<< 18-06-2023 >>
01:03:54*systemdsucks joined #nim
01:23:10*via quit (Server closed connection)
01:23:17*via joined #nim
01:28:36*derpydoo joined #nim
03:39:43*azimut quit (Remote host closed the connection)
03:40:10*azimut joined #nim
03:45:50*hernan quit (Server closed connection)
03:46:05*hernan joined #nim
03:54:20*xet7 quit (Remote host closed the connection)
04:16:20*rockcavera joined #nim
04:17:38FromDiscord<heysokam> sent a code paste, see https://play.nim-lang.org/#ix=4ywq
04:18:04FromDiscord<heysokam> (edit) "https://play.nim-lang.org/#ix=4ywq" => "https://paste.rs/1cOw4"
04:22:39FromDiscord<Elegantbeef> Iterate all elements creatiing a set
04:24:39FromDiscord<heysokam> but i don't know how to iterate the thing correctly in the first place. that's the question, really
04:25:20*azimut quit (Ping timeout: 240 seconds)
04:26:02FromDiscord<Elegantbeef> sent a code paste, see https://paste.rs/YiHeE
04:27:06FromDiscord<graveflo> Whats the deal with the `out` keyword. I cant find documentation on it
04:27:40FromDiscord<Elegantbeef> In devel it's used for definite assignment analysis and slightly more explicit C interop
04:28:43FromDiscord<graveflo> in this link: https://github.com/nim-lang/Nim/wiki/Nim-for-Java-programmers it is tied to the same weird java syntax that I used that one time
04:28:51FromDiscord<michaelb.eth> In reply to @graveflo "Whats the deal with": cf. https://discord.com/channels/371759389889003530/768367394547957761/1118909269782298715↵and Araq's replies that follow
04:28:53FromDiscord<graveflo> at least I think. I'm not a java guru either
04:29:51FromDiscord<Elegantbeef> It's purely `var` but ensure you assign it
04:30:38FromDiscord<graveflo> oh ic. I read that message in internals too and forgot about it
04:32:04FromDiscord<heysokam> sent a code paste, see https://play.nim-lang.org/#ix=4yws
04:32:27FromDiscord<Elegantbeef> Are you trying to fold this into a single list?
04:32:47FromDiscord<graveflo> use difference then and prune on the empty set
04:32:50FromDiscord<heysokam> multiple, its a filter to a bunch of -W flags etc
04:32:54FromDiscord<Elegantbeef> if so replace `HashSet[string]` with `seq[string]` and `incl` with `add`
04:32:58FromDiscord<Elegantbeef> Wait you want all the names shared?
04:33:06FromDiscord<heysokam> i need the flags that are in all of the entries, but also the differences
04:33:59FromDiscord<heysokam> In reply to @Elegantbeef "Wait you want all": yeah, the names that are in -all- of them, so i can do `baseflags = @[....]` and then `release_flags = baseflags & @[ ... ]`
04:35:01FromDiscord<heysokam> sent a code paste, see https://play.nim-lang.org/#ix=4ywt
04:35:41FromDiscord<heysokam> (edit) "https://play.nim-lang.org/#ix=4ywt" => "https://play.nim-lang.org/#ix=4ywu"
04:35:42FromDiscord<Elegantbeef> use a `seq[HashSet[string]]` instead is what i'd suggest
04:36:12FromDiscord<heysokam> In reply to @Elegantbeef "use a `seq[HashSet[string]]` instead": im not following at all 🤔
04:36:27FromDiscord<Elegantbeef> If you want all shared flags it's a union of all sets
04:37:09FromDiscord<graveflo> ye i messed that up its union. Then if you want the difference for each you diff their set with the union set
04:37:09FromDiscord<Elegantbeef> Sorry not a union but an intersection
04:37:13FromDiscord<Elegantbeef> A union gives you all flags
04:37:25FromDiscord<graveflo> god damn it
04:37:38FromDiscord<Elegantbeef> A union - intersect gives you the ones that are not shared
04:37:58FromDiscord<heysokam> i understand how to filter the things
04:38:03FromDiscord<heysokam> the question is how to iterate them
04:38:27FromDiscord<graveflo> you are currently doing that. You might need to just move them to more suitable containers
04:38:47FromDiscord<graveflo> you posted the code that iterates them
04:39:03FromDiscord<heysokam> but it feels hacky. im repeating checks agains the same elements
04:39:13FromDiscord<graveflo> not if you do what beef said
04:39:24FromDiscord<graveflo> just move them into has containers in the loop or before the loop
04:39:33FromDiscord<graveflo> (edit) "has" => "hash"
04:39:54FromDiscord<heysokam> i don't quite get how, that's the issue
04:40:06FromDiscord<graveflo> define a hash set that will be the intersection of all seqs
04:40:16FromDiscord<graveflo> iterate over the list of seqs and update the hashset
04:40:24FromDiscord<graveflo> then you have the set of all shared
04:40:46FromDiscord<graveflo> if you want the difference between a given subset of all seqs then take the intersection of that seq and the union set
04:40:56FromDiscord<graveflo> (edit) "union" => "intersection"
04:41:21FromDiscord<graveflo> (edit) "intersection" => "difference"
04:41:26FromDiscord<graveflo> damn this set terminology
04:41:30FromDiscord<Elegantbeef> sent a code paste, see https://paste.rs/8H2D8
04:41:36FromDiscord<Elegantbeef> I've saved you grave 😄
04:41:40FromDiscord<graveflo> thank you
04:42:30FromDiscord<Elegantbeef> We could even do an optimisation here and save the len of all elements and allocate those to allocate more than we ever will need
04:43:21FromDiscord<Elegantbeef> shucks `+=` isnt defined which I guess makes sense
04:43:25FromDiscord<heysokam> sent a code paste, see https://play.nim-lang.org/#ix=4yww
04:43:55FromDiscord<Elegantbeef> Not exactly
04:44:08FromDiscord<Elegantbeef> It's practically the same but not internally
04:44:47FromDiscord<Elegantbeef> sent a code paste, see https://paste.rs/s6IiT
04:45:09FromDiscord<heysokam> ah so without the extra allocs
04:45:13FromDiscord<Elegantbeef> I was just writing code quickly, to explain
04:45:40FromDiscord<Elegantbeef> `toSet` allocates a new set on top of the one we're using so it's not that great, no clue if RVO kicks in
04:45:54FromDiscord<Elegantbeef> It should kick in
04:45:57FromDiscord<heysokam> whats the intersection doing in the next part?
