03:26*nsf quit (Quit: WeeChat 1.7.1)
01:08:13*yglukhov quit (Ping timeout: 272 seconds)
01:43:55*chemist69 quit (Ping timeout: 246 seconds)
01:58:03*chemist69 joined #nim
02:16:29*dddddd quit (Quit: Hasta otra..)
02:17:03*bjz_ quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
03:05:47*pilne quit (Quit: Quitting!)
03:46:08*chemist69 quit (Ping timeout: 240 seconds)
03:48:47*krux02 quit (Remote host closed the connection)
03:50:41*chemist69 joined #nim
04:02:35*bjz joined #nim
04:15:50FromGitter<mratsim> @libman interesting, benchmark choice is so-so though, they should use a benchmark where icc shines like OpenCV.
04:51:59*rauss joined #nim
05:05:22*bjz quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
05:07:29*bjz joined #nim
05:26:13*libman quit (Quit: Connection closed for inactivity)
05:37:05*bjz quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
05:44:21*kier quit (Quit: No Ping reply in 180 seconds.)
05:49:45*kier joined #nim
05:50:46*Tiberium joined #nim
05:55:37*yglukhov joined #nim
05:59:36*yglukhov quit (Ping timeout: 240 seconds)
06:37:59*bjz joined #nim
06:54:47FromGitter<mratsim> Regarding, inlining, proc, templates. Following the discussion in https://forum.nim-lang.org/t/879#5218 ⏎ "Use the least powerful construct that works, which means in this order: ⏎ procs ⏎ templates ⏎ macros” ... [https://gitter.im/nim-lang/Nim?at=591fe83700efc2bb3e871644]
07:02:52FromGitter<Varriount> @mratsim I was just about to talk to you.
07:03:01FromGitter<Varriount> For your refactored code, did you consider doing this: https://gist.github.com/Varriount/352df3399e13f889d989f620d16fa762
07:04:18FromGitter<Varriount> Don't explicitly mark procedures as inline unless you are doing it for performance reasons. Use procedures rather than templates, because template overloading can be a bit finicky.
07:05:19FromGitter<mratsim> @Varriount: Actually yes, but I placed the parenthesis in the wrong place ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=591feaaf5e34568d5e7216ce]
07:06:02FromGitter<Varriount> I like boolean logic rather than inline if statements.
07:07:20FromGitter<mratsim> I wasn’t sure I wouldn’t get tripped by out of bounds exception
07:08:11FromGitter<Varriount> <3 short circuiting logic
07:08:45FromGitter<Varriount> @mratsim Have you taken a look at term-rewriting macros?
07:09:10*yglukhov joined #nim
07:09:11FromGitter<Varriount> https://nim-lang.org/docs/manual.html#term-rewriting-macros
07:10:13FromGitter<Varriount> You might petition Araq on instructions for modifying the compiler to expose a 'matchesAstPattern' procedures
07:11:51*bjz quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
07:13:32*yglukhov quit (Ping timeout: 260 seconds)
07:15:18*bjz joined #nim
07:15:37FromGitter<mratsim> I’ve looked into Term rewriting macros yes but to fuse operation like AX + B in a single optimized call. ⏎ For slicing I didn’t want to have side effects on a normal sequence slicing so I choose to directly manipulate the AST tree. It might be possible with TRM but since my code is working :P
07:18:03FromGitter<Varriount> Couldn't you use term-rewriting macros for all that desugering?
07:22:16*xet7 quit (Quit: Leaving)
07:23:19*xet7 joined #nim
07:25:49FromGitter<mratsim> Like I said, it might be possible, but I had 2 concerns, not impacting normal sequences, i.e. not rewriting seq[1..10] into seq[1..10|1] and second traversing the input as few times as possible. First concern is easy but for second concern it seemed easier to manipulate the AST tree. ⏎ And it’s obviously a good excuse to learn Nim AST manipulation ;)
07:28:21FromGitter<Varriount> True
07:31:17FromGitter<Varriount> @mratsim I'm working with a really nice data validation library at the moment
07:31:28FromGitter<Varriount> https://github.com/kolypto/py-good
07:31:36FromGitter<Varriount> I wonder if I could port it to Nim
07:43:34FromGitter<mratsim> @Varriount: interesting, what do you use it for?
