<< 21-01-2023 >>
00:09:09*argonica joined #nim
00:38:22*NimBot joined #nim
01:12:50*argonica quit (Quit: Leaving)
01:14:07*argonica joined #nim
01:43:24*NimBot joined #nim
01:50:02*NimBot joined #nim
02:50:14FromDiscord<exelotl> In reply to @4zv4l "well I don't know": Give this tutorial a read, you should find it quite familiar https://castle-engine.io/modern_pascal
02:51:06FromDiscord<exelotl> it's like Nim but a lot wordier xD
03:21:35*sagax joined #nim
03:23:57*ltriant quit (Ping timeout: 268 seconds)
04:13:31*arkurious quit (Quit: Leaving)
04:29:05FromDiscord<Boston> Pascal 😭
04:30:32FromDiscord<Boston> We have a test suite written in Delphi (most of the code was written in the late 90s) but because we sell to the military we have to maintain it, horrendous stuff
04:34:51*Jjp137 quit (Ping timeout: 248 seconds)
04:35:10*Jjp137 joined #nim
04:52:30*ltriant joined #nim
06:14:53*azimut_ quit (Ping timeout: 255 seconds)
06:17:09*tiorock joined #nim
06:17:09*tiorock quit (Changing host)
06:17:09*tiorock joined #nim
06:17:09*rockcavera quit (Killed (platinum.libera.chat (Nickname regained by services)))
06:17:09*tiorock is now known as rockcavera
06:35:40*argonica quit (Quit: Leaving)
06:44:30FromDiscord<qb> In reply to @enthus1ast "<@590220785029873674>\: a simple full": Well yea. All this hooking c libraries take pointers as parameters to install a hook. The one I'm using takes the function addresses. I've tried to use `parseHex(repr(someFunction), tointvar)` without success
06:52:14*ehmry quit (Quit: https://quassel-irc.org - Chat comfortably. Anywhere.)
06:53:25*ehmry joined #nim
07:08:55rockcaveraqb are you trying to capture the address of a nim proc?
07:13:35rockcaveraqb https://forum.nim-lang.org/t/1560
07:13:45rockcaverahttps://forum.nim-lang.org/t/1560#9667
07:32:56*argonica joined #nim
07:52:31FromDiscord<4zv4l> In reply to @exelotl "Give this tutorial a": I Will Check that !
08:04:30*ltriant quit (Ping timeout: 260 seconds)
09:09:12*argonica quit (Quit: Leaving)
09:09:50FromDiscord<michaelb.eth> sent a long message, see https://paste.rs/bwP
09:10:37FromDiscord<Elegantbeef> Import where you can, include where you cannot
09:11:20FromDiscord<Elegantbeef> The way i look at cross platform libraries is that the OS implementations should be modular, in that they should have the same API then i can just import them and use them how i want, but this depends on the library and the code you're using
09:11:49FromDiscord<michaelb.eth> (edit) "http://ix.io/4lON" => "http://ix.io/4lOM"
09:12:14FromDiscord<Elegantbeef> Mixins are for generics, so not really work for your purpose, as they're for moving the solution onto the user
09:12:15FromDiscord<Phil> In reply to @michaelb.eth "`include` statements in Nim": I'm struggling to imagine what part of dealing with OS code would make you have to deal with include
09:12:43FromDiscord<Elegantbeef> There isnt much of a reason it does
09:13:10FromDiscord<Phil> sent a code paste, see https://paste.rs/kMR
09:13:11FromDiscord<Elegantbeef> https://github.com/treeform/hottie/blob/master/src/hottie.nim is a prime example of doing imports 'properly'
09:14:10FromDiscord<Elegantbeef> Treat modules as an interface, could even do some CT magic to ensure a module implements what you want
09:15:02*pro joined #nim
09:15:15FromDiscord<michaelb.eth> The C lib author has done some very clever stuff that affects what `string` you want to pass to `{. ..., importc: [computed-string], ... .}`
09:16:14*pro left #nim (#nim)
09:16:59FromDiscord<Elegantbeef> Oh phil it was you that wanted that whole "ensure module implements X" proc right?
09:17:35FromDiscord<michaelb.eth> so beyond `include`, and nothing I could figure out with `import` and `mixin` (not that `mixin` is actually an option in this case, but I was experimenting), my option is to reproduce many lines of hand-written wrapper code and the include instructions in some context or another: "make sure you make changes in both `.nim` wrapper modules"
09:18:08FromDiscord<michaelb.eth> (edit) removed "the"
09:18:25FromDiscord<michaelb.eth> (edit) ""make" => ""make" | "modules"" => "modules""
09:19:43FromDiscord<Phil> In reply to @Elegantbeef "Oh phil it was": Yeah, but ideally in a manner that means I don't have to change my source code, only like import statements.↵In an ideal world, that'd be↵`import <someModule> implementing <Moduleinterface>` I think.↵I know ideally the entire "what interface a thing implements" would be within the module itself but that seems like even more of a pain
09:19:49FromDiscord<Elegantbeef> sent a code paste, see https://paste.rs/m3T
09:20:41FromDiscord<Phil> In reply to @michaelb.eth "so beyond `include`, and": I am also slightly confused as to the mention of mixins, as the target usecase for the mixin keywords is generics, which don't necessarily have anything to do with platform specific code
09:20:42FromDiscord<michaelb.eth> because the string infects a pragma that needs to be replicated throughout the wrapper module
09:21:07FromDiscord<Elegantbeef> I dont follow
09:21:35FromDiscord<Elegantbeef> macro time 😛↵(@Phil)
09:21:41FromDiscord<michaelb.eth> In reply to @Isofruit "I am also slightly": yeah, `mixin` was a stretch, I was just trying to find a way to have the dependent module inject a symbol into the dependency
09:21:49FromDiscord<Elegantbeef> Do you have a more concrete example of the issue?
09:22:28FromDiscord<Phil> Me?
09:22:36FromDiscord<Elegantbeef> No michael
09:29:53FromDiscord<Try2Cry> How come that @Elegantbeef#0000 is just a bot !
09:30:04FromDiscord<Elegantbeef> The matrix is a hell of a protocol
09:32:51FromDiscord<Phil> I programmed him
09:32:53FromDiscord<Phil> He is cheeky
09:32:55FromDiscord<Rika> Beef is from the matrix I see
09:32:57FromDiscord<Phil> Keeps denying it
09:33:34FromDiscord<Elegantbeef> I have as many cheeks as you
09:33:58FromDiscord<Try2Cry> Can i see his source code ?
09:34:03FromDiscord<Rika> In reply to @Elegantbeef "I have as many": Two up there and two down there?
09:34:06FromDiscord<Try2Cry> I really like it 🙂
09:34:12FromDiscord<Phil> In reply to @Try2Cry "How come that @Elegantbeef#0000": All kidding aside, this discord server is bridged to IRC and matrix (another way of chatting that is more privacy focussed)
09:34:14FromDiscord<Elegantbeef> How come it's okay to say i'm cheeky but when i say you're assey it's wrong?!
09:34:26Amun-Ra:P
09:34:43FromDiscord<Elegantbeef> Damn someone really likes me, my mom is going to be livid when i tell here I have a friend, she said no one likes me!
09:35:00FromDiscord<Phil> So what happens basically is that folks on IRC and matrix write stuff and their messages get forwarded to the discord server and re-posted here via bots/a bot
09:35:49FromDiscord<Phil> And at the same time our messages here from discord get forwarded back to the IRC/matrix servers
09:35:50FromDiscord<Try2Cry> So he just person using another software for privacy
09:35:59FromDiscord<Try2Cry> Hmm am the bot here then 😂
09:36:04FromDiscord<Phil> Not just another frontend, another frontend + another server
09:36:12FromDiscord<Elegantbeef> Privacy is a farce, given i'm talking to discord
09:36:22FromDiscord<Elegantbeef> They eat my data and then tell me "No you get nothing"
09:36:29FromDiscord<Elegantbeef> Even though i had a golden ticket
09:36:33FromDiscord<leetnewb> In reply to @Elegantbeef "They eat my data": You get to chat!
09:36:37FromDiscord<Phil> I'm playing around with webcord at the moment, its nice
09:36:56FromDiscord<Phil> Basically discord but you get to mess around with a few options that otherwise aren't exposed
09:37:14FromDiscord<Phil> In exchange you get the frontend crashing after around an hour of voice chat =/
09:37:22FromDiscord<Phil> And only then, fascinatingly
09:38:00FromDiscord<Try2Cry> Hmm nice
09:38:06FromDiscord<michaelb.eth> sent a long message, see http://ix.io/4lOU
09:38:07FromDiscord<Elegantbeef> I dont want to chat, i want to own things!
09:39:01*argonica joined #nim
09:39:09FromDiscord<Try2Cry> You won't own things without chatting
09:39:43FromDiscord<Elegantbeef> Have fun modifying discords font, or theme without breaking TOS! 😛
09:39:48FromDiscord<Phil> In reply to @michaelb.eth "Alright, I'll give a": I mean, that's basically what I've been doing with nimstoryfont the first 6 months
09:39:58FromDiscord<Try2Cry> Btw Since nim have a js backend why there isn't any repos that nim is being used with react , react native ?
09:40:00FromDiscord<Elegantbeef> Michael why can you not just move that conditional logic insied `abi/private`?
09:40:14FromDiscord<Elegantbeef> Cause karax exists
09:40:21FromDiscord<Phil> In reply to @Try2Cry "Btw Since nim have": For the most part because folks just use karax
09:40:23FromDiscord<michaelb.eth> karax?