04:46:07FromDiscord<Elegantbeef> Gives you all the shared flags
04:46:12FromDiscord<Elegantbeef> union gives you all the flags
04:46:22FromDiscord<Elegantbeef> difference holds all the flags that are not in all the seqs
04:47:19FromDiscord<heysokam> so difference = some, intersection = all, and union = irrelevant↵the names are giga-confusing
04:47:48FromDiscord<heysokam> but that's probably just my 🐠 brain
04:47:53FromDiscord<Elegantbeef> I'm using the set terms to describe it cause I have 0 clue what you're interested in
04:48:26FromDiscord<Elegantbeef> An intersection is like bitwise `and`, a union is like bitwise `and`
04:48:43FromDiscord<Elegantbeef> union is like bitwise or\
04:48:50FromDiscord<Elegantbeef> My brain no worky
04:50:05FromDiscord<Elegantbeef> As such the union variable has all the flags. The intersection only the ones every sequence has atleast one of. The difference the ones that come up 1..str.high - 1 times
04:50:46FromDiscord<graveflo> are there "inplace" set operations in nim? It looks like the `sets` module keeps making new sets
04:51:01FromDiscord<Elegantbeef> `incl` `excl`
04:51:44FromDiscord<graveflo> oh ok you just have to do the loops yourself. That's fair
04:51:56FromDiscord<Elegantbeef> No `incl` and `excl` take `HashSet[T]`
04:52:14FromDiscord<Elegantbeef> so the above `union +=` is really `union.incl st`
04:52:38FromDiscord<graveflo> yea but youll have to make your own say "difference update" routine using those procs if you want an inplace version of it
04:53:08FromDiscord<Elegantbeef> Right
04:53:14FromDiscord<Elegantbeef> "PRs welcom" 😄
04:53:18FromDiscord<toma400> Out of curiosity, is there any Ruby interop library for Nim?
04:53:49FromDiscord<Elegantbeef> Use crystal then interop using crystal as a system library 😛
04:53:56FromDiscord<Elegantbeef> https://nimble.directory/search?query=ruby
04:57:53FromDiscord<heysokam> your loop outputs the same text that i did, but without the seq hack. thx beef ✍️
05:00:03FromDiscord<Elegantbeef> Data types are magical
05:01:21FromDiscord<heysokam> indeed
05:03:52*derpydoo quit (Ping timeout: 240 seconds)
05:06:14*rockcavera quit (Remote host closed the connection)
05:30:40*ntat joined #nim
05:48:01FromDiscord<heysokam> 500 lines down to 90 with the sets trim 🙌
06:26:11FromDiscord<demotomohiro> Nim interop C + Ruby interop C = Nim interop Ruby ?
07:22:06FromDiscord<eerbium> Can I write as low level code as Rust in Nim
07:22:54FromDiscord<obi3112> In reply to @Elegantbeef "Data types are magical": is this a bot or user?
07:22:58FromDiscord<eerbium> like manual memory management on embedded system/baremetal, kernels, userspaces, etc
07:23:09FromDiscord<obi3112> nvm its a user
07:23:16FromDiscord<eerbium> In reply to @obi3112 "is this a bot": its an interop from the telegram or some other
07:23:29FromDiscord<obi3112> oh ok andrew ;D
07:23:38FromDiscord<Elegantbeef> Matrix to be precise
07:23:43FromDiscord<Elegantbeef> Yes Nim is as low level as Rust
07:23:43FromDiscord<eerbium> yes
07:23:45FromDiscord<heysokam> In reply to @eerbium "Can I write as": yes
07:23:50FromDiscord<obi3112> Breathe Air!
07:24:28FromDiscord<eerbium> So I suppose it should be possible to do anything in Nim?
07:24:33FromDiscord<heysokam> yep
07:24:33FromDiscord<Elegantbeef> Yes
07:24:37FromDiscord<eerbium> Like it covers all usecases?
07:24:55FromDiscord<heysokam> anything a systems language can do, yes
07:25:20FromDiscord<eerbium> no like i mean all usecases
07:25:27FromDiscord<eerbium> not only systems
07:25:29FromDiscord<heysokam> define -all-
07:25:52FromDiscord<heysokam> a systems language can do anything related to systems
07:26:05FromDiscord<heysokam> C, Rust, Cpp, are all systems languages
07:26:09FromDiscord<eerbium> In reply to @heysokam "define -all-": things that the mainstream languages are used for (python ts c# rust clojure elixir etc)
07:26:10FromDiscord<heysokam> python is not, its interpreted
07:26:55FromDiscord<heysokam> In reply to @eerbium "things that the mainstream": it can do everything python and rust can, if that helps solving your question from your frame of reference
07:27:25FromDiscord<heysokam> the question is a bit non-sensical, so its hard to answer it so that you get closure for it
07:27:54FromDiscord<heysokam> nim can do anything you want, saying "but can it do literally anything" after the first is answered, is what is losing me
07:27:57FromDiscord<eerbium> but can it be used for data processing like julia or concurrency handling like elixir
07:28:13FromDiscord<eerbium> In reply to @heysokam "nim can do anything": I meant just systems by that
07:28:20FromDiscord<eerbium> (edit) "that" => "the first questions"
07:28:21FromDiscord<eerbium> (edit) "questions" => "question"
07:28:27FromDiscord<heysokam> but a systems language can do literally anything already
07:28:31FromDiscord<eerbium> by the second question I mean everything in the realm of computer sicnece
07:28:40FromDiscord<heysokam> giving it another layer after that makes not much sense
07:29:04FromDiscord<eerbium> In reply to @heysokam "but a systems language": C is not used for an AI website
07:29:08FromDiscord<heysokam> the fact that they are not -usually- used for other things is not that it cant be done, just that for x or y reasons its -easier- to do in other langs
07:29:21FromDiscord<demotomohiro> It seems Nim can do anything C can do.
07:29:22FromDiscord<eerbium> (edit) "In reply to @heysokam "but a systems language": C is not used for an AI website ... " added "(eg)"
07:29:24FromDiscord<heysokam> In reply to @eerbium "C is not used": but it could
07:29:59FromDiscord<heysokam> you would write C that you compile to wasm or wasi and run that from a simple ts/js glue layer
07:30:08FromDiscord<heysokam> and you would have your ai website in C
07:30:27FromDiscord<eerbium> hm well then why is nim not widely adopted
07:30:47FromDiscord<heysokam> because its very niche, and people tend to not like python syntax
07:31:05FromDiscord<heysokam> but feature-wise, its up to par to cpp/c/rust/python/younameit
07:31:07FromDiscord<eerbium> the 2nd part is understandable but how is it niche?
07:31:16FromDiscord<heysokam> 🤷‍♂️
07:31:24FromDiscord<graveflo> are you asking if nim is turing complete?