07:43:48FromGitter<Varriount> Configuration normalization and validation
07:44:51FromGitter<Varriount> For things like JSON and YAML, I often prefer to give users as much freedom as possible (like accepting both a single values or a list of value for a given key)
07:49:55FromGitter<Varriount> @mratsim https://gist.github.com/Varriount/642802396582bee73c1065471d7b3e70
07:57:51FromGitter<mratsim> I see, that is indeed very useful.
08:06:48*Matthias247 joined #nim
08:08:21FromGitter<Varriount> @mratsim Did you ever manage to get arrays to support heterogenous types?
08:09:14FromGitter<mratsim> I didn’t try and it’s not a target, but my ndarrays support objects
08:09:39*Trustable joined #nim
08:09:51FromGitter<mratsim> You can do arbitrary transformations with slicing and fmap
08:11:56FromGitter<mratsim> The heterogeneous issue I had 2 months ago was with Andrea’s lining. Creating a list of BackProp closures with Matrix[2,3] and Matrix[3,3] 2, and 3 being static int part of the type
08:13:48FromGitter<mratsim> That was one of the reason I started my own + I needed 4D and 5D arrays anyway for Deep Learning
08:14:58FromGitter<Varriount> I think linalg is more for game dev stuff
08:16:04FromGitter<mratsim> Actually no, andrea published Neurotic, and experimental neural net library using linalg: https://github.com/unicredit/neurotic It can do digit recognition
08:17:54FromGitter<Varriount> Huh.
08:19:00FromGitter<mratsim> In hindsight, after playing with PyTorch a lot (new framework from Facebook from Jan this year) I think closures are not the best way, I should probably use objects with `forward` and `backward` method/proc instead of `forward`procs that creates a seq of closure.
08:23:09FromGitter<mratsim> Or the new and shiny vtref
08:26:09*kunev quit (Ping timeout: 268 seconds)
08:26:33FromGitter<stisa> @Varriount in nimble there's both linalg ( fast linear algebra, BLAS etc, some cuda I think? ) and linagl (OpenGL Matrix stuff)
08:26:38*kunev joined #nim
08:29:31*bjz quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
08:35:57*tigerfinch joined #nim
08:40:01Tiberiumehm, what does it mean if I add "--gc:none" and my project compiles with many warnings "x uses GC'ed memory"? So does it remove a GC or not?
08:41:36FromGitter<Varriount> Tiberium: '--gc:none' removes the GC from the code.
08:42:03Tiberiumoh I understand now :)
08:42:10Tiberiummemory usage of my app is just growing
08:42:20Tiberium(because my app is made for usual nim - with gc)
08:43:34*kunev quit (Ping timeout: 246 seconds)
08:45:08*kunev joined #nim
08:45:35*tigerfinch quit (Ping timeout: 240 seconds)
08:48:54*bjz joined #nim
08:58:23*couven92 joined #nim
09:01:28*Tiberium quit (Ping timeout: 260 seconds)
09:03:27*Tiberium joined #nim
09:09:19*Tiberium quit (Remote host closed the connection)
09:09:36*kunev quit (Ping timeout: 240 seconds)
09:11:09*kunev joined #nim
09:11:29*Arrrr joined #nim
09:11:29*Arrrr quit (Changing host)
09:11:29*Arrrr joined #nim
09:14:00Arrrrhttps://github.com/nim-lang/Nim/pull/5852/files I think everyone will look at it as a misspelled `if`. I would replace it with `if and only if`
09:17:20*tigerfinch joined #nim
09:19:55*zidsal joined #nim
09:20:06zidsalhi there
09:20:15*Tiberium joined #nim
09:20:56*bjz quit (Ping timeout: 272 seconds)
09:27:59*bjz joined #nim
09:32:23*yglukhov joined #nim
09:32:31*yglukhov quit (Remote host closed the connection)
09:32:55*chemist69 quit (Ping timeout: 246 seconds)
09:37:33*yglukhov joined #nim
09:39:56*couven92 quit (Ping timeout: 272 seconds)
09:41:01*yglukhov quit (Remote host closed the connection)
09:51:37*chemist69 joined #nim
09:53:04FromGitter<Varriount> hello!