09:40:33FromDiscord<Phil> Karax is nim's own SPA framework
09:40:38FromDiscord<Elegantbeef> Karax is a Single page application for web dev
09:40:43FromDiscord<Phil> the forum is written in karax IIRC
09:40:46FromDiscord<Elegantbeef> Almost wrote SPA properly
09:41:02FromDiscord<Elegantbeef> Sorry you fell trap to my dual conversation habit
09:41:10FromDiscord<michaelb.eth> eh, I think I got crosswise in the chat, nim-notcurses has nothing at all to do with jS
09:41:18FromDiscord<michaelb.eth> (edit) "jS" => "JS"
09:41:25FromDiscord<Elegantbeef> Yep of course
09:41:30FromDiscord<Try2Cry> Don't you think if that got some modified that will give a huge push to nim 🤔
09:41:46FromDiscord<Phil> Define "some modified"
09:41:52FromDiscord<Phil> Also define "that"
09:42:02FromDiscord<Elegantbeef> Cause people want to use a bastardised tedious alternative version of react
09:42:08FromDiscord<Elegantbeef> Instead of a new program
09:42:33FromDiscord<Elegantbeef> Anyway like i said michael i dont see why that conditional logic is inside of `abi.nim`
09:43:24FromDiscord<michaelb.eth> because the string passed to importc is prolific through derivative pragma-statements
09:43:45FromDiscord<Elegantbeef> Yes, but you declare the string inside abi for no reason afaict
09:43:58FromDiscord<Elegantbeef> Why can it not recide in `private.nim`?
09:44:03FromDiscord<Elegantbeef> the fuck is that spelling
09:44:09FromDiscord<Elegantbeef> reside\
09:45:18FromDiscord<michaelb.eth> `import notcurses` results in completely different compile/link-time behavior from `import notcurses/core`
09:46:06FromDiscord<michaelb.eth> and there are variants, help me to explain it, so that I can be rid of my.... misunderstandings or blindnesses
09:46:14Amun-RaI wrote my own [n]curses module for my image viewer, I use dlopen/dlsym for everything
09:46:30FromDiscord<michaelb.eth> unrelated to n/curses
09:46:35FromDiscord<Elegantbeef> You also could use a delayed import
09:46:50FromDiscord<michaelb.eth> In reply to @Elegantbeef "You also could use": do tell
09:46:53FromDiscord<Elegantbeef> I might be tired and missing the issue completely
09:48:23FromDiscord<Elegantbeef> sent a code paste, see https://play.nim-lang.org/#ix=4lOV
09:48:49FromDiscord<Elegantbeef> But i still dont quite see why you cannot just move the conditional logic from `abi.nim` to `private.nim`
09:49:10FromDiscord<Elegantbeef> As far as i can see the only usage is inside `private.nim`
09:49:41FromDiscord<michaelb.eth> `import notcurses` vs. `import notcurses/core`
09:49:45FromDiscord<michaelb.eth> or...
09:50:00FromDiscord<michaelb.eth> `import notcurses/cli` vs. `import notcurses/core/cli`
09:50:48FromDiscord<michaelb.eth> the C lib author is exceptionally clever; how you import the lib largely affects what gets linked
09:51:05FromDiscord<Elegantbeef> Ok so `abi` vs `cli` and `core` are different entry methods
09:51:10FromDiscord<michaelb.eth> yes
09:52:16FromDiscord<michaelb.eth> note: ignore what's on the `master` branch entirely, the only action is on the `version_3_revamp` branch
09:54:18FromDiscord<Elegantbeef> Some of your includes could just be a `import x` followed by `export x`
09:54:28FromDiscord<Elegantbeef> `include ./api/private/constants` for instance
09:54:41FromDiscord<Elegantbeef> Oh wait never mind
09:54:45FromDiscord<Rika> Oh shit someone wrapping not cursed
09:54:46FromDiscord<Elegantbeef> tired brain strikes again
09:54:49FromDiscord<Rika> (edit) "cursed" => "curses"
09:54:51FromDiscord<michaelb.eth> Even then, there's a big mess on `version_3_revamp`, but it's all artifact of an unexperienced Nim programmer inspired by Nick Black and rushing to make a wrapper. Since my start of it, I learned a lot and am cleaning it up, but figuring out the best way to deal with the `include` vs `import` + `export` stuff seemed like the best thing to tackl.e
09:54:57FromDiscord<michaelb.eth> (edit) "tackl.e" => "tackle now."
09:57:44FromDiscord<michaelb.eth> In reply to @Rika "Oh shit someone wrapping": My personal goals: successfully wrap it (finally), then provide a UI lib inspired by and rivaling https://github.com/gansm/finalcut (but using the wrapper and Nim concepts or whatever), then conquer the world
09:57:57FromDiscord<Elegantbeef> Yea seems include is the only solution here since you're using the module as a template practically
09:58:19FromDiscord<Elegantbeef> Ideally nim would treat modules as instantiateable and allow replacing symbols
09:59:47FromDiscord<Elegantbeef> There is not even a hacky i can think of to change logic based off importing a file aside from `{.define: notCursesUsingCLI.}`
10:00:22FromDiscord<michaelb.eth> alright, `include` it is
10:00:37FromDiscord<Elegantbeef> Not that anyone overly uses the define pragma, but it exists 😄
10:13:57FromDiscord<michaelb.eth> In reply to @Isofruit "I mean, that's basically": 🙏
10:28:14FromDiscord<4zv4l> sent a code paste, see https://play.nim-lang.org/#ix=4lP4
10:30:11*jjido joined #nim
10:35:04FromDiscord<Phil> In reply to @4zv4l "is there a better": You want to execute execShellCmd either way, right?
10:35:38FromDiscord<Phil> Wait, why is that a template not a proc? What benefit do you gain from it being a template?
10:36:23FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lP6
10:38:04FromDiscord<4zv4l> well template is code copy pasted in the place you call it right ?
10:38:40FromDiscord<Phil> Correct, but in the end all it achieves is a call to `execShellCmd` with a string, that same side-effect can be achieved by just a simple proc
10:41:22FromDiscord<4zv4l> sent a code paste, see https://play.nim-lang.org/#ix=4lP8
10:41:33FromDiscord<4zv4l> In reply to @Isofruit "Correct, but in the": I'll change it to be a proc then if it doesn't change much
10:41:59FromDiscord<Phil> Imo procs are always preferrable over templates unless you do with a template stuff that a proc can't do, like context managers
10:43:20FromDiscord<4zv4l> alright, I'll keep that in mind, thanks ! ^^
10:44:08FromDiscord<4zv4l> sent a code paste, see https://play.nim-lang.org/#ix=4lP9
10:44:31FromDiscord<enthus1ast> @4zv4l\: https://nim-lang.org/docs/terminal.html#eraseScreen%2CFile
10:46:35FromDiscord<4zv4l> In reply to @enthus1ast "<@329196212282458112>\: https://nim-lang.org/docs/t": yeah I already checked this one but it never works as expected
10:47:06FromDiscord<4zv4l> it makes the time (what is printed) go down and down
10:47:34FromDiscord<4zv4l> like if putting the cursor at the right pos didn't work for this proc
10:48:17FromDiscord<enthus1ast> do you want to overwrite a line?
10:48:27FromDiscord<enthus1ast> for a progress indicator?
10:48:47FromDiscord<4zv4l> oh it's multiple lines
10:49:00FromDiscord<4zv4l> basically like this https://media.discordapp.net/attachments/371759389889003532/1066308426541584405/6c8d8207-6d4c-41fb-b227-b07d14ff9a42.png
10:49:09FromDiscord<4zv4l> supposed to clear and update the time
10:49:33FromDiscord<enthus1ast> i see+
10:50:19FromDiscord<4zv4l> In reply to @4zv4l "basically like this": I tried↵`cursorUp` to get rid of the final newline↵but it didn't work either xD
10:51:22FromDiscord<enthus1ast> i looked at old code of mine how i did it back then
10:51:27FromDiscord<enthus1ast> but it used illwill
10:51:29FromDiscord<enthus1ast> sent a code paste, see https://play.nim-lang.org/#ix=
10:51:52FromDiscord<enthus1ast> maybe illwill is an option for you as well
10:52:12FromDiscord<4zv4l> what is illwill ?
10:52:14FromDiscord<enthus1ast> https://github.com/enthus1ast/muk/blob/bb65fa5d44adc62944d3240d83eefaa510b5e5de/src/muk.nim#L583
10:52:16FromDiscord<4zv4l> a library ?
10:52:33FromDiscord<enthus1ast> its a little like curses
10:52:46FromDiscord<enthus1ast> https://github.com/johnnovak/illwill
10:53:14FromDiscord<enthus1ast> the code i linked above, tries to detect a terminal resize, and reprints the whole tui
10:53:23FromDiscord<4zv4l> oh yeah right↵I didn't plan to do something with curses↵I mean sure it would be like `tty-clock` but I wanted to do something simple xD
10:54:09FromDiscord<enthus1ast> yeah illwill is quite simple
10:54:14FromDiscord<4zv4l> In reply to @enthus1ast "https://github.com/johnnovak/illwill": actually I used it for another project, really nice lib
10:55:30FromDiscord<4zv4l> sent a code paste, see https://play.nim-lang.org/#ix=4lPc
10:55:39FromDiscord<4zv4l> because even if I call `defer: deinitIwill()`
10:55:51FromDiscord<4zv4l> if the guy hit `ctrl-c` or a kill signal or whatever
10:55:56FromDiscord<4zv4l> the cursor will be hidden
10:56:20FromDiscord<enthus1ast> yes i think you should put this in the ctrl-c handler
10:56:53FromDiscord<enthus1ast> https://nim-lang.org/docs/system.html#setControlCHook%2Cproc%29
10:59:16FromDiscord<enthus1ast> i'm not a terminal expert, but it could be that some terminal (in some modes) clear the screen by scrolling up (or emitting newlines)
10:59:34FromDiscord<enthus1ast> noticed this myself aswell
11:04:18FromDiscord<Phil> Its kinda annoying I currently can't build my project because ndb isn't useable with nim 2.0 =/
11:04:47FromDiscord<Phil> And I can't figure out how to "go back" to a nim 1.7 version
11:04:53FromDiscord<Phil> (edit) "And I can't figure out how to "go back" to a nim 1.7 version ... " added "before the db_connector split"
11:05:22FromDiscord<enthus1ast> can you choosenim the correct version?
