<< 12-01-2023 >>
00:00:41FromDiscord<pyolyokh> it's not magically empty, but it has the last value it was assigned in the loop, and the procs that are called all reference that same location. If your last col is "x" instead, they'll all be "x" instead of ""
00:01:39FromDiscord<pyolyokh> c.f. <https://nim-lang.org/docs/manual.html#closures-creating-closures-in-loops>
00:02:39FromDiscord<pyolyokh> sent a code paste, see https://play.nim-lang.org/#ix=4kWy
00:08:36FromDiscord<ted__> sent a long message, see http://ix.io/4kWv
00:09:03*ltriant joined #nim
00:33:47FromDiscord<Nilts> sent a code paste, see https://play.nim-lang.org/#ix=4kWE
00:36:56FromDiscord<Girvo> sent a long message, see https://paste.rs/3lf
00:38:04FromDiscord<Girvo> To add to that, it would be ideal if I could store the thread ID next to it as well, but that's less of an issue -- the threads are always unique/theres only ever one X thread, so I can store that in standard vars anyway
00:38:50FromDiscord<Elegantbeef> `set[MyThreadNames]`?
00:40:43FromDiscord<Girvo> Yeah thats basically the conclusion I came to
00:40:56FromDiscord<Elegantbeef> So why did you mention `std/sets`?
00:41:18FromDiscord<Girvo> Because I have the docs open right now and need another coffee? šŸ˜„
00:44:13FromDiscord<Girvo> Built-in set doesn't require unique/only-one-of-this-enum hey? Not a great stress, the way the config handling works it shouldn't be an issue
00:45:46FromDiscord<Elegantbeef> It's a bitset for the ordinal range
00:48:29FromDiscord<Girvo> šŸ‘ sweet
00:52:35*ltriant quit (Ping timeout: 255 seconds)
00:52:47FromDiscord<Girvo> sent a code paste, see https://paste.rs/emL
00:56:55FromDiscord<Elegantbeef> Without any hashing or similar! šŸ˜›
01:00:58*jjido quit (Quit: My laptop has gone to sleep. ZZZzzzā€¦)
01:12:43FromDiscord<jos> i just ran into an issue where i'm passing around a lot of `type X = pointer` types and i had the order swapped in an argument list
01:12:55FromDiscord<jos> two types defined like that are considered covariant by the nim compiler i guess
01:13:04FromDiscord<jos> so i was passing them in lieu of each other and it wasn't giving me an error, took me forever to figure out
01:13:11FromDiscord<jos> how can i avoid that
01:13:40FromDiscord<pyolyokh> `type X = distinct pointer`
01:14:36FromDiscord<jos> ok cool
01:14:44FromDiscord<jos> a lot of my c ffi dependencies aren't doing that
01:14:51FromDiscord<jos> šŸ˜¦
01:18:41FromDiscord<Elegantbeef> Or `type X = ptr object`
01:18:42FromDiscord<Elegantbeef> `type X = pointer` is just an alias
01:25:49FromDiscord<T0lk1en> Is there a reason why random would generate the same numbers over and over again?
01:27:34FromDiscord<Elegantbeef> You didnt call randomise
01:29:28FromDiscord<T0lk1en> Wdym
01:30:31FromDiscord<T0lk1en> Nvm
01:30:34FromDiscord<T0lk1en> Ty
01:30:50FromDiscord<Elegantbeef> Like did you even look at `std/random`'s manual?
01:37:08*crem quit (Ping timeout: 260 seconds)
02:05:11FromDiscord<jtv> If you're using it for crypto, it's going to be easily broken.
02:05:32FromDiscord<jtv> Instead, you want std/sysrand
02:16:44FromDiscord<!!sharpcdf!!> sent a code paste, see https://paste.rs/GOo
02:17:28FromDiscord<Elegantbeef> do you deinit illwill?
02:17:44FromDiscord<!!sharpcdf!!> at the end of my program yes i do
02:18:21FromDiscord<!!sharpcdf!!> oh wait not in all cases
02:23:06FromDiscord<!!sharpcdf!!> yea no it doesnt show the second prompt still
02:36:19*ltriant joined #nim
02:42:10FromDiscord<Elegantbeef> well the code you shared doesnt give much insight
03:12:34*crem joined #nim
03:55:12FromDiscord<T0lk1en> What is the best way to shift an array
03:55:26FromDiscord<T0lk1en> In reply to @Elegantbeef "Like did you even": Overlooked it
04:32:33FromDiscord<Gumbercules> In reply to @sharpcdf "yea no it doesnt": I'd get rid of `choosing` and instead do `while true` with `break`
04:33:26FromDiscord<Gumbercules> not that I think it will fix your problem, but then at least you know that `choosing` isn't being modified elsewhere
04:34:00FromDiscord<Gumbercules> unless you need to modify `choosing` outside of the `case` statement
04:34:45FromDiscord<Gumbercules> posting a more complete example would certainly help though, as beef pointed out, there's not much to go on
04:37:40FromDiscord<pyolyokh> In reply to @T0lk1en "What is the best": shift how? probably you want algorithm.rotateLeft or the like.ā†µbest how? if you're doing this a lot, and if you're rotating all the elements, the best way to use another data structure because this will be very costly.
04:38:21FromDiscord<T0lk1en> I want to shift array right. For example
04:38:39FromDiscord<T0lk1en> 12345 ā€”>
04:38:46FromDiscord<T0lk1en> 51234
04:40:11FromDiscord<pyolyokh> then, algorithm.
04:52:51FromDiscord<T0lk1en> Got it
04:59:34FromDiscord<Mustache Man> it's late and brain isn't working well. how should I delete a sequence of items, or known non-contiguous indices, from a sequence?
05:01:21FromDiscord<Mustache Man> sent a code paste, see https://play.nim-lang.org/#ix=4kX9
05:51:56*azimut quit (Ping timeout: 255 seconds)
06:10:48FromDiscord<iffy (Matt Haggard)> How do I truncate a file I'm writing to?
06:26:25FromDiscord<demotomohiro> @iffy Read and save the content of a file you want to keep, then open that file with `fmWrite` (this clear the file) and write the saved content.
06:29:50FromDiscord<demotomohiro> It seems there is no procedure that remove a part of file content. You can only clear file content by opening it with `fmWrite`.
06:48:21FromDiscord<Gumbercules> you could wrap posix's `truncate` / `ftruncate`
06:50:10FromDiscord<Gumbercules> sent a code paste, see https://play.nim-lang.org/#ix=
07:03:02*kenran joined #nim
07:20:18*kenran quit (Remote host closed the connection)
07:37:24*ltriant quit (Ping timeout: 260 seconds)
08:03:46*ltriant joined #nim
08:04:02FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW!
08:04:05FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:05FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:07FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:08FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW!
08:04:08FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:09FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:10FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW!
08:04:13FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:15FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:15FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:16FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:16FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:18FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:18FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:21FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:21FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:24FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:25FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:27FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:28FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:29FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyoneJOIN discord.gg/ yunjin NOW! @everyoneJOIN discord.gg/ yunjin NOW! @everyoneJOIN discord.gg/ yunjin NOW! @everyoneJOIN discord.gg/ yunjin NOW! @everyoneJOIN discord.gg/ yunjin NOW! @everyoneJOIN discord.gg/ yunjin NOW! @everyoneJOIN discord.gg/ yunjin NOW! @everyoneJOIN discord.gg/ yunjin NOW! @everyoneJOIN discord.gg/ yunjin NOW! @everyone
08:04:29FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @everyone
08:04:36FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @everyone
08:04:38FromDiscord<rĀ”n> JOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @not_everyoneeJOIN discord.gg/ yunjin NOW! @everyone
08:04:40FromDiscord<rĀ”n> sent a long message, see http://ix.io/4kXJ
08:04:45FromDiscord<rĀ”n> sent a long message, see http://ix.io/4kXJ
08:04:48FromDiscord<rĀ”n> sent a long message, see http://ix.io/4kXJ
08:04:48FromDiscord<rĀ”n> sent a long message, see http://ix.io/4kXJ
08:04:51FromDiscord<rĀ”n> sent a long message, see http://ix.io/4kXJ
08:04:51FromDiscord<rĀ”n> sent a long message, see http://ix.io/4kXK
08:04:52FromDiscord<rĀ”n> sent a long message, see https://paste.rs/s9L
08:04:52FromDiscord<rĀ”n> sent a long message, see http://ix.io/4kXJ
08:04:59FromDiscord<rĀ”n> sent a long message, see http://ix.io/4kXL
08:04:59FromDiscord<rĀ”n> sent a long message, see http://ix.io/4kXJ
08:05:00FromDiscord<rĀ”n> sent a long message, see http://ix.io/4kXJ
08:05:00FromDiscord<rĀ”n> sent a long message, see https://paste.rs/rAs
08:05:04FromDiscord<rĀ”n> sent a long message, see http://ix.io/4kXM
08:05:06FromDiscord<rĀ”n> sent a long message, see http://ix.io/4kXL
08:05:15FromDiscord<rĀ”n> who the hell is not everyonne
08:05:23FromDiscord<rĀ”n> idk
08:05:24FromDiscord<rĀ”n> sent a long message, see http://ix.io/4kXL
08:05:38FromDiscord<rĀ”n> sent a long message, see http://ix.io/4kXJ
08:05:48FromDiscord<ElegantBeef> <@&371760044473319454> spam isnt from matrix this time
08:05:53FromDiscord<ElegantBeef> Welp...
08:06:16FromDiscord<impbox [ftsf]> nice surprise when it's from discord and we can ban+delete easily
08:08:03FromDiscord<Elegantbeef> It's rare
08:08:17FromDiscord<Elegantbeef> Though matrix spam has gone down tremendously
08:21:12FromDiscord<Phil> Maybe enough of the bot-spammer-networks count our server as unprofitable enough to not be worth the CPU cycles
08:21:18FromDiscord<Phil> We can only hope
08:41:50*PMunch joined #nim
08:41:57*crem quit (Ping timeout: 255 seconds)
08:45:06*crem joined #nim
08:46:15FromDiscord<PMunch> Federico set up some automatic measures against spam as well.
09:12:28*cyraxjoe quit (Quit: No Ping reply in 180 seconds.)
