<< 03-04-2018 >>
00:05:33*jrbrt joined #nim
00:21:49*jrbrt quit (Quit: Oíche mhaith)
00:30:13*jrbrt joined #nim
00:31:29*jrbrt quit (Client Quit)
00:51:39*dddddd quit (Remote host closed the connection)
00:52:38*enthus1ast joined #nim
00:54:25*enthus1a1t quit (Ping timeout: 248 seconds)
00:57:38FromGitter<AjBreidenbach> is there a syntax for instantiating a `ptr object`
01:01:01*vlad1777d joined #nim
01:10:33*vlad1777d quit (Ping timeout: 265 seconds)
01:12:11*S1t1Schu joined #nim
01:14:53FromGitter<AjBreidenbach> nvm it's probably better to just use ref
01:15:27*S1tiSchu quit (Ping timeout: 240 seconds)
01:38:09*tiocavera joined #nim
01:38:09*tiocavera quit (Changing host)
01:38:09*tiocavera joined #nim
01:38:09*rockcavera is now known as Guest93566
01:38:09*Guest93566 quit (Killed (rajaniemi.freenode.net (Nickname regained by services)))
01:38:09*tiocavera is now known as rockcavera
02:04:38*smt` joined #nim
02:04:54*smt` quit (Max SendQ exceeded)
02:05:28*smt` joined #nim
02:06:21*smt` quit (Remote host closed the connection)
02:06:48*smt` joined #nim
02:08:45*smt quit (Ping timeout: 264 seconds)
02:43:29*smt_ joined #nim
02:47:15*smt` quit (Ping timeout: 256 seconds)
03:13:31*NimBot joined #nim
03:21:08*vivus quit (Quit: Leaving)
03:21:26*endragor joined #nim
04:08:07*Tanger joined #nim
04:53:49*leorize joined #nim
05:03:18TangerHey folks! I'm looking at nim's Emerald DSL to see if I can fix up it's issue on 0.18.0. I think the issue lies in calling `quit` within a {.compiletime.} pagrama'd proc
05:05:11TangerError message suggests quit gets evaluated to a NimNode at compile time
05:06:24*smt_ quit (Read error: Connection reset by peer)
05:06:43*smt_ joined #nim
05:07:55*smt` joined #nim
05:11:01*smt joined #nim
05:11:17TangerThought about replacing the quit's with .fatal. pragmas, but I gotta look more into getting node data at compile time
05:12:04*smt_ quit (Ping timeout: 256 seconds)
05:13:12*smt` quit (Ping timeout: 256 seconds)
05:13:33*smt` joined #nim
05:16:02*smt quit (Ping timeout: 256 seconds)
05:17:49*r3d9u11 joined #nim
05:32:22*xet7 joined #nim
05:42:41*yglukhov quit (Read error: Connection reset by peer)
05:43:33*yglukhov joined #nim
05:58:47*leorize quit (Ping timeout: 256 seconds)
06:06:45*nsf joined #nim
06:07:06*r3d9u11 quit (Remote host closed the connection)
06:12:23*leorize joined #nim
06:14:05FromGitter<diegogub> Hi guys, any way to to avoid null values when marshaling ref objects JSON ?
06:14:19FromGitter<diegogub> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5ac31bbb270d7d370888a202]
06:25:37*smt_ joined #nim
06:29:21*smt` quit (Ping timeout: 248 seconds)
06:33:03*user1101 quit (Quit: user1101)
06:52:54*Vladar joined #nim
07:05:19*PMunch joined #nim
07:29:25*yglukhov quit (Remote host closed the connection)
07:29:45*yglukhov joined #nim
07:36:19*floppydh joined #nim
07:37:03*Lysenkoism joined #nim
07:37:20*gokr joined #nim
07:38:12*rokups joined #nim
07:40:06FromGitter<Varriount> @diegogub Not off the top of my head, you could try looking at the marshalling code
07:46:31*smt` joined #nim
07:49:03*yglukhov quit (Read error: Connection reset by peer)
07:49:37*yglukhov joined #nim
07:50:25*smt_ quit (Ping timeout: 248 seconds)
07:53:07FromGitter<survivorm> Hrm. Just read an article about truely hellish management. Glad, I'm not working here: https://projectfailures.wordpress.com/2008/06/24/project-from-hell/
08:00:05FromGitter<diegogub> @survivorm sounds like hell
08:00:40FromGitter<diegogub> @Varriount I'm on it..but first week of nim, still struggling
08:01:21*jaco60 joined #nim
08:03:03*gmpreussner quit (Ping timeout: 268 seconds)
08:04:02PMunchWhat is the proper way to deprecate a proc?
08:04:24*smt joined #nim
08:05:18*gmpreussner joined #nim
08:07:44*smt` quit (Ping timeout: 256 seconds)
08:11:18FromGitter<mratsim> {.deprecated: “your reason”.}
08:12:12FromGitter<mratsim> @survivorm this website also has a good selection of projects from hell: https://thedailywtf.com/ :)
08:14:44FromGitter<survivorm> @mratsim Thanks god i'm not in one of them. My work problems suddenly seem so little.
08:15:23FromGitter<mratsim> actually dailywtf is mentionned at the bottom of project from hell, I’m pretty sure I read that piece on some other website without a black background
08:15:59FromGitter<mratsim> You should also read tales from tech support: https://www.reddit.com/r/talesfromtechsupport/ :D
08:16:24FromGitter<survivorm> Actually, i've read a translation in russian. It just contained a link there
08:17:08FromGitter<survivorm> Ok, i'll do. But support is not only about depressing stories, it also is funny sometimes
08:17:45FromGitter<survivorm> I may retell my sister's husband support story here if people are interested
08:21:08PMunchmratsim, oh nice you can give a reason as well. The documentation for deprecated is a bit lacking: https://nim-lang.org/docs/manual.html#pragmas-deprecated-pragma
08:21:23PMunchI tried to use the list of renamings for a proc, but that didn't work well
08:21:43FromGitter<mratsim> I actually learned that yesterday as well ;)
08:22:03FromGitter<mratsim> https://github.com/nim-lang/Nim/issues/7468
08:27:59*smt` joined #nim
08:30:37*AndroUser2 joined #nim
08:32:01*smt quit (Ping timeout: 248 seconds)
08:33:53*Lysenkoism quit (Ping timeout: 276 seconds)
08:38:24*Arrrr joined #nim
08:41:00PMunchMan, this timothecourr guy is all over the place lately
08:43:05FromGitter<Varriount> @mratsim Have you tried the new toOpenArray proc yet?
08:43:06*dddddd joined #nim
08:43:23FromGitter<mratsim> Not yet
08:43:28FromGitter<Varriount> I'm having great fun with it for string manipulation.
08:43:35FromGitter<mratsim> @Pmunch I’m pretty sure I was the same when i started :P
08:44:38PMunchHaha, I kept my head down a bit more. Really liking his ideas with the generic stuff by the way
08:45:22FromGitter<mratsim> Well, I wanted to use “static” right from the get go, so ….
08:45:40*xkapastel quit (Quit: Connection closed for inactivity)
08:47:46*AndroUser2 quit (Ping timeout: 264 seconds)
09:34:38FromGitter<alehander42> that feel when you have 20 `of` branches and nim says "not all cases are covered"
09:34:47FromGitter<alehander42> do we have a way to show which one ?
09:35:20FromGitter<survivorm> Yeah, funny
09:37:08FromGitter<survivorm> I recall a task our team had a task of re-writing perl program which processed a data on mobile operators short-numbers and it's routing and it was about 1000 cases of if's
09:37:36FromGitter<survivorm> which became a hell to maintain, of cause
09:37:55FromGitter<alehander42> well it's a simplish REPL so it doesn't seem too bad
09:40:04FromGitter<survivorm> The program original developers thought so too, i think. And it was, then the cases were 20. May be 50. Or even 100. But then there were more then few hundreds...