11:06:11FromDiscord<Phil> Nah because they're both devel versions and I can't go `choosenim 1.7.3`
11:06:39FromDiscord<enthus1ast> mh then check out the correct version and build it yourself
11:06:45FromDiscord<Phil> Can't use 1.6.10 because my project needs various fixes to the std lib that are only on devel.↵Can't use the most recent devel because 1.9 broke 2 of my dependencies
11:08:45FromDiscord<Phil> That's more knowledge limited on my end than anything else. ↵Aka the good old "how" question.↵I guess I can take a look at the compiler docs and hope I understand anything
11:09:39FromDiscord<enthus1ast> afaik you can use koch boot
11:12:36FromDiscord<enthus1ast> https://nim-lang.org/docs/koch.html#commands-boot-command
11:12:48FromDiscord<Phil> I mean, it's basically ↵step 1) Figure out the commit hash of some commit that was valid before 1.7↵step 2) Figure out how to use koch to build the compiler from that hash
11:12:55FromDiscord<Phil> Sadly devel versions aren't tagged
11:13:30FromDiscord<enthus1ast> yeah annoying, one reason i stay on stable or even old stable....
11:14:11*jjido quit (Quit: My laptop has gone to sleep. ZZZzzz…)
11:14:27FromDiscord<Phil> but then stable is missing so many important fixes
11:14:47FromDiscord<Phil> Its just uuuuggh until we get a decent nim 2.0 release and the libraries catch up
11:14:57FromDiscord<enthus1ast> maybe you can fix the issues in your dependencies?
11:16:32FromDiscord<Phil> I already have, but I also test my shit in CI/CD pipelines so it doesn't just work locally.↵And the PR for the fix on ndb is open since yesterday, after that it should solely be a matter of norm updating its ndb dependency and I'm good
11:17:17FromDiscord<enthus1ast> you could do a branch where you temporarly change the nimble dependencies on your fork
11:17:22FromDiscord<Phil> I guess I'm just somewhat frustrated because I finally have some spare time on my hands and my hands are basically tied behind my back unless I start dealing with building the compiler myself for an older devel version
11:18:08FromDiscord<Phil> Can you actually have nim requires clauses that point to a specific github repo instead of a published nimble package?
11:18:18FromDiscord<enthus1ast> yes
11:18:24FromDiscord<enthus1ast> you can put the whole url
11:18:49FromDiscord<enthus1ast> AAAND you can even pinpoint to a branch, but this i always must google \:)
11:19:07FromDiscord<enthus1ast> `url@#yourbranch` or similar
11:20:02FromDiscord<enthus1ast> if i recall correctly the syntax is a little different than on the commandline
11:20:45FromDiscord<Phil> Okay, so ↵step 1, make an ndb fork with an update custom ndb dependency... well I have that already, it's what I use for the current PR↵step 2, make a norm fork and update it to use the ndb fork↵step 3, use the norm fork instead of the real norm package
11:21:00FromDiscord<enthus1ast> yes
11:24:14*ltriant joined #nim
11:45:10FromDiscord<Try2Cry> In reply to @Isofruit "Define "some modified"": I meant some modification ↵And that was referring for the modification '''dont you think if nim got these modification that will end up with good impact on nim matket'''
11:46:13FromDiscord<enthus1ast> this discussion again? \:D
11:46:30FromDiscord<enthus1ast> do x then nim would be better recognized
11:46:31FromDiscord<Phil> You mean some kind of lib that makes react/react-native useable in nim?
11:46:57FromDiscord<Try2Cry> Yes
11:46:58FromDiscord<enthus1ast> last time it was gui
11:47:19FromDiscord<enthus1ast> the thing is someone must do it
11:47:20FromDiscord<Try2Cry> What last time ?🤔
11:47:23FromDiscord<enthus1ast> must do it properly
11:47:38FromDiscord<enthus1ast> we had this discussion various times already↵(@Try2Cry)
11:47:43FromDiscord<Phil> It might, but it also might not. Either way I personally am not interested in writing a JS framework so 🤷
11:48:18FromDiscord<Try2Cry> I mean while we can already compile it down to js so why do we even need a lib ?
11:48:30FromDiscord<Phil> It's the good old matter of "who writes it?".↵Because I'm writing for fun and myself, such a lib doesn't seem fun to me
11:48:55FromDiscord<Try2Cry> In reply to @enthus1ast "we had this discussion": Was i exist in that conversation 😅
11:48:56FromDiscord<Phil> You'd need to ask that people that actually use the JS backend, I mostly use nim for the backend and Angular for the frontend
11:49:08FromDiscord<Phil> In reply to @Try2Cry "Was i exist in": Nah, was a different convo with a different person
11:49:16FromDiscord<Try2Cry> Why not to make a topics so if anyone have the same question so he can read it
11:49:37FromDiscord<enthus1ast> this idk, but the thing is, someone must write it (and do it properly) and someone must support it for various years to come↵(@Try2Cry)
11:49:47FromDiscord<Phil> But the topic was the same, "Why don't you have bindings for this lib/these frameworks".↵Well, because nobody made them yet, nobody is getting paid to make them and the rest doesn't find it fun to make them.
11:49:48FromDiscord<enthus1ast> and people have other interests as far as i can see
11:50:29FromDiscord<enthus1ast> and i can also imagine that this is a fast moving target
11:50:43FromDiscord<enthus1ast> when even js libs cannot keep up with other js libs
11:50:47FromDiscord<Phil> any JS framework would be a fast moving target
11:50:57FromDiscord<enthus1ast> yes
11:51:21FromDiscord<Try2Cry> In reply to @Isofruit "But the topic was": Is there a tutorial in how to make such a binding ?
11:51:41FromDiscord<Phil> My workplace is investing hundreds of thousands into developers each month for a webpage and even they only barely manage to get somewhat up to date on the angular half-year-per-version cycle
11:52:13FromDiscord<Phil> "investing" as in "paying salary to"
11:52:44FromDiscord<Phil> In reply to @Try2Cry "Is there a tutorial": You'd need to ask folks actually using the JS backend, @hotdog maybe since I recall him using karax (?)
11:52:46FromDiscord<Try2Cry> Hmmm yes it's a fast moving target but worth a try i can bit if it's done or even started there will be alot of interested ppl
11:53:30FromDiscord<Phil> I can tell you how to set up a backend webserver any day of the week, when it comes to JS backend or C bindings I'm out
11:54:05FromDiscord<Try2Cry> In reply to @Isofruit "My workplace is investing": Same goes here still in django 2.2 and react 1 year old i think ↵Ionic was 2 years old and we had problems in many projects
11:55:06FromDiscord<enthus1ast> and the other thing is, people want to write nim
11:55:08FromDiscord<Phil> `/home/philipp/.nimble/pkgs2/ndb-0.19.9-8227e7d747c90b21c282c4db64a3464aa356b31f/ndb/sqlite.nim(528, 29) Error: expression 'connection' has no type (or is ambiguous)`↵What the flying fish flipper
11:55:48FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lPo
11:57:56FromDiscord<Phil> God damnit, why does this have to be harder than just "change imports"
12:03:41*jkl quit (Quit: Gone.)
12:05:15*jkl joined #nim
12:07:30FromDiscord<Phil> What exactly is `RootEffect`?
12:08:50FromDiscord<enthus1ast> https://nim-lang.org/docs/system.html#RootEffect
12:11:29FromDiscord<Phil> Okay, wth, did the move to db_connector also fundamentally change the contents of db_common or sth?
12:13:05FromDiscord<keks> Is there an equivalent in Nim of doing assignment of multiple variables at once, like in Python? This is useful for a swap function: `x, y = y, x`.
12:13:46FromDiscord<enthus1ast> there is swap
12:13:46FromDiscord<enthus1ast> https://nim-lang.org/docs/system.html#swap%2CT%2CT
12:18:26FromDiscord<keks> I apologise - missed that in the search, thanks!
12:18:46FromDiscord<enthus1ast> let (a, b) = ("foo", "baa")
12:20:11Amun-Raor: (a, b) = (b, a)
12:24:45FromDiscord<keks> right, I didn't try the parentheses
12:30:12FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lPD
12:31:23FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lPE
12:32:05FromDiscord<Phil> This all works flawlessly under 1.6.10
12:39:34*sagax quit (Ping timeout: 268 seconds)
12:42:20FromDiscord<Phil> How do I show all symbols a module expports, I want to stare at this damn module and what it gives or not gives me
12:42:59FromDiscord<Phil> (edit) "expports," => "exports,"
12:43:01FromDiscord<enthus1ast> maybe nim doc
12:44:35FromDiscord<Phil> I'm about this close to just go with the plan of compiling a 1.7, never moving to nim 2.0 and just fuck ndb alltogether, what the hell is this module, how does it not have types that db_common clearly exports, grrrrr
12:44:45FromDiscord<Phil> (edit) "1.7," => "1.7 compiler,"
12:45:17FromDiscord<enthus1ast> was ndb not quite "abadonned" ?
12:45:31FromDiscord<Phil> It is, but norm uses it
12:45:50FromDiscord<enthus1ast> mh yeah
12:45:55FromDiscord<Phil> moigagoo has been contemplating a fork but god damn even then just trying to move this to 2.0 is a GIGANTIC pain in myass
12:46:57FromDiscord<ringabout> What's your Nim version and its compilation date?
12:47:22FromDiscord<Phil> jadghksfjldhglfjkdhgsfdüpskgfhägshkäfgf
12:47:35FromDiscord<Phil> It was because the damn dependency was missing in the effing nimble file
12:47:39FromDiscord<enthus1ast> cat on keyboard?