07:31:32FromDiscord<heysokam> i heard different opinions. but nobody knows is the only real one i assume
07:31:50FromDiscord<heysokam> (edit) "nobody knows" => ""nobody knows""
07:32:21FromDiscord<eerbium> languages like Zig are not even 1.0 yet they are seeing widespread adoption but not nim
07:34:15FromDiscord<heysokam> sent a long message, see http://ix.io/4ywV
07:35:12FromDiscord<heysokam> imo, nim is better than those just because prototyping speed is ridiculously faster, due to the syntax and stdlib↵but... everybody has their preferences
07:35:12FromDiscord<evissim> Popularity is usually more complicated than good = popular, with how a language markets itself, what projects use it and get exposure etc
07:35:21FromDiscord<demotomohiro> Maybe there are many people don`t like python-like syntax?
07:35:41FromDiscord<graveflo> yea I think that content creators could do a lot to increase nims popularity
07:36:05FromDiscord<heysokam> ^ in my experience, a lot of them. like a lot
07:36:34FromDiscord<heysokam> imo its more rare to find a systems programmer that likes python syntax than the opposite
07:36:43FromDiscord<heysokam> (edit) "syntax" => "syntax,"
07:36:50FromDiscord<demotomohiro> https://github.com/nim-lang/Nim/wiki/Organizations-using-Nim
07:37:12FromDiscord<toma400> In reply to @heysokam "because its very niche,": That was one of the reasons I so fell in love with Nim... it's so clean language to write code in, unlike C-syntaxed ones
07:37:21FromDiscord<heysokam> yeah agreed
07:38:08FromDiscord<heysokam> its so hard to write ugly nim. like you have try really hard, and it still looks somewhat clean
07:38:44FromDiscord<Elegantbeef> You can just read sokam's code for ugly nim
07:39:02FromDiscord<Elegantbeef> `a : t` is just wrong 😛
07:40:31FromDiscord<demotomohiro> Nim is detected by antivirus softwares can be another reason:↵https://internet-of-tomohiro.netlify.app/nim/faq.en.html#nim-compiler-is-nimslashnimble-virus-or-malwareqmark
07:40:54FromDiscord<graveflo> oh yea I forgot about that one
07:40:59FromDiscord<heysokam> tru 🙈
07:41:12FromDiscord<heysokam> but it would be `a :Type` though
07:41:44FromDiscord<demotomohiro> I heard Nim is detected by antivirus soft and some companys dont allow to use Nim.
07:44:05FromDiscord<graveflo> couple goofballs write some malware and now the standard library has a false positive fingerprint. Again an exposure problem. Clean examples in the wild would improve the heuristic scanners
07:44:31FromDiscord<eerbium> well I mean the thing is Nim doesnt have any unique reason to use it
07:44:52FromDiscord<eerbium> Most people in a job know python, js, java
07:45:13FromDiscord<eerbium> They will just use that they dont have any need to take the effort to use nim
07:45:31FromDiscord<graveflo> I understand the reasoning. Sometimes you want to know what the gimmick is. Its essentially syntax, meta-programming and feature implementation
07:46:26FromDiscord<graveflo> there is a way to do things in nim that exposes high level design patterns in a simple low level way. It's pretty remarkable when you start to learn the ropes
07:49:17FromDiscord<toma400> In reply to @eerbium "well I mean the": If you love Python, but are annoyed by lack of private access modifiers, poor strict typing or performance, Nim is absolutely has all reasons to use it
07:49:31FromDiscord<eerbium> I dont mean for production
07:49:32FromDiscord<toma400> And in my case, it's also gateway to low-level, as it is optional in Nim
07:49:34FromDiscord<eerbium> I mean for prototyping
07:50:07FromDiscord<graveflo> performance is pretty much the only issue in python tbh
07:50:10FromDiscord<Elegantbeef> Nim is great for prototyping
07:50:14FromDiscord<eerbium> during production most companies use languages they developed themselves
07:50:24FromDiscord<Elegantbeef> Yea like C/C++/Rust
07:50:33FromDiscord<Elegantbeef> Totally languages developed in house
07:50:49FromDiscord<Elegantbeef> There are only a few places that use in house languages
07:50:52FromDiscord<Elegantbeef> Shiro games being one
07:50:53FromDiscord<toma400> In reply to @graveflo "performance is pretty much": The only issue that is very solid. Dynamic system is two-sided sword, so it's huge issue, but also huge benefit.
07:50:58FromDiscord<eerbium> I dont mean like the top 10 companies
07:51:26FromDiscord<Elegantbeef> Also left out distributing python isnt great
07:51:28FromDiscord<eerbium> Almost every person I know with a job in the tech sector says that they code in some language you wont even find on google
07:51:46FromDiscord<Elegantbeef> Why don't I believe that
07:52:29FromDiscord<graveflo> because it would be a nightmare. Your friends work for companies that makes questionable decisions
07:53:15FromDiscord<toma400> In reply to @eerbium "Almost every person I": This doesn't make sense honestly, it's a suicidal decision to take
07:53:34FromDiscord<eerbium> well a lot of them are in TCS and Infosys which are the largest tech companies in India with
07:53:46FromDiscord<eerbium> (edit) removed "with"
07:54:01FromDiscord<graveflo> unless we are talking about weird scripting langs that are basically bloated config files made for people who don't know how to code
07:54:08FromDiscord<toma400> Like, even picking Nim (googlable thing) is hard because of how niche it is, thus making use of libraries harder, as there isn't too big range of them to choose↵That pushes you back a lot when creating whatever service
07:54:43FromDiscord<toma400> In reply to @graveflo "unless we are talking": You reminded me of times I "coded" mods for Battle for Wesnoth ❤️
07:54:43FromDiscord<eerbium> but doesnt Nim have interop with c/cpp/js?
07:54:44FromDiscord<emanresu3> Anyone else sad about Nim usage in stackoverflow's 2023 survey?
07:54:55FromDiscord<Elegantbeef> Hey it's in SO's survey
07:54:58FromDiscord<Elegantbeef> That's a first
07:55:01FromDiscord<graveflo> In reply to @eerbium "but doesnt Nim have": yes
07:55:06FromDiscord<toma400> In reply to @eerbium "but doesnt Nim have": It has, yeah. I got to learn using C libs in my code indeed.