09:41:20FromGitter<survivorm> So think future and try not to make life harder for people wich will come after you ;)
09:43:51FromGitter<alehander42> the problem is Nim doesn't have good support for "polymorphism" on variant objects
09:44:17FromGitter<alehander42> so if I decide to make a different proc for each kind of "operation"
09:44:47FromGitter<alehander42> I have to use some macro magic to dispatch (basically to generate a `case .. of : ..` on compile to or to write it manually)
09:45:14FromGitter<alehander42> and then I have to still do `expectKind` or `assert` in each procedure to help Nim recognize which fields are valid there
09:45:25FromGitter<alehander42> if we could do stuff like
09:47:17FromGitter<alehander42> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5ac34da5e3d0b1ff2c97601a]
09:47:22FromGitter<alehander42> which expands to ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5ac34da9e4ff28713aa15207]
09:47:27FromGitter<alehander42> that would be interesting
09:47:51FromGitter<alehander42> but maybe that can be solved with a good pattern matching lib again
09:50:06FromGitter<alehander42> of course, you can also use inheritance & methods but variant objects are better for some things
09:56:54FromGitter<alehander42> you can do it with macros
10:02:51*leorize quit (Ping timeout: 240 seconds)
10:20:59FromGitter<mratsim> @dom96 @Araq Can we reopen #6700, I think we should keep #7250 about package users and #6700 about package providers, see my last comment: https://github.com/nim-lang/Nim/issues/7250#issuecomment-378200483
10:51:17FromGitter<alehander42> ok i did it
11:04:56*smt_ joined #nim
11:08:57*smt` quit (Ping timeout: 256 seconds)
11:19:13*skelett joined #nim
11:21:53FromGitter<krux02> @alehander42 I just started a project do do ast pattern matching here: https://github.com/krux02/ast-pattern-matching/blob/master/matching.nim
11:22:22FromGitter<krux02> keep in mind I just started this. It is not a nimble package.
11:23:27FromGitter<krux02> All it does at the moment is, it generates a lot of `expectKind` and `expectLen` statements
11:33:26girvoWhat I would love to see is zero-overhead pattern matching ;) Mmmmm pipe-dreams...
11:34:06*xet7 quit (Quit: Leaving)
11:34:29*Vladar quit (Quit: Leaving)
11:35:34FromGitter<alehander42> @girvo I don't see why zero-overhead pattern matching should be impossible
11:37:36*xet7 joined #nim
11:38:06*leorize joined #nim
11:39:46dom96mratsim: #6700 has turned out quite long already though
11:40:41dom96but okay, if you'll repost your comment there, I think it will be better.
11:40:47FromGitter<mratsim> Should i start a new one otherwise?
11:41:38dom96Yeah, start a fresh one.
11:41:53dom96I think it'll be easier that way for others to contribute
11:42:21*xet7 quit (Read error: Connection reset by peer)
11:42:28*xet7_ joined #nim
11:42:35dom96But I already have an answer for you
11:42:44*xet7_ is now known as xet7
11:44:20dom96brb
11:50:45*user1101 joined #nim
11:55:27*yglukhov quit (Remote host closed the connection)
11:55:45FromGitter<alehander42> @survivorm I published my macro for overloading based on variant, thanks for the idea
11:56:14FromGitter<alehander42> @Araq , @dom96 do you think the syntax in https://github.com/alehander42/poly or something similar would be useful for stdlib ?
11:57:08FromGitter<alehander42> @krux02 looks great, I can't wait to use something like this in my macros
11:57:22FromGitter<mratsim> @dom96, cleaned up, opened #7485
11:59:20FromGitter<alehander42> @krux02 btw I had the idea to generalize some kind of matching syntax on proc level as in poly (the previous link): something like ⏎ ⏎ ```poly: ⏎ proc display*(a: A{e: 4, h > 2}) = ⏎ echo a.e``` ⏎ ⏎ so you can have many overloads with different conditions directly in the signature and the dsl will generate a single proc with `case`/`ifs` ... [https://gitt
11:59:20FromGitter... er.im/nim-lang/Nim?at=5ac36c98270d7d37088a284f]
12:00:53*SenasOzys quit (Ping timeout: 256 seconds)
12:03:18*Vladar joined #nim
12:04:44girvoalehander42: Me either! Which is why I'd love to see it
12:08:50FromGitter<alehander42> at least in my pattern matching DSL-s I am trying to generate optimal code
12:12:23*SenasOzys joined #nim
12:12:51*smt` joined #nim
12:17:04*smt_ quit (Ping timeout: 256 seconds)
12:17:55*SenasOzys quit (Read error: Connection reset by peer)
12:18:10*SenasOzys joined #nim
12:19:14dom96alehander42: I dunno, I'd rather keep it out of the stdlib until it stabilises and is used by more people.
12:19:31*Lysenkoism joined #nim
12:21:01dom96mratsim: replied
12:22:35Araqproblem is that many do not want a nimble dependency for pattern matching
12:25:46FromGitter<alehander42> @dom96 absolutely, I wanted to just demonstrate how the syntax would work with poly (and to use it in my projects)
12:26:18FromGitter<alehander42> I just wondered why Nim itself doesn't have such syntax (but maybe it's not too often needed)
12:28:31FromGitter<alehander42> @Araq I agree, if eventually a Nim lib manages to provide good solutions for most pattern matching usecases, it might be better to include it as several competing 3rd party pattern matching syntaxes can be confusing (this sounded like some kind of anti-market soc speech hahaha)
12:29:02FromGitter<alehander42> btw is there a reason `{.push.}` doesn't really work with all pragmas
12:29:49Araq.push is messy to implement and it's yet another of these features I wish we lacked.
12:30:01Araqshould have be done with a macro
12:30:12*donotturnoff joined #nim
12:30:59FromGitter<alehander42> so you'd prefer something like ⏎ ⏎ ```push(pragma): ⏎ proc .. ⏎ ⏎ proc ..``` ⏎ ⏎ ? [https://gitter.im/nim-lang/Nim?at=5ac37403bb1018b37a54127d]
12:31:31Araqsure
12:31:33FromGitter<alehander42> I can't see how can you do with a macro otherwise
12:34:28FromGitter<alehander42> I am still using `using` proudly .. :D
12:35:06Araqhmm I removed this feature, how can you use it?
12:36:29FromGitter<alehander42> o.O when did you remove it
12:36:46Araqone year ago or something
12:36:47FromGitter<alehander42> my devel should be ~= middle of march
12:37:07YardanicoAraq, you didn't remove `using`, and why would you?
12:37:07Araqso you mean the new 'using' statement?
12:37:09FromGitter<alehander42> you mean, know one needs to use `{.experimental.}` ?
12:37:16Araqwell
12:37:25Araq'using' is two unrelated features
12:37:28FromGitter<alehander42> are there two different `using` ? :D
12:37:49Araqit once was a Pascal like 'with' statement
12:37:55Yardanicowow
12:38:01Araqnow it's used for parameter type inference
12:38:14FromGitter<alehander42> yeah yeah I use the parameter type thing
12:38:24FromGitter<alehander42> haha, I didn't know about the `with` thing
12:38:27Araqoh ok.
12:38:57FromGitter<alehander42> hahahah I read `with` definition now
12:39:05FromGitter<alehander42> I actually implemented that myself this year
12:39:24FromGitter<alehander42> and I use it but with a different name, I even suggested including it in `sugar.nim`
12:39:57*sendell joined #nim
12:40:33FromGitter<alehander42> ok, my version goes only one level deep
12:40:33*smt` quit (Ping timeout: 248 seconds)
12:42:55FromGitter<zacharycarter> anyone good with nginx / Jester?
12:43:23dom96certainly not me :P
12:43:45FromGitter<zacharycarter> :P well... I was kind of hoping you were around anyway because I wanted to ask you something Jester related
12:44:11FromGitter<zacharycarter> I set up a post route of "/compile" - and I noticed that if I post to "/compile/" the request isn't handled - is Jester really that picky?