12:47:41FromDiscord<enthus1ast> \:)
12:48:03FromDiscord<Phil> Face roll of frustration over keyboard
12:48:19FromDiscord<Phil> a Frof if you will
12:50:25FromDiscord<Phil> Okay, I can compile the docs for the ndb package on 1.9.1 now, I take that as a sign it works there
12:50:37FromDiscord<Phil> Now to see if I can get norm to the point of compatibility
12:55:07FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lPL
12:55:17FromDiscord<Phil> (edit) "https://play.nim-lang.org/#ix=4lPL" => "https://play.nim-lang.org/#ix=4lPM"
12:55:59FromDiscord<enthus1ast> should also start to test my stuff on nim v2
12:56:02FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lPN
12:56:47FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lPO
12:57:10FromDiscord<enthus1ast> never used the effect system before so no clue
12:58:02FromDiscord<Phil> @ringabout Do you know anything about the effects system? The above feels like a bad thing, because it reminds of doing a try-catch with a base-exception class, effectively catching all Exceptions ever, which is generally an anti-pattern
13:05:42FromDiscord<ringabout> In reply to @Isofruit "<@658563905425244160> Do you know": I rarely use effect tracking system. Perhaps it is caused by stricteffects. What about using `--legacy:laxEffect`?
13:07:11FromDiscord<ringabout> I'm on the phone so I cannot look up the exact name of the legacy feature, something like laxEffect.
13:07:21FromDiscord<Phil> Check, I'll try one min
13:13:27FromDiscord<Phil> In reply to @ringabout "I rarely use effect": Didn't change the outcome.↵I ran:↵` exec "nimble doc --outdir:docs/apidocs --legacy:laxEffect --project --index:on src/norm"`
13:13:44*jjido joined #nim
13:13:51FromDiscord<Phil> Its currently what I use to test if the lib works at all or not
13:14:02FromDiscord<Phil> Once that works I figure out if the tests run
13:15:07FromDiscord<ShalokShalom> Do strictEffects still have to be activated in 2.0?
13:15:12FromDiscord<ShalokShalom> Or are they default then?
13:15:51FromDiscord<Phil> In reply to @ringabout "I rarely use effect": Wait! It was `legacy:laxEffects`, I don't think it's documented anywhere, I found via github search in a commit of yours:↵https://github.com/nim-lang/Nim/commit/090661d46cefdb62ff376fd2ce316c5a052752f3
13:15:56FromDiscord<Phil> That one works
13:17:11FromDiscord<Phil> Now I'm not sure what to do with that information.↵Should all users be required to compile with `--legacy:laxEffects` (How long will that option even exist? Is it planned to have long-term support?), or do I have to start talking with somebody that can tell me if adding `RootEffect` everywhere is the right way to go?
13:17:21FromDiscord<Phil> (edit) "Now I'm not sure what to do with that information.↵Should all users ... be" added "of norm"
13:17:57FromDiscord<ringabout> I don't know the detail. It is probably a bug of stricteffects.
13:19:54FromDiscord<ringabout> I mean laxEffects are used to identify whether the problem is caused by stricteffects. It is not a fix.
13:20:41*jjido quit (Quit: My laptop has gone to sleep. ZZZzzz…)
13:24:38FromDiscord<ringabout> Probably start from tagging https://github.com/xzfc/ndb.nim/blob/d942f882c63e9abebe9b2dd8fa5be597c46763b1/ndb/postgres.nim#LL371 ReadDbEffect
13:28:04*argonica quit (Ping timeout: 256 seconds)
13:29:17FromDiscord<Require Support> are queues considered cyclical ? 🤔 idk what cyclical means
13:36:02FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lPW
13:36:05FromDiscord<Phil> What is "Effect" here? Like SideEffect?
13:44:22FromDiscord<hotdog> In reply to @Isofruit "You'd need to ask": Hey 👋 @Try2Cry no tutorials that I know of, but if you have trouble please write on #webdev and I will try to help later on
13:49:20FromDiscord<Phil> Ohhhh so Effects are kinda like an arbitrary "Only procs from those specific groups are allowed in here, none from any of the other groups"
13:49:35FromDiscord<Phil> (edit) "Ohhhh so Effects ... are" added "in tags"
13:59:25FromDiscord<Phil> Alrighty, so `parseDate` introduced `TimeEffect`, which comes from the standard lib. All I had to do was correctly introduce that effect to all procs that make use of it
14:01:41FromDiscord<Phil> Does anyone actually use the effects system? If so, what's your usecase? What benefit does it provide you?
14:02:35FromDiscord<Phil> This actually seems like the kind of thing that could be useful for domain driven design I guess @ShalokShalom , since you can basically define a range of "acceptable procs" within your proc by defining which effects in a proc are allowed and which aren't.
14:02:56FromDiscord<Phil> (edit) "procs"" => "topics"" | "topics"within your proc ... by" added "whose procs are allowed to be used here"
14:03:13FromDiscord<Phil> (edit) "by defining which effects in a proc are allowed and which aren't." => "within said proc."
14:03:56FromDiscord<Rika> effects system is like a more granular version of func iirc
14:05:28FromDiscord<Phil> Isn't it mostly a mechanism of "excluding" unwanted effects?↵Like, they don't really have any component of actually running code.↵It's just "If your tag pragma defines you are allowed procs with effects X and Y, then the proc that has defined the effect Z is forbidden to be used in here"
14:05:58FromDiscord<Phil> And apparently you use objects for some reason to define a tag, which I don't fully get why that one is syntactically necessary
14:06:56FromDiscord<Rika> isnt func also technically meant for excluding side effects?
14:08:39FromDiscord<Phil> kiiiiiiiiiiiiind of?↵I dunno, in my head those side-effects feel more like "this does things changes state" which doesn't seem strictly necessary to have an effect ... it's just... wait so the effects system is just used to express that and basically is capable of way more than just that?
14:08:49FromDiscord<Phil> (edit) "kiiiiiiiiiiiiind of?↵I dunno, in my head those side-effects feel more like "this does things ... changes" added "that"
14:09:21FromDiscord<Phil> That seems like a powerful mechanism that I'm still not entirely sure what to do with ^^
14:11:31FromDiscord<Phil> Anyway, what you could basically do is add tags to procs to mark them as part of a specific "domain" and thusly control that no interference between domains for example is possible
14:11:34FromDiscord<Rika> i think nowadays its mainly abandoned though, the system
14:11:49FromDiscord<Rika> In reply to @Isofruit "Anyway, what you could": "control that no interference between domains"?
14:12:07FromDiscord<Rika> no, its meant like raises:[] but not exceptions
14:12:15FromDiscord<Rika> so basically a more granular func
14:12:17FromDiscord<Rika> iirc
14:12:59FromDiscord<Phil> Ah, so if you used it like that it would mostly be a bastardization of the system. It'd be using it, but not as intended
14:14:06FromDiscord<Rika> basically you define what effects can be in a proc and if its not there the compiler would complain, just like raises
14:14:09FromDiscord<Phil> In reply to @Rika ""control that no interference": Domains as in Domain Driven Design. Where you divide your problem into specific "domains" between which you basically have limited to absolutely no communication. And what little communication you have is just stuff like passing a message from one to the other, that's it
14:14:35FromDiscord<Phil> Which is basically OO but thought in larger units of code than individual modules/classes
14:16:12FromDiscord<Phil> In reply to @Rika "basically you define what": Which you can kinda use to say "You shouldn't use this proc, this proc shouldn't have additional effects like this"
14:16:34FromDiscord<Rika> what exactly an effect is is fuzzy to me
14:16:37FromDiscord<Phil> But eh, I'm semi iffy on DDD either way not like I'm going to actually use it like that.
14:16:56FromDiscord<Phil> In reply to @Rika "what exactly an effect": TBH given what I've seen, an effect is pretty arbitrary. Just define an object and you can use it as effect.
14:17:15FromDiscord<Rika> yes
14:17:16FromDiscord<Phil> So far I've seen "TimeEffect" and "DbReadEffect"
14:17:17FromDiscord<Rika> i know
14:17:22FromDiscord<Rika> but the intention is what i mean
14:17:29FromDiscord<Rika> what an effect is supposed to intend, i do not know
14:24:25FromDiscord<Dexterdy> anyone here with some gintro experience? How do I make a gio.ListStore using a datatype defined by me? I can't find sufficient documentation about this. In the rust gtk bindings that is a wrapper class that inherits from GObject that allows you to put your own data in a ListStore, but I can't find anything like that, and you can't seem to be able to inherit from GObject directly yourself in the gintro library. any ideas?
14:31:53*jjido joined #nim
14:43:59*argonica joined #nim
14:45:20FromDiscord<Phil> No ginto experience on my end, sorry
14:46:35FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lQb
14:47:07FromDiscord<Rika> `closure` likely
14:47:36FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lQc
14:48:31FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lQd
14:49:01FromDiscord<Phil> (edit) "https://play.nim-lang.org/#ix=4lQd" => "https://play.nim-lang.org/#ix=4lQe"
14:49:11FromDiscord<Phil> (edit) "https://play.nim-lang.org/#ix=4lQe" => "https://play.nim-lang.org/#ix=4lQf"
14:49:13FromDiscord<ringabout> Add `gcsafe`?
14:49:38FromDiscord<ringabout> sent a code paste, see https://play.nim-lang.org/#ix=4lQg
14:50:21FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lQh
14:50:32FromDiscord<Phil> The bigger question is: Does `{.cast(gcsafe).}` still work?