10:07:44*vlad1777d joined #nim
10:34:30FromGitter<cooldome> Guys, i am trying to debug typerel from the sigmatch in the compiler. The problem I am facing that too much goes this function, I trying to find a way to filter only type relations from my test file and every type pair. With NimNode you can use info field to filter but types don't have it. Any ideas?
10:37:52*Ven joined #nim
10:38:16*Ven is now known as Guest55917
10:39:43Tiberiumlol, rust has 1.17.0 version (released in april)
10:44:02*zidsal quit (Remote host closed the connection)
10:44:53*nsf joined #nim
10:46:07FromGitter<ephja> :o
10:46:30*zidsal joined #nim
10:47:16*Guest55917 quit (Ping timeout: 260 seconds)
10:47:51*Ven_ joined #nim
10:52:22*Ven_ quit (Ping timeout: 246 seconds)
10:54:06*Ven_ joined #nim
10:59:05*Ven_ quit (Ping timeout: 268 seconds)
11:01:02*Ven_ joined #nim
11:03:29*Snircle joined #nim
11:05:40*Ven_ quit (Ping timeout: 246 seconds)
11:07:21*Ven_ joined #nim
11:14:33FromGitter<mratsim> I’m happy I’m not coding in Rust anymore. Looking at a (well coded) library, I see things like this: ⏎ ⏎ ```code paste, see link``` ⏎ ⏎ Can you count how many `unwrap`there is? [https://gitter.im/nim-lang/Nim?at=59202519eec422e415cf73ed]
11:16:00*planhths joined #nim
11:17:46ArrrrThat's sick.
11:17:58ArrrrMaybe is not that well coded
11:18:36FromGitter<zacharycarter> that looks like a test
11:18:48FromGitter<zacharycarter> or an example or something
11:19:07FromGitter<zacharycarter> still it doesn't defeat the point that Rust has syntax problems
11:20:03*Ven_ quit (Ping timeout: 268 seconds)
11:20:13FromGitter<mratsim> It’s an example in the README for https://github.com/autumnai/collenchyma. It was a machine learning library that was faster or at least as fast as Facebook’s and Google's back in 2016. (They stopped working on it though)
11:21:29*Ven_ joined #nim
11:21:33FromGitter<zacharycarter> it's more than a syntax issue it's a verbosity issue, Rust stumbles in a number of areas IMO. The lifecycle feature is difficult to grasp / get right, it's too verbose, syntax is ugly.
11:21:39FromGitter<mratsim> The `unwrap` is because everything is wrapped in an Option type that can have either Nothing or an actual value. I guess this is one way to avoid null pointers but the boiler plate ...
11:21:58FromGitter<zacharycarter> yes
11:22:31*dddddd joined #nim
11:22:37FromGitter<zacharycarter> I considered Rust for a moment, went through a few OpenGL tutorials, looked at gfx-rs and then came here
11:22:53FromGitter<mratsim> I’m not even talking about all the sigils `::` `<Vec<f32>>`… I have cognitive overload
11:23:05FromGitter<zacharycarter> :P
11:23:51FromGitter<zacharycarter> it is a prime example of how ugly the language can be
11:24:06FromGitter<ephja> yeah ownership semantics etc seems too verbose for such tasks
11:24:12FromGitter<mratsim> I wrote a go bot in Rust last year but the way arrays works in Rust is just crazy. Beyong 32 elements in your array you are on your own. You have to implement Copy, Clone, Display, everything ...
