<< 13-08-2021 >>
00:57:37*auxym_ quit (Ping timeout: 248 seconds)
01:25:35*happycorsair[m] joined #nim
01:37:03FromDiscord<vevirit105> https://media.discordapp.net/attachments/371759389889003532/875553502158528532/fd503064d92655844644f29f77fb4bbb.png
01:37:17FromDiscord<vevirit105> https://media.discordapp.net/attachments/371759389889003532/875553559997980802/81067760_p0.png
01:38:31FromDiscord<vevirit105> https://media.discordapp.net/attachments/371759389889003532/875553873727729664/sample-a5a114c6ddc515d6b861faf3813865c7.png
02:10:30*Helios joined #nim
02:22:51FromDiscord<impbox [ftsf]> if you want to mutate a variable passed in, but not mutate it outside, you can just declare in the function `var a = a` for example
02:31:35*nixfreak_nim[m] joined #nim
02:46:38*arkurious quit (Quit: Leaving)
03:32:04madpropswhat is {'a', 'b'} ? seems to work like an array
03:32:24*rockcavera quit (Remote host closed the connection)
03:33:25nrds<Prestige99> A set
03:34:23madpropsoh
03:34:47madpropsso unique values
03:38:30madpropswhat could be the nicest way to get an array of a..z ?
03:41:47madpropsI'm just doing a simple 'collect'
03:42:04FromDiscord<Elegantbeef> int -\> char or char -\> array?
03:44:06FromDiscord<Elegantbeef> cause you can do `var a: array['a'..'z', int]`
03:46:00madpropsoh nice
03:47:55madpropsdon't think that works
03:48:11madpropsoh nvm
03:49:42madpropshmm
03:49:52madpropsnot sure how to initialize that
03:50:16madpropswith a..z
03:50:44madprops'a', 'b', 'c' ...
03:50:59FromDiscord<Elegantbeef> `for c in 'a'..'z'`
03:51:13madpropsyeah
03:51:24madpropswas wondering if there was a oneliner
03:51:38FromDiscord<Elegantbeef> Init it to what though?
03:52:47madpropsa seq of chars
03:52:50madpropsor array i mean
03:52:59FromDiscord<lamersc.com 🐧> In reply to @Elegantbeef "Init it to what": abcdefghijklmnopqrstuvwxyz 🕶️
03:53:16FromDiscord<Elegantbeef> I dont quite know what you mean, you want an array of characters?
03:53:25madpropsyeah from a to z
03:53:52madpropsoh i guess i could just use a string
03:53:54madprops:)
03:54:08madpropsbut i would still want to generate that
03:54:12madpropsinstead of copying a long line
03:54:48madpropsarray is better though
03:55:40madpropsalso seems 'collect' is new
03:55:43madpropssince it's not available on 1.0
03:58:10madpropshttps://play.nim-lang.org/#ix=3vNL
04:00:28FromDiscord<Elegantbeef> There isnt a range to array method so this is pretty much the best you can do https://play.nim-lang.org/#ix=3vNM
04:03:19madpropsdidn't know you could return from a block like that, interesting
04:03:32FromDiscord<Elegantbeef> All flow control can be an expression
04:05:20madpropsI knew about if/else
04:05:26madpropsbut didn't think of block
04:06:01*supakeen quit (Quit: WeeChat 3.2)
04:06:30*supakeen joined #nim
05:03:15*flynn quit (Read error: Connection reset by peer)
05:04:21*flynn joined #nim
06:32:48*max22- joined #nim
06:44:49*PMunch joined #nim
07:22:32PMunchHmm, started looking at eink monitors again
07:35:22*xet7 joined #nim
07:42:08PMunchEvery vacation I find myself wanting a small eink laptop which would work in direct sunlight, have insane battery life, and would allow me to code and use simple services on the go
07:43:04FromDiscord<Elegantbeef> Didnt realize there we eink monitors for external screens
07:43:27FromDiscord<Rika> There’s plenty
07:43:49FromDiscord<Rika> I own an e ink tablet and it’s pretty amazing for reading manga or books
07:44:27PMunchYeah I have a small Kinda that's great for reading books. But I'd like something with an input :P
07:45:49FromDiscord<Rika> Kinda hahaha
07:46:19FromDiscord<Rika> They’re too niche and being usurped by other technologies with drawbacks other people are more fine with
07:46:49FromDiscord<Rika> Like faster refresh and colour but worse battery life and daylight visibility
07:47:15FromDiscord<Rika> But I think the manufacturer of the displays is called Dasung
07:47:38FromDiscord<Rika> No laptop for the aforementioned issues
07:47:56FromDiscord<Rika> Probably just easier to get a display and turn the main one off lol
07:48:26FromDiscord<Elegantbeef> pmunch has a 3D printer so clearly he'll modify his laptop to work with one of them
07:53:33FromDiscord<System64 ~ Flandre Scarlet> > SIGSEGV: Illegal storage access. (Attempt to read from nil?)↵File not found or null pointer?
07:55:50FromDiscord<Rika> Null pointer
07:55:55FromDiscord<Rika> Which could mean file not found lol
07:57:07PMunch@ElegantBeef, there is even a full 25.3" screen from Dasung that's coming at the end of the year. And the new ones have surprisingly good refresh rates, but still no colours
07:57:22FromDiscord<Rika> Well is colour a concern for you
07:57:23PMunchThere are small ones with 7 colour displays, which would be great for syntax highlighting and such
07:57:52PMunchColour would be nice, but definitely not a necessity
07:58:04FromDiscord<Rika> Well you could just go with underline bold italics and permutations
07:58:40FromDiscord<Rika> If you’re a bit crazy maybe also strikethrough
07:58:45PMunchElegantBeef, I don't think I'd want to try and sacrifice my laptop for something like this. But there is a guy who is in the process of replacing the panel on his T480 (same laptop I have) with an eink display
07:59:09PMunch@Rika, for sure. I guess with a special purpose editor I could even do different fonts
07:59:48FromDiscord<Rika> Sounds like it’s time to make one
07:59:50PMunch@System64, hard to tell what's wrong without any code
08:00:11PMunchThere are some small eink screens on ebay that's RPi compatible
08:00:29PMunch9.7" and 10.3", which might be big enough for what I have in mind
08:01:07FromDiscord<Rika> That’s just as big as my tablet though
08:01:21PMunchWhich tablet?