14:50:33FromDiscord<Rika> just ctrl-f SQLITE_POOL
14:50:39FromDiscord<Rika> sure it does
14:52:14FromDiscord<ringabout> sent a code paste, see https://play.nim-lang.org/#ix=4lQi
14:57:01FromDiscord<ShalokShalom> sent a long message, see https://paste.rs/L3L
14:58:19FromDiscord<hdbg> How can I write term rewriting macro, which targets every line? Doing some internal things, so had to turn off stack tracing to avoid crashes. Now I wanna insert echo after every line for debugging purposes
15:06:17FromDiscord<Phil> In reply to @ShalokShalom "The ddd I mean,": More security maybe, but I did not really think it through that much
15:06:31FromDiscord<Phil> Also, yehaw, I can compile again!
15:15:36*azimut joined #nim
15:26:08*krux02 joined #nim
15:29:37*arkurious joined #nim
15:34:50FromDiscord<Phil> I'm running into `could not import: EVP_MD_size`, not sure what that is about, is this an openssl 3 issue?
15:37:05FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lQu
15:38:46FromDiscord<Phil> Ugh, it got renamed with openssl 3 (https://github.com/dotnet/runtime/issues/67375#issuecomment-1084741700): ↵`EVP_MD_get_size`
16:01:17*jjido quit (Quit: My laptop has gone to sleep. ZZZzzz…)
16:06:16FromDiscord<Require Support> why when I add an object with a field of type `Deque` from `std/deques` I get `'myobect' not a concrete type` error
16:10:16FromDiscord<ShalokShalom> In reply to @Isofruit "More security maybe, but": Still recommend you that book 📖
16:10:27FromDiscord<ShalokShalom> In reply to @Isofruit "Also, yehaw, I can": 🥳
16:26:15FromDiscord<Require Support> how come with `--stackTrace:off --stackTraceMsgs:off` message boxes still popup when errors happen 🤔
16:26:52FromDiscord<Require Support> (edit) "how come with `--stackTrace:off --stackTraceMsgs:off` message boxes still popup when errors happen 🤔 ... " added "(in windows)"
16:27:08FromDiscord<Require Support> (edit) "how come ... with`-d:danger" added "building" | "`--stackTrace:off" => "`-d:danger --stackTrace:off"
16:51:30*jjido joined #nim
17:11:52*termer quit (Quit: ZNC 1.7.5+deb4 - https://znc.in)
17:13:04*termer joined #nim
17:35:50*termer quit (Quit: ZNC 1.7.5+deb4 - https://znc.in)
17:39:32*termer joined #nim
18:10:18FromDiscord<ShalokShalom> @Phil didn't you look for a debugger lately?↵↵https://github.com/epasveer/seer
18:18:12*nmz left #nim (Ploop)
19:08:40*Guest38 joined #nim
19:09:50*Guest38 quit (Client Quit)
19:13:16*krux02 quit (Remote host closed the connection)
19:25:20FromDiscord<Leftbones> FINALLY figured out how to get non-blocking input including special characters on both Windows and Unix
19:25:37FromDiscord<Leftbones> By reading and learning how Illwill and Snip do it, basically
19:26:16FromDiscord<Leftbones> Is it normal that `stdout.write` seems to act slightly different on different platforms?
19:27:11FromDiscord<michaelb.eth> Are you sure it’s not the terminal behavior that is different?
19:29:31FromDiscord<michaelb.eth> In reply to @Leftbones "FINALLY figured out how": As a fun test to check if it’s working correctly, try copy/pasting many KB of emoji mixed with other text
19:32:14FromDiscord<Leftbones> In reply to @michaelb.eth "As a fun test": Oh god I hadn't even considered unicode and emojis yet
19:32:44FromDiscord<Leftbones> In reply to @michaelb.eth "Are you sure it’s": I suppose it could be, now that you mention it. I've been testing using Windows Terminal on my PC, but using vscode's terminal on Mac. Though I do get the same behaviour in iTerm2 on mac.
19:34:28FromDiscord<michaelb.eth> Maybe also try in Kitty and Alacritty on macOS
19:35:17FromDiscord<Leftbones> sent a code paste, see https://play.nim-lang.org/#ix=4lRn
19:36:11FromDiscord<Leftbones> Full code is here https://github.com/leftbones/nancy
19:38:28FromDiscord<Leftbones> I haven't really looked too far into it yet but I don't yet understand why `stdout.write` and `echo` would work differently depending platform, I suppose it could have to do with the generated C code?
19:38:37FromDiscord<Leftbones> In reply to @michaelb.eth "Maybe also try in": I'll try this too
19:41:46FromDiscord<Ntsékees> Hello, I'm new to Nim.↵Which way would you recommend me to convert a HashSet into a Sequence?
19:42:00FromDiscord<Ntsékees> (which implies imparting an arbitrary ordering)
19:42:25FromDiscord<Ntsékees> (the ordering doesn't matter to me)
19:44:28rockcaveraNtsékees sequtils
19:44:35FromDiscord<Phil> Generally, import sequtils and use `toSeq`
19:45:10FromDiscord<Leftbones> Gotta say after a week in Nim, I'm loving all the various 'utils' packages (libraries? modules?) available
19:45:22FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lRr
19:45:30FromDiscord<Ntsékees> Ah, I had found about `@` but it didn't support the Set type. I'll check out `toSeq`.
19:45:55FromDiscord<Phil> In reply to @Ntsékees "Ah, I had found": @ generally is for instantiating sequences, so for `let x: seq[int] = @[]`
19:47:14FromDiscord<Phil> If you want a shorthand for instantiating sets though, I honestly don't really know. I know `{}` works to instantiate a set based on nim-magic (so not based on a type from std/sets), but I'm not aware if that also works for Hashsets, never tried
19:47:49FromDiscord<Phil> In reply to @Ntsékees "Ah, I had found": As an aside, since I don't know if you're already aware of this:↵For quick prototyping a couple loc's or the like nim has a "repl-like"-piece of software called inim
19:48:36FromDiscord<huantian> if you want a "hashset literal", use `[1, 3, 4, 6].toHashSet()`
19:48:42FromDiscord<huantian> `{}` constructs a `set`
19:48:53FromDiscord<Phil> Helps with checking quickly if a particular syntax actually works or not. I also have a separate "playground" project for checking larger ideas and debugging larger code snippets, but for anything below 5 loc I tend to fire up inim real quick
19:49:41FromDiscord<Ntsékees> In reply to @Isofruit "As an aside, since": Ah, I didn't know about `inim`. I had web-searched “Nim REPL” but didn't find much (well, I found https://replit.com/languages/nim but I haven't tested it yet)
19:50:09FromDiscord<Phil> In reply to @Ntsékees "Ah, I didn't know": There's also "nim-secret", which is the VM that nim uses to evaluate macros and compile-time code etc.
19:50:34FromDiscord<Phil> It has like 80% of nim's capabilities, but various things like std/os and some other procs demanding interacting with the environment can't run in it
19:51:05FromDiscord<Phil> inim works around that by not actually being a repl, it secretly writes a file line by line and compiles it and prints the output.↵Not going to matter 99% of the time, just an interesting tidbit
19:52:40FromDiscord<Leftbones> In reply to @Isofruit "As an aside, since": Well this is just excellent. I was using repl.it for quick things, haha.
19:53:02FromDiscord<huantian> personally I just use `play.nim-lang.org` for prototyping lol
19:53:08FromDiscord<Phil> Happy to spam my intermediate knowledge here any day
19:53:16FromDiscord<Phil> play.nim-lang.org is really nice if you want to actually share your stuff
19:53:29FromDiscord<Phil> For troubleshooting in this chat
19:53:52FromDiscord<Phil> The tile it needs to output the compile results are too long for me to consider using it when I just want to check something for myself
19:53:57FromDiscord<Phil> (edit) "tile" => "time"
19:54:50*chmod222 joined #nim
19:55:15FromDiscord<Phil> In reply to @Leftbones "Gotta say after a": You mean the ones that basically give you new language constructs nim doesn't have by default?
19:55:31FromDiscord<Leftbones> Essentially
19:55:45FromDiscord<Leftbones> I like `strutils`
19:55:55FromDiscord<Phil> Yeh, that's pretty much one of nim's strengths.↵You can write stuff like custom context managers yourself pretty trivially
19:55:57FromDiscord<Leftbones> Like I said I've only been in Nim for a week so I'm still discovering a lot of stuff
19:56:04FromDiscord<Phil> Ohhh strutils I wouldn't say counts for that 😛
19:56:10FromDiscord<Ntsékees> sent a code paste, see https://play.nim-lang.org/#ix=4lRv
19:56:24FromDiscord<Ntsékees> > Error: 'toHashSet' doesn't have a concrete type, due to unspecified generic parameters.