11:24:23FromGitter<zacharycarter> what?
11:24:32FromGitter<zacharycarter> that doesn't seem right
11:24:34FromGitter<mratsim> My full rant with details: https://andre-ratsimbazafy.com/why-rust-fails-hard-at-scientific-computing/
11:24:56FromGitter<mratsim> https://andre-ratsimbazafy.com/wp-content/uploads/2017/04/2017-04-01_15-31-14.png
11:24:59*couven92 joined #nim
11:25:46FromGitter<zacharycarter> wow
11:26:06*Ven_ quit (Ping timeout: 240 seconds)
11:26:17FromGitter<mratsim> Rust still (?) doesn’t have generics with integer parameters so that’s how they implement Arrays in the official repo: ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=592027d9fa63ba2f766010dd]
11:27:15FromGitter<ephja> hmm
11:28:25FromGitter<mratsim> Sorry edited, pasted the code twice
11:30:43FromGitter<zacharycarter> I remember when i used to write a bit of JS and I loved browserify
11:30:58FromGitter<zacharycarter> now I'm not so sure
11:33:59FromGitter<zacharycarter> I'd like to leverage some modules in http://stack.gl/
11:34:26FromGitter<zacharycarter> and I've successfully gotten some modules to import but getting them to actually work is another bear all together
11:34:27*zidsal quit (Quit: Leaving)
11:42:37FromGitter<dom96> @mratsim nice, post it to Reddit/HN :)
11:43:12FromGitter<mratsim> If I post that to HN people will say I’m a Nim troll :D
11:58:22*nsf quit (Quit: WeeChat 1.7.1)
12:00:21FromGitter<zacharycarter> @dom96: any idea how you'd bind to something like this -
12:00:29*Ven joined #nim
12:00:31FromGitter<zacharycarter> https://github.com/mattdesl/gl-shader-core#api
12:00:52*Ven is now known as Guest16010
12:00:59FromGitter<zacharycarter> it's the extra parameters I have no idea how to pass
12:01:36FromGitter<TiberiumN> maybe "from x import y"? but I don't know :)
12:02:12FromGitter<zacharycarter> hrm well here's what I have so far
12:02:25FromGitter<zacharycarter> proc require(moduleName: string): JsObject {.importcpp: "require(@)".}
12:02:34FromGitter<zacharycarter> which works if there are no extra args
12:05:21*BennyElg quit (Ping timeout: 245 seconds)
12:06:07FromGitter<dom96> Overloading maybe?
12:06:21FromGitter<dom96> On phone now so can't help much
12:08:14FromGitter<zacharycarter> okay thanks
12:08:18FromGitter<zacharycarter> I'll give that a shot
12:14:05*Guest16010 quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
12:30:43*nsf joined #nim
12:40:42FromGitter<mratsim> Got an issue in a macro: ⏎ ⏎ Trying to typematch `foo[1,3] = 999` ⏎ ⏎ This works ... [https://gitter.im/nim-lang/Nim?at=5920394a00efc2bb3e8824ce]
12:46:12*BennyElg joined #nim
12:47:57FromGitter<cooldome> varargs[untyped] consumes all inputs including last one
12:51:34FromGitter<ephja> try an openarray instead
12:51:41FromGitter<mratsim> Actually it works now I messed up the order of my args inside the macro, and the error message didn’t match with that
12:52:20FromGitter<mratsim> argh no forget it I was still testing with varargs[int]
12:54:17FromGitter<ephja> val=999?