08:01:28FromDiscord<Rika> Some tablets can be used as monitors
08:01:36FromDiscord<Rika> Some Onyx Boox tablet
08:01:56PMunchHmm, interesting
08:02:20PMunchI have the 6" Kindle, that would be a bit small. But apparently there is some jailbreak software that can turn it into a terminal
08:02:31PMunchBut I want something I can customise a bit more
08:02:47FromDiscord<System64 ~ Flandre Scarlet> How can I disable the GC please?
08:03:02FromDiscord<Rika> Why?
08:03:14PMunch--gc:none
08:03:25PMunch@Rika, was that why for me or System64?
08:03:35FromDiscord<System64 ~ Flandre Scarlet> I try to figure out why I have a null pointer
08:03:35FromDiscord<Rika> System
08:03:41FromDiscord<System64 ~ Flandre Scarlet> ah thanks
08:03:55FromDiscord<Elegantbeef> GC doesnt magically give a null pointer
08:03:57PMunch@System64, hint: it's not the GC :P
08:04:25PMunchYeah it's definitely not the GC giving you a null pointer exception, unless you're doing something really exotic
08:04:35FromDiscord<Rika> It could be they are using the memory allocation APIs wrong
08:04:55FromDiscord<Elegantbeef> or you know
08:05:13FromDiscord<Elegantbeef> !eval var a\: ref int; echo a[]
08:05:14NimBotCompile failed: /usercode/in.nim(1, 9) Error: ':' or '=' expected, but got 'keyword ref'
08:05:37FromDiscord<Rika> Yes but I assume that would have been weeded out early on the debugging process
08:05:39PMunchI would call that exotic in Nim-land
08:05:46PMunchMost people never has to deref anything
08:05:49FromDiscord<System64 ~ Flandre Scarlet> In reply to @Elegantbeef "!eval var a\: ref": what's that?
08:06:05FromDiscord<Rika> I guess I’m not part of most people
08:06:51FromDiscord<proud linux user> :susPog:
08:08:26FromDiscord<Elegantbeef> That's purposely dereferencing a nil reference
08:14:16FromDiscord<System64 ~ Flandre Scarlet> even with GC off it crashes
08:14:39FromDiscord<Elegantbeef> It's a nil reference the GC isnt causing it
08:15:30PMunchYeah, no big surprise there
08:15:34FromDiscord<System64 ~ Flandre Scarlet> and can a nil reference be caused by an extern lib?
08:15:51FromDiscord<Elegantbeef> It can be caused by anything that attempts to dereference a nil pointer
08:15:52PMunchOf course, if it's doing something weird, or you're passing it weird data
08:24:04PMunchHmm, how easy would it be to interface with a keyboard in an old laptop?
08:24:31PMunchI've got a bunch of old laptops in the basement, so I could definitely reuse one of those as a case
08:24:33FromDiscord<Elegantbeef> They're usually designed to be interfaced with human hands, so easily! 😜
08:25:39PMunchHaha :P
08:26:04PMunchHmm, I mean I have already written keyboard firmware. So if I just get the pins of the matrix I should be able to fairly easily do it :P
08:26:25FromDiscord<Elegantbeef> depends on the model of course but some old laptops just used ps/2 for trackpad/keyboard
08:26:55PMunchHmm, I have some really old ones as well
08:27:11PMunchAnd I think they even have 4:3 monitors which is the same as the e-ink screen
08:28:41PMunchProblem with old laptops is that they are a bit chunky..
08:33:10PMunchI mean I have two old Lifebook B2154 laptops that might be perfect for this
08:39:44PMunchIt has a 10.4" screen
08:40:58*xet7 quit (Ping timeout: 272 seconds)
08:41:17FromDiscord<haxscramper> https://github.com/haxscramper/hack/blob/master/testing/nim/compilerapi/nims_template.nim#L62
08:42:58FromDiscord<haxscramper> [enthus1ast](https://matrix.to/#/@sn0re:matrix.code0.xyz)\: Code before that is just a boilerplate that does not really have to change much, except for `conf.structuredErrorHook =` part - that's where you add your error handler, and I assume you would want something better than just `assert false`
08:43:32FromDiscord<haxscramper> Then you just pass ast to `processModule3` and it evaluates things in nimscript
08:45:03FromDiscord<haxscramper> `include compiler/passes` is needed due to `processTopLevelStmt` - it is not exported. There is an older version of this script where I tested things from code script and from node, (so you can eval text or eval AST) - https://github.com/haxscramper/hack/commit/3881ae5be6b63e7843299061a81a30b970f4b9bf#diff-b92c209186bfda7a25aae8b0d1fb3d876b9fe7e5f28007fbe2e9fcd4804d6937R131-R183
08:46:01FromDiscord<haxscramper> [Edit](https://discord.com/channels/371759389889003530/371759389889003532/875660267764187136): https://github.com/haxscramper/hack/blob/d2324554cff3d9c3401715700e67383bb4474771/testing/nim/compilerapi/nims_template.nim#L62
08:52:59*xet7 joined #nim
09:02:27FromDiscord<System64 ~ Flandre Scarlet> @haxscramper even with the rewritten version of my program it crashes
09:04:41FromDiscord<haxscramper> https://github.com/haxscramper/hack/commit/ac6941b3058d6cc36a4205c575e889ebd13ea30f also example of adding custom proc
09:06:33FromDiscord<haxscramper> In reply to @System64 "<@!608382355454951435> even with the": I didn't say rewriting it would make crash go away. Instead I suggested you to rewrite it in order to make it easier to spot the difference
09:06:55FromDiscord<haxscramper> So, if it crashes, then rewrite is not exact, and you need to figure out what changed
09:07:37FromDiscord<System64 ~ Flandre Scarlet> maybe the problem are from my bindings so I'll inspect that
09:08:44FromDiscord<System64 ~ Flandre Scarlet> I don't think it's from SDL (it's not from me and it's supposed to be stable)
09:10:12FromDiscord<haxscramper> You used `.cdecl, importc, dynlib: tilenginedll.` for bindings, so it makes sense to just look at the C code
09:10:54FromDiscord<haxscramper> You can "inspect" your binings however you like, but if you want to really see what is going on, and what is different, you should just take end result and compare them
09:12:24FromDiscord<System64 ~ Flandre Scarlet> the binding is based on the header file
09:14:21FromDiscord<haxscramper> whatever
09:15:38FromDiscord<System64 ~ Flandre Scarlet> sent a code paste, see https://play.nim-lang.org/#ix=3vRP
09:15:54FromDiscord<haxscramper> yes
09:15:55FromDiscord<System64 ~ Flandre Scarlet> (edit) "https://play.nim-lang.org/#ix=3vRP" => "https://play.nim-lang.org/#ix=3vRQ"
09:16:50FromDiscord<System64 ~ Flandre Scarlet> the thing is TLN_VideoCallback is a pointer in C
09:17:48FromDiscord<haxscramper> `typedef void(TLN_VideoCallback)(int scanline);` means `declare TLN_VideoCallback as pointer to function (int) returning void`
09:18:09FromDiscord<haxscramper> Nim `TLN_VideoCallback = proc(cint) {.cdecl.}` has the same effect
09:18:23FromDiscord<haxscramper> cdecl procvar is a pointer to proc
09:18:41FromDiscord<haxscramper> closure procvar is different, but if you have `{.cdecl.}` it would mean the same thing
09:20:53FromDiscord<System64 ~ Flandre Scarlet> sent a code paste, see https://play.nim-lang.org/#ix=3vRR
09:21:50FromDiscord<haxscramper> yes
09:24:07FromDiscord<System64 ~ Flandre Scarlet> So I don't know what's going on :/
09:24:38FromDiscord<haxscramper> So C code looks identical?
