<< 03-07-2025 >>
00:14:07*xtr00 quit (Ping timeout: 244 seconds)
00:29:39*beholders_eye quit (Ping timeout: 276 seconds)
00:31:20*xtr00 joined #nim
01:11:24*jjido quit (Quit: My laptop has gone to sleep. ZZZzzz…)
01:16:28FromDiscord<.tokyovigilante> sent a code paste, see https://play.nim-lang.org/#pasty=gMOsTpHX
01:17:24FromDiscord<Elegantbeef> If they're not getting the value set, nor have `{.importc.}` the value isn't going to populate
01:26:06FromDiscord<.tokyovigilante> I'm compiling `volk.c` which ultimately includes `vulkan_core.h` (which has both declaration and definition as above) as part of my build, but do I still need to have `{.importc.}` in the generated Nim wrapper?
01:26:56FromDiscord<Elegantbeef> It does if you're not getting a `var a = b`
01:28:06FromDiscord<.tokyovigilante> It does need an `importc` you mean? I'm not sure if this is how it should work, but looking at that C include, I'd hope Futhark would generate `const VK_PIPELINE_STAGE_2_COLOR_ATTACHMENT_OUTPUT_BIT_KHR: VkPipelineStageFlagBits2 = 0x00000400.uint64`.
01:37:06FromDiscord<.tokyovigilante> sent a code paste, see https://play.nim-lang.org/#pasty=bIuhgbsj
01:37:21FromDiscord<.tokyovigilante> with a `{.importc.}` so I may have bigger problems.
01:40:09FromDiscord<.tokyovigilante> Also relevant - https://github.com/PMunch/futhark/issues/144
01:40:20FromDiscord<Elegantbeef> I said it does not
01:40:21FromDiscord<Elegantbeef> It's either or
01:40:35FromDiscord<Elegantbeef> It's supposed to generate `cosnt name = value`
01:50:45*xtr00 quit (Read error: Connection reset by peer)
01:56:41*xtr00 joined #nim
02:13:06*thunder quit (Remote host closed the connection)
02:13:40*thunder joined #nim
02:17:57FromDiscord<.tokyovigilante> Sorry I don't follow, but I feel like we're agreeing this is a Futhark bug?
02:18:04FromDiscord<Elegantbeef> Yes
03:31:27*Goodbye_Vincent1 joined #nim
03:48:30*rockcavera quit (Remote host closed the connection)
04:36:24*thunder quit (Killed (NickServ (GHOST command used by metasintactic)))
04:36:31*thunder joined #nim
04:45:31*thunder quit (Ping timeout: 244 seconds)
04:47:48*amadaluzia quit (Quit: ZNC 1.10.0 - https://znc.in)
04:52:56*thunder joined #nim
04:59:59*thunder quit (Ping timeout: 244 seconds)
06:06:17*skippy8 joined #nim
06:10:01*nils` quit (Quit: nils`)
06:28:51*alexdaguy joined #nim
06:31:13*derpydoo joined #nim
07:19:44*FromDiscord quit (Ping timeout: 265 seconds)
07:20:13*FromDiscord joined #nim
07:30:03*thunder joined #nim
07:52:18*thunder quit (Remote host closed the connection)
07:53:14*thunder joined #nim
07:53:31*thunder quit (Client Quit)
07:59:06*derpydoo quit (Quit: derpydoo)
08:13:39*lucerne quit (Ping timeout: 260 seconds)
08:21:55*lucerne joined #nim
08:30:26*beholders_eye joined #nim
08:39:59FromDiscord<tauruuuuuus> sent a code paste, see https://play.nim-lang.org/#pasty=dQkeXfat
08:43:59*tokyovigilante quit (Read error: Connection reset by peer)
08:44:14*tokyovigilante joined #nim
08:45:53Amun-Ratauruuuuuus: he can't use it in that case
08:46:10Amun-Ra(uint32)
08:46:58Amun-Rahmm, would HashSet counterpart work in that case?