12:44:18FromGitter<zacharycarter> and if it is - should it be?
12:47:02dom96yes
12:47:07dom96"/compile/?"
12:47:09dom96is what you want
12:47:18FromGitter<zacharycarter> ah okay thanks
12:47:45FromGitter<zacharycarter> I'll flail around a bit more before asking my next question - I'm sure I'm doing something stupid
12:48:16dom96just ask :)
12:52:20FromGitter<zacharycarter> okay well - I have this nginx config - ⏎ ⏎ ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5ac379042b9dfdbc3a4f392e]
12:52:42FromGitter<zacharycarter> but when I try to hit http://localhost/compile
12:52:50FromGitter<zacharycarter> with a post request
12:52:58FromGitter<zacharycarter> my jester handler isn't hit :/
12:56:06*kier quit (Remote host closed the connection)
12:56:54FromGitter<zacharycarter> I'm running all of this with docker compose (attempting to re-write the playground to be easier to maintain / deploy and manage) - so it might be an issue w/ that as well
12:57:39*kier joined #nim
13:04:00FromGitter<Yardanico> strange
13:04:06FromGitter<Yardanico> https://github.com/samdmarshall/pewpewthespells-com/blob/master/configuration/etc/nginx/nginx.conf it looks the same as yours
13:08:26FromGitter<zacharycarter> yeah - I'm guessing it's something I'm doing wrong w/ docker or docker-compose
13:09:22*leorize quit (Ping timeout: 260 seconds)
13:10:04*leorize joined #nim
13:13:39dom96zacharycarter: hrm, since you're working on this, maybe you could deploy it on our VPS?
13:13:57FromGitter<zacharycarter> https://github.com/nemui-fujiu/docker_compose_nim_jester - should help me
13:13:59dom96I can make you an account
13:14:06FromGitter<zacharycarter> dom96 - I think you already did and yeah that's my plan
13:14:11FromGitter<zacharycarter> I want to just be able to ship a docker image though
13:14:21FromGitter<zacharycarter> so I'm not sure if I'll be able to use the current VPS
13:14:41dom96:/
13:15:07dom96It should just be a case of creating a systemd service file on the VPS
13:16:22*BitPuffin joined #nim
13:16:33FromGitter<zacharycarter> well... here's my thing - it's a pita to update the playground when the solution isn't dockerized
13:16:51dom96git pull && nim build?
13:16:55dom96*nimble build
13:17:03FromGitter<zacharycarter> it's not that simple when your solution relies on docker being installed on the machine
13:17:09FromGitter<zacharycarter> and files being written to directories that need permissions
13:17:10FromGitter<zacharycarter> etc
13:17:19FromGitter<zacharycarter> so like - the environment needs to be prepped first
13:17:22dom96hrm
13:17:28FromGitter<zacharycarter> and then I've run into situations where my verison of karax or something is so old
13:17:31FromGitter<zacharycarter> I go to make an update
13:17:46FromGitter<zacharycarter> and the app just stops working - and then I have to figure out what versinon of karax worked and roll back to that - it's just a headache
13:17:53dom96You should pin the commit: karax@commitHash
13:17:55FromGitter<zacharycarter> I'd rather just have one command to run that spins up the environment and the app in a container
13:18:02FromGitter<zacharycarter> that I can deploy
13:18:02dom96*karax@#commitHash
13:18:09FromGitter<zacharycarter> and run when I want to work on the app
13:18:16dom96but anyway, if you do that won't it be trivial to deploy on our VPS?
13:18:19FromGitter<zacharycarter> I know digitial ocean supports deploying containerized apps
13:18:27dom96also, how does a docker container within a docker container work?
13:18:27FromGitter<zacharycarter> if your VPS is running docker I imagine it would be
13:18:39FromGitter<zacharycarter> well I'm going to figure that one out today :)
13:18:53FromGitter<zacharycarter> the last time I attempted to get it to work - I couldn't - but I'm going to give it another shot
13:19:08dom96I'm happy to install docker on the VPS
13:19:34FromGitter<zacharycarter> alright - I'm going to finish getting the app in better shape and I'll let you know when I'm ready to deploy / start testing stuff
13:19:46dom96ok
13:20:00dom96I should make an account for federico3 too
13:20:14FromGitter<zacharycarter> I think you made myself and someone else an account at one point
13:20:18FromGitter<zacharycarter> it may very well have been federico3
13:20:20dom96federico3: PM me your public SSH key when you have a minute :)
13:20:31FromGitter<zacharycarter> also - my public ssh key most likely changed as I got a new laptop
13:20:32dom96I think Yardanico was the other person
13:20:37FromGitter<zacharycarter> yeah you're right
13:20:49dom96feel free to PM me your new SSH key
13:20:53FromGitter<zacharycarter> will do
13:20:58Yardanicodom96, yeah, it was me
13:21:58FromGitter<zacharycarter> dom96: going to PM you on gitter as I'm not on IRC and don't think I can PM from gitter -> irc
13:22:42dom96sure
13:23:02Yardanicoyeah, sadly you can't because there's no first-class gitter-freenode bridge (at IRC server level)
13:24:20livcdis there a performant web server pkg for nim that i could use without relying on reverse proxy ?
13:24:39Yardanicolivcd, probably httpbeast/mofuw
13:24:56YardanicoIDK really
13:26:02PMunchHmm, mratsim you on?
13:26:19FromGitter<mratsim> always
13:26:25FromGitter<mratsim> I work on Nim everyday now :P
13:26:38PMunchOh yeah that's true. Lucky you :P
13:26:52PMunchI'm trying to incorporate your genAccessors thing
13:27:03PMunchBut I've run into an issue
13:27:25PMunchSince it now generates a proc for each field I can't use those fields as names for say a variable
13:28:34livcdYardanico: mofuw has more commits :D
13:29:37PMunchhttp://ix.io/16a7/Nim
13:30:05dom96httpbeast isn't ready at all :)
13:30:17dom96the stdlib asynchttpserver isn't that bad though
13:30:22PMunchThat last line throws an error saying "Error: redefinition of 'fieldNull'"
13:31:32*arecaceae quit (Remote host closed the connection)
13:31:58*arecaceae joined #nim
13:32:37dom96zacharycarter: updated your key
13:35:25PMunchmratsim, have you run into similar issues?
13:36:13FromGitter<mratsim> wait I’m on the phone
13:36:41FromGitter<mratsim> It expects the fields to be named Ffield
13:36:51PMunchYeah I know
13:37:04PMunchI changed mine to expect "private_field"
13:37:24PMunchThe issue is that it creates a top level proc..
13:37:35PMunchWhich takes up a symbol
13:37:42*athenot joined #nim
13:37:56FromGitter<zacharycarter> thanks dom96
13:39:31FromGitter<mratsim> @PMunch, you can prefix with get and set. (I hear Java calling in the background btw).
13:39:49PMunchWell, then I'm back to square one
13:39:59PMunchI wanted to create a transparent interface for this
13:41:49FromGitter<mratsim> The only alternative I see is by using include, so that the private fields are usable within the package but not directly outside.
13:42:19FromGitter<mratsim> but I for me include is kind of a last resort
13:45:46PMunchThis is unfortunately going to be spit out from a macro, meaning it will all be inserted into the users code
13:47:03dom96yay for more people using Nim at work https://github.com/nim-lang/nimble/issues/456#issuecomment-378255753
13:47:06FromGitter<mratsim> I could see namespacing helping somehow. (i think) but it’s not there.