19:56:27FromDiscord<Phil> In reply to @Ntsékees "Something I didn't manage": Generic functions don't actually exist
19:56:54FromDiscord<Phil> They are blueprints for the compiler to copy around when it gets an explicit type for them at the site where the generic blueprint is used
19:57:13FromDiscord<Phil> Its a very elaborate "Control + C" command for the compiler
19:57:58FromDiscord<Phil> ~~It's also the reason why mocking generic procs is a complicated undertaking that I can only imagine folks like Elegantbeef could manage should they get the passion to want to do so, I tried that and failed~~
19:58:48FromDiscord<Phil> So if you want to assign that proc, you need to "properly" instantiate it by providing nim with concrete types, basically nailing down that this proc is one for creating e.q. a hashet for strings or the like
19:59:13FromDiscord<huantian> you could also just deifne your own proc that calls toHashSet
19:59:31FromDiscord<huantian> ~~though please don't rename `toHashSet` to `toS` that'll just make your code more unclear~~
19:59:39FromDiscord<Phil> Yeah, that solves the problem... maybe I should orient myself more towards looking for ways to solve the problem than explaining why they exist ^^'
19:59:56FromDiscord<Ntsékees> sent a code paste, see https://paste.rs/Cb8
20:00:03FromDiscord<Ntsékees> but it's wordy
20:00:04FromDiscord<Ntsékees> (edit) "https://play.nim-lang.org/#ix=4lRz" => "https://play.nim-lang.org/#ix=4lRA"
20:00:05*jjido quit (Quit: My laptop has gone to sleep. ZZZzzz…)
20:00:17FromDiscord<Phil> Yeah, because toS is also a blueprint. It doesn't exist, it only exists once it gets used and copied into a "useable" form
20:00:18FromDiscord<Ntsékees> (edit) "https://play.nim-lang.org/#ix=4lRA" => "https://play.nim-lang.org/#ix=4lRB"
20:00:29FromDiscord<Phil> (edit) "Yeah, because toS is also a blueprint. It doesn't exist, it only exists once it gets used and copied into a "useable" form ... " added "with explicit types"
20:00:50FromDiscord<huantian> In reply to @Ntsékees "but it's wordy": It's mostly fine though, you won't be aliasing procs to other names often
20:01:07FromDiscord<Leftbones> In reply to @michaelb.eth "Maybe also try in": Tried now, it's definitely not the terminal, it's the platform
20:01:35FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lRC
20:01:40FromDiscord<Leftbones> It's probably... fine? There doesn't seem to be a difference between `stdout.write` and `echo` at least in the way I'm using them, so I guess it's fine
20:02:52FromDiscord<Phil> sent a code paste, see https://play.nim-lang.org/#ix=4lRD
20:03:21FromDiscord<Phil> Some symbols are mostly taken by nim convention for the most part, $ is universally for "toString", % is universally for "toJson" for example
20:05:45FromDiscord<Ntsékees> Thanks for the help.
20:06:15FromDiscord<Phil> Oh and @ is for instantiating seq
20:06:25FromDiscord<Phil> % likely is modulo... hmmm whta's free
20:06:32FromDiscord<Phil> (edit) "whta's" => "what's"
20:06:52FromDiscord<huantian> In reply to @Isofruit "Oh and @ is": technically @ converts arrays to seqs
20:07:29FromDiscord<Phil> Oh hey, `?` is free
20:07:36FromDiscord<Phil> And of course 2-symbol combinations are as well
20:07:39FromDiscord<michaelb.eth> In reply to @Leftbones "It's probably... fine? There": So, I don't want to discourage you in any way, it's really good to experiment to learn how things work. That being said, terminals are tricky things, which is why historically and currently people put a lot of effort into software libraries such as ncurses (old and venerable) and notcurses (much newer), and there are many others as well for various languages, the point of which
20:08:06FromDiscord<Phil> (edit) "And of course 2-symbol combinations are as well ... " added "(not always but a lot of them aren't in use)"
20:08:28FromDiscord<michaelb.eth> Depending on your goals, you might consider using e.g. a Nim wrapper for ncurses
20:08:52FromDiscord<Phil> In reply to @huantian "technically @ converts arrays": I feel like you basically can't be wrong if you just associate @ with "a seq will get created in the process of this statement"
20:11:16FromDiscord<Leftbones> In reply to @michaelb.eth "So, I don't want": Honestly, I'm not really trying to do anything crazy. I'm just writing a nice wrapper around using ANSI escape codes, basically. Terminals are tricky, you're 100% correct.↵↵I was just going to implement true color support into Illwill and make a PR, but I ended up doing this instead.
20:12:23*argonica quit (Remote host closed the connection)
20:12:40*argonica joined #nim
20:13:09FromDiscord<Leftbones> I've used ncurses quite a bit myself, and as much as I am comfortable with it, I don't love the idea that we've been stuck with the same TUI library since the 90s just because "terminals are tricky"
20:13:54FromDiscord<pyolyokh> we haven't been, but slang died off
20:14:30FromDiscord<Leftbones> Plus, I'm a hobby programmer. I don't get paid to write code, I only do it for fun, and to learn. I'd like to know more about the inner workings of terminals, and this is a fun way to do it.
20:14:52FromDiscord<pyolyokh> there are lower-level libraries that know about terminal capabilities (termcap) and what codes you have to send for what feature. ncurses isn't about trickiness but efficiency: the main thing about terminals is that they are slow
20:15:30FromDiscord<Leftbones> This is true
20:15:45FromDiscord<Leftbones> Like I said though I'm just making nice ways to use escape codes, which isn't really slow by any means
20:15:59FromDiscord<Leftbones> My main gripe with curses is the way it handles color, which I despise
20:16:22FromDiscord<Leftbones> To the point where I wrote a whole-ass system around making the color system more usable in Python for a previous project of mine lol
20:16:24FromDiscord<michaelb.eth> slow, relatively, sure, but modern terminal can be made to do some crazy visual stuff that to human eyes isn't particularly slow feeling/looking
20:16:39FromDiscord<michaelb.eth> (edit) "terminal" => "terminals" | removed ""
20:17:05FromDiscord<michaelb.eth> e.g. https://www.youtube.com/watch?v=dcjkezf1ARY
20:17:30FromDiscord<michaelb.eth> demo in terminals starts about 0:15
20:19:12FromDiscord<Leftbones> I've actually never used notcurses, I remember looking at it for Python stuff before I learned ncurses and deciding to go with ncurses instead for whatever reason
20:25:10FromDiscord<pyolyokh> sent a long message, see https://paste.rs/M2o
20:26:44FromDiscord<Leftbones> My only experience with anything curses is ncurses in Python, and I was happy to learn that it supported color, and then sad and confused to see how it supported color
20:27:03FromDiscord<michaelb.eth> The python bindings for notcurses were (and still are, I think) a bit of a mess; there were two different sets of bindings, one done by the notcurses' author and another by a contributor. I remember seeing bug reports that neither was fully working, but maybe that's been remedied.
20:28:22FromDiscord<Leftbones> I'm watching a vide of the creator talking about notcurses
20:28:32FromDiscord<Leftbones> I find it a bit odd that I can't seem to find ANY examples of it by doing basic searches
20:28:57FromDiscord<michaelb.eth> well it hasn't been around very long (only a few years) and isn't super well known
20:28:58FromDiscord<michaelb.eth> so...
20:29:33FromDiscord<pyolyokh> what are you searching? <https://github.com/search?q=notcurses>
20:29:57FromDiscord<Leftbones> Just googling "notcurses examples"
20:29:59FromDiscord<Leftbones> I found one example in Zig
20:30:09FromDiscord<michaelb.eth> https://github.com/dankamongmen/notcurses/tree/master/src/demo
20:30:41FromDiscord<michaelb.eth> more basic examples are in https://github.com/dankamongmen/notcurses/tree/master/src/poc
20:32:18FromDiscord<Leftbones> Ah nice, thanks
20:32:53FromDiscord<Leftbones> Maybe I'll look into it someday. I'm definitely not super interested in writing any substantial amount of C these days though, so it'd probably be Python or some other wrapper of notcurses.
20:35:25*genpaku quit (Remote host closed the connection)
20:39:29*genpaku joined #nim
21:02:31FromDiscord<Verdagon> hey all! fellow language dev here, i'm going to be writing an article soon about the next generation of languages and their memory safety approaches
21:03:38FromDiscord<Verdagon> ive had the (possibly incorrect) impression that nim isn't as much focused on memory management/safety innovations, and mostly is innovating in developer experience, making a stellar ergonomic language
21:04:31FromDiscord<Elegantbeef> It does allow fine control of memory, and does have multiple safety features, though many are experimental
21:04:33FromDiscord<Verdagon> ORC _appears_ to be reference counting with a cycle collector, which appears in some other languages (python), so i wonder if theres some interesting details in there that make ORC really shine
21:04:50FromDiscord<Verdagon> im also revisiting my impression from re-reading up on destructors, =sink, and so on
21:04:58FromDiscord<Phil> I mean, what counts as focused?↵You have ORC/ARC, refc, no memory management at all, you have move semantics for easier memory control and I'm pretty sure I'm forgetting an option
21:05:01FromDiscord<Verdagon> i would love to know what yall's favorite innovations are from nim, so i could include them in my article =)
21:05:09FromDiscord<Verdagon> (edit) "i would love to know what yall's favorite ... innovations" added "memory-related"
21:05:34FromDiscord<Elegantbeef> The interesting detail is that it's deterministic for non cycles, so you get RAII style memory management for most types, but cycles free still
21:05:55FromDiscord<Phil> TBH, combining ORC and enabling me to do memory optimization and dealing with it when I want to / when it actually matters for performance
21:06:05FromDiscord<Verdagon> In reply to @Elegantbeef "The interesting detail is": thats a really interesting angle, especially since im a big fan of deterministic destruction and destructors in general
21:06:09FromDiscord<Phil> 99% of the times the performance differences to me do not matter
21:06:10FromDiscord<Verdagon> do people make use of that in practice, with nim?
21:06:14FromDiscord<Pájaro.rs> Hi 👋
21:06:33FromDiscord<Elegantbeef> If you use Orc your use that in practice
21:06:40FromDiscord<Elegantbeef> you use that\
21:07:04FromDiscord<Phil> Orc is the upcoming default in nim2.0 and has been decently useable since... well I've been using it for a year now consistently
21:07:15FromDiscord<pyolyokh> ergonomic advantages are retrospective, from people looking at zig and rust and noticing that Nim has better ergonomics. I wouldn't call it a focus. There are too many idiosyncratic limitations for that.
21:07:17FromDiscord<Phil> Not sure how old it is
21:07:29FromDiscord<Verdagon> in your opinion, whats the best example of nim code that makes use of these custom destructors? (by default, ill probably look at the file-closing code, thats usually a good example of this kind of thing)
21:08:00FromDiscord<Elegantbeef> Using it for CFFI is nice
21:08:12FromDiscord<pyolyokh> In reply to @Verdagon "ORC _appears_ to be": You should say "in some other languages (lobster)" here, because the innovation you noted it has is very similar to arc.