09:24:58PMunchWhat's the type of that callback?
09:25:00FromDiscord<haxscramper> `s/looks/is`
09:25:09FromDiscord<haxscramper> In reply to @PMunch "What's the type of": `void(TLN_VideoCallback)(int scanline);`
09:25:23FromDiscord<haxscramper> And it was wrapped as`proc (scanline: cint) {.cdecl, gcsafe.}`
09:25:25FromDiscord<System64 ~ Flandre Scarlet> my binding is accurate to the header so yeah
09:25:30FromDiscord<haxscramper> ffs
09:25:51FromDiscord<haxscramper> I'm sorry, but do you read what I just said? You need to look at the generated C
09:26:11PMunchThat looks like it's correct
09:26:11FromDiscord<haxscramper> not just visuall inspect bindings and say "aha, it is the same so everything else is also identical"
09:26:40FromDiscord<haxscramper> it is highly unlikely the error is in the bindings
09:28:01FromDiscord<System64 ~ Flandre Scarlet> This is the generated C file https://media.discordapp.net/attachments/371759389889003532/875672032245846046/generated.c
09:28:58*vicfred quit (Quit: Leaving)
09:31:16FromDiscord<haxscramper> yes, I know. I also mentioned you can use `--stacktrace:off` and `--linestrace:off` to get rid of `nimfr_("<%", "C:\\nim-1.4.8\\lib\\system\\comparisons.nim");` that pollutes everything
09:31:35FromDiscord<System64 ~ Flandre Scarlet> ah yeah, let me do that
09:39:08FromDiscord<System64 ~ Flandre Scarlet> normally it's cleaned up https://media.discordapp.net/attachments/371759389889003532/875674826466033745/generated.c
09:57:10FromDiscord<fae> Today has been fun. Figured out how to build enet, load up the dylib, connect a host and client using https://github.com/c0ntribut0r/cat-400/blob/master/c4/lib/enet/enet.nim and got peer messages sending back and forth in response to keypress with SDL. Now to continue with multiplayer pong 😄 . Nim makes things very easy.
09:58:06FromDiscord<fae> The biggest thing that tripped me up was figuring out how to cast a `pointer` back to a string
10:01:15*Vladar joined #nim
10:02:17FromDiscord<Paupi woodberg> In reply to @System64 "normally it's cleaned up": https://media.discordapp.net/attachments/371759389889003532/875680650894671902/Screenshot_20210813-110136.png
10:03:48FromDiscord<System64 ~ Flandre Scarlet> you can't simply execute the code
10:04:02FromDiscord<Paupi woodberg> In reply to @System64 "you can't simply execute": Why
10:04:05FromDiscord<Paupi woodberg> ?
10:04:34FromDiscord<System64 ~ Flandre Scarlet> it's not a main function
10:05:37FromDiscord<Paupi woodberg> O!↵Am not really good in c!
10:23:30PMunchfae, how are you casting a pointer to string, that's potentially something that could trip you up and give you segfaults
10:24:31PMunchThere is a command that generates a build script for the C output if you want to build the C sources. I think it's --genScript
10:25:17FromDiscord<Rika> Who is Fae
10:25:30FromDiscord<Rika> Ah I missed their message nvm
10:27:26FromDiscord<fae> https://media.discordapp.net/attachments/371759389889003532/875686976496492574/Screen_Shot_2021-08-13_at_3.27.07_AM.png
10:27:42FromDiscord<Rika> Ah that is fine
10:27:42FromDiscord<fae> I'm sure it's unsafe, but not sure what else to do with a `ptr uint8`
10:28:07FromDiscord<Rika> Well you cannot avoid unsafe when interfacing with C (unless you aren’t)
10:28:19FromDiscord<dom96> ideally you should write a safe wrapper around this to isolate the unsafety
10:28:34FromDiscord<dom96> depends what you're doing though
10:29:50FromDiscord<fae> Just tinkering really, started messing with Nim only recently. I was happy the enet wrapper even worked at all.
10:31:08FromDiscord<fae> I had never used enet before directly (only indirectly love2d and godot) so being able to get it working so easily says a lot about how easy Nim makes it.
10:31:23*auxym_ joined #nim
10:40:32emeryif I put a C++ object into a `ref` type, is the constructor called when the Nim `ref object` is created ? is there simple way to force the constructor to be called?
10:41:54FromDiscord<haxscramper> you need to use placement new
10:42:30emeryhaxscramper: what is that?
10:47:17emeryc++ is such a horrible language
10:47:20PMunchIt would be nice if Nim had a built-in object type for cstring with size that could work like a string, was assumed to not be freed while you're using it (manually managed by the user), and didn't require copying..
10:47:41PMunchSomething like a special case seq that didn't interact with the GC at all..