13:51:34PMunchPlaying around with the special `.` operator now
13:53:36*r2 joined #nim
13:54:29*yglukhov joined #nim
13:57:57*leorize quit (Quit: WeeChat 2.1)
13:58:56dom96man, so many RFCs these days
13:59:06dom96At this rate we'll never release v1
13:59:38*yglukhov quit (Ping timeout: 276 seconds)
14:00:48FromGitter<narimiran> thanks for reminding me i should write one too :)
14:00:55dom96hah
14:01:08dom96I mean, it's a good thing that we're discussing these things :)
14:05:36FromGitter<mratsim> Well the hard thing now is closing those RFCs
14:08:39FromGitter<survivorm> @alehander42 You're welcome. Though I don't remember giving you any ideas 😄
14:08:41FromGitter<mratsim> Should we schedule a regular review of RFCs (every x weeks) and a report on the forums? ⏎ ⏎ 1) More people will get exposed to them ⏎ 2) It makes sure that what is important doesn’t get stalled. [https://gitter.im/nim-lang/Nim?at=5ac38ae97685a046389d4370]
14:10:08FromGitter<gogolxdong> knows how to make this right？
14:10:19FromGitter<gogolxdong> ```code paste, see link``` [https://gitter.im/nim-lang/Nim?at=5ac38b4b27c509a774d21c03]
14:11:36dom96mratsim: sure, are you volunteering to perform this review? ;)
14:13:19FromGitter<mratsim> @dom96 Why not. I can pick 3 hot RFCs and 3 stale RFCs every 3 weeks for a start.
14:14:04dom96Extra points for writing an article on nim-lang.org ;)
14:14:22dom96although I suppose forum might be better to foster a discussion
14:14:57PMunchProque no les dos?
14:15:10PMunchPorque*
14:15:24PMunchWell, I guess there are more spelling mistakes there :P
14:15:49PMunchMy point is that it wouldn't be that much extra work to word it in such a way that they could be copy-pasted into both locations
14:16:49FromGitter<zacharycarter> maybe an article to bring attention to the forum post :P
14:17:07FromGitter<mratsim> I think more people read the forum than the blog
14:17:28federico3nothing wrong with posting on both
14:18:12FromGitter<Vindaar> People tend to read the forums more than the blog, simply because there's more activity there I'd wager. Both is imo the way to go (it's a good idea!)
14:19:12dom96It's more likely that blog will be posted on HN/Reddit
14:20:19FromGitter<narimiran> @mratsim instead of 3 RFC every 3 weeks - 1 RFC per week? ;)
14:21:26FromGitter<Vindaar> @dom96 Indeed. And cross posting to HN/Reddit is a good way to promote that Nim's development is indeed very active and ongoing
14:21:34FromGitter<mratsim> The thing is, when I’m in something it’s easier to do more in one go. But if multiple people alternate it can be done
14:22:18FromGitter<narimiran> @Vindaar no posting on HN, please, to not be accused of nim-whoring, or whatever @dom96 received the other day :D
14:22:38FromGitter<Vindaar> What, I missed that. What happened? :O
14:22:39dom96narimiran: that's only if you comment on Rust threads :)
14:22:43FromGitter<zacharycarter> only rustaceans seem to be annoyed about hearing about Nim
14:22:49FromGitter<zacharycarter> so IMO post more about Nim in rust threads
14:22:51FromGitter<zacharycarter> and in general
14:23:32FromGitter<narimiran> i usually post about nim in python threads, as this is how i found out about nim
14:23:54FromGitter<narimiran> and i think pythonistas are nice target group for nim
14:24:12FromGitter<Vindaar> agreed
14:24:25FromGitter<zacharycarter> so are people considering Rust :P
14:24:43FromGitter<mratsim> Btw, is there a Nim nightly builds for Windows? When trying to use koch in Appveyor, I have a strange issues where `nimble test`run the test files in for Nim instead of for my project …: https://ci.appveyor.com/project/jarradh/nim-ttmath/build/12/job/5um6l9idts1gl0ky#L1260
14:25:57FromGitter<zacharycarter> @mratsim - https://github.com/nim-lang/Nim/blob/devel/appveyor.yml
14:26:20FromGitter<zacharycarter> I don't think there are any nightly builds of Nim
14:26:21FromGitter<zacharycarter> atm
14:26:26dom96mratsim: why are you running `nimble test` in the compiler dir?
14:26:35FromGitter<mratsim> I’m not
14:26:49*yglukhov joined #nim
14:27:08FromGitter<mratsim> line 1255 there is a cd %CD% that takes me back to the project root and off the compiler dir (I think)
14:29:18FromGitter<mratsim> well I’ll put cd %CD% everywhere and pray
14:30:32FromGitter<narimiran> here you go @dom96 - another RFC :D
14:32:41dom96mratsim: you must be, the above line installs the 'compiler' package
14:32:46dom96I don't see how else it could do so
14:33:39*smt joined #nim
14:36:10*smt` joined #nim
14:37:02*athenot_ joined #nim
14:38:37*athenot quit (Ping timeout: 276 seconds)
14:40:01*smt quit (Ping timeout: 248 seconds)
14:46:18*endragor quit (Remote host closed the connection)
14:46:45*endragor joined #nim
14:47:05*skelett1 joined #nim
14:48:56*skelett quit (Ping timeout: 276 seconds)
14:50:57*endragor quit (Ping timeout: 240 seconds)
15:00:02*nsf quit (Quit: WeeChat 2.0.1)
15:01:01*miran joined #nim
15:01:45*PMunch quit (Quit: Leaving)
15:09:05YardanicoI started to seeing much more notifications from nim repo (in issues/pr's), that's a good sign!
15:09:48miranYardanico: and forum seems more lively to me
15:10:15Yardanicosadly I don't read nim forum anymore :/ should start reading it again
15:11:20federico3Yardanico: could this be of help? https://github.com/nim-lang/nimforum/issues/57
15:13:28Yardanicofederico3, well, I've meant reading threads on the forum, but OK :)
15:13:51YardanicoI've never really did any web development in nim :D
15:18:35*yglukhov quit (Ping timeout: 240 seconds)
15:20:46FromGitter<mratsim> I feel naked without a trusty castptr UncheckedArray[Foo (bar)
15:24:32*floppydh quit (Quit: WeeChat 2.1)
15:24:51*xkapastel joined #nim
15:25:19dom96FromGitter really butchers the brackets
15:26:44*aguspiza joined #nim
15:27:23FromGitter<mratsim> It tries to be helpful. Yesterday I used `s/foo/bar/` and it actually replaced foo in my previous message :O
15:27:48FromGitter<mratsim> it’s Gitter itself that does the weird formatting
15:28:21dom96yeah, it's a good bot
15:28:34dom96Many moons ago its author was a Nim developer :'(
15:30:12FromGitter<zacharycarter> who?
15:30:20dom96oprypin
15:30:26FromGitter<zacharycarter> oh yeah
15:30:26dom96(sorry for highlight)
15:30:40FromGitter<zacharycarter> I was always curious what happened to him / her as well as fowl
15:31:06dom96Hrm, I wonder what happened to fowl too
15:32:55FromGitter<zacharycarter> if - Billingsly Wetherfordshire - is fowl's real name
15:32:56FromGitter<zacharycarter> lolol
15:33:00FromGitter<zacharycarter> so awesome
15:42:22*Trustable joined #nim
15:42:56*yglukhov joined #nim
15:43:29*aguspiza quit (Ping timeout: 248 seconds)
15:53:41*smt joined #nim
15:56:49*smt` quit (Ping timeout: 248 seconds)
16:01:33*smt_ joined #nim
16:01:34*smt_ quit (Remote host closed the connection)
16:04:46*r3d9u11 joined #nim
16:05:38*smt quit (Ping timeout: 276 seconds)
16:19:56*sendell quit (Remote host closed the connection)
16:21:01*jrbrt joined #nim
16:26:46*SenasOzys quit (Ping timeout: 264 seconds)
16:38:24*r3d9u11 quit (Remote host closed the connection)
16:44:02*jaco60 quit (Ping timeout: 260 seconds)
16:45:53*Vladar quit (Ping timeout: 248 seconds)
16:46:19*Vladar joined #nim
16:49:45*Lysenkoism quit (Quit: ...is discredited)
16:56:45*nsf joined #nim
17:00:24*SenasOzys joined #nim
17:01:36*SenasOzys quit (Read error: Connection reset by peer)
17:01:37*SenasOzys_ joined #nim
17:19:53*Em_ joined #nim
17:22:49*Em_ quit (Client Quit)
17:29:14*aguspiza joined #nim
17:30:04subsetparkdom96: my understanding of `nimble test` is apparently imperfect. `nimble init` creates a test/ directory with test1.nim; however if i put something like `doAssert(false)` in that file and run `nimble test`, It still just prints out about verifying dependencies and then exits successfully.