09:13:45*cyraxjoe joined #nim
10:06:09FromDiscord<cow> more and more matrix home servers have added a verified email as a requirement for an account šŸ™‚
10:09:04FromDiscord<Elegantbeef> Dont tell the spammers they can host their own
10:10:10FromDiscord<cow> their homeserver will just get blacklisted
10:10:38FromDiscord<cow> and domain names cost money
10:11:05FromDiscord<cow> I think they'd rather get a million emails than a million domains
10:11:31FromDiscord<leetnewb> plenty of free ddns domains
10:12:35FromDiscord<cow> you can block the parent domain
10:12:55FromDiscord<cow> some email spam filters mark entire TLDs as spam
10:14:10FromDiscord<cow> especially that notorious one that asks you to pay to get unblocked
10:14:17FromDiscord<Elegantbeef> I'm not going to be pro spam, but ostensibly spamming has to be profitable enough to get a few $3 domains šŸ˜„
10:18:28FromDiscord<cow> then we're heading to a home server whitelist era soon šŸ˜”
10:18:42FromDiscord<cow> which would be sad
10:19:50FromDiscord<Elegantbeef> Would defeat the whole federated part
10:20:11FromDiscord<cow> doesn't mastodon already operate that way?
10:20:49FromDiscord<Elegantbeef> Sure but still feels like it defeats the point
10:20:53FromDiscord<cow> there's already a big blocklist of "bad" servers being maintained
10:22:13FromDiscord<cow> and extremism qualifies to get there
10:22:33FromDiscord<cow> so wouldn't be such a big step to make
10:22:57FromDiscord<cow> especially if good servers could share the list of good servers, and there would be an application form, where anyone could easily join
10:23:37FromDiscord<Elegantbeef> I think a blacklist and a whitelist are vastly different when it comes to federated communites
10:24:40FromDiscord<Elegantbeef> A vast majority of homeservers are not going to host large swathes of people that 'misbehave'
10:25:40FromDiscord<Elegantbeef> A small amount of homeservers will host the type of people most people would rather not interact with
10:26:41FromDiscord<Elegantbeef> I especially think mastadon is a bit different due to the fact everyone can see eachother
10:26:57FromDiscord<Elegantbeef> I mean there is a whole federated feed that you can see content
10:27:34FromDiscord<Elegantbeef> Anyway I really should put some effort into expanding the matrix sdk, just so much tedious work
10:37:02FromDiscord<federico3> please use #offtopic for this
11:16:07*azimut joined #nim
11:23:26FromDiscord<whosent> anyone else having the issue of choosenim downloading x32 distributions?
11:27:21FromDiscord<ShalokShalom> In reply to @cow "then we're heading to": Matrix.org makes up for the most accounts anyway
11:27:42FromDiscord<ShalokShalom> Sorry, #offtopic
11:31:50*jjido joined #nim
11:44:40Amun-Rais there nim template that returns the current proc/func name?
11:45:55FromDiscord<Phil> In reply to @Amun-Ra "is there nim template": I had a similar question a couple months ago, the answer was no back then
11:46:39FromDiscord<Phil> needed the name of the proc from inside the proc though I don't remember anymore for what purpose, wasn't doable
11:51:05FromDiscord<!!sharpcdf!!> In reply to @Gumbercules "I'd get rid of": OH WAIT
11:51:20FromDiscord<!!sharpcdf!!> CHOOSING IS A GLOBAL VARIABLE I DONT THINK ITS BEING RESET TO TRUE LMAO
11:53:48Amun-RaPhil: oh, thanks
12:00:10PMunch@Phil, well it is doable
12:00:20PMunchUse instantiationInfo and read the file manually
12:00:57PMunchYou can apparently also use some hacks: https://stackoverflow.com/questions/48686642/how-can-i-get-the-name-of-procedure-in-nim
12:09:19FromDiscord<Phil> Huh, wasn't aware of that one back then
12:15:26Amun-RaI ended up with this: https://play.nim-lang.org/#ix=4kYT
12:18:20Amun-RaI know there's logging module in std, but it lacks some features I needed
12:29:26PMunchWhat features would that be?
12:29:36PMunchAsking in case I ever get around to improve superlog
12:30:58Amun-RaI wanted another level of debugging, plus one of my target is non-posix non-windows one and I can't rely on std/os
12:31:12Amun-Ratargets*
12:32:03Amun-Raand I automagically add module:line to the log entry
12:32:47*ltriant quit (Ping timeout: 246 seconds)
12:33:03*ltriant joined #nim
12:33:52Amun-Raā€¦and I wanted every debug/info/trace/notice to have {.raises: [].}
12:43:07PMunchI see
12:43:26PMunchAll of that should be achievable with superlog
12:43:49FromDiscord<luteva> In reply to @PMunch "Basically if you set": ok as far as i understand, there should be a logfile, generated by nimlsp_debug. but i can't find it. where is it supposed to be?
12:44:31PMunchThe point of superlog is essentially that you can litter log statements everywhere, and log any type you'd like. Then in your main module you define what log statements should expand to, based on module and type
12:45:58PMunchSo e.g. your module could log a JsonNode and some tuples of LogLevel and string. Then in my module which uses your module I can say that only the log statements that log JsonNodes should actually do anything, and they should expand to whatever thing I want them to expand to (so you could have all your {.raises: [].} and os-aware stuff.
12:49:34Amun-Rahmm
12:50:07FromDiscord<iffy (Matt Haggard)> Okay, thanks. Guess it's not built-in to stdlib.ā†µ(@Gumbercules)
12:52:46PMunch@luteva, the logs are stored in `getTempDir() / "nimlsp-" & $getCurrentProcessId() & "/nimlsp.log"`
12:55:58FromDiscord<luteva> yes! that was exactly what i was searching for! thx
12:56:13FromDiscord<luteva> (edit) "yes! that was exactly what i was searching for! thx ... " added "@PMunch"
12:58:24FromDiscord<luteva> cool! can see it logging e.g. on hover šŸ™‚
12:58:52PMunchAmun-Ra, @Phil, here is some more stuff for getting the proc name: https://github.com/nim-lang/Nim/pull/17641
12:59:48PMunch@luteva, if you only want the communications going back and forth you can build manually with -d:debugCommunications
13:00:19PMunch(It's recommended to still use -d:release otherwise it's quite slow)
13:03:20*jjido quit (Quit: My laptop has gone to sleep. ZZZzzzā€¦)
13:21:43Amun-RaPMunch: great find, it works
13:23:55FromDiscord<luteva> thx again @PMunch i can now do, what i wanted to!ā†µI have one more question about the lsp: My understanding is that the editor and the lsp server open the file and they both keep a separate copy of the file in a buffer. The editor then sends each update, made to the file, directly the the lsp server which updates its own buffer. This is how the editor and the lsp server keep themselves up-to-date.ā†µIs that correct?
14:10:04PMunch@luteva, not quite. The editor opens the file, then sends a notification to the LSP server with the entire contents of the file
14:10:38PMunchThen the editor sends edited versions of the file at whichever interval it sees fit
14:10:57PMunchSo the server doesn't really read the input file at all
14:11:29PMunchOf course it still serves you the paths and such, so that the server can find imports and other things like that
14:35:52FromDiscord<luteva> but it only sends the changes, right? or does it send the whole file (including the changes)?
14:42:12PMunch@luteva, https://microsoft.github.io/language-server-protocol/specifications/lsp/3.17/specification/#textDocument_synchronization
14:42:26PMunchNimLSP uses TextocumentSyncKind.Full
14:47:51FromDiscord<luteva> ok. i see. there is a "range and text" OR a (full) "text only".
14:48:40*PMunch quit (Quit: Leaving)
15:11:23*arkurious joined #nim
15:57:47*kenran joined #nim
16:00:36*kenran quit (Remote host closed the connection)
16:16:16FromDiscord<pyryrin> if the rule is whenever you can use a proc instead of template, then why is `assert` a template?
16:17:46FromDiscord<Rika> because it needs the template?
16:18:01FromDiscord<pyryrin> why?
16:19:34FromDiscord<Rika> because it uses instantiationInfo to point to the correct line when an error occurs, or so
16:19:51FromDiscord<pyryrin> oh
16:23:27FromDiscord<pietroppeter> interesting bug in integer division found in bigints library thanks to a fuzzer: https://github.com/nim-lang/bigints/issues/123
16:24:06*fallback quit (Remote host closed the connection)
16:59:13FromDiscord<pietroppeter> @planetis you want to try DrChaos against bigints? šŸ˜‰
17:11:03*jjido joined #nim
17:22:12*fallback joined #nim
17:22:14FromDiscord<pyryrin> sent a code paste, see https://paste.rs/ljZ
17:23:47FromDiscord<Rika> `template šŸ¤¬(): File = stderr`
17:23:53FromDiscord<Rika> need to have return type
17:24:04FromDiscord<pyryrin> oh oops
17:24:10FromDiscord<pyryrin> forgot basic nim
17:24:18FromDiscord<Rika> thats fine
17:24:28FromDiscord<Rika> common forgotten thing i would say
17:26:30FromDiscord<pyolyokh> and Nim is a relatively bad language when half-remembered. It accepts nonsense instead of erroring on it, and it errors on bad code but not in a useful way (evergreen: fake indentation errors).
17:32:37FromDiscord<āš¶ Zeno> hello people, how can i change a node into another node ?ā†µi did `.body=` but that gives me "Invalid node king nnkCommand for macros.body="ā†µā†µi am trying to do var assignment of apl, so like `h ā† [1, 2]` is equivalent to `var h = [1, 2]` for fun https://media.discordapp.net/attachments/371759389889003532/1063148506182848643/image.png
17:32:52FromDiscord<āš¶ Zeno> i am new to macros by the way
17:32:56FromDiscord<āš¶ Zeno> (edit) "fun" => "fun, if theres a way to do this without macros, let me know"
17:33:46FromDiscord<pyolyokh> well, just like templates, macros have a return value
17:33:52FromDiscord<āš¶ Zeno> oh yeah
17:34:27FromDiscord<āš¶ Zeno> well i still get same result even if i `return ast` assuming `child.body=tree` does exactly what i want it to do
17:34:34FromDiscord<pyolyokh> although a void return is now the old `typed` return I guess. They return the new AST structure
17:34:55FromDiscord<pyolyokh> well, you don't get the same result, because you wouldn't be using body= anymore
17:35:10FromDiscord<āš¶ Zeno> well so i have to make my own ast structure without editing original ?