21:08:48FromDiscord<michaelb.eth> might be good to point out that while `arc`/`orc` are available today, they're still not the default (`refc` is still the default), so I think the long-term impact on the language and its users will start to be seen after Nim 2.0 is released
21:08:51FromDiscord<Verdagon> iirc, lobster uses borrow-checker-like static analysis under the hood, what kinds of static analysis does nim use?
21:09:01FromDiscord<pyolyokh> In reply to @Verdagon "iirc, lobster uses borrow-checker-like": exactly that.
21:09:05FromDiscord<Verdagon> oh wow
21:09:09FromDiscord<Verdagon> ...oh no
21:09:18FromDiscord<Verdagon> now i feel bad for mentioning lobster every other article and not nim
21:09:39FromDiscord<Verdagon> ive been looking for reasons to promote nim hahah
21:09:41FromDiscord<pyolyokh> I noted that, but blamed nim-lang.org for only discussing Arc in blogposts
21:09:48FromDiscord<pyolyokh> Lobster has prominent documentation
21:09:53FromDiscord<auxym> to your defense it's not very well publicized. marketing is not nim's strong suit haha
21:10:11FromDiscord<auxym> (it's not even well documented)
21:10:30FromDiscord<Phil> TBH I wouldn't be able to understand that point even if it was brought up
21:10:41FromDiscord<Ntsékees> Are there already available min/max functions in the standard library or should I write mine? I don't see any in the math module.
21:11:15FromDiscord<Phil> In reply to @Ntsékees "Are there already available": There are, they're either in sequtils (since they go over iterables) or semi hidden.↵Generally for that the search bar in the nim docs is really helpful, I use it constantly
21:11:15FromDiscord<Verdagon> how big of a deal is `=sink` in practice? i think nim's the only one that does that, perhaps i could focus on that
21:11:29FromDiscord<pyolyokh> In reply to @Ntsékees "Are there already available": <https://nim-lang.org/docs/system.html#max%2CopenArray%5BT%5D> , the search box at the top left will find that
21:11:29FromDiscord<Verdagon> (edit) "`=sink`" => "`sink`"
21:11:36FromDiscord<Verdagon> (edit) "how big of a deal is `sink` in practice? i think nim's the only one that does that, perhaps i could focus on that ... " added "too"
21:12:05FromDiscord<Verdagon> also, is nim's static analysis primarily based on sink parameter inference, or a lot past that?
21:12:24FromDiscord<Phil> In reply to @Ntsékees "Are there already available": Turns out it's part of the systems lib, which you have by default
21:12:32FromDiscord<Phil> Dangit, pyolokh was faster!
21:13:06FromDiscord<Elegantbeef> Nim implicitly copies when a move cannot be done
21:13:13FromDiscord<Phil> But yeah, miniexample:↵`max([1,2,3,4,2,412])`
21:13:38FromDiscord<Phil> or if you wish:↵`[1,2,3,4,5,6,41234].max`
21:14:52FromDiscord<ShalokShalom> In reply to @Verdagon "ORC _appears_ to be": @Verdagon this may help↵↵https://nim-lang.org/blog/2020/10/15/introduction-to-arc-orc-in-nim.html
21:17:38FromDiscord<Verdagon> its a great introduction, but its hard to find what really makes ORC special in that particular article
21:18:19FromDiscord<Verdagon> id love to be able to cite something that says nim uses static analysis to drastically eliminate RC ops
21:18:32FromDiscord<pyolyokh> In reply to @Verdagon "how big of a": it just lets you avoid the `=destroy`+`copyMem` that the compiler would insert in the alternate case,
21:18:37FromDiscord<Elegantbeef> It only uses RC ops on reference types
21:19:02FromDiscord<Verdagon> i also found this which will be a great quote:↵> The reference counting operations (= "RC ops") do not use atomic instructions and do not have to -- instead entire subgraphs are moved between threads.↵this puts it _way_ ahead of existing RC languages
21:19:28FromDiscord<michaelb.eth> @Verdagon maybe there's interesting/helpful info https://nim-lang.org/docs/destructors.html
21:19:32FromDiscord<pyolyokh> In reply to @Verdagon "id love to be": <https://nim-lang.org/blog/2020/12/08/introducing-orc.html> is the closest that I've found off-hand.
21:19:44FromDiscord<michaelb.eth> (edit) "@Verdagon maybe there's interesting/helpful info ... https://nim-lang.org/docs/destructors.html" added "in"
21:19:57FromDiscord<Phil> Yeh I'll check out of the discussion because at that point my knowledge is exhausted thrice over 😄
21:20:55FromDiscord<pyolyokh> if you talk to Araq directly I'm sure he'll have something better. I've heard lines like "if 80% of RC is optimized out, is it really reference counting?"
21:21:15FromDiscord<Verdagon> In reply to @pyolyokh "<https://nim-lang.org/blog/2020/12/08/introducing-o": this is _perfect_!
21:21:21FromDiscord<Verdagon> this is exactly the kind of thing i was looking for
21:24:16FromDiscord<Verdagon> In reply to @pyolyokh "if you talk to": is there a benchmark or something i can cite, with the 80% figure?
21:24:31FromDiscord<Verdagon> oh wait, nevermind, its further down that same page
21:25:07FromDiscord<Elegantbeef> One thing that can arguably be great, is like I said move semantics in Nim do not get in your way, so if something was supposed to be moved it is copied implicitly not bugging the programmer with "Error\: X was supposed to be moved but is reused at line Y"
21:25:46FromDiscord<Elegantbeef> Copying can be disabled on a type basis, or you can enable a hint for this implicit copy so it's not a big problem if you know what you're doing
21:26:15FromDiscord<michaelb.eth> sent a long message, see http://ix.io/4lS7
21:26:38FromDiscord<michaelb.eth> (edit) "http://ix.io/4lS7" => "http://ix.io/4lS8"
21:27:25FromDiscord<michaelb.eth> so instead you need to do deep copies and take an immutable message passing approach
21:27:26FromDiscord<Verdagon> might be a nonsense question but: is there a way to mimic something like a mutex in a shared-nothing heap with just isolated subgraphs and channels?
21:28:10FromDiscord<michaelb.eth> (edit) "do" => "make" | "take" => "use"
21:29:58FromDiscord<Elegantbeef> https://github.com/nim-lang/threading/blob/master/threading/channels.nim located here
21:30:13FromDiscord<Elegantbeef> https://github.com/nim-lang/RFCs/issues/244 explains the isolated idea
21:30:47FromDiscord<michaelb.eth> I don't think my understanding is strong enough to be able to answer your question. But, with ARC/ORC, if my basic understanding is correct, there's just one heap so hopefully it will some things simpler.
21:31:00FromDiscord<Verdagon> got it, many thanks
21:31:00FromDiscord<Elegantbeef> It is generally simpler
21:31:05FromDiscord<michaelb.eth> (edit) "I don't think my understanding is strong enough to be able to answer your question. But, with ARC/ORC, if my basic understanding is correct, there's just one heap so hopefully it will ... some" added "make"
21:31:15FromDiscord<Verdagon> i think i have enough to write my article promoting nim and other new languages' memory approaches 😃
21:31:20FromDiscord<Elegantbeef> There is still a complex issue with aliasing GC'd types
21:31:36FromDiscord<Verdagon> ill post a draft here and send it to araq before i publish, in case theres corrections or more things i can include in it
21:31:36FromDiscord<Elegantbeef> I say complex but it's resolved with one `{.cursor.}`
21:39:21FromDiscord<ShalokShalom> sent a long message, see http://ix.io/4lSb
21:39:46FromDiscord<ShalokShalom> I think Nim is a great showcase, since it has deep Oberon, and by that Pascal and Modula roots in it.
21:42:00FromDiscord<Verdagon> In reply to @ShalokShalom "<@337101146017497109> tbh, I always": thats a pretty good point
21:42:24FromDiscord<ted__> sent a code paste, see https://play.nim-lang.org/#ix=4lSd
21:42:26FromDiscord<Verdagon> mind if i DM you a draft when its done and you can tell me if any of the features seem familiar from past languages?
21:42:37FromDiscord<ShalokShalom> I am happy 😊
21:42:43FromDiscord<Verdagon> thanks 😃
21:43:41FromDiscord<Elegantbeef> Ted `[^1]`
21:43:42FromDiscord<ShalokShalom> Do you know Pharo?
21:43:53FromDiscord<Verdagon> nope, heard the name once or twice but thats all
21:43:56FromDiscord<ShalokShalom> Its a modern Smalltalk
21:44:05FromDiscord<ShalokShalom> And they have IDE and language rolled into one
21:44:10FromDiscord<ted__> In reply to @Elegantbeef "Ted `[^1]`": awesome, thanks! I knew there had to be something
21:44:20FromDiscord<ShalokShalom> The whole thing is VM that state you modify
21:44:41FromDiscord<ShalokShalom> So you don't save anything in the traditional sense. You live edit a mutable VM.
21:44:52FromDiscord<Elegantbeef> `^x` in Nim makes a `BackwardsIndex`, it's all in user space 😄
21:44:55FromDiscord<ShalokShalom> Lots of genius ideas from Alan Kay.
21:45:21FromDiscord<ShalokShalom> And my personal favorite is F#, its the most modern SML
21:46:21FromDiscord<Gumbercules> In reply to @Verdagon "nope, heard the name": @Verdagon Vale lang might be of interest to you
21:46:37FromDiscord<Verdagon> In reply to @Gumbercules "<@337101146017497109> Vale lang might": funny story that, i'm the developer behind vale hahah
21:46:41FromDiscord<Gumbercules> If you haven't already found it
21:46:48FromDiscord<Verdagon> ive come across it a few times ;)
21:46:49FromDiscord<Gumbercules> Hahaha fair enough!
21:47:13FromDiscord<Gumbercules> I like what you're doing with Vale btw
21:47:32FromDiscord<Gumbercules> And the blog posts have been great!