07:55:12FromDiscord<Elegantbeef> You still have to write the bindings
07:55:25FromDiscord<Elegantbeef> Futhark exists but no one wants to use Nim to just write C
07:56:35FromDiscord<graveflo> true that. One of the biggest things that I think nim could alleviate compared to C is in build systems. C and C++ build systems make me want to swallow pills
07:56:35FromDiscord<Elegantbeef> Also grave, grats on the first PR 😄
07:56:40FromDiscord<graveflo> ty
07:56:46FromDiscord<eerbium> I think I will still go with typescript and rust better to use the language better in their thing than an all rounder
07:57:01FromDiscord<Elegantbeef> Have fun
07:58:57FromDiscord<Elegantbeef> Who really expects the person with tate as a pfp to be sensible 😄
08:00:22FromDiscord<eerbium> yeah if that were the case nim would be the one in job searches not js
08:06:30FromDiscord<toma400> sent a long message, see http://ix.io/4yx3
08:07:25FromDiscord<Elegantbeef> Well the language might have some warts but it's a really good premise
08:08:07FromDiscord<Elegantbeef> I have very little interest in using anything else myself, as someone that technically is a hobbyist there just is not anything that really stands out
08:27:13FromDiscord<toma400> sent a long message, see http://ix.io/4yx8
08:47:36FromDiscord<heysokam> sent a code paste, see https://paste.rs/8kyBK
08:48:49FromDiscord<graveflo> you want to simply silence it or is it causing a breakage?
08:49:01FromDiscord<heysokam> just silence it, since its not relevant to this project
08:49:12FromDiscord<graveflo> redirect stdout and srderr
08:49:19FromDiscord<graveflo> (edit) "srderr" => "stderr"
08:49:33FromDiscord<heysokam> how? im just launching `nimble mytask` from console
08:50:21FromDiscord<heysokam> and the rest of the output is great, and useful for everything im doing. not all of stdout is useless, only those messages
08:50:37FromDiscord<graveflo> I'm not sure if nim has a boxed up way to do this but you should be able to give that nimble invokaction a pipe for stdout and stderr. You can filter the messages from this pipe and re-emit them to the real stdout and stderr
08:51:01FromDiscord<graveflo> or re-emit a filtered version
08:51:13FromDiscord<heysokam> wouldn't that be system-specific for my pc?
08:51:57FromDiscord<graveflo> yes but it's a common thing. It either exists already for nim or you can make it without too much grief. Dont quote me on that. There might be a simpler way but thats all I know
08:52:12FromDiscord<graveflo> what proc are you using to start the subprocess?
08:52:33FromDiscord<heysokam> then it defeats the purpose of using nimble at all. i would rather just write my own script than require every user of the buildsystem to do that as a pre-requisite 😔
08:53:05FromDiscord<graveflo> well I mean you can support nix and windows and pretty much have all your bases covered
08:53:08FromDiscord<heysokam> In reply to @graveflo "what proc are you": `nimble mytask`
08:53:26FromDiscord<graveflo> oh so they run nimble straight up not from a nim program?
08:53:30FromDiscord<heysokam> yeah
08:53:36FromDiscord<heysokam> no not from a nim program
08:53:39FromDiscord<graveflo> then you might have to modify nimble LOL
08:53:43FromDiscord<graveflo> or PR it or something
08:54:01FromDiscord<heysokam> its just nimble being there for solving the dependency, and launching the buildsystem app
08:54:52FromDiscord<graveflo> you could make your system such that there is a valid nimble path and that nimble path runs the bootstrap for your program
08:55:08FromDiscord<heysokam> sent a code paste, see https://play.nim-lang.org/#ix=4yxe
08:55:33FromDiscord<heysokam> its just easier to self-contain everything with nimble solving the dependency auto
08:56:42FromDiscord<heysokam> In reply to @graveflo "you could make your": i thought about distributing nim like zig does, but i think that might be overkill 🤔
08:56:48FromDiscord<graveflo> well I idk how to change the way nimble works without modifying nimble or messing with it using code beforehand. Maybe it's ok that confy must be installed. That is very normal requirement in these scenarios
08:57:07FromDiscord<graveflo> unless you require a launcher but thats even more complicated
08:57:46FromDiscord<heysokam> In reply to @graveflo "well I idk how": but the user is already installing cmake, rust, and nim to build this project specifically↵i think one less friction point should be good
08:58:22FromDiscord<heysokam> confy does not require those, so its fine for other stuff. but this project is already a bit dependency-heavy
08:58:27FromDiscord<graveflo> hey installers and package managers make this really easy for users on the OS level
08:58:50FromDiscord<graveflo> not that supporting every linux distro is easy though 😆
08:58:56FromDiscord<heysokam> yeah, the same can be argued about anything. you can do all of it manually. its just that removing one step is never a bad thing
08:59:12FromDiscord<heysokam> the less steps the less friction
09:00:38FromDiscord<graveflo> yea I can only think that modifying nimble to have a warning suppression flag or having a launcher are your two ways to make that happen. I can't see how it would be possible otherwise
09:01:01FromDiscord<heysokam> yeah i figure. i was just wondering if that flag already exists in nimble
09:01:42FromDiscord<graveflo> there is `--silent`
09:01:58FromDiscord<graveflo> but that has no filter
09:02:37FromDiscord<graveflo> wait a minute how are you going to pass the flag in the first place
09:02:45FromDiscord<heysokam> and the user needs to add it themselves, otherwise it would be good enough i guess
09:02:47FromDiscord<graveflo> something isn't adding up here
09:03:11FromDiscord<graveflo> oh if the user has to add it then that is even worse then requiring the install i say
09:03:20FromDiscord<heysokam> yeah
09:03:58FromDiscord<graveflo> I wouldn't worry about it. A tool that solves a problem is worth a single command
09:04:14FromDiscord<graveflo> especially if you have a build script or an installer / package
09:04:19FromDiscord<graveflo> no one will care
09:04:52FromDiscord<heysokam> yeah for sure. but im just looking for ways to improve the usability. and those messages there are irrelevant noise to the person who doesn't know they are irrelevant noise
09:05:11FromDiscord<graveflo> understandable
09:06:19FromDiscord<heysokam> i could change my `build.nim` file to make it have the same name as the project, and that would solve the warning. but that would require an extra step for setting up the basic structure. and the whole point is to not have any setup unless needed
09:07:41FromDiscord<heysokam> well, only the hint would be solved, though i believe
09:07:48FromDiscord<heysokam> the warning would still be there if im not mistaken
09:08:04FromDiscord<heysokam> (edit) "solved, though" => "solved though,"
09:08:43FromDiscord<graveflo> I imagine that just pulling up the folder and running a single command would be ideal regardless of the setup. Like if someone could just go to the folder and type `confy build` and then it just works. It doesnt matter what needs to be installed really
09:09:01FromDiscord<graveflo> a simple setup script might be the key if you want prime usability
09:09:07FromDiscord<heysokam> that would mean i have to distribute a binary
09:09:15FromDiscord<heysokam> but the buildsystem is a library
09:09:27FromDiscord<heysokam> i thought about that, but im not liking the launcher idea so far 🤔
09:09:35FromDiscord<graveflo> or you could use nimble. It could build the cli tool that drops the file and auto installs in a single command
09:10:11FromDiscord<heysokam> In reply to @graveflo "a simple setup script": you mean something that initializes the project like git init or nimble init?