17:37:03FromDiscord<pyolyokh> yes. because stuff like body= aren't even defined
17:37:12FromDiscord<āš¶ Zeno> alright then, thanks for help
17:37:15FromDiscord<āš¶ Zeno> (edit) "help" => "info"
17:38:53FromDiscord<pyolyokh> Mastering Nim language is "A macro receives ASTs in its parameters and returns a new AST.", and even the manual examples that don't have an apparent return type, they use `result.add` to build an AST to return
17:40:31FromDiscord<demotomohiro> You can copy `ast` param, modify it and return
17:40:42FromDiscord<āš¶ Zeno> and how would i do modifying ?
17:40:45FromDiscord<āš¶ Zeno> thats what im trying to do
17:40:51FromDiscord<āš¶ Zeno> except its on original ast but eh
17:41:00FromDiscord<āš¶ Zeno> (edit) "its" => "i was modifying it"
17:42:34FromDiscord<demotomohiro> You can modify copy of `ast` just by adding new node or replace exisiting node.
17:43:17FromDiscord<āš¶ Zeno> i'd prefer if i could replace existing nnkCommand, can you tell me more about it ?
17:43:19*jjido quit (Quit: My laptop has gone to sleep. ZZZzzzā€¦)
17:47:24FromDiscord<Tanguy> sent a code paste, see https://play.nim-lang.org/#ix=4l0x
17:47:30FromDiscord<demotomohiro> You can set specific child of `NimNode` like `foo[i] = newNode`. https://nim-lang.org/docs/macros.html#%5B%5D%3D%2CNimNode%2Cint%2CNimNode
17:48:41FromDiscord<demotomohiro> There is `pairs` iterator to get the index of a NimNode: https://nim-lang.org/docs/macros.html#pairs.i%2CNimNode
17:50:15*xet7 quit (Quit: Leaving)
17:50:19FromDiscord<āš¶ Zeno> thanks tanguy and demotomohiro
17:50:23FromDiscord<āš¶ Zeno> i did it šŸ„³ https://media.discordapp.net/attachments/371759389889003532/1063152976820183040/image.png
17:51:32FromDiscord<pyolyokh> what's your final code?
17:51:36FromDiscord<āš¶ Zeno> oh and thanks to pyolyokh too
17:51:45*xet7 joined #nim
17:51:46FromDiscord<āš¶ Zeno> In reply to @pyolyokh "what's your final code?": https://media.discordapp.net/attachments/371759389889003532/1063153326591582318/image.png
17:52:00FromDiscord<pyolyokh> nah, that shows I was wrong
17:52:24FromDiscord<āš¶ Zeno> what do you mean you were wrong
17:52:28FromDiscord<āš¶ Zeno> im sure you were correct
17:52:38FromDiscord<pyolyokh> you're able to modify the tree in place like you wanted
17:52:48FromDiscord<āš¶ Zeno> ah
17:52:54FromDiscord<āš¶ Zeno> that
17:52:58FromDiscord<scipio> https://forum.nim-lang.org/t/4758
17:52:59FromDiscord<Tanguy> You can modify the tree, but you still need to return, AFAIK
17:53:03FromDiscord<āš¶ Zeno> In reply to @Tanguy "You can modify the": yup
17:53:05FromDiscord<scipio> How about V lang now???
17:53:27FromDiscord<āš¶ Zeno> code itself doesnt run if you dont return
17:58:53FromDiscord<scipio> https://media.discordapp.net/attachments/371759389889003532/1063155117911707688/image.png
17:59:08FromDiscord<scipio> this looks pretty cool to me tbh, V has a stdlib UI even
17:59:14FromDiscord<pyolyokh> sent a long message, see http://ix.io/4l0A
17:59:30*xet7 quit (Quit: Leaving)
17:59:55FromDiscord<scipio> https://media.discordapp.net/attachments/371759389889003532/1063155374401802250/image.png
18:00:20FromDiscord<scipio> I've spent 2 full days on Nim hunting down the best way to do this . Any idea how much more welcoming V comes across to new devs?
18:00:26FromDiscord<demotomohiro> sent a code paste, see https://paste.rs/cdy
18:00:46FromDiscord<scipio> The community here on Nim is absolutely one of a kind tho...
18:00:54FromDiscord<pyolyokh> I did a little flash card program in v, with its gui, before rewriting it in Nim and illwill. I think v's not bad at all as a language, and I prefer that kind of syntax to Nim's even.
18:01:12FromDiscord<scipio> But tbh, I fell off my chair multiple times seeing the Nim founder reply to community efforts
18:01:29FromDiscord<āš¶ Zeno> In reply to @demotomohiro "You can use `<-`": oh damn thanks for giving me non-macro way, which turned out to be simple
18:01:35FromDiscord<scipio> https://media.discordapp.net/attachments/371759389889003532/1063155794633298080/image.png
18:01:36FromDiscord<āš¶ Zeno> i'll try to use templates next time, thanks again
18:01:44FromDiscord<scipio> https://media.discordapp.net/attachments/371759389889003532/1063155836077232258/image.png
18:01:46FromDiscord<scipio> ehh.... wtf?
18:02:06FromDiscord<scipio> You think seeing this makes me want to immediately invest > 1 month of my time in Nim?
18:02:32FromDiscord<āš¶ Zeno> (edit) "In reply to @demotomohiro "You can use `<-`": oh damn thanks for giving me non-macro way, which turned out to be simple ... " added "(although not same identifier, but thats fine)"
18:02:51FromDiscord<Tanguy> https://nim-lang.org/docs/manual.html#syntax-precedence the "First character" column is the list of first char you can use for the kind of template that @demotomohiro sent
18:02:53FromDiscord<scipio> It doesnt matter if this comment is right or wrong, on a technical perspective: Owlkettle improves his product Nim. Responding like this to an announcement in public is "not done", period.
18:02:59FromDiscord<Tanguy> I think
18:03:18FromDiscord<pyolyokh> He thinks a particular design feature is ugly. Why does that bother you? If we're comparing dev community engagement, v is pretty scandalous as "vapoware" with a "lying dev" thanks to its initial handling. Anyway if that's the core issue, I don't care to discuss it.
18:03:56*xet7 joined #nim
18:04:00FromDiscord<Tanguy> Who cares about the language creator, if everything goes well, you will never have to talk to him
18:04:11FromDiscord<pyolyokh> I think I had a pretty useful reply above to the language, rather than to the personalities.
18:04:58FromDiscord<scipio> @pyolyokh I am equally neutral to V as I am to Nim at this point in time, I don't know either founders so I cannot tell if one is lying or not. What I am saying, is that I have been researching Nim UI development for 2 days straight and I only came across abandonded or half-baked solutions. And I am _very open_ to embracing Nim
18:05:26FromDiscord<scipio> But I'm not looking at a language from a technical perspective only
18:05:58*xet7 quit (Remote host closed the connection)
18:06:22FromDiscord<demotomohiro> Nim allows you to define operators with some combination of charators: https://nim-lang.org/docs/manual.html#lexical-analysis-operators
18:06:31FromDiscord<pyolyokh> that's fair. I haven't look at either language's GUI offerings, just terminal for both (decent in both cases) and raylib for Nim.
18:07:31FromDiscord<āš¶ Zeno> In reply to @demotomohiro "Nim allows you to": yeah i was aware of that, but i was mainly trying to do `ā†` thats why i resorted to not doing \`<-\`
18:07:34FromDiscord<scipio> In reply to @Tanguy "Who cares about the": I do. When I dedicate my time into mastering a language or framework, I want to assess the potential ROI on my time spent on it, commercially, as well.ā†µHow many job advertisements do you currently see on Nim? How many companies have embraced it into their tech stack? ā†µā†µ--- this is no Pun: I'd love Nim to succeed. It's an honest assessment.
18:07:52FromDiscord<pyolyokh> for GUIs I'd start with <https://nimble.directory/search?query=gui> and, unfortunately since that doesn't have any kind of way to sort by metrics, I'd check each repository until I found something that looks reasonably baked. It wouldn't take two days.
18:08:37FromDiscord<Tanguy> How does the commercial success of the language depends on the creator of the languages responses on a forum
18:09:29FromDiscord<scipio> If you need to ask that I don't have time to answer
18:09:38FromDiscord<scipio> Tl;DR: a lot
18:10:22FromDiscord<Tanguy> How often do you read the responses of the creator of python, or C++ on forums? Never, because the community took over
18:10:50FromDiscord<scipio> When I see Andre von Auck talk about Nim, I love it: want to dive into it immediately. When I read a toxic reply or see a book author getting bashed, I don't. See the difference?
18:11:46FromDiscord<scipio> There's a ton of competition among new languages
18:11:55FromDiscord<scipio> They're all fighting for developer attention
18:12:25FromDiscord<scipio> That was not the case when my fellow-countryman Guido van Rossum developed Python
18:13:35FromDiscord<scipio> In reply to @pyolyokh "for GUIs I'd start": I did exactly that. Check the repos mentioned in there. Most of them aren't actively being developed.
18:13:37FromDiscord<cow> dutch person detected
18:13:52FromDiscord<pyolyokh> indeed, Python was grossly inferior to Perl in both technical and social capital senses for the first 15 years of its life, so saying "only technicals matters" would be wrong. And although probably Tim Peters is more responsible for the propaganda than Guido, Guido still started it and set the direction of the language.
18:13:56FromDiscord<scipio> Absoluut, je kunt de Groeten uit Brabant krijgen šŸ˜‰
18:14:00FromDiscord<scipio> @cow
18:14:40*jjido joined #nim
18:15:17FromDiscord<Tanguy> Let's be honest, the reason rust or go took of compared to nim is because they are backed by massive companies, not for their syntax, features or creators activity on forums (imo)
18:15:48FromDiscord<scipio> Don't get me wrong here, Nim has a ton of very cool stuff baked in, I'd love Nim to succeed. I do. ā†µI do miss a lot of focused attention for UI development in Nim. Should be 1st class citizen. It's not, face the facts.
18:15:51FromDiscord<cow> they advertise themselves as solving a problem that pains many companies
18:15:58FromDiscord<pyolyokh> there's a lot to say there but none of it very interesting.
18:16:59FromDiscord<Tanguy> scipio: Nim tries to be "good at everything", thus not 1st class citizen for anything, thus almost everything is average šŸ™‚
18:18:13FromDiscord<mratsim> In reply to @scipio "<@803214225002463283> I am equally": TBH, apart from Delphi and JS, all languages have half-backed OR overengineered UI solutions.
