<< 24-04-2025 >>
01:56:05*zgasma quit (Ping timeout: 248 seconds)
02:10:28FromDiscord<lainlaylie> don't those surfaces need to be destroyed
02:30:24FromDiscord<devlop_gaming> That's exactly what i been trying to find out. I been on Google and stuff and I'm still unsure.
02:31:09FromDiscord<lainlaylie> i think you need to free the surfaces and the textures
02:31:18FromDiscord<lainlaylie> otherwise they just keep accumulating
02:32:02FromDiscord<devlop_gaming> That would make sense and it would explain why I been seeing massive increase in memory🤔
02:32:10FromDiscord<devlop_gaming> I'll try it in a few minutes
02:47:27*alexdaguy joined #nim
02:59:56FromDiscord<aryzen> is using `when true:` to allow a collapsable region a code smell?
03:01:13FromDiscord<threefour> Get a fold plugin that lets you fold at arbitrary spots
03:01:29FromDiscord<aryzen> I can do #region and #endregion but that's vscode specific
03:02:40FromDiscord<aryzen> ~~I could also do a `region` macro~~
03:14:14FromDiscord<aryzen> hooo-ray my first macro ever
03:14:19FromDiscord<aryzen> very exciting
03:16:03FromDiscord<devlop_gaming> In reply to @lainlaylie "otherwise they just keep": Memory still rising. I don't think destroy is being called.
03:17:59FromDiscord<devlop_gaming> Nvm it is. Then how is my memory still going up quickly??
03:35:05FromDiscord<lainlaylie> dunno, use a memory debugger
03:37:59FromDiscord<lainlaylie> you're already calling DestroySurface after each CreateTextureFromSurface and DestroyTexture in `=destroy`?
04:01:42FromDiscord<devlop_gaming> Yes
04:02:11FromDiscord<devlop_gaming> In reply to @lainlaylie "you're already calling DestroySurface": Tho I don't think all the objects are being destroyed. Either that or I'm tweakin🤷‍♂️
04:02:47FromDiscord<devlop_gaming> Looked like there are less objects in the console than there are in the sequence but I'll look again and confirm that
04:14:41FromDiscord<devlop_gaming> I'll check tomorrow
04:45:16*m5zs7k quit (Ping timeout: 244 seconds)
05:08:35*alexdaguy quit (Quit: w)
05:22:40*redj joined #nim
06:03:01*m5zs7k joined #nim
06:33:05FromDiscord<mkadd.sh> hi, could someone help me solve some doubts about the language?
06:37:11FromDiscord<aryzen> both nimsuggest and nimlangserver are inoperably slow, what's going on?
06:37:49FromDiscord<aryzen> In reply to @mkadd.sh "hi, could someone help": doubts?
06:52:35FromDiscord<lainlaylie> enumerate your doubts so that someone can respond to them
07:12:00*andy-turner joined #nim
07:32:51FromDiscord<nocturn9x> I figured out why my threaded init code was slow btw
07:32:58FromDiscord<nocturn9x> it was only ever spawning one thread cuz I'm a big dummie :3c
08:43:00FromDiscord<nnsee> In reply to @nocturn9x "it was only ever": technically still threaded
08:43:05FromDiscord<nnsee> just not multi
08:43:14FromDiscord<nocturn9x> xD yeah
08:50:18*przmk quit (Remote host closed the connection)
08:51:41*przmk joined #nim
08:52:23*przmk quit (Remote host closed the connection)
08:52:47*przmk joined #nim
09:27:10*bcksl quit (Ping timeout: 260 seconds)
09:27:34*end quit (Ping timeout: 260 seconds)
09:58:27*bcksl joined #nim
10:07:43FromDiscord<Robyn [She/Her]> In reply to @aryzen "is using `when true:`": why not use a `block`?
10:08:54*przmk_ joined #nim
10:09:42*przmk quit (Ping timeout: 252 seconds)
10:09:42*przmk_ is now known as przmk
10:09:45*end joined #nim
10:25:28FromDiscord<aryzen> blocks can't have exports in them
10:26:24FromDiscord<aryzen> is nimlangserver/nimsuggest really supposed to be so unreliable and slow?