12:54:46FromGitter<ephja> I wonder if it works, but it might be confusing in any case
12:54:47FromGitter<mratsim> I really need a varargs though, foo[@[1,3]] = 999 isn’t nice. val = 999 is just for testing
12:55:35FromGitter<cooldome> I don't have a better idea then macro `[]=`*B, T (t: var Tensor[B,T], args: varargs[untyped])
12:56:48FromGitter<cooldome> then unpack last argument manually and check it's type is T
12:57:02FromGitter<mratsim> yes that’s what I will do
12:57:07FromGitter<ephja> oh right, can't do val=999 with that syntax. perhaps that signature could be allowed as a special case for those operators
12:57:38FromGitter<ephja> and it might be unnecessary to pass in a sequence rather than an array
12:57:39FromGitter<mratsim> I’ll post a feature request
13:03:28*BennyElg_ joined #nim
13:03:35*BennyElg quit (Ping timeout: 240 seconds)
13:04:27*planhths quit (Ping timeout: 240 seconds)
13:05:59*Ven joined #nim
13:06:22*Ven is now known as Guest87036
13:07:05*skrylar joined #nim
13:09:03FromGitter<cooldome> that would be a useful feature
13:09:16*skrylar splashes water and arrives
13:10:48enthus1asthave you ever iterate a (Ordered) table in reverse order?
13:11:19enthus1asttables does not seem to be able to do it
13:18:17FromGitter<cooldome> Guys, did anyone tried or have seen a library in Nim that implements physics units of measurements. The idea is quite old, to use `distinct float` to represents things like velocity, acceleration, force and make sure at compile time that things that make sense only composable? So you can some force with velocity and you can't mess up the calculation. Ideally support differentiating, types like dt.
13:18:52skrylari am not aware of that but it could probably be done
13:19:07skrylarwell.. automatic differential stuff might be harder
13:20:31FromGitter<Varriount> @mratsim If you're really brave, submit your post to reddit and hacker news
13:22:18FromGitter<mratsim> @Varriount, lol, I don’t have a HN account so feel free ;)
13:22:39FromGitter<mratsim> Issue posted for the varargs[untyped] - https://github.com/nim-lang/Nim/issues/5855
13:23:48FromGitter<mratsim> @skrylar I wrote an autodifferentiation library, it only works for scalars for now but you might be able to retrofit it (or ideas) for your need: https://github.com/mratsim/nim-rmad
13:27:02FromGitter<cooldome> nim-rmad looks somewhat relevant, will have a look
13:28:32FromGitter<stisa> @zacharycarter just seen you got webgl to draw something, nice! Do you have a repo/branch I can look at?
13:29:03FromGitter<zacharycarter> @stisa one sec
13:29:17FromGitter<zacharycarter> http://imgur.com/a/jAO74
13:29:35FromGitter<zacharycarter> I'm kind of doing some crazy stuff @stisa - I'm only barely using your webgl wrapper now
13:29:50FromGitter<zacharycarter> most of the heavy lifting of the drawing is done by a sprite batch module I'm wrapping using requirejs
13:30:46*bjz_ joined #nim
13:31:37*bjz quit (Ping timeout: 246 seconds)
13:33:39*Trioxin quit (Ping timeout: 272 seconds)
13:33:43FromGitter<zacharycarter> @stisa I'm going to do a bit more work on this branch and then I'll commit it
13:36:54FromGitter<stisa> oh interesting,I'll wait for it then. I just finished building bgfx, I hope I got it right
13:37:03FromGitter<ephja> @cooldome it does?
13:37:11FromGitter<ephja> https://github.com/def-/nim-units
13:37:31FromGitter<ephja> it did not compile a couple of months ago. the error message was confusing
13:39:19FromGitter<ephja> apparently the samples at the bottom do compile now. great!
13:41:23*pilne joined #nim
13:43:39FromGitter<zacharycarter> @stisa I'm sure you did and if not it's easy to remedy
13:44:01FromGitter<zacharycarter> going to take a break for a bit, be back in a few
13:48:12skrylari think my power cable died. hmm
13:49:07skrylaroh good it didn't
13:49:29*krux02 joined #nim
13:50:16*Vladar joined #nim
13:51:33FromGitter<Varriount> @zacharycarter What's requirejs?