18:18:35FromDiscord<mratsim> because a good solution will tend to attract people that demand more and more features and then "bloat"
18:18:47FromDiscord<pyolyokh> In reply to @scipio "Don't get me wrong": everybody thinks that the thing they care about should be the #1 priority. This is a good difference that I already noted above: Araq doesn't care about selling the language himself on its third party support, but V's author has a wishlist that he's actively pursuing. But then Araq knows what type an array index should have. They're both decent examples of how language development might need mo
18:19:43FromDiscord<cow> In reply to @mratsim "TBH, apart from Delphi": Flutter and .NET UI platforms are pretty good
18:20:04FromDiscord<mratsim> I remember trying flutter in 2017, it was no man's land.
18:20:24FromDiscord<pyolyokh> I've had the most fun with J's UI support, which is a very limited DSL.
18:20:45FromDiscord<scipio> @mratsim which is EXACTLY why I am looking at V (and Jai) and bringing it up here in Nim.ā†µWhy did "the web" succeed? html, js .... ā†µ==> it's because if you perceive "a webpage" as an app, html/css/js is "UI first". New devs get immediate results from dabbling in it.. That's very rewarding to explore the language further.
18:21:37FromDiscord<Tanguy> The web succeeded because you can just go on a page and use a program, instead of having to install it. And it works "everywhere"ā†µNot because of dev confort. Dev confort came from the need of building web apps
18:21:47FromDiscord<cow> arguably Vala and Dart are languages that exist because of a good UI library written for the language specifically
18:22:02FromDiscord<cow> the former case shows that a language can be good for UI even if its selling point is just a binding
18:22:07FromDiscord<mratsim> In reply to @scipio "<@570268431522201601> which is EXACTLY": I think you are inverting causality. JS succeeded because it was the only language available on the web (if we ignored java applet and Flash/ActionScript).ā†µā†µThe web succeeded because it filled a need.
18:22:50FromDiscord<cow> In reply to @Tanguy "The web succeeded because": also, it's cross platform and JS devs are cheap
18:23:09FromDiscord<pyolyokh> and I've made very nice web apps with a Nim backend, using only the 'cgi' module to take care of some chores, and printing `"""<!doctype html>"""` strings. Not a very high effort-floor, to make something useful.
18:23:10FromDiscord<mratsim> In reply to @cow "also, it's cross platform": are they?
18:23:18FromDiscord<cow> compared to desktop devs, I think so
18:23:32FromDiscord<mratsim> what's a desktop dev?
18:23:52FromDiscord<cow> a .NET dev for example who makes windows native apps
18:23:56FromDiscord<pyolyokh> In reply to @mratsim "what's a desktop dev?": I believe those are called "electron devs" these days
18:24:12FromDiscord<cow> or a C++ dev using Qt
18:24:16FromDiscord<scipio> Irrelevant for this discussion.ā†µMy point is, language competition is intense now and without a modern native UI framework out of the box, the language will only attract the most hardcore of devs, not ideal for mainstream adoption.
18:24:22FromDiscord<mratsim> ah, I feel like no one makes native apps anymore :p
18:24:39FromDiscord<Tanguy> If JS devs are cheap, it's because everyone needs JS devs, so everyone learns JS, so the market is floodedā†µAgain, JS is only here because it has(had) monopoly on a very popular platform
18:25:06FromDiscord<pyolyokh> In reply to @scipio "Irrelevant for this discussion.": Nim isn't going to have a "modern native UI framework out of the box". It'll have the nimble libraries that people make for it. Those that are just wrappers might completely work for you despite not seeming to be actively developed, because wrappers don't need a lot of development.
18:25:11FromDiscord<cow> it's also a factor that JS is easy to get into (while might be painful in the long run)
18:25:25FromDiscord<mratsim> In reply to @scipio "Irrelevant for this discussion.": Python doesn't really have a nice UI story, many bindings all over the place as well.
18:25:30FromDiscord<cow> python'
18:25:36FromDiscord<scipio> JS is about the only language that becomes harder to write the better you get at it
18:25:48FromDiscord<mratsim> ==="foo"
18:25:59FromDiscord<mratsim> (but PHP has it too AFAIK)
18:26:00FromDiscord<cow> (edit) "python'" => "python found a niche for itself without a UI library"
18:26:17FromDiscord<pyolyokh> In reply to @scipio "JS is about the": with a strong willpower, you can write `function foo() {}` and not use semicolons, no matter what the crazy people add to it
18:26:18FromDiscord<mratsim> Rust doesn't have an UI library either
18:26:22FromDiscord<scipio> Iced
18:26:48FromDiscord<scipio> Well it's not in the std lib no if that's your point
18:26:49FromDiscord<cow> there's definitely languages that are successful without a very good UI developing experience
18:26:55FromDiscord<pyolyokh> Rust doesn't have a UI library in core, "out of the box". It's in the exact same position of Nim, where that stuff is something that a library developer wanted.
18:26:57FromDiscord<cow> stdlib or not
18:27:22FromDiscord<mratsim> What about Go?
18:27:26FromDiscord<scipio> Sure, but if nobody in the language really cares about a focused UI framework that's truly cross-platform and well maintained ... hmmm ... no bueno
18:27:43FromDiscord<cow> but it would be amazing if nim had an easy to use, mature UI library... but who has the time to make one
18:27:53FromDiscord<scipio> In reply to @cow "there's definitely languages that": But Rust does have commercial Mozilla backing. That is also lacking for Nim
18:27:56FromDiscord<T0lk1en> Gtk?
18:27:57FromDiscord<scipio> No UI, no commercial backing
18:28:10FromDiscord<scipio> Then how are you going to attract adoption?
18:28:22FromDiscord<mratsim> other way around again. You don't develop an UI if you don't have a commercial need.
18:28:23FromDiscord<nixfreak> Owlkettle is pretty fun to write in. I really want to learn more about Karax
18:28:27FromDiscord<nixfreak> The examples are ok , but I'm having to dig everywhere to find decent Vcomponent examples and ajax examples
18:28:38FromDiscord<scipio> Where the language founder doesnt want to publish a digital version of his book because people might pirate it
18:28:43FromDiscord<scipio> I mean wtf
18:28:46FromDiscord<mratsim> if it's only hobbyist, with no use-case you just don't know the problem you want to fix
18:28:53FromDiscord<T0lk1en> In reply to @scipio "Where the language founder": What lang is this
18:28:57FromDiscord<cow> I think there's a commercial need to move away from electron, it's just not big and it's only gonna happen if companies get to keep their cross platform uniform apps
18:29:09FromDiscord<scipio> In reply to @T0lk1en "What lang is this": this one
18:29:12FromDiscord<Tanguy> Anyway, it's all chicken and egg, if you are here for a language with big commercial capacity, just go look at the languages with highest salaries and opportunities or whateverā†µNim, V, and to some extent rust, are not gonna be on top of that list any time soon
18:29:17FromDiscord<T0lk1en> Ohā€¦
18:29:40FromDiscord<Tanguy> if ever
18:30:14FromDiscord<Tanguy> I'm here because I enjoy using Nim, and been lucky enough to find a job using Nim
18:30:28FromDiscord<scipio> In reply to @Tanguy "I'm here because I": care to elaborate on that?
18:31:02FromDiscord<Tanguy> [..] not because I did a business analysis
18:31:06FromDiscord<Tanguy> On what exactly?
18:31:13FromDiscord<scipio> On getting a job as Nim dev
18:31:24FromDiscord<mratsim> same. I liked writing Nim, so i wrote more and more Nim then found a Nim job
18:31:53FromDiscord<Tanguy> Well, there are basically two ways, you find one of the few companies that use nim, or you try to introduce a bit of nim at your companyā†µHad success with both
18:32:21FromDiscord<Tanguy> But ngl, very few companies use Nim
18:32:35FromDiscord<scipio> And what was your angle to introduce Nim at your current company? Which primary use case does Nim solve for your company?
18:33:29FromDiscord<pyolyokh> sent a long message, see http://ix.io/4l0O
18:33:56FromDiscord<Tanguy> That was at a previous company, I started something that was supposed to be a small project, wanted to try it out in Nim, so I didā†µThe project grew, and at no point we had to rewrite it in something else (was still a fairly small project, with ~1.5 FTEs)
18:34:06*jjido quit (Quit: My laptop has gone to sleep. ZZZzzzā€¦)
18:35:09FromDiscord<scipio> sent a long message, see http://ix.io/4l0P
18:35:33FromDiscord<Tanguy> Everything becomes a technical challenge once it reaches a certain scale & number of users (including libraries). If it was that straightforward, it would already exist
18:36:02FromDiscord<scipio> I need to quickly buy some groceries before the store closes, brb
18:41:48FromDiscord<cow> In reply to @scipio "> _Nim is a": my impression is that the godot and qt bindings are the most mature way to make UI apps in nim
18:42:07FromDiscord<cow> or possibly other bindings, but I haven't seen those used in production
18:42:38FromDiscord<cow> maybe just reducing the entry barrier and improving those bindings is the best way to go
18:43:30FromDiscord<nixfreak> Gintro is very mature for GK3/4
18:44:18FromDiscord<nixfreak> Its really too bad nimx isn't getting the help or documentation it needs.
18:45:15FromDiscord<nixfreak> Fidget was going to be nice because it was (declarative).
18:47:38FromDiscord<pyryrin> when you do `nimble run` can you just put falgs like `-d:release --gc:orc` after it
18:48:23FromDiscord<pyolyokh> In reply to @pyryrin "when you do `nimble": yeah.
18:48:25FromDiscord<scipio> sent a long message, see http://ix.io/4l0U
18:48:54FromDiscord<cow> In reply to @scipio "just stand still for": no there is definitely a barrier
18:49:18FromDiscord<scipio> In reply to @nixfreak "Gintro is very mature": Exactly my point: there's a ton of half-baked, in essence very promising but still not finished, attempts, but NONE is mature / production ready
18:49:21FromDiscord<pyolyokh> >accept your GUI won't look nativeā†µwe live in the post-electron world. Nobody cares.ā†µOr rather, the only people who care are Tk devs who are spinning in their graves about how much their #1 enemy no longer matters.