10:49:50FromDiscord<haxscramper> In reply to @emery "<@608382355454951435>: what is that?": C++ object can call constructor on already allocated memory
10:49:51FromDiscord<haxscramper> https://wandbox.org/permlink/qqpPSMM468EpLIpL
10:50:05FromDiscord<haxscramper> Nim allocates memory with `ref` and you call placement new here
10:50:18FromDiscord<haxscramper> Without placement new it will just allocate, but constructor won't be called
10:51:06FromDiscord<haxscramper> Though it seems like `new(result)` is sufficient, and destructor is called correctly
10:52:10FromDiscord<haxscramper> Nim `new` just does ` result = (Struct)newObj( (&NTIDzqc1yILohvRHG0NLsQ9aXw_), sizeof(Struct));`
10:53:12FromDiscord<haxscramper> In reply to @haxscramper "Though it seems like": Though I would not be so sure about that, and it is probably safer to still use destructor,
10:53:17FromDiscord<haxscramper> Or look into it in more detailk
10:55:54FromDiscord<haxscramper> `proc newImportAux() {.header: "<new>", importcpp: "//".}` is needed because placement new is declared in this header, but I can't use header and provide implementation in `newT`
10:56:09FromDiscord<haxscramper> This is a hack, i don't know what would be the correct workaround
10:57:33FromDiscord<haxscramper> `importcpp: "((void)0)".` is probably better
10:59:22emeryhaxscramper: I'll work something out, thanks
10:59:45*PMunch quit (Read error: Connection reset by peer)
11:01:49*PMunch joined #nim
11:06:16*PMunch quit (Remote host closed the connection)
11:07:21*auxym_ quit (Read error: No route to host)
11:07:41*auxym_ joined #nim
11:11:43*PMunch joined #nim
11:20:00FromDiscord<enthus1ast> sent a code paste, see https://play.nim-lang.org/#ix=3vSi
11:20:01*auxym_ quit (Ping timeout: 248 seconds)
11:20:50FromDiscord<enthus1ast> i must carefully study the code you sent, have just skimmed through it yet.
11:21:49FromDiscord<haxscramper> Yes, I also wanted to addapt jinja templates, so it is naturally similar
11:22:13FromDiscord<enthus1ast> but you want to do it dynamic?
11:22:29FromDiscord<haxscramper> It would be possible to write an implementation that allows to have both dynamic and static templates
11:22:40FromDiscord<haxscramper> It is more involved though
11:23:03FromDiscord<haxscramper> There are actually three ways to do this
11:23:40FromDiscord<haxscramper> ● Nim macro that generates template code↵● Compiler-based logic that generates template code (not really different from macro)↵● Compiler-based, but directly evaluates code - this one allows for no-recompile template changes
11:25:08FromDiscord<haxscramper> I'm after latter one because it does not require user to import whole haxdoc source code in their template (\~200k sloc including all dependencies)
11:25:35FromDiscord<haxscramper> But at the same time compiled template is faster, so I can provide couple of default presets
11:25:53FromDiscord<haxscramper> For now I just wanted to show you an example of how dynamic one can be done
11:25:58FromDiscord<haxscramper> So you would have full overview of the possibilities
11:27:12FromDiscord<haxscramper> If you want I can think hyperhybrid one that allows to switch between all three solutions in a single code base
11:28:02FromDiscord<haxscramper> At least right now I feel like it would be possible
11:29:08FromDiscord<enthus1ast> yes i also had nr3 in mind, but it must be possible to use any nim functions also the ffi ones, i guess it could be possible to make ffi functions available in a custom compiler
11:30:24FromDiscord<enthus1ast> so one must be able to define functions that are later available in the templates, then maybe the compiler lib must be then compiled once and then is able to evaluate the templates dynamically
11:31:41FromDiscord<enthus1ast> yes, i'm looking forward to study your code
11:35:33FromDiscord<PsychoClay> is there a way to make `db_sqlite.exec` go faster, currently its takes ~17 seconds for about 7k inserts, on python this takes like half a second
11:35:54FromDiscord<haxscramper> Yes, that's how main nim vm does it - just tons of `registerCallback()` calls for all procs https://github.com/nim-lang/Nim/blob/devel/compiler/vmops.nim#L69You can do https://en.wikipedia.org/wiki/Opaque_pointer by exposing opaque handlers to the objects, and making all template logic function via proc calls
11:36:14FromDiscord<haxscramper> nim just binds to sqlite3, so all optimizations apply equally https://avi.im/blag/2021/fast-sqlite-inserts/↵(@PsychoClay)
11:36:31FromDiscord<haxscramper> You can also use prepared queries etc.
11:36:48FromDiscord<haxscramper> I would start with prepared queries
11:36:56FromDiscord<PsychoClay> il try that
11:37:01FromDiscord<haxscramper> I suppose python implementation does this implicitly, or has default connection configurations
11:38:49FromDiscord<haxscramper> sent a code paste, see https://play.nim-lang.org/#ix=3vSo
11:39:39FromDiscord<haxscramper> or `evalTemplateWithDb(db, getPrecompiledTemplate(hptNimDocStyle #[ or hptRustDocStype, or hptDoxygenStyle ]#)`
11:40:02FromDiscord<haxscramper> Latter one is already possible with your current implementation
11:40:32FromDiscord<haxscramper> I just need to find some time to test it
11:41:15FromDiscord<enthus1ast> @PsychoClay\:↵And maybe disable all foreign key checks\:↵PRAGMA ignore\_check\_constraints = boolean;and also insert your stuff in a transaction, then the index will be generated only once i guess
11:44:30*auxym_ joined #nim
11:52:01*auxym_ quit (Ping timeout: 248 seconds)
11:53:45FromDiscord<PsychoClay> got it down to 8 secs
11:54:13*auxym_ joined #nim
11:57:40FromDiscord<PsychoClay> now the slow part seems to be `strutils.replace`
11:59:02FromDiscord<PsychoClay> ok nvm
11:59:27FromDiscord<PsychoClay> it was 8 sec with vallgrind 0.6 without
12:04:50*max22- quit (Ping timeout: 252 seconds)
12:05:41*rockcavera joined #nim
12:05:41*rockcavera quit (Changing host)
12:05:41*rockcavera joined #nim
12:06:02*supakeen quit (Quit: WeeChat 3.2)
12:06:31*supakeen joined #nim
12:14:41*rockcavera quit (Remote host closed the connection)
12:16:10FromDiscord<System64 ~ Flandre Scarlet> @haxscramper Alright I found something very weird↵with ``--stacktrace:off --linetrace:off``, no crash, if I remove theses arguments, crashes
12:16:48FromDiscord<haxscramper> Crashes as in `SEGFAULT`, or you just get an exception?
12:17:19FromDiscord<System64 ~ Flandre Scarlet> ``Error: execution of an external program failed: 'C:\Users\nicol\Documents\programmation\testNim\newTest.exe '``
12:17:24FromDiscord<System64 ~ Flandre Scarlet> just that
12:18:52FromDiscord<System64 ~ Flandre Scarlet> and sometime the ``SIGSEGV: Illegal storage access. (Attempt to read from nil?)`` error
12:22:57FromDiscord<System64 ~ Flandre Scarlet> ``nim compile -r --nimcache:"./cache" --stacktrace:off --linetrace:off newTest.nim`` keeps runtime checks right?