13:53:01FromGitter<cooldome> Wow, nim-utils really cool
13:53:30FromGitter<cooldome> Oops, I meant nim-units
13:53:59FromGitter<cooldome> thanks ephja
14:01:50FromGitter<zacharycarter> @Varriount I'm not using requirejs I'm using browserify but requirejs is a way to load javascript modules
14:02:12FromGitter<zacharycarter> browserify lets you create node modules and require them in the browser
14:02:13FromGitter<Varriount> Ah, it's like webpack
14:02:20FromGitter<zacharycarter> similar yes
14:04:50FromGitter<zacharycarter> I'd say webpack is better but this framework uses browserify / gulp so I'm at it's mercy
14:07:46Tiberiumzacharycarter: did you found a solution to your require problem?
14:07:57Tiberiumbecause you can just emit JS code AFAIK
14:09:01FromGitter<zacharycarter> I did
14:09:38FromGitter<zacharycarter> thank you for asking though
14:09:42FromGitter<zacharycarter> I didn't have to resort to that
14:15:06*tigerfinch quit (Quit: Leaving.)
14:30:01FromGitter<zacharycarter> @stisa I just pushed up the html5 branchhttps://github.com/fragworks/frag/tree/html5
14:31:40FromGitter<mratsim> There is no proc in the standard lib that creates a closure iterator from an openarray right?
14:35:32*skrylar quit (Ping timeout: 268 seconds)
14:38:28*Arrrr quit (Ping timeout: 240 seconds)
14:38:46FromGitter<Varriount> Not that I can recall
14:41:02*skrylar joined #nim
14:42:21FromGitter<mratsim> okay
14:50:52*benoliver999 quit (Ping timeout: 260 seconds)
14:52:22FromGitter<mratsim> build mine, but I get illegal capture when I try to use it so that may be why it doesn’t exist.
14:52:51*Matthias247 quit (Read error: Connection reset by peer)
14:53:27*skrylar quit (Ping timeout: 272 seconds)
15:10:04*Guest87036 quit (Quit: My MacBook has gone to sleep. ZZZzzz…)
15:15:21*bjz_ quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
15:26:17*nsf quit (Quit: WeeChat 1.7.1)
15:27:58*jsgrant_ joined #nim
15:31:15FromGitter<mratsim> Finished slice assignments with syntactic sugar and arbitrary steps: http://img.ctrlv.in/img/17/05/20/5920611548d00.png
15:38:21*bjz joined #nim
15:44:52*adeohluwa joined #nim
15:46:04FromGitter<Varriount> @mratsim You can't capture openarrays, as they might be located on the stack
15:48:10*skrylar joined #nim
15:50:20*bjz quit (Quit: My MacBook Pro has gone to sleep. ZZZzzz…)
15:55:58*Arrrr joined #nim
15:55:58*Arrrr quit (Changing host)
15:55:58*Arrrr joined #nim
15:59:19*Trioxin joined #nim
16:09:06*Tiberium quit (Remote host closed the connection)
16:15:47Arrrrhttps://github.com/nim-lang/Nim/issues/5827 +1
16:23:33*Trustable quit (Read error: Connection reset by peer)
16:32:52*JHonaker joined #nim
16:33:34JHonakerIs there any way to use a tuple as the arguments to a function call?
16:33:44couven92yup
16:33:59JHonakerdo tell
16:34:06couven92exactly as you would with a regular argument
16:34:24couven92proc foo(t: tuple[f1, f2: int])
16:34:37JHonakerAh, not what I meant
16:34:50JHonakersay I have proc(r, g, b: int)
16:35:03JHonakerand I have a tuple, (r, g, b: int)
16:35:21JHonakercan I say thatproc(thattuple)
16:35:27JHonakeror something similar
16:35:32couven92ah, deconstruct the tuple into three variables and use them as arguments
16:35:35JHonakerYea
16:35:52couven92I don't think you can auto-deconstruct a tuple in a call...