18:49:34FromDiscord<nixfreak> In reply to @scipio "just stand still for": Yeah if your new at godot and nim and especially bindings , I would not use godot
18:49:35FromDiscord<cow> unless you're on mobile
18:49:51FromDiscord<cow> I think iOS users still want a iOS touch to their apps
18:50:23FromDiscord<pyryrin> what does gui "not looking native" even mean
18:51:11FromDiscord<cow> that it doesn't look like your file explorer and settings app šŸ˜›
18:51:15FromDiscord<pyolyokh> In reply to @pyryrin "what does gui "not": back in 1995, Windows apps had a particular look and feel, and Tcl/Tk apps (and python using Tkinter, having all of scipio's problems above, but mitigated with good documentation) didn't have that look and feel
18:51:17FromDiscord<nixfreak> In reply to @pyryrin "what does gui "not": Good point , to me its , ios native GUI framework , cocoa , windows winforms and linux Qt, or GTK.
18:51:47FromDiscord<ambient> I guess there's some UX design principles that some OS in theory have as goal. But on Windows and Linux there are at least 3 different, often more, so it doesn't really matter.
18:51:52FromDiscord<cow> I think WPF deprecated winforms by now šŸ˜›
18:52:26FromDiscord<nixfreak> lol I couldn't remember the windows framework , sorry I don't use windows
18:53:07FromDiscord<ambient> You could build perfectly functional applications with visual basic 3 in minutes. People just want the latest new shiny things, even when they don't make any practical sense
18:53:40FromDiscord<nixfreak> yeah the visual builder made it pretty simple to create something
18:54:14FromDiscord<scipio> Show me one user friendly TUTORIAL on making a decent Nim GUI app from scratch
18:54:31FromDiscord<pyolyokh> I've had <https://nappgui.com/en/start/hello.html> open in a tab for the last week. I invite you to read its starting quote, which is pretty good, and then to read the code that follows.
18:54:54FromDiscord<nixfreak> Well , lol it depends how in debt you want the application to do
18:55:02FromDiscord<planetis> Slightly offtopic, I had to write a GUI in java swing, I am curious how owlkettle and Gtk compares.
18:55:13FromDiscord<ambient> In reply to @scipio "Show me one user": https://github.com/StefanSalewski/gintro#a-few-basic-examples
18:55:27FromDiscord<nixfreak> (edit) "in debt" => "much" | "do" => "do."
18:55:50FromDiscord<planetis> Easy https://youtu.be/O44x_la1o_Mā†µ(@scipio)
18:55:52FromDiscord<nixfreak> OwlKettle isn't bad to work in , accept when you can't do something
18:56:12FromDiscord<nixfreak> niup is pretty easy to get into
18:56:21FromDiscord<nixfreak> OwlKettle, NIgui
18:56:41FromDiscord<nixfreak> Even NimX you can create things
18:57:29FromDiscord<planetis> You mean this? https://github.com/dariolah/niup I will check it out.ā†µ(@nixfreak)
18:57:55FromDiscord<scipio> In reply to @ambient "https://github.com/StefanSalewski/gintro#a-few-basi": decent attempt, looks promising, but I'm not seeing installation steps targeting macOS
18:57:56FromDiscord<nixfreak> yep , I created some stuff
18:58:25FromDiscord<ambient> In reply to @scipio "decent attempt, looks promising,": On Windows, it was just about installing GTK and letting the C compiler know where it was located IIRC
18:58:59FromDiscord<ambient> Mac should be easier, for example with brew
19:00:16FromDiscord<planetis> Come on don't be like that, he's donating his free time to make something and share it with the community. If you want macos support work on it.ā†µ(@scipio)
19:00:17FromDiscord<scipio> But don't you realize what's happening here?ā†µ- how does Gintro compare to Owlkettle? It both targets GTK3/4 right? It's confusing to have _too many options_. Without concerted efforts to creating one _de facto solution_ that is agreed on should be "standard", even these two similar solutions rival eachother
19:00:39FromDiscord<pyolyokh> what I realize what just happened, in the last moment, is goal shifting
19:01:12FromDiscord<nixfreak> OwlKettle is "declarative"
19:01:14FromDiscord<scipio> In reply to @planetis "Come on don't be": I don't understand why you'd conclude I'm dissing his attempts; I'm not. I'm applauding them. I'm only pointing out both solutions are re-inventing the wheel and none target macOS
19:01:22FromDiscord<ambient> yes, one great solution would be excellent, however outside somebody writing industry leading Qt bindings, I don't think that will happen
19:01:26FromDiscord<pyolyokh> and this is a chilling dilemma that someone has to have already overcome in order to choose Nim at all, which is a language that is not the "one defacto solution" in a world with "too many options"
19:01:49FromDiscord<ambient> In reply to @pyolyokh "and this is a": Both Rust and Python have/had serious issues with desktop GUI workflows
19:02:29FromDiscord<scipio> In reply to @pyolyokh "what I realize what": its not. I'm on macOS, I want to build cool stuff including modern UI apps in Nim. Show me the way.
19:02:47FromDiscord<nixfreak> Yep , in python I really like pyjamas and it died
19:03:03FromDiscord<ambient> well, both Apple and Microsoft make their desktop application build process very much a walled garden. You should go knock on their door
19:03:24FromDiscord<pyolyokh> no, what you want to happen is for an official announcement that Nim is backing a single GUI framework as the defacto one, and that's not going to happen, so you should be happy when you ask a for a good tutorial and are given one.
19:03:26FromDiscord<scipio> It is absolutely mindboggling Python never had a native proper UI framework. Tkinter on steroids.
19:03:48FromDiscord<ambient> it's not mind boggling at all, refer to my previous comment
19:04:21FromDiscord<scipio> In reply to @pyolyokh "no, what you want": You have it completely backwards if you want Nim to become mainstream. If you want NIm to remain obscure, a footnote in computer science 25 years from now, then you\re on track
19:04:56FromDiscord<ambient> most OSS communities aren't businesses, it's just a bunch of guys hacking together things they want to exist
19:05:02FromDiscord<pyolyokh> what I have is a good sense of what will happen and what will not happen, and the thing you want just isn't going to happen, so there's no need to keep goal-shifting to get there.
19:05:14FromDiscord<ambient> therefore there's no customer department that you can really target your complaints at
19:05:20FromDiscord<scipio> I'm merely asking the right questions for Nim to succeed. And you assume I am looking for a dev job in Nim. You're wrong. I'm a serial entrepreneur in tech and a coding auto-didact.
19:05:46FromDiscord<scipio> I'm asking why there aren't any jobs in Nim if it's so good
19:06:03FromDiscord<scipio> I'm asking why there are like 40 half baked UI framework attempts
19:06:10FromDiscord<planetis> I had a summer job in nim last summer
19:06:17FromDiscord<jtv> I'm new to the community too. Great language, some helpful people, but in general, seems like it's not culturally geared toward giving engineers a good onramp.
19:06:34FromDiscord<pyolyokh> OK, the answer to that question is: no, Nim won't have a "nim ui" in the same way that V has a "v ui". This is an interesting difference between the two languages that I was interested to talk about above, and I think it's overshadowed by core competency as mentioned above.ā†µThe jobs question I don't care to talk about.
19:06:36FromDiscord<planetis> Will keep working for them when i find time
19:06:42FromDiscord<scipio> Nim is the product any dev would want to buy, you're approaching things from the other side, as if devs should kneel down and thank Nim for gracing them with its presence
19:06:46FromDiscord<jtv> It really seems to be a community issue, which has many facets.
19:06:51FromDiscord<scipio> It's borderline narcissistc to think like that
19:07:33FromDiscord<pyolyokh> No, narcissism is "wow I can't believe that my personal interest isn't the #1 priority of this community!"
19:07:36FromDiscord<planetis> What? ...does it meanā†µ(@scipio)
19:08:00FromDiscord<nixfreak> I'm confused also
19:08:50FromDiscord<jtv> And that's exactly part of it, there's a lot of treating people like they need to catch up or get run over the bus, when generally good OSS projects are full of humility and grace. Not everyone is like that, but it's plenty.
19:09:05FromDiscord<scipio> In reply to @ambient "most OSS communities aren't": Exactly, my alma mater is on Eindhoven Tech University (tue.nl ). There's a computer club called Stack there. In 1995 when I was in my first year there at Industrial Management, you should have seen the types of people at Stack. And arguably those people are exceptionally good in writing working code, but not in devleoping products people actually want to use and become a huge success.
19:10:05FromDiscord<jtv> Oh yeah, I have a friend from Eindhoven, pretty sure he was in that computer club around that time frame. Rings a huge bell
19:10:08FromDiscord<nixfreak> If you want to support Nim , by the books, create something and publish it, help others. If you want a "GUI" library that you can use , then create it or start developing and show others why you think your GUI library will be the best? ... Its really easy to talk about this and talk about that but the end of the day its action.
19:10:28FromDiscord<planetis> Can't agree more with pyolyokh on that one. If you're an entrepreneur as you claim and won't get your hands dirty but want the community make libraries for free then whatever
19:10:38FromDiscord<scipio> In reply to @jtv "Oh yeah, I have": We created Ilse and later Vindex pre-Google.
19:10:45FromDiscord<ambient> here, go create Nim bindings for this https://github.com/rochus-keller/LeanQt šŸ˜‰
19:10:53FromDiscord<planetis> I mean everyone has their goals.. I understand it but good luck
19:11:04FromDiscord<nixfreak> Talk is cheap
19:11:37FromDiscord<nixfreak> Create decent bindings for "Nuklear" gui library
19:11:54FromDiscord<scipio> In reply to @planetis "Can't agree more with": Apparently you completely misinterpret me or my intentions. Read back my chat history in here the past 3 days
19:12:06FromDiscord<jtv> And do you all even feel like you need to say anything negative, instead of appreciating that more people are here wanting to use Nim?? Seriously, might this person not have a day job? And if they intro Nim into it, might help it spread which will improve it? The whole, "participate or we're going to answer every question with 'you should be participating' is hardly welcoming.
19:12:12FromDiscord<cow> In reply to @nixfreak "Create decent bindings for": hell yeah nuklear seems so fun
19:12:24FromDiscord<nixfreak> I really like it
19:12:46FromDiscord<scipio> In reply to @nixfreak "If you want to": Having a decent thought provoking discussion is among that. It's highly needed I think
19:13:04FromDiscord<scipio> That doesnt mean I would like to trashtalk Nim, it's the exact opposite
19:13:16FromDiscord<scipio> I'm trying to figure out what's missing to turn it into a success
19:13:18FromDiscord<ambient> In reply to @scipio "Having a decent thought": Complaining that there are no great GUI libraries isn't exactly that šŸ˜“
19:13:18FromDiscord<nixfreak> but I can't create bindings to save my life and I would like to look for more tutorials on that to help out create cool bindings rather then complain about the "half-baked" repos out there.