10:28:07FromDiscord<enthus1ast.> we have problems with nimsuggest for a long time unfortunately
10:28:41FromDiscord<enthus1ast.> some say it gets better, some say everything will get much better with incremental compilation and or nimony
10:29:20FromDiscord<enthus1ast.> to be honest, (in nvim) i just use treestitter for colors, thats it
10:30:59FromDiscord<aryzen> I mean I get it being slow cuz it basically compiles to get the info it needs, but for some reason my files just straight up don't load sometimes
10:31:39FromDiscord<enthus1ast.> vscode?
10:32:16FromDiscord<aryzen> yes... I also noticed my CPU is at 100%
10:32:21FromDiscord<enthus1ast.> with vscode + vim plugin, vscode stucks sometimes and i cannot type until this is done, sometimes its never done
10:32:39FromDiscord<enthus1ast.> yes, recommend disable it
10:32:45FromDiscord<enthus1ast.> i
10:33:34FromDiscord<enthus1ast.> it would be awesome if we have a 100% working intellisense BUT one can write good nim withouth it
10:34:15FromDiscord<aryzen> I mean, will I have better luck with zed?
10:34:38FromDiscord<enthus1ast.> idk i think this would use the same langsever etc?!
10:35:07FromDiscord<aryzen> yeah it uses nimlangserver too
10:35:07FromDiscord<enthus1ast.> afaik in the nim plugin you can disable nimsuggest
10:35:35FromDiscord<enthus1ast.> then you still have colors + shortcuts
10:46:44*derpydoo quit (Ping timeout: 252 seconds)
11:19:51FromDiscord<dawidek.2137> For some reason I have had less slowdown in zed, but it doesn’t make sense so it might have been a fluke
11:20:04FromDiscord<dawidek.2137> Compared to vscode
11:27:20*ntat joined #nim
12:47:46*ntat quit (Read error: Connection reset by peer)
12:47:56*ntat joined #nim
12:52:49*ntat quit (Quit: leaving)
13:29:09*ntat joined #nim
14:06:52*ntat quit (Quit: leaving)
14:07:59*ntat joined #nim
14:15:43FromDiscord<threefour> In my experience, nimlangserver in VS Code was actually usably responsive. nimlangserver in Emacs, with all the proper tweaks I can think of, is incredibly slow.
14:22:58FromDiscord<threefour> Like, stop typing and wait for 2 seconds every time you want to get suggestions, slow. Which doesn't sound like much when I say it, but that's just enough of a delay that it's rarely useful to me.
15:26:07*ntat quit (Ping timeout: 252 seconds)
15:27:53*ntat joined #nim
16:19:52*zgasma joined #nim
17:46:29*Lord_Nightmare quit (Quit: ZNC - http://znc.in)
17:49:24*Lord_Nightmare joined #nim
18:47:28FromDiscord<nocturn9x> I was wondering
18:47:43FromDiscord<nocturn9x> since it seems `createThread` can be called from pure functions
18:47:53FromDiscord<nocturn9x> wouldn't that allow calling impure functions from a pure function
18:47:58FromDiscord<nocturn9x> and then awaiting for the result
18:48:04FromDiscord<nocturn9x> thereby bypassing the checks of the compiler
18:48:46FromDiscord<nocturn9x> (I tried this and it works)
18:52:53*Guest5 joined #nim
18:53:00*Guest5 left #nim (#nim)
19:14:15FromDiscord<Robyn [She/Her]> Kinda wanna make a crappy CLI library ngl
19:34:55FromDiscord<griffith1deadly> In reply to @battery.acid.bubblegum "Kinda wanna make a": when asm (java bytecode manipulation library) in nim??
19:44:30*przmk quit (Remote host closed the connection)
19:45:36FromDiscord<threefour> In reply to @nocturn9x "thereby bypassing the checks": 😬
19:46:19*przmk joined #nim
19:51:41*CypherCat quit (Remote host closed the connection)
20:08:24FromDiscord<dawidek.2137> does someone know how is default allocator different from malloc?