12:24:06FromDiscord<haxscramper> yes, but C library does not have any
12:24:14FromDiscord<haxscramper> tilengine
12:24:34FromDiscord<System64 ~ Flandre Scarlet> yeah? what happens with Tilengine?
12:25:01*rockcavera joined #nim
12:26:04*auxym_ quit (Ping timeout: 268 seconds)
12:26:34FromDiscord<haxscramper> nothing
12:26:46FromDiscord<haxscramper> you use dynamic C dll with no runtime checks
12:26:57FromDiscord<haxscramper> Therefore it won't magically add runtime checks
12:31:34FromDiscord<System64 ~ Flandre Scarlet> ah I'm conscious of that, but it's weird it crashes if I remove theses 2 args
12:35:23FromDiscord<System64 ~ Flandre Scarlet> But I like how well Nim interops with C
12:36:16FromDiscord<System64 ~ Flandre Scarlet> you can even import raw C code
12:38:36FromDiscord<System64 ~ Flandre Scarlet> what'ss the difference between stacktrace and linetrace?
12:42:35FromDiscord<haxscramper> `stacktrace` adds calls to `nimfr_` and `popFrame()` at the start/end of each function↵`linetrace` without stracktrace does not do anyting, but if both are enabled calls to `nimln_(` are added so C code could provide exception positions that correspond to the nim code
12:43:21FromDiscord<haxscramper> sent a code paste, see https://play.nim-lang.org/#ix=3vSB
12:44:18FromDiscord<System64 ~ Flandre Scarlet> Ah I see, but I dunno why they make my program crashing
12:49:03*arkurious joined #nim
12:51:06*auxym_ joined #nim
12:56:04*auxym_ quit (Read error: No route to host)
12:56:25*auxym_ joined #nim
13:06:20*auxym_ quit (Ping timeout: 272 seconds)
13:10:02*PMunch quit (Remote host closed the connection)
13:14:44*PMunch joined #nim
13:52:48FromDiscord<Rika> might be a timing thing
14:02:30FromDiscord<System64 ~ Flandre Scarlet> In reply to @Rika "might be a timing": wdym?
14:03:25FromDiscord<Rika> wait this isnt threaded so maybe a memory layout thing
14:03:28FromDiscord<Rika> ?
14:03:40FromDiscord<Rika> well its one of those "debug works release doesnt" kinda issues
14:04:19FromDiscord<System64 ~ Flandre Scarlet> ah so it's Nim's fault?
14:05:28PMunchI had a weird issue like that with using a crypto library to read Keepass files
14:05:40PMunchIt works perfectly in debug mode, but release mode breaks it
14:07:03*max22- joined #nim
14:07:16*PMunch quit (Quit: Leaving)
14:07:19FromDiscord<System64 ~ Flandre Scarlet> In reply to @PMunch "It works perfectly in": Ah let me try in release
14:10:07FromDiscord<Rika> In reply to @System64 "ah so it's Nim's": no, its just one of those odd bugs that happen in pretty much any lang maybe save for rust
14:10:21FromDiscord<Rika> no, im sure rust has these issues as well
14:10:39FromDiscord<System64 ~ Flandre Scarlet> but why this bug occurs?
14:10:56FromDiscord<System64 ~ Flandre Scarlet> and I use callback functions, it's supposed to be threaded no?
14:11:09FromDiscord<konsumlamm> "it"?
14:11:38FromDiscord<System64 ~ Flandre Scarlet> callbacks
14:16:39FromDiscord<konsumlamm> and wdym by "supposed to be threaded"?
14:23:51FromDiscord<System64 ~ Flandre Scarlet> executed in parallel
14:26:10FromDiscord<Yardanico> In reply to @System64 "and I use callback": if tilengine doesn't run the callbacks on the same thread as your nim program and you don't use arc/orc you'll need to call setupForeignThreadGC at the start of any callback you pass to tilengine
14:26:48FromDiscord<Yardanico> although nowadays i'd just recommend you to use orc
14:27:14FromDiscord<System64 ~ Flandre Scarlet> what's the difference between Arc and Orc?
14:27:25FromDiscord<Yardanico> orc is arc with a cycle collector (so it can deal with cycles)
14:27:53FromDiscord<Yardanico> https://nim-lang.org/blog/2020/10/15/introduction-to-arc-orc-in-nim.html↵https://nim-lang.org/blog/2020/12/08/introducing-orc.html
14:31:36FromDiscord<System64 ~ Flandre Scarlet> I'll try that
15:07:16FromDiscord<System64 ~ Flandre Scarlet> But why it works without tracestack and traceline ?
15:19:49FromDiscord<haxscramper> probably due to some strackframe context that is not properly handled when callback is executed by another thread
15:20:24*stkrdknmibalz quit (Quit: WeeChat 3.0.1)
15:37:53*Vladar quit (Remote host closed the connection)
15:38:56*KurtGodel joined #nim
16:02:27*Pyautogui joined #nim
16:18:15FromDiscord<System64 ~ Flandre Scarlet> It crashes with --gc:orc
16:28:45FromDiscord<Ayy Lmao> sent a code paste, see https://play.nim-lang.org/#ix=3vTW
16:28:58*Pyautogui quit (Quit: Connection closed)
16:29:47*Pyautogui joined #nim
16:30:19FromDiscord<Rika> use `WideCString`
16:32:03FromDiscord<Ayy Lmao> Interesting, didn't know about that type. How do I go about assigning a nim string to it? `var test: LPCWSTR = "Test"` doesn't work when I change it
16:32:37FromDiscord<Rika> `var test: LPCWSTR = newWideCString "Test"`?
16:33:23FromDiscord<Ayy Lmao> Sweet that seems to work. Thanks!
16:33:59FromDiscord<Ayy Lmao> Out of curiosity, would making a converter for that be considered a bad idea?
16:35:00FromDiscord<Rika> that's your call
16:36:22FromDiscord<Ayy Lmao> Also I've been wondering, there are some types that are imported with `system`, like `UncheckedArray` and `WideCString` that disobey the convention of nim builtin types starting with lower case. Is there a reason for this?