16:36:24JHonakerThat's disappointing
16:36:43JHonakerI guess I can do it with a template
16:36:49couven92yup
16:37:12JHonakerErr, would it have to be a macro
16:37:35couven92nope, a template should do fine
16:40:13FromGitter<mratsim> (a,b,c) = tuple ⏎ proc(a,b,c)
16:40:31FromGitter<mratsim> with a let of course
16:41:31couven92@mratsim, I think he wanted to deconstruct the tuple in the argument side of the proc call, not on the return side
16:41:56*Tiberium joined #nim
16:43:20FromGitter<mratsim> Yeah, I understand, a,b,c are argument here ;)
16:49:03JHonakerYea, I wanted to unpack the tuple as the arguments
16:51:55*nsf joined #nim
16:57:50*fredrik92 joined #nim
16:58:12*couven92 quit (Disconnected by services)
16:58:19*fredrik92 is now known as couven92
17:09:36*skrylar quit (Ping timeout: 240 seconds)
17:32:08*yglukhov joined #nim
17:36:28*yglukhov quit (Ping timeout: 240 seconds)
17:55:46mindBHey everyone. Just starting trying to learn nim. Anyone have a slackbuild for Nim or other advice on getting Nim installed on slackware?
17:56:58couven92mindB, what's slackware? (That's a no btw :P )
17:57:18dom96mindB: You might wanna try choosenim https://github.com/dom96/choosenim#installation
17:58:47*Matthias247 joined #nim
18:01:24*libman joined #nim
18:03:52TiberiummindB, nim is actually changing very fast at the moment, so you would like to stay at devel and update every few days :) you can do that via choosenim that dom96 suggested
18:07:54mindBWill check choosenim out. Sounds better than my current solution of just adding things to my PATH manually anyway.
18:09:29mindBMost of the articles I've seen on nim are from several years ago. Interesting to hear that things are still fairly unstable; is the language not nearing a 1.0.0 release?
18:10:21mindB(though I guess Rust isn't really stable yet either, so I shouldn't be too surprised.)
18:15:16*skrylar joined #nim
18:24:31*nsf quit (Quit: WeeChat 1.7.1)
18:28:19*skrylar quit (Ping timeout: 246 seconds)
18:28:33FromGitter<zacharycarter> @stisa I think I'm going to have to scrap the html5 backend - it was a fun experiment but I don't think I'll be able to make it very usable without having a totally different API from the desktop
18:28:41FromGitter<mratsim> In my newbie experience, I find Nim stable enough but rough on the edges, edges being static parameters, generics. Can’t comment on multiprocessing as I only used `||` but I think it’s rough as well.
18:36:47*skrylar joined #nim
18:36:58FromGitter<zacharycarter> maybe I keep it and just say that you can't use this functionality with it -
18:41:12*vlad1777d_ joined #nim
18:42:08*vlad1777d quit (Ping timeout: 240 seconds)
18:49:47*Jesin quit (Quit: Leaving)
18:55:42*Jesin joined #nim
18:58:53*vlad1777d__ joined #nim
18:59:27*vlad1777d_ quit (Ping timeout: 240 seconds)
19:01:26*tigerfinch joined #nim
19:01:54*tigerfinch quit (Remote host closed the connection)
19:10:25*skrylar quit (Quit: Leaving)
19:11:27JHonakerHow do I create a new object that is of type World[N] = array[N, bool]?
19:11:37JHonakerI'm not quite sure how new works
19:12:13JHonakersay I'm wanting to make a proc newWorld(n: int): World[n]?
19:12:32JHonakerIs that even possible? Using the n in an argument and as the generic types return type?
19:15:33def-JHonaker: well, if n is determined at runtime, then how can the compiler know what it will be?