08:48:42FromDiscord<TӨMΛ ☠> In reply to @tauruuuuuus "FYI nim supports bitfields": Oooh, that can be handy, thank you 🙏
08:50:16Amun-Rait works with types of size up to 16-bits
08:50:29Amun-Raand it won't work on JS target
08:50:40FromDiscord<TӨMΛ ☠> Oh shoot, somewhat missed that response
08:50:55FromDiscord<TӨMΛ ☠> I target C, but it still sucks kind of
08:51:03Amun-Rayou may test it with HashSet from std/sets
08:51:38FromDiscord<TӨMΛ ☠> I will try 🫡
08:51:49Amun-Ra:>
08:59:05FromDiscord<tauruuuuuus> sent a code paste, see https://play.nim-lang.org/#pasty=NjpbLSOs
09:00:03*xet7 joined #nim
09:00:11Amun-Ratauruuuuuus: set a value of > 0xffff
09:00:57Amun-Ratauruuuuuus: https://play.nim-lang.org/#pasty=liRsIcif
09:06:26FromDiscord<tauruuuuuus> TIL
09:07:19Amun-Rabeen there before ;)
09:07:21FromDiscord<tauruuuuuus> But this is not a issue if using plain enums and set bitfields instead of assigning pow 2 values to the enum entrie itself, that was my point
09:07:38FromDiscord<tauruuuuuus> (edit) "entrie itself," => "entries themselves,"
09:07:55Amun-Ratrue, but in this case it has to have specific values
09:07:58FromDiscord<lainlaylie> dont we just need the set to fit in a uint32, i.e. ensure the enum has 32 elements or fewer, which is the case here?
09:14:19Amun-RaI usually do that using distinct integer, as in: https://play.nim-lang.org/#pasty=bSPLpoKI
09:17:02Amun-Rabugfixed* https://play.nim-lang.org/#pasty=sVbatAti
09:17:44Amun-RaI have to have specific flag values, as defined in a given format I parse
09:22:48Amun-Rahmm, I can use `in` instead of just `and`, it's prettier: https://play.nim-lang.org/#pasty=ZsgGPRpF
09:35:18*jjido joined #nim
09:58:20*jjido quit (Quit: My laptop has gone to sleep. ZZZzzz…)
10:55:04*jjido joined #nim
11:41:41*jjido quit (Quit: My laptop has gone to sleep. ZZZzzz…)
11:52:10*jjido joined #nim
11:52:39*beholders_eye quit (Ping timeout: 244 seconds)
12:00:30*jjido quit (Quit: My laptop has gone to sleep. ZZZzzz…)
12:21:47*fallback quit (Quit: IRCNow and Forever!)
12:38:07*fallback joined #nim
12:50:24*beholders_eye joined #nim
13:23:43*nils` joined #nim
13:39:58*beholders_eye quit (Ping timeout: 276 seconds)
13:53:52FromDiscord<alehander92> > generic methods are deprecated
13:54:00FromDiscord<alehander92> is this valid for nim 2 as well?
13:54:12FromDiscord<alehander92> how do people workaround it, or is it just in some problematic cases?