21:47:36FromDiscord<Verdagon> thanks! hearing that always brings a smile to my face
21:47:46FromDiscord<Verdagon> hah and thank you for that too, there's been _quite_ the backlash against some of them
21:48:14FromDiscord<Verdagon> in my research today, i was delighted to hear that nim has shared-nothing heaps, we're did that approach in vale too and i think that decision can really launch languages into some interesting realms
21:48:33FromDiscord<leorize> nim is moving away from that unfortunately
21:48:43FromDiscord<Verdagon> oh no =\
21:49:23FromDiscord<Verdagon> it enables some pretty stellar optimizations, especially when combined with a (java-like) `final`
21:49:37FromDiscord<Phil> java PTSD
21:49:53FromDiscord<Verdagon> it also enables the region borrow checker, something that could greatly benefit RC languages, not just gen-ref-based ones
21:50:17FromDiscord<Gumbercules> Nim at one point had an experimental region based memory management model but it never made it out of the experimental phase and was abandoned
21:50:22FromDiscord<Verdagon> when i talk to araq, i'll bring it up in case any of these things havent been considered
21:51:15FromDiscord<Gumbercules> https://github.com/nim-lang/Nim/blob/devel/lib/system/gc_regions.nim
21:55:35FromDiscord<Ntsékees> sent a code paste, see https://play.nim-lang.org/#ix=4lSh
21:55:40FromDiscord<Ntsékees> That's odd
21:55:45FromDiscord<Ntsékees> (edit) "https://play.nim-lang.org/#ix=4lSh" => "https://paste.rs/wRs"
21:55:57FromDiscord<Ntsékees> hmm maybe it's not ascii?
21:56:48FromDiscord<Ntsékees> (edit) "https://play.nim-lang.org/#ix=4lSi" => "https://play.nim-lang.org/#ix=4lSj"
21:57:36FromDiscord<Phil> I don't even know what character that is 😄
21:57:44FromDiscord<Ntsékees> 0xB4
21:58:32FromDiscord<Phil> Yeh no clue, accent going upwards?
21:58:46FromDiscord<leorize> it's probably multi-byte↵(@Ntsékees)
21:59:02FromDiscord<leorize> use a string to store that
21:59:12FromDiscord<Ntsékees> probably; the error message is a bit off
21:59:37FromDiscord<Ntsékees> "invalid character literal" would have been better
22:00:30FromDiscord<Elegantbeef> sent a code paste, see https://play.nim-lang.org/#ix=4lSl
22:00:33FromDiscord<Elegantbeef> So i think the error message is 'correct'
22:01:05FromDiscord<ShalokShalom> In reply to @Verdagon "in my research today,": Why that?
22:03:11FromDiscord<ShalokShalom> @Verdagon how far goes Vale's type inference?
22:14:27FromDiscord<Verdagon> In reply to @ShalokShalom "<@337101146017497109> how far goes": im planning on doing full bidirectional type inference within the year, but right now its only one-way (think c++ or c#)
22:14:47FromDiscord<Verdagon> ive been doing gradual upgrades to the solver to eventually support it
22:14:57FromDiscord<ShalokShalom> In reply to @Verdagon "im planning on doing": Nice
22:15:03FromDiscord<Verdagon> you wouldnt _believe_ the nonsense involved in bidirectional lol
22:15:12FromDiscord<ShalokShalom> That's like in Crystal then
22:15:17FromDiscord<Verdagon> i assume nim has bidirectional?
22:15:19FromDiscord<ShalokShalom> And did you consider HM
22:15:32FromDiscord<Verdagon> i think im heading towards HM, yep
22:15:51FromDiscord<ShalokShalom> I heard sarcastic things being said about Nims type system
22:15:54FromDiscord<jtv> And "generational references" look like epoch-based memory management, what's the difference?
22:15:58FromDiscord<ShalokShalom> I think its one way
22:16:36FromDiscord<ShalokShalom> In reply to @Verdagon "i think im heading": And can you just implement that on the existing code base, by breaking little to nothing
22:16:46FromDiscord<ShalokShalom> Or do you just break, since its pre 1.0
22:17:10FromDiscord<Verdagon> In reply to @jtv "And "generational references" look": i vageuly recall that term, link me?
22:17:16FromDiscord<ShalokShalom> I would be interested in your language, if you do HM
22:17:37FromDiscord<Verdagon> yeah i agree that HM is a huge boon to usability
22:17:45FromDiscord<jtv> It's one of the many techniques for memory management in the lock-free data structure world
22:19:59FromDiscord<jtv> For instance, if you look at my high performance lock-free data structure lib for my C code (https://github.com/viega/hatrack) it's the underlying memory management to ensure safety.
22:28:15FromDiscord<Gumbercules> https://github.com/viega/hatrack/blob/b5113e087b4ae4104803ff0969998438d39537b1/include/hatrack/mmm.h
22:28:30FromDiscord<Gumbercules> (edit) "https://github.com/viega/hatrack/blob/b5113e087b4ae4104803ff0969998438d39537b1/include/hatrack/mmm.h" => "https://github.com/viega/hatrack/blob/main/include/hatrack/mmm.h"
22:31:53FromDiscord<jtv> Yeah, most of the doc and impl is in the header, but the actual freeing happens here: https://github.com/viega/hatrack/blob/main/src/support/mmm.c
22:32:07*jjido joined #nim
22:32:48FromDiscord<Gumbercules> was looking for an example of region based memory management in C and of course DOOM has one
22:33:10FromDiscord<Gumbercules> https://github.com/id-Software/DOOM/blob/master/linuxdoom-1.10/z_zone.h
22:33:20FromDiscord<Gumbercules> https://github.com/id-Software/DOOM/blob/master/linuxdoom-1.10/z_zone.c
22:36:31FromDiscord<Gumbercules> https://github.com/id-Software/DOOM/blob/master/linuxdoom-1.10/i_system.h↵and↵https://github.com/id-Software/DOOM/blob/master/linuxdoom-1.10/i_system.c are also relevant
22:36:32FromDiscord<jtv> Yeah, there's been lots and lots and lots of work in memory management in the last 30 years 🙂
22:36:42*chmod222 quit (Quit: leaving)
22:37:28FromDiscord<jtv> Yet most languages like Go still have really poor GC behavior 🙂
22:37:32FromDiscord<Gumbercules> Yeah, Nim just needs custom allocators
22:37:43FromDiscord<Gumbercules> well, without having to ditch seq/string
22:38:38FromDiscord<ShalokShalom> In reply to @Verdagon "yeah i agree that": Rust uses it afaik for its borrow checker
22:39:33FromDiscord<Gumbercules> Cone's approach sounds interesting
22:39:45FromDiscord<Gumbercules> sent a code paste, see https://play.nim-lang.org/#ix=
22:40:09FromDiscord<ShalokShalom> If you need help with HM, try to reach OvermindDL1. He is busy, and in case you catch him, you can find someone who is very knowledgeable about implementing it.
22:45:39FromDiscord<Gumbercules> @Verdagon sorry I'm not a PL designer / contributor so some of my questions may come off as naive - what's the plan for Vale and interop? Is it possible to write a Vale program today and interop with a C library?
22:46:41FromDiscord<Gumbercules> I know you blogged about the whole fearless FFI thing but with so many portions of the language in the design phase and the language development moving fast, it's difficult to know when I should pick up and start playing with the lanugage
22:56:08FromDiscord<sOkam!> If I have an enum like `type Kinds {.pure.} = enum One, Two`↵Is there a way to get the names of the values in string form? ↵aka `"One"` and `"Two"`
22:58:46FromDiscord<sOkam!> In reply to @Gumbercules "was looking for an": You might want to check out:↵https://github.com/heysokam/idtech3-modules/tree/master/src/mem↵Its the id-tech3 code, but extracted from the engine to use in any other project
22:59:44FromDiscord<michaelb.eth> In reply to @sOkam! "If I have an": try `$Kind.One`
23:01:40FromDiscord<Ntsékees> And now I'm stumbling on a compiler crash…↵> Error: internal error: genTypeInfoV1(tyNone)↵> No stack traceback available↵> To create a stacktrace, rerun compilation with './koch temp c <file>', see https://nim-lang.github.io/Nim/intern.html#debugging-the-compiler for details
23:02:13FromDiscord<arkanoid> how can I perform in nim same op as https://registry.khronos.org/OpenGL-Refpages/gl4/html/bitfieldInsert.xhtml ? basically I need something like bitops.bitsliced, but for writing and not for reading
23:02:47FromDiscord<sOkam!> In reply to @michaelb.eth "try `$Kinds.One`": oh, shoot! i thought that was going to say the value itself. that's handy af, ty
23:02:58FromDiscord<Elegantbeef> I'd say use bitsets, but idk
23:04:17FromDiscord<sOkam!> @Gumbercules The memory model is explained in `core.h`
23:06:43FromDiscord<sOkam!> And these are the zones where data is stored when allocated↵https://github.com/heysokam/idtech3-modules/blob/a5cd0d55a32f27fc97799ac46095933d1453f013/src/mem/state.h#L22-L32
23:12:56*clemens3 joined #nim
23:16:45*argonica quit (Read error: Connection reset by peer)
23:39:36FromDiscord<Verdagon> In reply to @Gumbercules "I know you blogged": yeah, it is a hard thing to know when a language is ready to experiment with, especially one as new as vale
23:40:23FromDiscord<Verdagon> i wish i could say it's stable and rock solid, but we arent there yet, especially compared to great languages like nim ;)
23:41:23FromDiscord<Verdagon> one can successfully make a project with vale, but they might run into the occasional corner or hole, such as our lack of `match` statement, or unsupported linux distros
23:41:46FromDiscord<Verdagon> once i finish the first regions prototype, the next version's goal will be another ergonomics+stability sprint
23:46:15FromDiscord<arkanoid> solved with loop over testBit/setBit. Possibly the worst solution possible, but it passes my tests