20:12:06FromDiscord<aryzen> In reply to @threefour "In my experience, nimlangserver": Then I'm wondering what's wrong wrong with my setup :S
20:27:28FromDiscord<Robyn [She/Her]> In reply to @griffith1deadly "when asm (java bytecode": Tbf that's possible with native image
20:32:41*void09 quit ()
20:33:18*void09 joined #nim
20:35:39FromDiscord<griffith1deadly> In reply to @battery.acid.bubblegum "Tbf that's possible with": i mean class parser & writer
20:37:48FromDiscord<Robyn [She/Her]> In reply to @griffith1deadly "i mean class parser": yeah, use native image on the asm lib for Java and use the shared library
20:38:14FromDiscord<Robyn [She/Her]> i need to actually complete some projects 😭
20:38:43FromDiscord<griffith1deadly> In reply to @battery.acid.bubblegum "yeah, use native image": 💀
20:40:05FromDiscord<Robyn [She/Her]> In reply to @griffith1deadly "💀": better than nothing :p
20:41:32FromDiscord<griffith1deadly> i'm already make some class parser that fully parses class but with only method calls (can only detect some calls from opcodes, dont have more time for writing full opcode parser) https://media.discordapp.net/attachments/371759389889003532/1365065143100248185/2C4B4E58-1A31-4340-A69C-6A7FDF926DCF.png?ex=680bf37b&is=680aa1fb&hm=4e23bb416e201bbcdef4d9c3ebc2310068e7be4b9a2796299ae25139334b9ddd&
20:42:12FromDiscord<griffith1deadly> and output to html https://media.discordapp.net/attachments/371759389889003532/1365065309391687761/D5DDF84B-4CDA-4081-9874-F75BD99CFD59.png?ex=680bf3a3&is=680aa223&hm=7d2555f8b800080700d796df33ae5b37095565f3276a2ae7c83c41b4420d26ac&
20:42:18FromDiscord<Robyn [She/Her]> Oooh fair
20:42:41*xet7 quit (Remote host closed the connection)
20:42:56FromDiscord<Robyn [She/Her]> You may be interested in this tho? https://github.com/The-Ticking-Clockwork/JvTypeGen
20:43:03FromDiscord<Robyn [She/Her]> uses reflection on the Java side
20:43:21FromDiscord<Robyn [She/Her]> And outputs JSON files
20:43:54*xet7 joined #nim
20:45:04FromDiscord<griffith1deadly> why use class -> json when can use class -> nim with macros 🙃
20:45:28FromDiscord<Robyn [She/Her]> Fair enough
20:45:37FromDiscord<griffith1deadly> something like generating jni bindings like futhark for modding
20:47:26FromDiscord<Robyn [She/Her]> In reply to @griffith1deadly "something like generating jni": That's fair but in that case, you're in an environment where reflection is available anyway
21:20:20*CypherCat joined #nim
21:27:00*ntat quit (Quit: leaving)
21:39:58*andy-turner quit (Quit: Leaving)
22:01:28FromDiscord<nocturn9x> In reply to @threefour "😬": yeahh
22:01:38FromDiscord<nocturn9x> nbd I guess
22:27:32FromDiscord<dawidek.2137> In reply to @dawidek.2137 "does someone know how": 😇
22:29:23FromDiscord<.bobbbob> In reply to @dawidek.2137 "😇": somebody correct me if I'm wrong, the default allocator holds on to allocated memory for reuse later, so your program will use a bit more memory than you expected, whereas using malloc frees the memory when it's done which uses less ram but is potentially slower since you have to allocate more memory again if you need it. and also I think the nim allocator is more likely to have bugs?
22:38:32FromDiscord<dawidek.2137> In reply to @.bobbbob "somebody correct me if": Thanks, I recall reading that too. I am having serious slowdowns when allocating/deallocating when compiled to WASM which seems to require malloc, so it makes sense
22:39:07FromDiscord<dawidek.2137> using emscripten of course
22:41:47FromDiscord<dawidek.2137> btw lots of small allocations are the thing people complain about wrt RAII, funny if solving that problem is as simple as this
23:40:02*skippy8 joined #nim