16:36:57*Pyautogui quit (Quit: Connection closed)
16:38:49FromDiscord<Rika> because theyre not "builtin" types
16:40:07FromDiscord<Ayy Lmao> So they are essentially made up of builtin types instead of being builtin themselves?
16:40:56FromDiscord<Ayy Lmao> Oh I found it in the nim source as `WideCString = ptr UncheckedArray[Utf16Char]`
16:41:00FromDiscord<Ayy Lmao> So that makes sense
17:18:16madpropsdoes nim has that "feature" of rust where it caches stuff after builds, making some directories huge?
17:18:30madpropshave to do 'cargo clean' after a while
17:18:35madpropsor it grows to gigabytes
17:19:07madpropsso far it seems only rust projects were huge like that
17:19:30FromDiscord<Rika> nim binaries are not characteristically large so even if it did it probably wouldnt grow as large
17:19:47FromDiscord<Rika> i believe the cache actually gets wiped periodically
17:21:40FromDiscord<islem> hi
17:21:51FromDiscord<rxx> @islem @islem hi cupcakes
17:22:04FromDiscord<islem> you know were in a professional discord server, you should meme here
17:22:11FromDiscord<islem> (edit) "were" => "we're"
17:22:17FromDiscord<islem> (edit) "should" => "shouldn't"
17:22:18FromDiscord<rxx> What
17:22:23FromDiscord<rxx> Idc
17:22:43FromDiscord<islem> ok lemme just free() your life real quick
17:25:08FromDiscord<Rika> okay
17:25:12FromDiscord<Gabryx86_64> In reply to @islem "you know we're in": he did in the osdev server too
17:25:46FromDiscord<rxx> Hi cupcakes @Gabryx86_64 😘 😘 😘 🥰
17:25:56FromDiscord<Gabryx86_64> please stop
17:27:25FromDiscord<islem> oh god
17:40:38*auxym_ joined #nim
17:48:40FromDiscord<haxscramper> `Error: typechecked nodes may not be modified` how can I get `typed` proc in the pragma, and then change its implementation?
17:49:15FromDiscord<haxscramper> Should I just `copyNimTree` whole implementation or there is simpler way around this?
18:05:53*lain quit (Ping timeout: 248 seconds)
18:07:06*lain joined #nim
18:19:13FromDiscord<Sabena Sema> even with experimental:views I can't have a view type as a subobject of a record right?
18:21:11madpropsis 1.4.0 still the stable version?
18:21:36FromDiscord<haxscramper> 1.4.8
18:21:46FromDiscord<haxscramper> I think you should be able to↵(@Sabena Sema)
18:21:50madpropshmm if I do 'choosenim stable' it says it's already at stable
18:21:58FromDiscord<haxscramper> (when they are ready)
18:22:18FromDiscord<Sabena Sema> and if I do I can't rebind the view type after creating said record? (I can't reassign to the variable)
18:23:21FromDiscord<haxscramper> `choosenim update stable`↵(<@709044657232936960_madprops=5b=49=52=43=5d>)
18:23:23FromDiscord<haxscramper> I'm not sure what you mean. If you have `field: openarray[T]` then you would be able to assign to it, like any other field
18:23:47FromDiscord<haxscramper> `openarray[T]` would be a start/end pointer, so nothing prevents you from assigning to it
18:23:50madpropsoh I needed 'choosenim update stable'
18:26:29FromDiscord<haxscramper> Actually it works now
18:26:32FromDiscord<haxscramper> sent a code paste, see https://play.nim-lang.org/#ix=3vUx
18:27:06FromDiscord<haxscramper> Not with `var openarray[int]` ` error: request for member ‘Field0’ in something not a structure or union test_prog_13.oa.Field0 = (NI)s1_prog_7->data+(((NI) 0)); test_prog_13.oa.Field1 = (((NI) 2))-(((NI) 0))+1;`
18:27:37FromDiscord<Elegantbeef> try with `--newRuntime` or `arc/orc`
18:27:55FromDiscord<Elegantbeef> I know in my experience toying with them the old runtime did not add enjoyment
18:28:53FromDiscord<haxscramper> `nim r --newRuntime --gc:arc test.nim` same issue
18:30:01FromDiscord<haxscramper> sent a code paste, see https://play.nim-lang.org/#ix=3vUz
18:30:03FromDiscord<haxscramper> But not object with a variable
18:30:20FromDiscord<haxscramper> Fails with codegen
18:30:23FromDiscord<haxscramper> sent a code paste, see https://play.nim-lang.org/#ix=3vUA
18:30:40FromDiscord<haxscramper> Probably a bug, and I think I've seen this somewhere
18:31:51FromDiscord<haxscramper> nope, at least nothing tagged as "view types"
18:34:19FromDiscord<undersquire> is it possible to `raise` values that are not objects of Exception ?
18:35:09FromDiscord<leorize> nope
18:35:18FromDiscord<undersquire> i wish it supported that
18:35:32FromDiscord<undersquire> it would be nice to be able to just raise any value
18:35:32*max22- quit (Ping timeout: 245 seconds)
18:35:48FromDiscord<haxscramper> You can write own exception
18:36:01FromDiscord<undersquire> or something
18:36:03FromDiscord<haxscramper> But it is a horrible idea to raise non-exception IMO
18:36:09FromDiscord<undersquire> sent a code paste, see https://paste.rs/AeP
18:36:27FromDiscord<haxscramper> And strings are no reference types, so
18:36:40FromDiscord<undersquire> wait it has to be a reference type
18:36:41FromDiscord<undersquire> (edit) "type" => "type?"
18:37:15FromDiscord<undersquire> wait so how do you obtain data from an exception type?