17:38:46*qwertfisch joined #nim
17:39:01dom96subsetpark: bug? Dunno
17:42:29*ofelas quit (Quit: shutdown -h now)
17:43:09FromGitter<Vindaar> @subsetpark: I was pretty confused about `nimble test` too. I also thought that there should be a default `test` task, which runs all tests in `./tests`, but it didn't work for me either. So in all my projects I simply define the `test` task myself as mentioned on the github page of Nimble https://github.com/nim-lang/nimble#tests
17:43:47FromGitter<Vindaar> So @dom96 it should actually still work? I thought it was maybe changed recently
17:44:11dom96it should
17:44:25FromGitter<Vindaar> Hm, interesting
17:44:39dom96Here is the code: https://github.com/nim-lang/nimble/blob/master/src/nimble.nim#L1019
17:59:11FromGitter<mratsim> it tests all files that start with a `t` iirc, like koch
17:59:46FromGitter<Vindaar> Huh, now I'm really confused. Just compiled nimble again from source and it works with that version. Tried again with the version in my PATH and suddenly that works too? I don't get it. @mratsim yep, that's true
18:01:28FromGitter<Vindaar> Ahh, no! It seems to be a problem with some projects
18:01:34FromGitter<mratsim> otherwise I use this test proc as a template: https://github.com/status-im/nim-rocksdb/blob/eb2bd02c5adf128fe9fe08b83dea0908257078c8/rocksdb.nimble#L11-L22
18:01:39FromGitter<Vindaar> In one it works, in the other it doesn't
18:02:12FromGitter<mratsim> this way you call `test “your_test_file_without_extension” and it will compile it (and put the binary and nimcache in their proper place)
18:03:20FromGitter<Vindaar> ahh, thanks. that's nice!
18:12:15FromGitter<Vindaar> For some reason the `execBackend(optsCopy)` line breaks from the for loop in one of my projects
18:12:41*endragor joined #nim
18:12:48*jrbrt quit (Quit: Oíche mhaith)
18:14:31FromGitter<Vindaar> Hm, I have to leave now. I'll try to find out what's causing it tomorrow :)
18:17:09*endragor quit (Ping timeout: 256 seconds)
18:19:53*r3d9u11 joined #nim
18:21:56*jrbrt joined #nim
18:24:16*yglukhov quit (Remote host closed the connection)
18:24:53*yglukhov joined #nim
18:29:45*yglukhov quit (Ping timeout: 264 seconds)
18:32:43FromGitter<zacharycarter> interview with CSE went well - they said if I get the job I can't work on any side projects related to gaming
18:32:48FromGitter<zacharycarter> so no more nim game dev :/
18:35:04Yardanicosome NDA-related stuff?
18:35:16*ofelas joined #nim
18:35:22FromGitter<zacharycarter> I guess it's just a company policy
18:35:32FromGitter<alehander42> but why related to gaming
18:35:36FromGitter<alehander42> are they a gaming company?
18:35:37FromGitter<zacharycarter> it's a game dev studio
18:35:38FromGitter<zacharycarter> yeah
18:35:41FromGitter<alehander42> ah
18:35:52FromGitter<zacharycarter> http://citystateentertainment.com/
18:36:23FromGitter<krux02> sorry to hear that
18:36:33FromGitter<zacharycarter> well - I have to get the job first :P
18:37:35FromGitter<zacharycarter> but thank you for the optimisim!
18:38:28FromGitter<alehander42> that's just @zacharycarter 's time to shine in compiler side projects!
18:38:50FromGitter<zacharycarter> +1 ^
18:39:10FromGitter<krux02> well good luck with it. dont let yourself get exploited.
18:39:19FromGitter<zacharycarter> thank you! I'll try not to
18:39:45*skelett2 joined #nim
18:40:56FromGitter<krux02> well maybe people there will be like "wait in Nim you can do that. It is so simple? We need to change to Nim." and then it is a Nim full time job related to game development :P
18:42:51FromGitter<zacharycarter> this is the pipe dream
18:42:57*hogeland_ joined #nim
18:44:19FromGitter<wu-lee> newbie question: seems like a) You can't take a ref of a var. b) Object method polymorphism requires a ref. Therefore c) you can't have polymorphic var objects. Correct?
18:45:12FromGitter<krux02> well yea, but it is unrealistic.
18:45:35FromGitter<krux02> I have never heared of camelot unchained, but it reminds of of "Everquest Next"
18:46:12*avsej_ joined #nim
18:46:12*avsej_ quit (Changing host)
18:46:12*avsej_ joined #nim
18:46:14*vegax87_ joined #nim
18:47:12FromGitter<krux02> @wu-lee a ref is a pointer to something on the GC heap. a ref can only point to something that has been allocated on the GC heap.
18:47:25FromGitter<krux02> a var is on the stack
18:47:45FromGitter<krux02> therefore a var can't be transformed into a ref
18:47:49*skelett1 quit (*.net *.split)
18:47:49*dddddd quit (*.net *.split)
18:47:50*pqflx3[m] quit (*.net *.split)
18:47:55*vegax87 quit (*.net *.split)
18:47:55*avsej quit (*.net *.split)
18:47:55*hogeland quit (*.net *.split)
18:47:55Yardanicowu-lee why would you need a var for a ref though? you can modify ref values
18:47:56FromGitter<wu-lee> yep, what I thought
18:48:04YardanicoI mean ref variables :)
18:48:06FromGitter<krux02> but you can have polymorphic var objects.
18:48:06*avsej_ is now known as avsej
18:48:27FromGitter<krux02> it is just a different kind of polymorphism
18:49:09FromGitter<krux02> that pattern is known from C as tagged union.
18:49:48FromGitter<krux02> but it's not like virtual methods where the language automatically solves the molymorphism for you
18:49:59FromGitter<krux02> you would have to write your own branching statement
18:52:00FromGitter<wu-lee> @krux02 object variants, right?
18:52:04FromGitter<alehander42> or to write a macro that generates the branching for you
18:52:52FromGitter<krux02> @alehander42 let macros be for later
18:53:01FromGitter<krux02> @wu-lee yes
18:53:15FromGitter<wu-lee> @Yardanico, I might have an object I want to pass to a function accepting an immutable parameter
18:53:58*dddddd joined #nim
18:54:42FromGitter<wu-lee> (I think I can deref when passing, but not sure if that has a copying overhead)
18:54:45miranin CountTable, i have to manually check if a key exists, and if not - initialize it to zero?
18:54:52*pqflx3[m] joined #nim
18:55:24FromGitter<krux02> @wu-lee it is a hard problem to understand what is going on in your head. And it is hard for you to describe it. The most precise way to do it, post your code
18:55:34FromGitter<krux02> I prefer posting code on ix.io
18:55:47*nsf quit (Quit: WeeChat 2.0.1)
18:56:06FromGitter<krux02> but there are alternatives, a very good one is directly on github called gist
18:56:12FromGitter<alehander42> @krux02 I just did such a macro for object variants today, exactly because I didn't want to use methods & inh https://github.com/alehander42/poly#example ⏎ but still I think often one can relax the need for `var` and use methods
18:56:15FromGitter<wu-lee> Mainly I'm just trying to clear up the rules, when defining data. I am discovering I need to be more careful with my method signatures than I expected.