09:10:27FromDiscord<graveflo> yea basically
09:10:45FromDiscord<heysokam> not a bad idea. but how do you do that with just source code and no binary?
09:11:16FromDiscord<heysokam> would be a useful feature if i went the `nimble install confy` route, and distribute some basic launcher functionality
09:11:20FromDiscord<graveflo> nimble can create the binary on install I think.. Ive never used it so I might be totally wrong
09:12:01*marcus quit (Remote host closed the connection)
09:12:07FromDiscord<heysokam> yeah totally. but the idea about confy is that it is not a binary, so i don't have to hardcode what you will do. i can keep it configurable thanks to the code just being a set of functions that are called from the scripts
09:12:34FromDiscord<heysokam> (edit) "the" => "your"
09:13:17FromDiscord<heysokam> so while in buildsystems you usually setup a script that is run from a harcoded binary... in confy your script is actually a true app that you build
09:13:32FromDiscord<heysokam> (edit) "run" => "run/interpreted"
09:13:45FromDiscord<heysokam> (edit) "so while in buildsystems you usually setup a script that is run/interpreted from a harcoded binary... in confy your script is actually a true app that you build ... " added "yourself"
09:13:58FromDiscord<graveflo> I guess what I'm saying is you can make a cli tool that automates the setup process to make nimble happy. It even has the potential to branch out into bigger things like auto downloading known working install configs. Ah well maybe this is just getting complicated. I'm sure just installing confy through nimble is best for now
09:14:02*marcus joined #nim
09:14:08FromDiscord<heysokam> (edit) "yourself" => "yourself. the usability im looking for is making this last step easy to do"
09:14:48FromDiscord<heysokam> In reply to @graveflo "I guess what I'm": for the launcher route it definitely is the simplest way, yeah
09:23:39*robertmeta quit (Server closed connection)
09:24:15*robertmeta joined #nim
09:53:55NimEventerNew thread by alexeypetrushin: Why json hooks doesn't work for CountTable[int], see https://forum.nim-lang.org/t/10281
10:07:49FromDiscord<Require Support> iirc nim ny default builds a debug build but using `when defined(debug)` does not run the code within. what is wrong?
10:08:28FromDiscord<Require Support> (edit) "ny" => "by"
10:12:56FromDiscord<graveflo> manually pass `-d:debug` to define it
10:13:10FromDiscord<graveflo> idk if that is intentional or not on nims side
10:25:27FromDiscord<spotlightkid> Or use `when not defined(release)`
10:25:55FromDiscord<graveflo> is there a proc that dumps all defined variables? I would have used that multiple times by now
11:04:33FromDiscord<graveflo> It's the `dump` cli option. How do I always misread those
11:08:36*ixmpp quit (Server closed connection)
11:12:39FromDiscord<gogolxdong666> Does Nim support macOS m2?
11:17:39FromDiscord<graveflo> it builds from system compiler so I would guess so. I don't know for sure though
11:18:32FromDiscord<spotlightkid> why wouldn't it?
11:20:00FromDiscord<spotlightkid> If you want to build for macos arm64 on macos intel, you probably have to use the right `--passC:` / `--passL:` options.
11:20:27FromDiscord<graveflo> M2s still have objectivec support right?
11:20:54FromDiscord<graveflo> or more accurately, the other way around
11:32:14FromDiscord<tauruuuuuus> does anyone know if there is a way to check for cygwin/mingw within a nim.cfg file? It seems like `@if windows:` is not enough
11:51:45FromDiscord<Saint> Does the nim lsp support listing all symbols?
11:52:13FromDiscord<Saint> Also does anyone have thoughts on mojo compared to nim?
11:58:22FromDiscord<toma400> In reply to @Saint "Also does anyone have": Mojo isn't even publicly available, right? It feels very speculative at the moment to me, so I wait for real implementation
12:02:36FromDiscord<emanresu3> hype is everything
12:03:40FromDiscord<Saint> https://github.com/nim-lang/langserver
12:03:49FromDiscord<Saint> Anyone use this vs https://github.com/PMunch/nimlsp
12:03:53FromDiscord<Saint> @pmunch
12:04:14FromDiscord<Saint> Is there a lot of work to get workspace/symbol in nimlsp?
12:04:39*azimut joined #nim
12:07:22FromDiscord<Saint> Oooh nim-langserver already is in nvim-lspconfig
12:07:26FromDiscord<Saint> Will give it a spin
12:24:35FromDiscord<pmunch> They're pretty similar, mostly different architectures around the same base tool (nimsuggest). Workspace/symbol shouldn't be too hard I believe, I think nimsuggest already has that information on offer
12:31:43FromDiscord<salt rock lamp> I have been meaning to try adding nimlsp to nvim-lspconfig for a while
12:32:23FromDiscord<salt rock lamp> It's not that hard to add a new config
12:49:32*Onionhammer quit (Ping timeout: 240 seconds)
12:49:44FromDiscord<michaelb.eth> In reply to @tauruuuuuus "does anyone know if": one approach would be to use an MSYS2 environment and check if env var `MSYSTEM` is set, assuming nim.cfg can check env vars (config.nims can do that)
12:52:40*Onionhammer joined #nim
13:34:43FromDiscord<krab4t> what the difference between langserver and nimlsp?
13:43:09*derpydoo joined #nim
13:44:52FromDiscord<krab4t> ` Both nimlangserver and nimlsp are based on nimsuggest, but the main difference is that nimlsp has a single specific version of nimsuggest embedded in the server executable, while nimlangserver launches nimsuggest as an external process. This allows nimlangserver to handle any nimsuggest crashes more gracefully. `
14:26:51NimEventerNew thread by mantielero: Macro - global var behaviour, see https://forum.nim-lang.org/t/10282
14:26:53FromDiscord<krab4t> https://pastebin.com/ZaVWet1g something fishy happens here
14:27:11FromDiscord<krab4t> it prints
14:27:37FromDiscord<krab4t> result is sequence of 2 strings
14:27:52FromDiscord<krab4t> of first iteration: `1. result1`
14:28:12FromDiscord<krab4t> on second: `1. result1\n 1.result1\n 2.result2\n`
14:28:43FromDiscord<krab4t> result1 result2 just a text
14:28:51NimEventerNew thread by mantielero: Play nim - not working?, see https://forum.nim-lang.org/t/10283
14:34:10FromDiscord<Yardanico> In reply to @krab4t "https://pastebin.com/ZaVWet1g something fishy happe": `myText.add("$1$2. $3\n" % [myText, $pos, r])`
14:34:12FromDiscord<Yardanico> you're adding myText to itself
14:40:14FromDiscord<krab4t> yes
14:40:42FromDiscord<Yardanico> so what is the issue?