19:15:58TiberiumJHonaker, you can create arrays only with length known at compile time
19:17:03TiberiumJHonaker, if you want something like arrays but with dynamic creating at runtime - use sequences
19:18:16*Tiberium quit (Remote host closed the connection)
19:20:48def-JHonaker: but wit works if N is known at compiletime
19:20:51def-proc newWorld(N: static[int]): World[N] = discard
19:22:58JHonakerAh, that makes sense
19:23:36JHonakerI keep trying to make arrays fit work, instead of just using seqs
19:24:02def-careful with arrays though or you copy around large structures
19:24:08JHonakerI didn't think about needing to know at compile time for type, which makes total sense. Thanks Tiberium, def-
19:24:57JHonakerYea, I'm making an implementation of Wireworld (a cellular automata)
19:25:16JHonakerOnly one array in the whole program, minus the copy I have to make for updating each cycle
19:25:31def-sounds more like a seq to me for the World
19:25:42def-since you might want the user to change the size of the world without recompiling
19:26:34JHonakerI wasn't going to make it resizable ^_^
19:26:45JHonakerBut that's a good point I'll keep in mind for the future
19:28:36*JHonaker quit (Remote host closed the connection)
19:30:48*Arrrr quit (Read error: Connection reset by peer)
19:39:27*gangstacat quit (Quit: Ĝis!)
19:42:20*rosshadden quit (Quit: WeeChat 1.7)
19:46:15*yglukhov joined #nim
19:54:07*gangstacat joined #nim
20:15:56libmanhttp://excel.fit.vutbr.cz/submissions/2017/018/18.pdf - This paper mentions Nim but not enough.
20:29:07SusWombatlibman, its about their own language why should they focus on a different one more?
20:29:12*Sentreen quit (Ping timeout: 260 seconds)
20:31:56libmanNot a big deal. It references "D programming language, c1999-2017" and "Nim programming language, c2015".
20:34:13*yglukhov quit (Remote host closed the connection)
20:35:40libmanWeird how D gets a year range from when it was first conceived (and wasn't released until two years later) to the present.
20:43:04*Sentreen joined #nim
20:48:10*Matthias247 quit (Read error: Connection reset by peer)
20:56:28couven92Me and PMunch have actually tried to spread the word more and more at the Computer Science department... We're planning on doing a workshop/presentation with Nim next semester :)
20:57:15*yglukhov joined #nim
20:59:05couven92But IMHO Nim deserves much more attention than it currently gets, it's a real awesome language! :)
21:05:15*jsgrant_ quit (Read error: Connection reset by peer)
21:34:46*yglukhov quit (Remote host closed the connection)
21:38:10*krux02 quit (Remote host closed the connection)
21:38:26*krux02 joined #nim
21:38:55krux02is there interest to remove stmt and expr from the standard library?
21:39:30couven92they are deprecated, aren't they?
21:41:19*Trioxin quit (Ping timeout: 272 seconds)
21:49:37*Vladar quit (Remote host closed the connection)
21:50:10krux02yes they are deprecated sice 0.15
21:50:41krux02well I removed all occurences of them and wanted to do a pull request
21:50:52krux02problem I did a wrong git command and everything is gone
21:51:07couven92ah... my guess is that is wouldn't hurt :P
21:52:43*bjz joined #nim
22:01:03*BennyElg joined #nim
22:02:04*BennyElg_ quit (Ping timeout: 260 seconds)
22:34:37*adeohluwa quit (Quit: Connection closed for inactivity)
22:36:26krux02well fixing all those use cases I see inconsistencies in the typed untyped stuff
22:36:36krux02generally for macros it's weird
22:37:15krux02this typed/untyped for return values of macros/templates doesn't make a lot of sense
22:39:22*Kingsquee joined #nim
22:42:29krux02the test case tusingstatement doesn't pass when I port it to untyped/typed
22:47:01*libman quit (Quit: Connection closed for inactivity)
23:04:32*Ven joined #nim
23:04:57*Ven is now known as Guest81488
23:12:20*BennyElg_ joined #nim
23:12:27*BennyElg quit (Ping timeout: 240 seconds)
23:38:40krux02well I made the pull request
23:38:48krux02I know there is a test that won't pass
23:38:56krux02it is basically I bug that I detected.
23:39:40krux02something that did work with the old syntax, but not with the new syntax
23:51:43*chemist69 quit (Ping timeout: 246 seconds)
23:55:24*bjz_ joined #nim
23:56:22*bjz quit (Ping timeout: 246 seconds)