18:56:31Yardanicowu-lee: you can modify ref variable even if you've declared it with `let`
18:56:41Yardanicobecause `let` means that reference is immutable, not the value
18:56:46Yardanico(for ref type)
18:57:20FromGitter<alehander42> exactly ^
18:57:45FromGitter<alehander42> I did use `var A` for that before because I just assumed it needs it to mutate fields
18:57:48FromGitter<wu-lee> @Yardanico, can a function declare it won't modify an object parameter?
18:58:22FromGitter<wu-lee> (function or method)
18:59:48FromGitter<krux02> @alehander42 that is interesting
19:00:04FromGitter<krux02> but I think I would prefer to have full pattern matching
19:00:32FromGitter<krux02> I mean a really good pattern matching library
19:00:47FromGitter<krux02> I think I have to write it on my own
19:00:57FromGitter<krux02> currently I write ast pattern matching
19:04:36FromGitter<data-man> @miran: You can just do ```yourtable.inc(value)```
19:11:58mirandata-man: nice! i've been trying to use `inc mytable[key]` and this throws an error, so i had made `if not mytable.haskey(key): ....` - your version is much nicer!
19:13:38subsetparkdom96: looks like the `nimble test` issue is a bug... here's the tail output with --verbose:
19:13:41subsetpark Setting Nim stdlib prefix to
19:13:43subsetpark Setting Nim stdlib path to /home/zax/.choosenim/toolchains/nim-0.18.0/lib
19:13:45subsetpark Info ???(???, 0) Error: internal error: n is not nil
19:16:20dom96subsetpark: see here https://github.com/nim-lang/nimble/issues/456
19:16:31dom96weird issue
19:17:05subsetparkHey, speak of the devil...
19:18:29FromGitter<krux02> @alehander42 did you know that this works: http://ix.io/16fu
19:20:45*Arrrr quit (Ping timeout: 264 seconds)
19:23:14*Pwntus quit (Ping timeout: 268 seconds)
19:23:31*Pwntus joined #nim
19:33:58*jrbrt quit (Quit: Oíche mhaith)
19:48:10*jrbrt joined #nim
19:55:01*athenot_ quit (Read error: Connection reset by peer)
19:55:32*rokups quit (Quit: Connection closed for inactivity)
19:56:07*athenot joined #nim
20:02:18*SenasOzys_ quit (Ping timeout: 256 seconds)
20:05:16*SenasOzys_ joined #nim
20:21:15FromGitter<zetashift> @zacharycarter good luck and if youg et in try to sneak in some Nim tools
20:24:33FromGitter<krux02> I heared it helps to say "it's just a prototype" and then in the end you would not want to port it to anything.
20:28:21*gokr quit (Ping timeout: 240 seconds)
20:28:29*r3d9u11 quit (Remote host closed the connection)
20:29:09FromGitter<zacharycarter> 😃 thanks! you know I will! 😉
20:37:33*miran quit (Ping timeout: 264 seconds)
20:37:36*miran_ joined #nim
20:41:38*nsf joined #nim
20:43:53*yglukhov joined #nim
20:44:13*miran_ quit (Quit: Konversation terminated!)
20:45:20*Trustable quit (Remote host closed the connection)
20:54:05*r2 quit (Ping timeout: 240 seconds)
21:01:17*xet7 quit (Ping timeout: 260 seconds)
21:02:30*jrbrt quit (Quit: Oíche mhaith)
21:06:55*jrbrt joined #nim
21:08:42*Vladar quit (Quit: Leaving)
21:10:20*jrbrt quit (Client Quit)
21:11:14*xet7 joined #nim
21:24:52*donotturnoff quit (Ping timeout: 276 seconds)
21:42:57*jrbrt joined #nim
21:43:43FromGitter<krux02> What is with the builtin optional type? is that still relevant?
21:50:56*jrbrt quit (Quit: Oíche mhaith)
21:52:07*libman_ joined #nim
21:53:41Araqno.
22:00:03subsetparkNo because we gave up on it, or no because it got superceded by something more general?
22:03:02*jrbrt joined #nim
22:03:41*nsf quit (Quit: WeeChat 2.0.1)
22:08:03*jrbrt quit (Quit: Oíche mhaith)
22:08:44Araqsuperceded by something more general, I guess.
22:08:52Araqa bit complex to explain
22:09:58FromGitter<zacharycarter> errr is this the Option type being referred to>
22:10:38FromGitter<krux02> nope, it is the system.opt type
22:10:58FromGitter<krux02> that has as special type kind
22:11:00FromGitter<krux02> tyOpt
22:11:54FromGitter<zacharycarter> oh
22:12:05FromGitter<krux02> Araq: in the macros package, why is there no way to get the symbol kind of a symbol?
22:12:52FromGitter<krux02> @zacharycarter I am just curious, what was for you the reason to learn and use Nim instead of Rust?
22:12:56FromGitter<mratsim> @krux02 Please vote: https://github.com/nim-lang/Nim/issues/7366
22:14:19*yglukhov quit (Read error: Connection reset by peer)
22:14:22FromGitter<krux02> @mratsim I actually just started to create a pull request to fix that, because it is bothering me for quite a while now
22:14:51*yglukhov joined #nim
22:14:59FromGitter<krux02> I just forgot again how that `when` statement looked, so that I don't break bootstrapping
22:16:03FromGitter<zacharycarter> @krux02 - I wanted to write a game / game engine and do something beyond using UE4 / Unity / etc... ⏎ ⏎ So my options basically boiled down to C++ and SDL2 or using something like Allegro / Coscos etc... and I wasn't happy with those either.
22:16:37FromGitter<zacharycarter> I had heard / written hello world with Nim in the past, and when I started exploring Rust as an alternative I felt like I was writing C++ again
22:17:20FromGitter<zacharycarter> I was venting my frustrations with rust in a roguelike gamedev channel and someone brought up Nim - and that was the catalyst I guess
22:18:05FromGitter<zacharycarter> I was coming from other high level programming languages so, but I had experience with C/C++ programming in the past
22:18:29FromGitter<zacharycarter> Rust felt like I was swimming in complexity of C/C++ with extra layers added on top
22:18:47FromGitter<zacharycarter> Nim felt like something I was unfamiliar with but at the same time kind of used to
22:19:23FromGitter<krux02> that is interesting, do you still remember what your frustration with Rust was?
22:22:20FromGitter<zacharycarter> I was just now trying to find the tutorial that I had read originally when exploring Rust so I could try to figure out what was bugging me - but obviously it just gets harder and harder to find such a thing with the popularity
22:23:05FromGitter<zacharycarter> I have to guess - but I'm fairly confident it was the fact that a lot of concerns Rust addresses you don't deal with in higher level programming languages
22:23:42FromGitter<zacharycarter> so when I was directed to read a manual about a bunch of concepts that didn't necessarily apply to me getting started with a new language
22:24:08FromGitter<zacharycarter> I felt overwhelmed - and I also felt like I was just transitioning from struggling with C++ to Rust
22:24:16FromGitter<krux02> well rust has basically just one concept, borrowing and move semantics
22:24:36FromGitter<zacharycarter> yes but the explanation isn't that concise or clear is it?
22:24:40FromGitter<krux02> and different defaults than c++ (but still the wrong defaults)
22:24:44FromGitter<zacharycarter> instead we have to learn about the borrow checker
22:25:16FromGitter<zacharycarter> @krux02 - keep in mind I was a history major in college
22:25:27FromGitter<krux02> well I did not have that problem at all in Rust, because I knew c++ with move semantics already
22:25:34FromGitter<zacharycarter> so move semantics and borrowing weren't necessarily clear to me when I stumbled upon Rust / Nim
22:26:36FromGitter<krux02> there is something in Rust that I do not like, I can't implement an interface for a forein type.