14:28:56*amadaluzia joined #nim
14:29:47*amadaluzia quit (Client Quit)
14:30:55*amadaluzia joined #nim
14:31:59*jjido joined #nim
14:43:20*jjido quit (Quit: My laptop has gone to sleep. ZZZzzz…)
14:52:07*jjido joined #nim
15:15:44*jjido quit (Quit: My laptop has gone to sleep. ZZZzzz…)
15:30:07*przmk4 quit (Quit: The Lounge - https://thelounge.chat)
15:30:45*przmk4 joined #nim
15:31:02FromDiscord<Phil> In reply to @alehander92 "> generic methods are": By not having generic methods. ↵In rxnim where I provided the base implementation for subjects and observables I assign a closure to a field on the object. ↵↵That was the only time I ever needed generic methods, other patterns usually work better
15:31:53*alexdaguy quit (Quit: w)
16:08:29FromDiscord<alehander92> i wanted to impl mediator pattern
16:08:44FromDiscord<alehander92> e.g. things like `method emit[T](kind: Event, value: T, ..) = ..`
16:08:52FromDiscord<alehander92> which can be overriden eventually
16:09:16FromDiscord<alehander92> for now indeed i decided not to mix them, and just have a single type with all the fields/flags needed currently
16:09:52FromDiscord<alehander92> but for some other types i need inheritance because e.g. i need some types to have different fields which are available only in some cases (with different logic)
16:10:00FromDiscord<alehander92> (edit) "cases" => "cases/environments"
16:10:23FromDiscord<alehander92> for now in the other cases, i made them non-generic (using `JsObject` for the values)
16:10:48FromDiscord<alehander92> but if i need generic-ness i wonder about using e.g. conditional compilation instead and replacing the type itself
16:11:07FromDiscord<alehander92> (edit) "itself" => "and some of the procs(methods) instead"
16:11:29FromDiscord<alehander92> i do think there are valid usecases, even if i usually try to not use inheritance as well
16:12:02FromDiscord<alehander92> maybe a generic closure as a field is also ok, but this means the whole object must become generic
16:12:20FromDiscord<alehander92> and if you have various possible methods/ways to call it, it wouldn't be ok
16:13:39FromDiscord<alehander92> ( reply to @Phil thanks for the rxnim example )
16:19:32*beholders_eye joined #nim
17:20:57*nils` quit (Quit: nils`)
17:25:10*xet7 quit (Remote host closed the connection)
17:31:01*beholders_eye quit (Ping timeout: 248 seconds)
17:43:38*nils` joined #nim
18:03:45*jjido joined #nim
19:29:04*jjido quit (Quit: My laptop has gone to sleep. ZZZzzz…)
19:30:42*thunder joined #nim
20:18:02*jjido joined #nim
20:27:21FromDiscord<morgan (ping with reply)> sent a code paste, see https://play.nim-lang.org/#pasty=NrvXyVva
20:34:16FromDiscord<leorize> because converters are a buggy mess with no defined semantics
20:34:31FromDiscord<leorize> it is matched wherever it pleases
20:35:14FromDiscord<leorize> I'd recommend not using converters and especially not exporting them if possible
20:35:19FromDiscord<Elegantbeef> Hey the semantics are fine here
20:35:31FromDiscord<Elegantbeef> Converters are not fired for converters
20:35:47FromDiscord<Elegantbeef> openarray requires a converter so you need another converter
20:38:20FromDiscord<Elegantbeef> Just make a `ByteMapping.init(oa: openArray[int])` and call it a day
20:56:22FromDiscord<morgan (ping with reply)> interesting
20:58:20FromDiscord<morgan (ping with reply)> i had initially coded this with arrays and had bytemapping be generic, but decided that was going to be more of a problem than a benefit and today converted their code all over to uncheckedarrays and seqs
20:58:54FromDiscord<morgan (ping with reply)> having to take in a seq specifically for this way of constructing one is fine
21:12:45*xet7 joined #nim
21:14:35*jjido quit (Quit: My laptop has gone to sleep. ZZZzzz…)
21:17:46*thunder quit (Killed (NickServ (GHOST command used by metasintactic)))
21:17:46*metasintactic joined #nim
21:28:05*metasintactic is now known as thunder
21:37:14*beholders_eye joined #nim
21:46:59*xet7 quit (Remote host closed the connection)
21:58:42*beholders_eye quit (Ping timeout: 248 seconds)
21:59:27*skippy8 quit (Ping timeout: 252 seconds)
22:17:00*jjido joined #nim
22:24:51FromDiscord<demotomohiro> In reply to @alehander92 "> generic methods are": It seems generic methods are hard to implement correctly.↵Even C++ give up supporting virtual template function (C++ feature similar to Nim's generic methods).↵https://stackoverflow.com/questions/2354210/can-a-class-member-function-template-be-virtual
23:36:46FromDiscord<alehander92> In reply to @demotomohiro "It seems generic methods": Interesting thanks