19:13:30FromDiscord<scipio> What would make Rust and Go and FLutter devs join Nim
19:13:31FromDiscord<pyolyokh> sent a code paste, see https://paste.rs/p3j
19:13:56FromDiscord<nixfreak> I really like OwnKettle also
19:14:01FromDiscord<ambient> In reply to @scipio "What would make Rust": The biggest Rust GUI atm seems to be Tauri, which is just lets do Electron with Rust. It doesn't not fill me with joy
19:14:04FromDiscord<scipio> I also like Owlkettle
19:14:06FromDiscord<jtv> Well, Rust is pretty hard to write and master, Nim's the opposite. Go isn't a real systems language.
19:14:14FromDiscord<hector> In reply to @scipio "What would make Rust": Bribery?
19:14:23FromDiscord<ambient> (edit) "doesn't" => "does"
19:14:37FromDiscord<jtv> But Go's got a massive ecosystem, and that generally is a big advantage
19:14:39FromDiscord<āš¶ Zeno> sent a code paste, see https://paste.rs/aHx
19:15:04FromDiscord<hector> I'm still mad at Rob Pike. Or whoever told him to make Go
19:15:10FromDiscord<cow> In reply to @jtv "Well, Rust is pretty": many people just want statically compiled binaries with the impression that their language is "fast"
19:15:11FromDiscord<leetnewb> In reply to @jtv "But Go's got a": That's really the core of everything discussed here, right? Nim is a great language with a thin community.
19:15:17FromDiscord<āš¶ Zeno> (edit) "https://play.nim-lang.org/#ix=4l13" => "https://play.nim-lang.org/#ix=4l12"
19:15:28FromDiscord<āš¶ Zeno> (edit) "https://play.nim-lang.org/#ix=4l12" => "https://play.nim-lang.org/#ix=4l14"
19:15:40FromDiscord<jtv> Yeah, agreed, @cow. But MORE people don't gaf until it matters, which is why lack of performance never slowed down Python's adoption
19:15:42FromDiscord<nixfreak> sent a code paste, see https://paste.rs/FBA
19:15:56FromDiscord<pyolyokh> \`\`\`nimā†µ\`\`\`ā†µto syntax-highlight
19:16:09FromDiscord<nixfreak> Hell I remember when python was hated
19:16:28FromDiscord<jtv> Yup, and my point about the Nim community is, the existing community is responsible for attracting the future community, and right now it's not doing a good job of that
19:16:36FromDiscord<jtv> I don't, and I first used it in 1994 šŸ™‚
19:16:43FromDiscord<nixfreak> How about a great GUI component library LCL (Lazarus Component Library) its created from Open Pascal.
19:16:56FromDiscord<cow> why do we even worry about adoption
19:17:04FromDiscord<nixfreak> good question
19:17:08FromDiscord<hector> "Avoid success at all costs" is a good motto
19:17:12FromDiscord<jtv> Well, it could be a non-goal.
19:17:24FromDiscord<scipio> In reply to @ambient "The biggest Rust GUI": https://media.discordapp.net/attachments/371759389889003532/1063174878255132752/image.png
19:17:25FromDiscord<pyolyokh> In reply to @leetnewb "That's really the core": the relevant thinness though is more documentation and direction. To the "v ui" vs. Nim's options, Nim would get 90% of the practical result of having a "v ui" nim-lang.org had a "want a GUI?" link to some discussion about what's available and some guidance on which to go for.
19:17:37FromDiscord<jtv> But for people trying to use it for real work, no ecosystem worth noting is a challenge
19:17:40FromDiscord<nixfreak> probably because of the "press" about a language ... lol stop looking at reddit, tickTok, twitter , youtube
19:18:01FromDiscord<pyolyokh> (edit) "In reply to @leetnewb "That's really the core": the relevant thinness though is more documentation and direction. To the "v ui" vs. Nim's options, Nim would get 90% of the practical result of having a "v ui" ... nim-lang.org" added "if"
19:18:34FromDiscord<nixfreak> It doesn't matter what language you use , because you need to spend a ton of time learning it if you want to do anything usefull
19:18:37FromDiscord<nixfreak> (edit) "usefull" => "useful"
19:18:50FromDiscord<scipio> In reply to @ambient "The biggest Rust GUI": Tauri is a lot more as just some webview wrapper. It has "sidecar" but Nim bindings are not high priority right now for Tauri and why would it?
19:19:06FromDiscord<nixfreak> lol I hate electron
19:19:20FromDiscord<jtv> The whole reason Python did so well is because it made it pretty EASY to do interesting stuff w/o a lot of programming experience
19:19:27FromDiscord<nixfreak> make system's slow and blah
19:19:31FromDiscord<jtv> And some of that was the syntax, but mostly it was the ecosystem
19:19:47FromDiscord<scipio> "Electron / Tauri" is primary aimed at JS frontend devs who want a better backend language as nodeJS
19:19:48FromDiscord<jtv> Lots and lots of modules in the standard libraries before there were package managers
19:19:51FromDiscord<cow> In reply to @jtv "And some of that": how did it get popular before its ecosystem existed though?
19:19:55FromDiscord<ambient> Python got big because of Numpy is 50% correct statement
19:20:00FromDiscord<cow> In reply to @jtv "Lots and lots of": ah okay yeah
19:20:13FromDiscord<nixfreak> python go adopted because of the "press it got it" from regular users
19:20:32FromDiscord<nixfreak> like "Watch me do this! in python" , Python is so easy to use ...
19:20:46FromDiscord<jtv> Yeah, I used Python 1.0 in 1994, and got tons and tons of people hooked on it
19:20:56FromDiscord<scipio> In reply to @jtv "The whole reason Python": Yup. And Python devs would IMHO be the #1 audience to port over to Nim.
19:21:11FromDiscord<nixfreak> what?
19:21:22FromDiscord<cow> and that would set the bar as low as having better UI thank tkinter šŸ˜„
19:21:35FromDiscord<nixfreak> I would C / C++ devs are more to follow Nim then python devs
19:21:36FromDiscord<jtv> IDK, I have done more C than Python in my career, it replaces both quite well for me
19:21:47FromDiscord<ambient> I don't really think this is the right way to think about things. It's not language A vs. language B zero-sum games. It's what tradeoffs are the best for your usecase and communicate them effectively.
19:22:16FromDiscord<scipio> In reply to @cow "and that would set": No. Because Pythonistas with an interest in UI have accepted to write html/css/js frontend BECAUSE Tkinter fucked up / isnt modern
19:22:23FromDiscord<jtv> But I cannot recommend it to anyone else. The people who I have raved about it to who checked it out think it's cool but no ecosystem, and won't adopt it for that reason
19:22:47FromDiscord<nixfreak> In reply to @scipio "No. Because Pythonistas with": What about all the other UI's you can use with python?
19:22:52FromDiscord<jtv> Anyone remember tcl? Tk was originally written in tcl
19:23:09FromDiscord<scipio> I'm also involved with the "Flet" https://flet.dev/ project and literally eveyr day new Python devs flow in.
19:23:19FromDiscord<nixfreak> wxPython, kivy, Qt, GTK, FLTK, EFL, etc....
19:23:29FromDiscord<scipio> Flet is a JSON serialized declarative Flutter lib/framework, targeted at Python and/or other languages
19:23:35FromDiscord<hector> Don't worry people, our soon-to-be AI globalist overlords will handle all the nitty-gritty programming details anyway.
19:23:46FromDiscord<pyolyokh> there should really be a #python chat for all the discussion that's inevitably about it
19:23:47FromDiscord<nixfreak> I don't take about flutter
19:23:51FromDiscord<scipio> Why is Flet so successful? Because Python devs are STILL struggling with a decent Python UI
19:23:54FromDiscord<nixfreak> (edit) "take" => "care"
19:24:24FromDiscord<pyolyokh> although the meme would be, "what are your opinions on v?" -> "ask in #python"
19:24:29FromDiscord<nixfreak> Google promisses a lot and usually disappoints
19:24:34FromDiscord<hector> Former coder humans will be entertained by plugging in colored blocks to other colored blocks to get animated cats to move around
19:24:35FromDiscord<scipio> In reply to @nixfreak "What about all the": Can use? Or are actively being used?
19:24:40FromDiscord<ambient> Python has no GUIs? PyQt is fine. FastAPI is fine.
19:24:55FromDiscord<āš¶ Zeno> In reply to @ambient "Python has no GUIs?": since when was fastapi a gui-
19:24:56FromDiscord<Phil> At this point I've talked nearly as much about wanting to use nimqml than actually spend time learning qml, but that's basically what I'm looking into in that regard, particularly because I think the scope where nim-code matters is smaller and thus easier to achieve, so I have higher hopes.
19:24:57FromDiscord<nixfreak> What do you mean "has no gui's"
19:25:09FromDiscord<ambient> In reply to @āš¶ Zeno "since when was fastapi": Since when you could graphically use it as an interface
19:25:23FromDiscord<nixfreak> You mean "built-in"
19:25:27FromDiscord<āš¶ Zeno> confused screaming
19:25:42FromDiscord<āš¶ Zeno> how does one use a webserver
19:25:45FromDiscord<āš¶ Zeno> as gui
19:25:56FromDiscord<nixfreak> what lol
19:26:09FromDiscord<pyolyokh> In reply to @āš¶ Zeno "how does one use": you make it serve HTML pages. A form that hits CGI, or a little bit of JS, and you've got an application.
19:26:33FromDiscord<āš¶ Zeno> yea that would make sense, i was more of confused at wording
19:27:54FromDiscord<Phil> In reply to @āš¶ Zeno "how does one use": Easy, you make the frontend a PWA and make it installable. Then you provide a script to "install" the webserver and let it all run over localhost, boom
19:28:18FromDiscord<āš¶ Zeno> sir yes sir
19:28:22FromDiscord<scipio> In reply to @pyolyokh "you make it serve": this is what I've been doing myself with Python (Flask) for desktop apps
19:28:49FromDiscord<scipio> In reply to @pyolyokh "you make it serve": i.e. (as I dropped in here yesterday): https://media.discordapp.net/attachments/371759389889003532/1063177748027289701/image.png
19:29:04FromDiscord<Phil> ~~What I just proposed may not be a good install experience, but a functional one~~
19:29:32FromDiscord<scipio> In reply to @pyolyokh "you make it serve": one issue however, is this app's "Charting engine", which I'n still developing (in KonvaJS, using canvas)
19:29:39FromDiscord<pyolyokh> sent a long message, see https://paste.rs/cbh
19:30:13FromDiscord<ambient> In reply to @pyolyokh "some fun things about": How did you do the sqlite static linking part?