22:27:41FromGitter<zacharycarter> I hadn't heard of a forein type before you mentioned it - just looking it up now
22:27:57FromGitter<mratsim> For me it was trying to use a tree data structure in Rust
22:28:05FromGitter<krux02> what I mean is a type that you do not own
22:28:11FromGitter<krux02> a type from a different package
22:28:13FromGitter<zacharycarter> gotcha
22:28:27*aguspiza quit (Ping timeout: 240 seconds)
22:28:36FromGitter<krux02> in nim I can just add procs to it, and generics will then eat the type and stop complaining
22:28:41FromGitter<krux02> in rust that's not possible
22:29:21FromGitter<zacharycarter> maybe we should create a website that highlights these issues w/ Rust and evangelizes Nim's solutions
22:29:32FromGitter<krux02> yea
22:29:39FromGitter<zacharycarter> then we'd really piss off the rustaceans on HN :P
22:29:50FromGitter<krux02> best approach: watch rust introduction videos, and show the Nim alternative
22:29:58FromGitter<krux02> btw that also works very well with python
22:30:25FromGitter<zacharycarter> yeah - if we could get someone producing videos for Nim comparing it to Rust / Python that'd be heaven for Nim I think
22:30:42FromGitter<mratsim> I would just focus on people pain points.
22:30:54FromGitter<mratsim> Python slow? —> Nim
22:31:05FromGitter<krux02> I recently saw a python talk that showcased new features of python3 with metaclasses and how awesome it is. And all I though was, good that I am in nim where I don't need metaclasses and all that crap in the video just to get a little bit of type safety.
22:31:16FromGitter<zacharycarter> I think the problem also is like - you can't just slap Nim in as a replacement for Python
22:31:23FromGitter<mratsim> Tired of C++ verbosity but want to keep memory under control —> Nim
22:31:29FromGitter<zacharycarter> currently I can't use Nim at work, because we can't talk to AWS w/ Nim
22:31:39FromGitter<krux02> it wasn't about performance even though it is at least as important
22:31:46FromGitter<data-man> D flexible? —> Nim :)
22:31:47FromGitter<krux02> AWS?
22:31:53FromGitter<zacharycarter> amazon web services
22:32:04FromGitter<mratsim> subsetpark has a good talk on how to use Nim on the critical path, and there is blog post about using Nim instead of Ruby for the hot path as well
22:32:22FromGitter<zacharycarter> what is hot vs critical path?
22:32:27FromGitter<zacharycarter> sorry <- newb
22:32:28FromGitter<mratsim> same*
22:32:36FromGitter<zacharycarter> oh okay
22:32:55FromGitter<mratsim> http://pyvideo.org/pygotham-2017/nim-a-new-option-for-optimizing-inner-loops.html
22:33:04FromGitter<krux02> well I think the most important construction sites in Nim are the bridges to other languages.
22:33:08FromGitter<mratsim> http://www.rubyflow.com/p/1j08ef-nim-for-the-discerning-rubyist
22:33:25FromGitter<krux02> I can imagine that at some point in time people can just use Rust packages from Nim and vice versa
22:33:30FromGitter<mratsim> Or a killer app that Nim does really well compared to other language
22:33:50FromGitter<mratsim> we need a better exportc for generics then
22:33:50FromGitter<zacharycarter> I don't know - people bitch about maintaining Python codebases
22:34:13FromGitter<krux02> I am developing a killer app: shaders in nim with closure
22:34:14FromGitter<zacharycarter> I think Nim really shines in that space
22:34:15FromGitter<mratsim> I agree, Python is a pain in production or to develop a test suite for
22:34:39FromGitter<data-man> killer app: shell written on Nim
22:34:41FromGitter<zacharycarter> I think bridging w/ other programming languages is good
22:34:42FromGitter<mratsim> the scientific packages are littered with “if isInstance(int)
22:34:51FromGitter<zacharycarter> but I think if you build a SDK like boto3 w/ Nim ppl will come
22:34:55FromGitter<krux02> I don't really understand why people started to like python in the first place
22:35:13FromGitter<mratsim> Because it’s rapid to prototype in
22:35:18FromGitter<mratsim> you get results fast
22:35:28FromGitter<zacharycarter> it's why nodejs has taken over
22:35:49FromGitter<mratsim> and a lot of people using Python are not software engineer but mechanical engineer or researcher in academia
22:35:56FromGitter<zacharycarter> yeah that too
22:36:21FromGitter<mratsim> Python is much better than Matlab once it got the equivalent packages
22:36:25FromGitter<krux02> nodejs has taken over because people are tired of using two programming languages at the same time and they thought: "which one can we throw out?"
22:36:56FromGitter<mratsim> I’m amazed at the number of CSS derivatives or compilers as well ...
22:37:05federico3nah
22:37:05FromGitter<mratsim> and now the “CSS-in-JS” movement
22:37:38FromGitter<krux02> but Matlab has a really great documentation. The documentation was always a weak spot to open source matlab alternatives, but I never used the python matrix libraries.
22:38:10FromGitter<krux02> I just used matlab for a while. And then I saw some fortran functions and thought: "hmm, this matlab really got some inspiration from fortran"
22:38:22FromGitter<mratsim> docs bring people through organic Google search
22:38:44FromGitter<krux02> in the web word there is also this framework hell
22:38:58FromGitter<krux02> thousands of frameworks that all want to be the next big thing.
22:39:08FromGitter<mratsim> Matlab manage to get as mandatary in many engineering school. I know I had to learn Matlab for mechanical engineering
22:39:16FromGitter<krux02> but frameworks are all crap, because they are frameworks.
22:39:22FromGitter<mratsim> managed to get as mandatory*
22:39:49FromGitter<krux02> I had to use it for "Computer Vision" and "Machine Learning" courses.
22:40:05FromGitter<krux02> horrible university licenso that only worked when you were online
22:40:10FromGitter<krux02> no work on the train possible.
22:40:26FromGitter<krux02> but yea enough bashing matlab i think. this is the Nim char :P
22:40:41FromGitter<zacharycarter> you'd be surprised - very little JS code gets reused b/w the client and server w/ Node apps
22:41:12FromGitter<zacharycarter> I mean, thinking about it critically - how much could really be reused? type definitions? helper functions - maybe?
22:41:20FromGitter<krux02> @zacharycarter I have no idea what b/w and w/ means
22:41:29FromGitter<zacharycarter> b/w = between
22:41:31FromGitter<zacharycarter> w/ = with
22:41:46FromGitter<krux02> where do these come from?
22:41:49FromGitter<zacharycarter> my head
22:42:05FromGitter<krux02> I think it is really hard to read that
22:42:08FromGitter<zacharycarter> but I'm not that creative - so I'm sure I've stolen them from elsewhere
22:42:53FromGitter<krux02> yea if you want to use them, write a browser plugin for you that automatically replaces them for you with the corresponding words ;)
22:42:58FromGitter<zacharycarter> hahaha
22:44:06FromGitter<krux02> and honestly I have no idea how much code between client and server could be shared.
22:44:39FromGitter<krux02> but from gpu programming I know that sometimes I just want to move some calculation from client to server and vice versa.
22:45:26FromGitter<krux02> i would still just want to use that functionality at one position in the code base.
22:45:28FromGitter<zacharycarter> very little can - that's useful anyway
22:45:58FromGitter<krux02> but then there is also the reduction of coplexity when you only need to think in one language.
22:46:18FromGitter<krux02> just think about it, how cool is it in Nim that you can write marcros in nim as well?
22:46:20FromGitter<zacharycarter> also you fall into the pit of each browser using its own JS engine - and then having to use something like babel to produce ES to overcome that
22:46:27FromGitter<krux02> and then soon you can also write shaders in Nim
22:46:35FromGitter<zacharycarter> and then - node js supports different features than browser based JS
22:46:40FromGitter<krux02> and everything else, too.
22:46:46FromGitter<krux02> one language to program it all.
22:46:54FromGitter<zacharycarter> going to program myself in Nim
22:47:08FromGitter<mratsim> oh I saw a blog post from the guy who wrote babel. He was underage at the time
22:47:11FromGitter<krux02> well if you have a programming interface, then go ahead.