14:41:25FromDiscord<krab4t> twice result1
14:41:28FromDiscord<odexine> Because you’re adding my text to itself
14:41:39FromDiscord<Yardanico> In reply to @krab4t "twice result1": after first iteration your myText will have 1. result1
14:41:45FromDiscord<Yardanico> and on the second iteration you're adding myText to myText
14:41:47FromDiscord<odexine> You’re doing result = result + result + other stuff
14:41:48FromDiscord<krab4t> https://pastebin.com/1dZpeR4H same in python works fine
14:41:57FromDiscord<Yardanico> In reply to @krab4t "https://pastebin.com/1dZpeR4H same in python": it is not the same
14:42:05FromDiscord<Yardanico> in Nim you're `.add` ing to the string
14:42:06FromDiscord<odexine> You’re doing a set = not an add
14:42:09FromDiscord<Yardanico> in Python you're changing the string itself
14:42:26FromDiscord<krab4t> rip
14:42:27FromDiscord<odexine> Nim add is Python append
14:42:54FromDiscord<Yardanico> also can't you rework this to actually use `add` without having to reassign text itself?
14:42:57FromDiscord<Yardanico> it should be simpler too
14:43:35FromDiscord<krab4t> i see i did assignment in python
14:44:43FromDiscord<krab4t> nimlangserver https://i.imgur.com/jieGeXG.png
14:45:49FromDiscord<Yardanico> is that with vscode?
14:46:06FromDiscord<Yardanico> maybe your IDE is opening a different lsp "project" for each file
14:46:22FromDiscord<Yardanico> I'm happy with nimsaem in vscode without lsp, but I've used nimlsp before (not nimlangserver) and it didn't lead to a lot of processes like this
14:46:26FromDiscord<Yardanico> (edit) "I'm happy with nimsaem in vscode without lsp, but I've used nimlsp before (not nimlangserver) ... and" added "in other editors"
14:49:13FromDiscord<Yardanico> oh nvm i see, it's nimlangserver itself doing this
14:49:32FromDiscord<Yardanico> well, nimlsp bundles nimsuggest as a library so it might be better, give it a try :)
14:52:24FromDiscord<krab4t> now just one with projectMapping configuration as described on github page
14:53:48FromDiscord<krab4t> my telegram bot in python variant uses 50mb, and just 1mb in nim ...
15:03:21FromDiscord<krab4t> working with sqlite relations in norm https://pastebin.com/B1pecktW most pepega thing ive ever seen
15:05:01FromDiscord<krab4t> in ponyorm https://pastebin.com/GeZ1Nu17
15:16:06*rockcavera joined #nim
15:54:48*antranigv quit (Quit: ZNC 1.8.2 - https://znc.in)
15:55:15*antranigv joined #nim
16:12:29FromDiscord<tauruuuuuus> In reply to @michaelb.eth "one approach would be": I can't believe I've been working without knowing that config.nims was a thing...
16:12:37FromDiscord<tauruuuuuus> so much easier and clean, thanks a lot!
16:39:54FromDiscord<Ras> sent a code paste, see https://play.nim-lang.org/#ix=4yz1
16:40:37FromDiscord<Ras> i tried narrowing it down and removing cruft which i thought might be causing this, like removing the global var seq etc, but nothing seemed to help at all
16:44:31FromDiscord<Ras> sent a code paste, see https://play.nim-lang.org/#ix=4yz2
16:44:56*azimut quit (Ping timeout: 240 seconds)
16:49:52FromDiscord<Ras> testing it now, it seems to always happen on the fifth opened connection
16:55:23FromDiscord<yg0068> Hey guys. Never thought I'd be asking this, but does nimble develop work for anyone since it was reimplemented? How do I make project A see dependency B, which is only available locally?
17:13:13FromDiscord<Ras> In reply to @Ras "testing it now, it": interestingly enough, when i add something like printing the `seq[Player]` after every connection, the second connection already crashes it
17:14:00FromDiscord<Ras> if i change things around so that less is printed, then the third connection crashes it
17:14:15FromDiscord<Ras> is this some sort of gc thing? it kind of smells like it
17:15:57FromDiscord<k0ts> In reply to @yg0068 "Hey guys. Never thought": I've heard that `nimble develop --global` makes it work similarly to the old behavior
17:15:59FromDiscord<Ras> yeah, if i do `--gc:none`, i don't get any crashes
17:16:02FromDiscord<Ras> how do i appease the gc gods?
17:20:56FromDiscord<Ras> that doesn't look good
17:20:58FromDiscord<Ras> sent a code paste, see https://play.nim-lang.org/#ix=4yz8
17:21:30FromDiscord<Ras> (this is me experimenting with `useMalloc`, which also crashes)
17:26:02FromDiscord<Ras> so what am i doing wrong?
17:26:03FromDiscord<Ras> sent a code paste, see https://paste.rs/Fp86n
17:33:38FromDiscord<yg0068> In reply to @k0ts "I've heard that `nimble": Yeah, that doesn't work for me. The link is created, but Nim just doesn't see it :(. That's why I wonder if anyone really uses it, or just "heard of it" 🙂
17:34:29FromDiscord<k0ts> I just tried it with a dummy project and local dependency and it worked
17:35:57FromDiscord<Yardanico> In reply to @Ras "yeah, if i do": how long ago you installed devel?
17:36:12FromDiscord<Yardanico> it might be a but with orc, meanwhile try `--mm:refc`, it's the old GC
17:36:14FromDiscord<Yardanico> (edit) "but" => "bug"
17:36:16FromDiscord<Ras> about 30 minutes ago, `choosenim devel --latest`
17:36:37FromDiscord<Ras> but it was also happening with the one that's installed without `--latest`
17:36:43FromDiscord<Yardanico> `--mm:arc` won't really work for you because async has cycles which `arc` doesn't deal with (only `orc` does)
17:37:33FromDiscord<Yardanico> In reply to @Ras "but it was also": it'll probably work with `--mm:refc`, but please report it on github issues anyway, orc issues are pretty important
17:38:26FromDiscord<Ras> you're right, refc works great
17:39:12FromDiscord<Ras> i'll try to report it, problem is making a minimal proof of concept...
17:39:23FromDiscord<Ras> it doesn't _seem_ to be happening with something minimal like a chat server
17:39:28FromDiscord<Ras> and i'm not really sure what is triggering it
17:39:41FromDiscord<Yardanico> In reply to @Ras "i'll try to report": i think it's worth to report it even in the current state, your project doesn't look too complex
17:39:49FromDiscord<Yardanico> seems like you're trying to write a minecraft server?