19:30:39FromDiscord<ambient> I've been meaning to look into it
19:31:13FromDiscord<Phil> In reply to @scipio "i.e. (as I dropped": Touch UI?
19:31:41FromDiscord<guttural666> In reply to @jtv "Yeah, I used Python": sounds like heroin
19:31:42FromDiscord<scipio> Just clicks and bindings to an Elgato StreamDeck
19:31:51FromDiscord<scipio> @Phil
19:31:59FromDiscord<pyolyokh> sent a code paste, see https://play.nim-lang.org/#ix=4l1a
19:32:11FromDiscord<pyolyokh> with sqlite3.c sqlite3ext.h sqlite3.h in the directory
19:33:01FromDiscord<Phil> In reply to @scipio "<@180601887916163073>": Ahhhh, streamdeck explains the button size
19:33:32FromDiscord<Phil> The shape just looked odd to me for general design so I was puzzling over what made it necessary, so my first assumption was this was for a touch surface for easy tapping
19:34:15FromDiscord<scipio> In reply to @Isofruit "Touch UI?": it's a multi-exchange trading app, and the UI concept I created here is to eliminate human typing for order processing
19:34:30FromDiscord<pyolyokh> In reply to @ambient "How did you do": ... and I just realized that ringabout ripped all of that out, so it's not as easy anymore
19:35:10FromDiscord<scipio> In reply to @Isofruit "Ahhhh, streamdeck explains the": https://media.discordapp.net/attachments/371759389889003532/1063179349051846746/image.png
19:35:18FromDiscord<pyolyokh> that was a 6Dec2022 commit, so maybe by the time it's released you'll be able to get the same thing from nimble
19:35:58FromDiscord<Phil> In reply to @scipio "": So it's fully reliant on clicking with a mouse as the usecase I assume
19:37:45FromDiscord<scipio> In reply to @Isofruit "So it's fully reliant": No, this is just one frontend (html, css, js) module you're seeing here, the current backend is Python Flask & Dask (= distributed parallel Pandas, Numpy, Scipy, etc), i.e. there's a backend process connected to both the frontend and various exchange backend websockets assisting / computing.
19:38:37FromDiscord<Phil> In reply to @scipio "No, this is just": I wasn't talking about the backend at all so far, just about the frontend and what it's primary mode of interaction are
19:38:50FromDiscord<Phil> Which I thiiiiiiiink is clicking with a mouse?
19:39:19FromDiscord<scipio> this module here is for order processing (LB = limit buy, LS = Limit Sell, MB = Market buy, Amend = order ladder offsetting/shifting for batch orders etc)
19:39:21FromDiscord<Phil> Wait no, streamdeck is like a button thing that this is for
19:39:50FromDiscord<scipio> I use the streamdeck version as a secondary "controller" doing the same thing
19:39:58FromDiscord<Phil> So button pressing based?
19:40:01FromDiscord<scipio> both
19:40:12FromDiscord<Phil> Check
19:41:03FromDiscord<nixfreak> To make a true nim based UI framework you would have to create with on bindings only using nim to create the core components
19:41:30FromDiscord<nixfreak> (edit) "on" => "no"
19:42:04FromDiscord<Phil> Actually, might make sense to move this to offtopic
19:42:11FromDiscord<Phil> We're talking about a product and UI Experience, not nim code
19:42:22FromDiscord<cow> In reply to @nixfreak "To make a true": that's an immense task
19:42:31FromDiscord<nixfreak> I know
19:42:45FromDiscord<scipio> So, let's take this as a Nim app use case for a moment ...ā†µ- I could in essence replace the Python backend with Nim and drop in Nimviewā†µ- or rewrite some components with faster Nim binariesā†µ- but if I'd want to include a live realtime charting engine with Nim UI too, then how to proceed?
19:42:47FromDiscord<ambient> In reply to @cow "that's an immense task": eh.. https://github.com/rxi/microui
19:42:56FromDiscord<Phil> ~~The only reason Iced is getting there is out of System76's spite for gnome~~
19:43:16FromDiscord<jtv> Generally, Nim's got a decent foreign function interface
19:43:55FromDiscord<jtv> So it depends on what language, but you could incorporate something in C++, C or JavaScript super easily, and things like Go aren't too hard
19:44:48FromDiscord<Phil> In reply to @ambient "eh.. https://github.com/rxi/microui": Without having any decent knowledge about it, one hard thing that are the main reason why imo bindings are preferrable over rewrites are accessibility features
19:45:12FromDiscord<nixfreak> The library expects the user to provide input and handle the resultant drawing commands, it does not do any drawing itself.
19:45:13FromDiscord<ambient> In reply to @Isofruit "Without having any decent": Yes, the corner cases and OS integration are 99.999% of the work
19:45:26FromDiscord<scipio> In reply to @Isofruit "~~The only reason Iced": Iced: HĆ©ctor RamĆ³n main dev, Artur Sapek Kraken tech lead, main product https://cryptowat.ch/ (however, the crypto watch code is NOT open sourced / included in Iced)
19:45:30FromDiscord<cow> In reply to @nixfreak "The library expects the": it's immediate mode GUI
19:45:35FromDiscord<nixfreak> yep
19:45:53FromDiscord<nixfreak> Just like nuklear now
19:45:57FromDiscord<jtv> Agreed, would love to have a lot more mature, well documented wrappings of important software
19:46:00FromDiscord<scipio> and egui in Rust
19:46:00FromDiscord<nixfreak> and nanoVG
19:47:15FromDiscord<sealmove> could it make sense in some context to export an object's field but not the object type itself?
19:48:08FromDiscord<scipio> To round this (interesting, to me at least) discussion off for now: what I think is Nim's main issue, is that it tries to solve a lot of stuff instead of focusing on something specific. I couldn't easily define in one sentence the Nim's USP vs other languages. ā†µā†µ==> if you'd try to "pitch" to somebody "Why Nim??, what would you answer?
19:48:45FromDiscord<ambient> Well, give me alternative for: fast exploratory programming that generates fast executables
19:48:58FromDiscord<scipio> Slower than V
19:49:02FromDiscord<scipio> Next attempt?
19:49:05FromDiscord<scipio> šŸ˜
19:49:13FromDiscord<jtv> Well, it's the best of C and Python, without their downsides as languages, with some amazingly useful bits like its macro system.
19:49:28FromDiscord<ambient> V has effortless C and C++ FFI?
19:49:38FromDiscord<nixfreak> No not really
19:49:42FromDiscord<nixfreak> just like Zig
19:50:02FromDiscord<ambient> curly braces violate the DRY principle
19:50:06FromDiscord<āš¶ Zeno> sent a long message, see http://ix.io/4l1g
19:50:23FromDiscord<scipio> In reply to @ambient "V has effortless C": That could be an angle for Nim, but then a better language name would be GLUE
19:50:35FromDiscord<āš¶ Zeno> (edit) "http://ix.io/4l1g" => "http://ix.io/4l1h"
19:50:44FromDiscord<nixfreak> what ?
19:51:13FromDiscord<nixfreak> Ok , lets take all of this #offtopic
19:51:31FromDiscord<cow> In reply to @scipio "Slower than V": doesn't V leak memory
19:53:02FromDiscord<pyolyokh> In reply to @cow "doesn't V leak memory": it doesn't do that inherently; it links with boehmgc for garbage collection. That might be a relatively recent change: one of the updates is "V no longer leaks memory".
19:55:05FromDiscord<pyolyokh> the memory management picture and the array index=int=32bits are my go-tos for explaining why I think v's development has a core-language weakness to go with its outside-language (v ui, tooling, etc.) energy
19:56:04*jjido joined #nim
19:56:22*deadmarshal_ quit (Quit: IRCNow and Forever!)
20:17:46*deadmarshal_ joined #nim
20:33:56FromDiscord<scipio> In reply to @pyolyokh "the memory management picture": If you invert this statement, you could distill Nim's strengths from it
20:35:18FromDiscord<pyolyokh> I think those are pretty generic strengths, or rather v is exceptionally poor here.
20:36:39FromDiscord<pyolyokh> D, Odin, Zig, Nim, even Go - they all get the basic stuff right. At the core they're all decent languages that you could probably live with. Nim's strengths vs. this group are more subtle.
20:37:54FromDiscord<pyolyokh> <https://nim-lang.org/blog/2022/11/11/a-cost-model-for-nim.html> and the like.
20:46:09*deadmarshal_ quit (Quit: IRCNow and Forever!)
20:55:51*dv^_^ quit (Remote host closed the connection)
21:09:39*deadmarshal_ joined #nim
21:24:53FromDiscord<deech> In reply to @scipio "To round this (interesting,": I gave a talk on just that: https://www.youtube.com/watch?v=d2VRuZo2pdA
21:25:59FromDiscord<Elegantbeef> What?ā†µ(@scipio)
21:26:22FromDiscord<Elegantbeef> In what world is Nim objectively slower than V?
21:26:36FromDiscord<scipio> In reply to @deech "I gave a talk": Oh hi! I literally watched your talk yesterday! Nice to meet you!
21:27:13FromDiscord<deech> In reply to @scipio "Oh hi! I literally": šŸ‘‹ Hope you liked it!
21:27:21FromDiscord<scipio> I did! A lot!
21:27:49FromDiscord<scipio> But I wouldn't per se distill an answer from that on the question I posed what would be Nim's USP
21:27:53FromDiscord<scipio> a glue language maybe?
21:28:15FromDiscord<Elegantbeef> A modern system programming language that is ergonomic and fast
21:28:59FromDiscord<pyolyokh> this "Nim has good FFI so it can do the FFI for another language like V" take is very, very strange.
21:29:29FromDiscord<Elegantbeef> Well the whole "V is faster than Nim" is a dumb take
21:29:33FromDiscord<scipio> Still, my question is left unanswered @ElegantBeef , and it's a serious question; what UI solution I should use if I wanted to code something like CryptoWatch Desktop with Nim only
21:29:41FromDiscord<scipio> @deech how would you?