18:37:32FromDiscord<haxscramper> just like any other reference type?↵(@undersquire)
18:37:42FromDiscord<Elegantbeef> `except YourException as e`
18:37:44FromDiscord<haxscramper> reference type only means it is allocated on heap
18:37:48FromDiscord<undersquire> In reply to @Elegantbeef "`except YourException as e`": ah ty
18:37:53FromDiscord<haxscramper> and behaves like it is a pointer/reference
18:38:11FromDiscord<haxscramper> Nim string buffer is allocated on heap, but it does not behave like a reference
18:38:13FromDiscord<undersquire> In reply to @haxscramper "reference type only means": yeah i meant how do u get it from the `except` statement, but beef answered that
18:38:23FromDiscord<Elegantbeef> Explained here https://nim-lang.org/docs/manual.html#exception-handling-except-clauses
18:38:34FromDiscord<undersquire> so you can technically store any values in your exception
18:38:54FromDiscord<undersquire> thats all I wanted to be able to do, I thought you could only store a string with an exception
18:39:55FromDiscord<haxscramper> you can create a `VoidStarError` which has `pointer: payload`
18:40:17FromDiscord<haxscramper> Basically exceptions are `ref` to guarantee they outlive stack unwinding. I think
18:40:21FromDiscord<undersquire> sent a code paste, see https://play.nim-lang.org/#ix=3vUE
18:40:36FromDiscord<haxscramper> Just `object of CatchableError` is enoug↵(@undersquire)
18:40:39FromDiscord<undersquire> oh ok
18:40:48FromDiscord<haxscramper> For some veird reason `newException` requires just `object`
18:40:48FromDiscord<haxscramper> idk why
18:41:11FromDiscord<haxscramper> But internally it does `(ref Exception)(msg: "....")`
18:41:20FromDiscord<haxscramper> I personally write constructors for my exceptions, so it goes like `raise newArgumentError( ... )`
18:41:38FromDiscord<undersquire> ah ok
18:41:44FromDiscord<undersquire> ty this helps a lot
18:42:08FromDiscord<haxscramper> Oh, and your code is not fully valid as `newException` accepts second argument as exception message
18:42:15FromDiscord<haxscramper> So `10` is not going to go into `errorcode`
18:43:12FromDiscord<haxscramper> sent a code paste, see https://play.nim-lang.org/#ix=3vUF
18:44:31FromDiscord<undersquire> yeah
18:44:42FromDiscord<undersquire> sent a code paste, see https://play.nim-lang.org/#ix=3vUI
18:44:59FromDiscord<undersquire> oh you can just do ref MyException () ?
18:45:49FromDiscord<undersquire> sent a code paste, see https://play.nim-lang.org/#ix=3vUJ
19:05:10*max22- joined #nim
19:28:01*auxym_ quit (Ping timeout: 248 seconds)
20:00:38*auxym_ joined #nim
20:01:38*Yardanico joined #nim
20:12:00*NimEventer joined #nim
20:12:29FromDiscord<Yardanico> moved eventer bot to libera chat
20:14:57*auxym_ quit (Ping timeout: 248 seconds)
20:48:43madpropsEditor shows an error but it still compiles https://i.imgur.com/M7mVd6r.png
20:58:07FromDiscord<Yardanico> how are you compiling it?
20:58:12FromDiscord<Yardanico> i mean the command
21:02:29madpropsnim js file.nim
21:02:54madpropsah i guess it's just not an error for js
21:04:53FromDiscord<Yardanico> you might try to specify --backend:js in your .nim.cfg or .nims file but I'm not certain it'll work
21:20:48*Lord_Nightmare quit (Quit: ZNC - http://znc.in)
21:24:29*Lord_Nightmare joined #nim
22:08:54FromDiscord<retkid> i know the answer is no but
22:09:11FromDiscord<retkid> do you think theres any known way to bundle efficient c binaries into a jvm enviroment
22:09:15FromDiscord<retkid> (edit) "enviroment" => "environment"
22:09:57FromDiscord<retkid> I would like to integrate nim into some java stuff because im afraid physics calculations might be too expensive for java
22:19:15FromDiscord<retkid> yea no
22:19:19FromDiscord<retkid> this is currently not a thing
22:19:20FromDiscord<retkid> :\
22:24:49*max22- quit (Quit: Leaving)
22:27:25FromDiscord<undersquire> In reply to @retkid "do you think theres": it is possible
22:27:35FromDiscord<undersquire> take a look at JNI
22:27:56FromDiscord<retkid> but idk if that would help preformance at all
22:27:56FromDiscord<treeform> In reply to @retkid "do you think theres": Don't ask me why I looked into it, regular java api looked really hard but I found this https://github.com/jnr/jnr-ffi
22:27:58nrds<prestigebot99> itHub: 7"Java Abstracted Foreign Function Layer"
22:27:59FromDiscord<retkid> (edit) "preformance" => "performance"
22:28:44FromDiscord<treeform> Android Java has the NDK which feels better.
22:29:15FromDiscord<retkid> Well, the only reason i would do this is for making my code run as fast as i can
22:29:15FromDiscord<treeform> Desktop java JNI looks kind of hard
22:29:18FromDiscord<retkid> but
22:29:25FromDiscord<retkid> I don't think this would help at all
22:29:35FromDiscord<undersquire> In reply to @retkid "Well, the only reason": just dont use Java then 😉
22:29:51FromDiscord<retkid> In reply to @undersquire "just dont use Java": CANT
22:29:53FromDiscord<treeform> In my experience java is ok at speed, depending how you write it.
22:29:53FromDiscord<undersquire> Java itself is not that slow
22:30:12FromDiscord<retkid> so my dumb idea requires a ton of java work
22:30:12FromDiscord<undersquire> you just have to avoid the GC as much as possible to avoid serious performance drops
22:30:27FromDiscord<undersquire> In reply to @retkid "so my dumb idea": what are you trying to make that needs Java (or is it an assignment)?
22:30:28FromDiscord<treeform> My ex-coworkers where very good at writing java-xml-java-deeply nested monsters that ran pretty slow.
22:30:33FromDiscord<treeform> so don't do that and your java should be fast
22:30:39FromDiscord<retkid> In reply to @undersquire "what are you trying": im doing some physics stuff in minecraftr
22:30:40FromDiscord<retkid> (edit) "minecraftr" => "minecraft"
22:30:42FromDiscord<undersquire> Ah
22:30:49FromDiscord<retkid> involving fluid dynamics and stuff
22:30:52FromDiscord<undersquire> Java is fine for it, just do the physics in java
22:31:10FromDiscord<retkid> well im afraid all of the calculations will just make it run like molasis
22:31:23FromDiscord<retkid> the game im basing it off runs like bytecode and still struggles
22:31:25FromDiscord<undersquire> 😂 trust me Java is really good at optimizing stuff
22:31:30FromDiscord<undersquire> math should be fine
22:32:04FromDiscord<undersquire> it probably isnt worth the effort to try to use C with Java (trust me it can be a pain)
22:32:07FromDiscord<treeform> The key to making simulation stuff in java is to avoid OOP and use flat arrays as much as possible. SOA vs AOS
22:32:11FromDiscord<undersquire> ^
22:32:24FromDiscord<undersquire> this helps avoid the GC
22:32:26FromDiscord<retkid> In reply to @treeform "The key to making": im learing to ue hashmaps
22:32:29FromDiscord<treeform> But that actually a key in any language 🙂
22:32:31FromDiscord<retkid> (edit) "ue" => "use"
22:32:34FromDiscord<undersquire> ye
22:33:35FromDiscord<retkid> for some reason hasmaps run like 186x faster
22:33:40FromDiscord<retkid> (edit) "hasmaps" => "hashmaps"
22:33:41FromDiscord<treeform> Many java (and non Java) games allocate most things upfront and have some thing used in-use and not-in-use flags, to get around GC.