17:40:11FromDiscord<Ras> yep, for classic (initially, then implement the classic extensions as well)
17:40:26FromDiscord<Ras> good guess :D
17:41:45FromDiscord<Yardanico> In reply to @Ras "more complete stack trace:": btw, the usual way to debug memory stuff with orc is with `-d:useMalloc --mm:orc` and then run that with `valgrind --leak-check=full ./binary`
17:45:02FromDiscord<k0ts> In reply to @yg0068 "Yeah, that doesn't work": After doing nimble develop, do you then compile your project through nimble (with `nimble build` or something) or directly with `nim c`? If the latter, you might need to `nimble setup` first so Nim knows where to find the local dependency.
17:53:11FromDiscord<Ras> In reply to @yardanico "btw, the usual way": hmmm, it doesn't seem to crash when running it with valgrind, is this.. normal?
17:58:47FromDiscord<yg0068> In reply to @k0ts "After doing nimble develop,": Interesting. I'm using `nim c`. Just did `nimble develop -g` in project A, went to project B, and did `nimble setup`. New files appeared, but they don't mention A, and B doesn't compile. Am I doing it wrong?
17:59:10FromDiscord<Yardanico> In reply to @Ras "hmmm, it doesn't seem": It might not crash but it should show errors
18:00:15FromDiscord<Ras> oh yeah, `ERROR SUMMARY: 3257 errors from 183 contexts (suppressed: 0 from 0)`
18:01:42FromDiscord<k0ts> In reply to @yg0068 "Interesting. I'm using `nim": Odd, I'm doing the exact same thing and it works for me.
18:04:30FromDiscord<k0ts> Right, did you add requires "A" to B.nimble?
18:04:51FromDiscord<yg0068> In reply to @k0ts "Odd, I'm doing the": Ah, one of the dependencies was missing from the nimble file. It seems to do better now. Thanks @k0ts
18:05:32FromDiscord<k0ts> glad I could help
18:11:35FromDiscord<basilajith> In reply to @Elegantbeef "I have very little": I'm willing to stick to front-end development (I want to do back-end dev more), just so that I won't have to learn other languages like TypeScript, Go, PHP, C# (.NET) etc. However, I am a bit wary about the lack of contributors in the web dev space for Nim. I'm a bit sad that people like dom96 left the Nim core team (his writing was really good).↵↵Of course, if I had the expertise, I
18:11:52FromDiscord<michaelb.eth> In reply to @tauruuuuuus "so much easier and": Glad to help!
18:30:57*krux02 joined #nim
18:39:13*nyeaa492842301 quit (Ping timeout: 265 seconds)
18:42:07*nyeaa492842301 joined #nim
18:42:39*Ekho quit (Server closed connection)
18:55:30*derpydoo quit (Ping timeout: 240 seconds)
18:57:23*Ekho joined #nim
19:20:00*ntat quit (Quit: Leaving)
19:20:08FromDiscord<Elegantbeef> @basilajith as someone that doesnt do web dev, I can only joke "Good you can now work on better things" 😄
19:25:44FromDiscord<voidwalker> are there some known issues with argument matching for generic procs? Or maybe I am just stupid
19:25:58FromDiscord<Elegantbeef> Depends, likely the latter
19:26:16FromDiscord<voidwalker> sent a code paste, see https://play.nim-lang.org/#ix=4yzu
19:27:06FromDiscord<voidwalker> sent a code paste, see https://play.nim-lang.org/#ix=4yzv
19:27:26FromDiscord<voidwalker> If I call with `udpTrackerScrape("125b77979ba4b183eb702f4ff00df9ff22c452d7", "udp://tracker.opentrackr.org:1337/announce")` - it works
19:27:30FromDiscord<Elegantbeef> `[1]: seq[string]`
19:27:48FromDiscord<Elegantbeef> People really need to stop using generics for overloading
19:28:33FromDiscord<voidwalker> yeah I get a seq[string] there but I shouldn't have a seq[string], hextoByte returns `seq[byte]`
19:30:15FromDiscord<Elegantbeef> Wait it's erroring that it's a `@[...]` which means this is erroing at line 10
19:30:53FromDiscord<Elegantbeef> Or you've provided incorrect code
19:30:57FromDiscord<Elegantbeef> Cause there is no `@[...]`
19:30:59FromDiscord<voidwalker> it errors at the last line, when I call
19:31:10FromDiscord<Elegantbeef> But it's created a sequence
19:31:15FromDiscord<Elegantbeef> You are not passing a sequence
19:31:21FromDiscord<Elegantbeef> `udpTrackerScrape("125b77979ba4b183eb702f4ff00df9ff22c452d7", parseUri("udp://tracker.opentrackr.org:1337/announce"))`
19:31:23FromDiscord<Elegantbeef> Does not have a sequence
19:31:32FromDiscord<Elegantbeef> So where the hell does the sequence get injected?!
19:32:49FromDiscord<voidwalker> well the second generic/overload proc should make it into a sequence, cause the main proc is written to take a sequence. But it should not be a `seq[string]`, but `seq[seq[byte]`
19:33:00FromDiscord<voidwalker> ` await udpTrackerScrape(@[infoHash], tracker)`
19:33:04FromDiscord<Elegantbeef> Cool, but that's not what's happening
19:33:28FromDiscord<Elegantbeef> We're not even getting into that proc
19:33:29FromDiscord<Elegantbeef> It's converting your string into a sequence then erroring
19:34:57FromDiscord<Elegantbeef> Do you accidentally have another usage somewhere with `@["mystring"]` inside it
19:34:59FromDiscord<Elegantbeef> Cause that's what it looks like
19:35:13FromDiscord<Elegantbeef> You either forgot to save, forgot to delete a line, or forgot to implement something
19:36:44FromDiscord<voidwalker> welp 😦 that was it. hidden between the forest of code, was a template that I written long ago to do the overload
19:38:09FromDiscord<voidwalker> time for more overloads now : P
19:38:28FromDiscord<Elegantbeef> As much as people complain about Nim's errors, many times the code they're compiling is worse 😛
19:45:34FromDiscord<voidwalker> what is a good strategy here ? I made the main function to work with seq cause the protocol is supposed to allow n hashes to query. And I want to call it with hexstring(s), binary data, string url, Uri
19:45:54FromDiscord<Elegantbeef> I'd use distinct strings myself, but you do you
19:47:10FromDiscord<voidwalker> distinct strings for hex hash params ?
19:47:28FromDiscord<Elegantbeef> Yep
19:48:18FromDiscord<voidwalker> oh like type casted string in the calling proc ?
19:48:28FromDiscord<Elegantbeef> `type HexString = distinct string`
19:48:44FromDiscord<Elegantbeef> then if you want to use hex string you do `yourProc(HexString myString)`
19:49:09*mad_toothbrush joined #nim
19:50:46*derpydoo joined #nim
19:52:27mad_toothbrushHi folks. A question. Is it possible to get formatted output using echo in a --os=standalone environment ? I was hoping I could provide a low level 'putc' like implementation and register it somehow with Nim. Kind of like Rust's fmt::Debug trait. Is that possible ? Thanks.