22:47:23FromGitter<zacharycarter> babel is probably my most hated enemy at this point
22:47:27subsetparkBTW y'all - at work I'm translating a simple pubsub system we're building into Nim. Check it out: https://github.com/makingspace/pubnub
22:47:31FromGitter<zacharycarter> I kind of want to do battle with this underage kid you're referring to
22:47:39FromGitter<krux02> I am not going to replace my neural network with a CPU and main memory :P
22:48:24FromGitter<krux02> I have no idea what pubnub is
22:48:43FromGitter<mratsim> @zacharycarter here you go: https://medium.com/@sebmck/2015-in-review-51ac7035e272
22:50:45dom96subsetpark: oooh, looks beautiful
22:52:09FromGitter<zacharycarter> subsetpark: looks sweet - what communication protocols will you end up supporting do you think?
22:53:18FromGitter<zacharycarter> @mratsim - babel is like a giant band-aid on top of JS
22:53:30FromGitter<zacharycarter> and it's a shitty off-brand one at that
22:55:09FromGitter<mratsim> it’s a band-aid because companies are stuck with Firefox 4 while devs want to move to the new shiny thing
22:56:18FromGitter<zacharycarter> I just don't think it's the right solution for the problem at hand - Nim, Typescript, etc... fine
22:56:20FromGitter<zacharycarter> babel is hell
22:56:26*jrbrt joined #nim
22:57:05FromGitter<zacharycarter> let's just all admin JavaScript is the common denominator and it sucks and there's no making it better
22:59:22*jrbrt quit (Client Quit)
23:00:33FromGitter<mratsim> WASM !
23:00:37FromGitter<mratsim> the holy grail
23:04:25FromGitter<zacharycarter> heh
23:04:58*jrbrt joined #nim
23:06:06*jrbrt quit (Client Quit)
23:06:09*Sentreen quit (Ping timeout: 248 seconds)
23:06:10FromGitter<wu-lee> is there a way to clone a ref? like deepCopy but creates a destination ref of the right type?
23:06:20*xet7 quit (Remote host closed the connection)
23:11:49FromGitter<zacharycarter> subsetpark: not to kill the name for your project, but there already is a sass offering with the same name: https://www.pubnub.com/ - I thought I recognized pubnub from somewhere...
23:12:16FromGitter<zacharycarter> unless this is an implementation for their API - in that case disregard my message
23:19:05*jrbrt joined #nim
23:19:48shashlickwow majority of messages are from gitter
23:21:40FromGitter<zacharycarter> yah gitter!
23:23:26*Sentreen joined #nim
23:23:59*jrbrt quit (Client Quit)
23:25:33FromGitter<krux02> yes I used irc
23:25:38FromGitter<krux02> it really feels a bit old
23:25:55FromGitter<krux02> I preferr to have pictures and some embedding of stuff
23:26:13shashlickI've been using weechat for the last few days since matrix/riot is delayed and unreliable off late
23:26:21FromGitter<krux02> and the usage and setup of IRC is just horrible
23:26:30shashlickglowing-bear.org is pretty nice
23:27:04FromGitter<krux02> @mratsim WASM is not yet suitable for NIM.
23:27:37FromGitter<krux02> wasm is not for GC languages. So my suggestion is to improve usability of nim when GC is disabled
23:29:28dom96IRC4Life
23:29:42dom96I'm going to get that tattoo at some point
23:30:00FromGitter<krux02> by the way for everybody who cares I just created the pull request the exposes the kind of nim symbols
23:30:57FromGitter<zacharycarter> you can still use the GC w/ Nim
23:31:02FromGitter<zacharycarter> you just have to manually cycle it
23:31:14FromGitter<zacharycarter> nim and wasm I mean
23:31:32FromGitter<zacharycarter> you don't have to disable the GC - you just need to enable realtime gc support
23:31:44FromGitter<zacharycarter> @krux02 ^
23:33:54FromGitter<zacharycarter> dom96: I'm going to get the ICQ4Lyfe tat and 1 up you
23:34:34dom96:o
23:34:57dom96I guess i'll have to get a BBS4LYFE tattoo then
23:35:02FromGitter<zacharycarter> hahaha
23:36:44FromGitter<krux02> https://github.com/nim-lang/Nim/pull/7491
23:38:13FromGitter<krux02> am I correct that every thread has it's own GC and that when I want one thread to be realtime I can just make that one not allocate anything on the heap and when the other threads need a gc pause that one thread will just be able to contiue?
23:38:43FromGitter<krux02> I mean that's the advantage of thread local storage, isn't it?
23:39:20dom96yep
23:39:26dom96no stop-the-world :D
23:39:59FromGitter<krux02> well stop the thread
23:40:12FromGitter<krux02> but that's much better than stop the world
23:41:13*user1101 quit (Ping timeout: 260 seconds)
23:41:56dom96Yeah, but you can explicitly call the GC
23:42:07dom96so you know when the stop happens
23:43:41FromGitter<krux02> i mean there is still a way to create shared objects
23:43:54dom96yeah, but they aren't controlled by the GC
23:44:04dom96so you have to manage that memory yourself
23:44:14FromGitter<zacharycarter> https://github.com/yglukhov/nimx/blob/22289de7252d39e9be77f698dfa9aec522ec7fa8/nimx/private/windows/emscripten_window.nim#L404-L421
23:45:22FromGitter<zacharycarter> I find the whole - you can rely on shared memory to alleviate the local thread heap / gc thing to be confusing
23:46:11FromGitter<krux02> yea I think that needs to be a bit more explicit.
23:46:32FromGitter<Varriount> @zacharycarter You up?
23:46:45dom96'night guys
23:46:46FromGitter<krux02> with explicit I mean better documented
23:46:55FromGitter<krux02> night to you
23:46:58FromGitter<krux02> thanks
23:47:07FromGitter<zacharycarter> like let's say I have the following code - ⏎ ⏎ ```code paste, see link``` ⏎ ⏎ and I want to allocate this in shared memory [https://gitter.im/nim-lang/Nim?at=5ac4127a92f5d620578ee89a]
23:48:05FromGitter<zacharycarter> does that mean I'm calling `createShared` for foo.bar too?
23:48:09FromGitter<zacharycarter> and the string parameters?
23:48:14FromGitter<zacharycarter> Hi @Varriount - yes I am
23:51:25FromGitter<Varriount> @zacharycarter https://gist.github.com/Varriount/b4c817b46ea5c1b141af7791ff2dbcf7
23:51:55FromGitter<Varriount> @zacharycarter It's an interesting challenge, trying to do the v4 signing process with as little memory allocation as possible.
23:52:13FromGitter<zacharycarter> I was about to say - no type defs
23:52:27FromGitter<zacharycarter> wow I'm impressed :D
23:52:46FromGitter<Varriount> It's not finished, but I'm getting there.
23:53:03FromGitter<zacharycarter> it's definitely looking good
23:55:06FromGitter<zacharycarter> @Varriount: what about https://github.com/nim-lang/Nim/blob/master/lib/system.nim#L3241 ?
23:55:52FromGitter<zacharycarter> https://github.com/nim-lang/Nim/blob/d55e02ddf12662781cf89e2fd91473dbf7552e5a/lib/system/ansi_c.nim#L20
23:57:51FromGitter<zacharycarter> also - I don't know if you like this style better or not but, you can do (I'm sure you already know this) - ⏎ ⏎ ```import ⏎ strutils ⏎ tables ⏎ algorithm ⏎ httpcore``` [https://gitter.im/nim-lang/Nim?at=5ac414ffc574b1aa3e63fa65]
23:58:05FromGitter<zacharycarter> I think anyway
23:58:11FromGitter<Varriount> @zacharycarter I'm using memcmp, not momcpy
23:59:09FromGitter<zacharycarter> I thought I linked to memcopy
23:59:12FromGitter<zacharycarter> oh
23:59:14FromGitter<zacharycarter> memcpm
23:59:14FromGitter<zacharycarter> sorry
23:59:20FromGitter<zacharycarter> > memcmp