22:34:06FromDiscord<treeform> So there is no allocations per frame.
22:34:11NimEventerNew Nimble package! tkrzw - Wrappers over the Tkrzw Database Manager C++ library., see https://git.sr.ht/~ehmry/nim-tkrzw
22:34:32*auxym_ joined #nim
22:34:34FromDiscord<retkid> with a game like minecraft where you might break a hole in a wall
22:34:42FromDiscord<retkid> and you gotta depressurize a part of a space station
22:34:49FromDiscord<retkid> you cant really prebake stuff
22:35:37FromDiscord<retkid> and you can obviously do generalizations, like if a large area of 0 pressure encounters a small area of preasure, just delete the area with preasureb
22:35:45FromDiscord<retkid> (edit) "preasureb" => "preasure"
22:36:09FromDiscord<retkid> Idk, i've been procrastinating starting this because its kinda scary to me lol
22:36:12FromDiscord<treeform> Is that minecraft or ss13?
22:36:22FromDiscord<retkid> In reply to @treeform "Is that minecraft or": porting ss13 to minecraft for no reason
22:36:31FromDiscord<retkid> at least thats my goal
22:36:39FromDiscord<treeform> ss13 has two models handling pressures
22:36:47FromDiscord<treeform> one is pre "voxel"
22:36:50FromDiscord<retkid> I might start digging around in gonecode
22:36:55FromDiscord<retkid> (edit) "gonecode" => "gooncode"
22:37:21FromDiscord<treeform> and one is per "region" where regions are created and destroyed and have nodes telling you how much goes from one to the other.
22:37:35FromDiscord<treeform> region is much faster and much more predictable
22:37:47FromDiscord<treeform> voxel is super slow, but can be more "fun".
22:38:06FromDiscord<treeform> Different servers choose one or the other and that sets up the mood.
22:38:33FromDiscord<retkid> so, my idea is currently that all sealed areas will be regions until something else is found in them
22:38:38FromDiscord<retkid> then they split into voxels or something
22:38:55FromDiscord<retkid> i need to start writing the code to get a solid understaning of it
22:38:57FromDiscord<retkid> (edit) "understaning" => "understanding"
22:38:59FromDiscord<treeform> yeah
22:39:05FromDiscord<retkid> i also have to polish up my fluid dynamics stuff
22:39:07FromDiscord<treeform> Minecraft is also 3d
22:39:10FromDiscord<treeform> its like much harder
22:39:21FromDiscord<treeform> if you like ss13 checkout my blog post: https://blog.istrolid.com/blog/space-station-13-remakes.html
22:39:27FromDiscord<retkid> 1^2 vs 1^3
22:39:31FromDiscord<retkid> exponentially harder
22:40:09FromDiscord<treeform> You probably want regions, some sort of quad tree
22:40:10*flynn quit (Read error: Connection reset by peer)
22:40:19FromDiscord<treeform> then breakout into voxels when its interesting
22:40:28FromDiscord<retkid> so what im doing is, im never announcing anything to anyone, im just gonna work and see what happens
22:40:30FromDiscord<treeform> door or a hole opened up
22:40:41FromDiscord<retkid> if i announce anything publicly it'll be a finished or semi-finished product
22:40:48FromDiscord<retkid> but it may just go nowhere
22:40:51FromDiscord<retkid> /sh\
22:40:53FromDiscord<retkid> (edit) "/sh\" => "¯\_(ツ)_/¯"
22:41:17*flynn joined #nim
22:41:23FromDiscord<treeform> You have seen: https://minecraft-spacestation13.weebly.com/
22:41:30FromDiscord<retkid> I have, im pretty sure
22:41:38FromDiscord<retkid> that one got abandoned
22:41:54FromDiscord<retkid> they have a github of unusable code because forge sucks
22:42:19FromDiscord<retkid> https://github.com/birdbrainswagtrain/MineStation13
22:42:21nrds<prestigebot99> itHub: 7"A Minecraft mod intended to replicate the popular BYOND game, Space Station 13."
22:42:53FromDiscord<retkid> I have other people who may or may not help me so im doing it all alone
22:55:04*auxym_ quit (Ping timeout: 258 seconds)
22:57:11FromDiscord<retkid> :nim2:
23:05:29*auxym_ joined #nim
23:06:00FromDiscord<dain> sent a code paste, see https://play.nim-lang.org/#ix=3vW2
23:06:10*auxym_ quit (Client Quit)
23:08:07FromDiscord<dain> like i want to parse the json without having to write it all in a busy nested line, but break it up into blocks like they do in that section
23:08:16FromDiscord<dain> (edit) "parse" => "deal with"
23:16:46*Pyautogui joined #nim
23:21:19PyautoguiIs it good Nim to use references like you would use pointers in C? Like https://play.nim-lang.org/#ix=3vWy.
23:22:04FromDiscord<Elegantbeef> You use references when you want references, if you want a pointer you use a pointer
23:22:12FromDiscord<dain> sent a code paste, see https://play.nim-lang.org/#ix=3vWz
23:22:22FromDiscord<Elegantbeef> You also dont need to use cast for simple conversions
23:22:30PyautoguiGot it.
23:33:42*Pyautogui quit (Quit: Connection closed)
23:48:34FromDiscord<dain> sent a code paste, see https://play.nim-lang.org/#ix=3vWD
23:51:57FromDiscord<haxscramper> Not right now, kv matching requires {} syntax
23:52:39FromDiscord<dain> oh :(
23:52:53FromDiscord<haxscramper> So for json you can only use it
23:53:02FromDiscord<dain> oh btw, are the docs correct here? https://nim-lang.github.io/fusion/src/fusion/matching.html#matching-different-things-kvminuspairs-matching
23:53:08FromDiscord<dain> there is no "file" key in the json
23:54:51FromDiscord<haxscramper> It does not show output, so technically it is a correct example as it only describes syntax. But this should be clarified, yes
23:55:02FromDiscord<haxscramper> And added to tests