19:53:02FromDiscord<Elegantbeef> You could implement your own echo variation
19:53:52FromDiscord<Elegantbeef> https://hatebin.com/asjyrgralz
19:54:03FromDiscord<Elegantbeef> replace `yourPrinter` with whatever you decide for your `putc` variation
19:54:51FromDiscord<Elegantbeef> Nim is designed quite nicely in a way that 90% of things are implemented in userspace which means you can pretty much replicate most things the system module does
19:55:26FromDiscord<voidwalker> is this too awful ? : D
19:55:28FromDiscord<voidwalker> sent a code paste, see https://play.nim-lang.org/#ix=4yzB
19:55:46FromDiscord<Elegantbeef> Yes you've successfully re-implemented overloading
19:56:05FromDiscord<Elegantbeef> God you like repeating yourself
19:56:44mad_toothbrushThanks Elegantbeef! :) I'm looking at your hatebin example. For some reason play.nim-lang.org won't open.
19:57:45mad_toothbrushI'm a complete Nim n00b so please bear with me but wouldn't I need to implement the business logic for different format specifiers on a case by case basis ? (I must be missing something very obvious - apologies!)
19:58:05FromDiscord<Elegantbeef> you could use `std/strformat` but Nim's echo doesnt do any formatting
19:58:14mad_toothbrushAh, I see.
19:58:52FromDiscord<Elegantbeef> `myEcho fmt"{a}, {b}, {c}"` or `myEcho "$#, $#, $#" % [$a, $b, $c]` from strutils
20:03:44FromDiscord<toma400> sent a code paste, see https://play.nim-lang.org/#ix=4yzD
20:04:25FromDiscord<Elegantbeef> If you want a plugin system you'll want to use system libraries or wasm
20:04:38FromDiscord<Elegantbeef> I'm pretty preferential to using wasm, but system libraries work aswell
20:06:06FromDiscord<toma400> In reply to @Elegantbeef "I'm pretty preferential to": Could you guide me to some resources? ^^
20:06:36FromDiscord<Elegantbeef> https://nim-lang.org/docs/dynlib.html
20:06:42FromDiscord<Elegantbeef> https://github.com/beef331/wasm3
20:06:55FromDiscord<Elegantbeef> I'd say that unless you really really need a plugin system this is not wise
20:07:03FromDiscord<Elegantbeef> My entire game compiles in like 4 seconds
20:08:32FromDiscord<dissolved.girl> What's the performance like using wasm? Much overhead when calling stuff from wasm?
20:08:47FromDiscord<Elegantbeef> No clue don't really care
20:09:12FromDiscord<Elegantbeef> It's meant for agnostic scripting, wasm3 is a fast enough vm for that
20:09:47FromDiscord<toma400> In reply to @Elegantbeef "I'd say that unless": Well, I'm working on playground right now, so it's more on experimental side anyway. And I want to see how much python-like systems I can bring into Nim, it gives a lot of space for users to dynamically embbed scripts.↵Thanks a lot for the links ^^
20:10:12FromDiscord<Elegantbeef> Well proper Nim doesnt allow what you did so you have to do what you'd do in C/C++
20:10:59FromDiscord<Elegantbeef> In wasm's case you'd compile your modules to `wasm` then load them and interact with them, with system libraries you'd compile to `.so` `.dll` or `dynlbi`
20:11:02FromDiscord<Elegantbeef> `dynlib` even
20:11:34FromDiscord<Elegantbeef> The major downside is that the system library isnt sandboxed and if you want to ship it to another OS you have to recompile for that platform
20:11:45FromDiscord<Elegantbeef> Wasm is really nice that it is quite literally compile once, run anywhere
20:23:36FromDiscord<jmgomez> NimScript may also work for you. I’m loading 100K+ loc sub 100ms at runtime. There are a lot of limitations compared to native Nim oc
20:25:04FromDiscord<jmgomez> If your goal is to tweak things quickly rather than shipping it may work. But it does require a lot of work to build all the necessary stuff for it to be useful though
20:29:33FromDiscord<toma400> sent a long message, see http://ix.io/4yzK
20:30:55FromDiscord<Andreas> In reply to @toma400 "I wanted to create": maybe `include` can help ?
20:32:17FromDiscord<Elegantbeef> They could make a macro to include all their directories, but it's not dynamic at runtime
20:32:26FromDiscord<jmgomez> For shipping, the real consideration to have in mind is that it performs really bad. Theoretically it can be improved but doing so it’s a project in itself. I would take a look a wasm as beef said if you want to ship it and requires decent perf
20:33:31FromDiscord<jmgomez> There are more trade off besides the perf in NimScript as it’s a subset of nim but depending on what you want to do it may no really matter
20:33:56FromDiscord<Elegantbeef> Plus there are 3000 different runtimes to choose for wasm some have JIT which can make performance even less of a concern
20:34:18*mad_toothbrush quit (Ping timeout: 252 seconds)
20:45:24FromDiscord<toma400> I will look into those resources given, so thanks a lot for those ideas from you both ^^
20:45:44FromDiscord<toma400> Also this could be neat nudge for me to remind myself some C, I haven't spent enough time with this language yet
22:08:29*disso-peach joined #nim
22:08:36FromDiscord<demotomohiro> sent a code paste, see https://play.nim-lang.org/#ix=4yA1
22:24:22FromDiscord<arathanis> i think const needs to be knowlable at compile time but before anything in the vm runs
22:37:36FromDiscord<voidwalker> should a proc that takes a `seq[seq[uint8]]` also be able to take a `seq[array[20, uint8]]` ? Cause i have no compile error, but I get `execution of an external program failed` in this scenario
22:38:31FromDiscord<arathanis> those types are not compatible as far as i know
22:38:36FromDiscord<arathanis> i would expect that to not work
22:38:43FromDiscord<arathanis> weird that the compiler says nothing though
22:43:37FromDiscord<voidwalker> oh I think I know what happens, it goes into infinite recursion, not actually calling the proc I wanted but the overloading one
22:44:14FromDiscord<demotomohiro> In reply to @arathanis "i think const needs": Maybe initializing const variables before running compile time code supposed to be doesn't cause strange result.
23:23:08*rockcavera quit (Remote host closed the connection)
23:24:20*rockcavera joined #nim
23:34:19*krux02 quit (Remote host closed the connection)
23:38:06FromDiscord<voidwalker> https://www.youtube.com/watch?v=iiHpi0ZkX_I&t=194s
23:38:08FromDiscord<voidwalker> this is so funny