21:30:01FromDiscord<scipio> https://media.discordapp.net/attachments/371759389889003532/1063208249375076443/image.png
21:30:34FromDiscord<scipio> this uses Iced btw, and Kraken Exchange (the owner of CryptoWatched) is also the main sponsor of Iced for Rust
21:30:51FromDiscord<scipio> (edit) "CryptoWatched)" => "CryptoWatch)"
21:31:04FromDiscord<pyolyokh> the good FFI lets you write code in Nim. To make the code available to another language you'd have to target the C ABI and you may as well have written a C++ header file (cor C++) and have used whatever language's C FFI. This is not a good use of Nim, you can show that by just trying it.
21:32:20FromDiscord<deech> @scipio I'm working on some GUI bindings right now. Repo's private until it's somewhat more that half baked. They are native (no Chrome runtime) and will for sure work seamlessly on Win/Lin/MacOS but no mobile support is planned.
21:32:52FromDiscord<scipio> In reply to @deech "<@376004371256836096> I'm working on": extremely interested!
21:33:08FromDiscord<deech> (edit) "that" => "than"
21:36:25FromDiscord<scipio> In reply to @pyolyokh "this "Nim has good": I've literally spent 3 days on the Nim ecosystem. I'm trying to assess how usable it could be commercially. Having no battle-hardened UI framework and seeing no job market on Nim after years, raises questions for me. ā†µYou only seem to approach it from a technical perspective; I take a multidimensional stance on it.
21:37:18FromDiscord<pyolyokh> What I have are things I'm interested in talking about and things I'm not interested in talking about, rather than a particular perspective.
21:38:01FromDiscord<scipio> Yet you immediately label me suggesting things as "strange". Not very openminded imho
21:38:20FromDiscord<pyolyokh> No, I said that that particular take was strange. You clearly haven't thought it through.
21:38:25FromDiscord<scipio> LOL
21:38:29FromDiscord<scipio> Okay
21:38:51FromDiscord<deech> There is no Nim job market except possibly Status. I wouldn't expect to see one for a long time if at all. So still definitely in hobbyist bucket.
21:38:53FromDiscord<scipio> You'[d be amazed what I think through but I'm not interested in a pissing contest, I've nothing to prive
21:38:56FromDiscord<scipio> (edit) "prive" => "prove"
21:39:02FromDiscord<Elegantbeef> I'm still stuck on "V is faster than Nim"
21:39:13FromDiscord<pyolyokh> "Nim has good FFI", "can Nim do the FFI for V?" are not questions you've put a lot of thought into.
21:39:23FromDiscord<Elegantbeef> As bad as V is they're both system languages so the speed is generally on the driver
21:39:34FromDiscord<scipio> In reply to @Elegantbeef "I'm still stuck on": I was referring to compile times and it's listed as such on the V website too
21:39:48FromDiscord<Elegantbeef> Ok but vlang lies on their speeds
21:39:52FromDiscord<scipio> In reply to @Elegantbeef "As bad as V": Why is V bad?
21:39:56FromDiscord<Elegantbeef> And also uses an arbitrary benchmark
21:40:09FromDiscord<Elegantbeef> They compare to C using gcc vs vlang + tcc
21:40:14*rockcavera joined #nim
21:40:42FromDiscord<scipio> In reply to @deech "There is no Nim": You'd say any CIO watching your Nim talk would get enthusiastic
21:40:44FromDiscord<pyolyokh> eh, V's compiletimes slow down a great deal with -prod, and if you set up Nim to use tcc like v does by default it'd probably be competitive. But yes the normal v experience has fast compilation.
21:40:44FromDiscord<scipio> I am
21:41:02*tiorock joined #nim
21:41:02*tiorock quit (Changing host)
21:41:02*tiorock joined #nim
21:41:02*rockcavera is now known as Guest9845
21:41:02*tiorock is now known as rockcavera
21:41:09FromDiscord<scipio> Then why is there no job market?
21:41:17FromDiscord<deech> In reply to @scipio "You'd say any CIO": I would? šŸ™‚ I honestly have no idea what a CIOs thought process is.
21:41:55FromDiscord<Elegantbeef> A lot of vlang was built on a promise that didnt exist, "There is no GC and no leaks" was a big thing the touted early on
21:41:55FromDiscord<Elegantbeef> It still hasnt come to fruition afaik
21:41:59*tiorock joined #nim
21:41:59*tiorock quit (Changing host)
21:41:59*tiorock joined #nim
21:41:59*rockcavera quit (Killed (tantalum.libera.chat (Nickname regained by services)))
21:41:59*tiorock is now known as rockcavera
21:42:11FromDiscord<pyolyokh> CTO thought process is "I want this tech stack", and languages are a distraction. You don't want different kinds of lego brands, you want a whole toy.
21:42:32FromDiscord<scipio> For example, I'm friends with Anatoly Yakovenko, the Solana Labs founder, a Rust advocate. Solana being a successful crypto project, an ETH challenger, built with Rust, has definitely generated a lot of attention to Rust itself. Such a driver could also spark Nim attention, yet there isn't one/ Why not?
21:43:23FromDiscord<planetis> There isn't one? Better check your sources first. there is nimbus
21:43:28FromDiscord<deech> AFAICT Nim's only major use is in the crypto space so it's already there.
21:44:38FromDiscord<scipio> In reply to @deech "I would? šŸ™‚ I": I think there are a ton of companies struggling with their legacy C/C++ systems in need of constant patching that a language like Nim (with its meta programming facilities and cross-transpiling) would solve immediatelyā†µIn this regard the 2 things I'm asking myself are the above (-1- why no solid market adoption yet for Nim?) as well as "-2- why is there no "de facto / standard" UI framework for Nim
21:44:40FromDiscord<deech> That's not a space I'm personally interested in so I don't have any more info on why it hasn't penetrated deeper.
21:44:43*Guest9845 quit (Ping timeout: 248 seconds)
21:44:58FromDiscord<enthus1ast> crypto space and malware šŸ™ƒ
21:45:25FromDiscord<scipio> "Red Team Offensive Security", sure
21:45:50*tiorock joined #nim
21:45:50*tiorock quit (Changing host)
21:45:50*tiorock joined #nim
21:45:50*rockcavera is now known as Guest9985
21:45:51*Guest9985 quit (Killed (sodium.libera.chat (Nickname regained by services)))
21:45:51*tiorock is now known as rockcavera
21:45:56FromDiscord<scipio> In reply to @deech "That's not a space": If Nim "glues" so easily, why not ML/AI / data science hopping on the Python toolchain?
21:46:22FromDiscord<scipio> i.e. https://www.dask.org/
21:46:24FromDiscord<deech> Good q, no idea. Does seem like a natural fit.
21:46:50FromDiscord<deech> Marketing? I'm frequently the only one in a group who's even heard of it.
21:47:40FromDiscord<auxym> zig does seem to have strong marketing compared to nim
21:47:42FromDiscord<scipio> If a language has no de facto UI framework, but rather 30 half-baked competing or abandoned repos, and zero tutorials, the entry barrier is raised higher and higher to even try it
21:49:03FromDiscord<pyolyokh> "hey guys, if you just cared about this one thing that's coincidentally something I'm interested in, you would see these wholly unrelated benefits: jobs, marketing"
21:49:55FromDiscord<enthus1ast> a good ui for nim would still be good, i think @scipio is right here
21:50:14FromDiscord<scipio> In reply to @deech "Marketing? I'm frequently the": Do you know https://flet.dev/ ? It's JSON serialized Flutter piped to Python (as first SDK). Feodor Fitsner created it, I'm active with it too.ā†µā†µLiterally daily we get new people in the Discord
21:50:15FromDiscord<pyolyokh> of course it would be good. Lots of things would be good. It is not the thing.
21:50:38FromDiscord<scipio> https://media.discordapp.net/attachments/371759389889003532/1063213441025851412/image.png
21:50:55FromDiscord<ambient> is there any language outside javascript that in any way fills the desktop GUI demand outside the native Windows and Apple development tools?
21:51:01FromDiscord<deech> Depending how what "good" means the bar is so high that you probably need company backing. Bindings are possible but they'll always feel foreign.
21:51:05FromDiscord<ambient> I mean, it's either Electron, Qt or the native tools
21:51:26FromDiscord<deech> (edit) removed "how"
21:51:32FromDiscord<scipio> I suspect it's devs liking the Python toolchain, not versed enough for PySide6 / Qt or Flutter/Dart, looking for a decent easy modern Python UI
21:51:33FromDiscord<enthus1ast> dearimgui
21:52:07FromDiscord<scipio> i.e. a clear UI framework is absolutely essential for a language's success
21:52:34FromDiscord<scipio> ==> if Nim solves this whole, I predict a massive influx of Python devs into the Nim ecosystem
21:52:40FromDiscord<ambient> only clear UI framework is Qt, Windows or Apple. I don't see how that's connected to language
21:53:28FromDiscord<pyolyokh> In reply to @deech "Depending what "good" means": documentation's always more of an issue with bindings than any foreign feel. People don't like having to look up docs in the original project, to see how to do things with the binding. Tcl/Tk's an example that came up earlier, and there you had whole O'Reilly books on how to do things entirely from the perspective of "Perl/tk" and later Tkinter. Tcl? implementation detail, don't worry a
21:53:40FromDiscord<enthus1ast> with an ui framework you can write/sell software for consumers
21:54:32FromDiscord<scipio> A UI framework allows for schools to teach the language as it gives immediate feedback
21:54:39FromDiscord<scipio> See the success of Processing / P5.js
21:54:45FromDiscord<pyolyokh> In reply to @scipio "==> if Nim solves": thanks for at least being very direct with these predictions. That's not the norm.
21:55:03FromDiscord<ambient> Processing is not a UI framework
21:55:57FromDiscord<scipio> In reply to @pyolyokh "thanks for at least": Thanks too. I guess we both have disjunct competences and yours probably isn't with business or social skills. I told you already I'm not interested in a pissing contest. But don't push me. Understood?
21:56:11FromDiscord<scipio> I'/m extremely friendly yet asking questions that need answering
21:56:14FromDiscord<pyolyokh> In reply to @scipio "Thanks too. I guess": No, go fuck yourself.
21:56:25FromDiscord<scipio> šŸ˜
21:56:33FromDiscord<ShalokShalom> In reply to @ambient "is there any language": https://medium.com/swlh/what-makes-godot-engine-great-for-advance-gui-applications-b1cfb941df3b
21:56:41FromDiscord<ShalokShalom> I am going to use it with Kotlin
21:56:50FromDiscord<ambient> In reply to @ShalokShalom "https://medium.com/swlh/what-makes-godot-engine-gre": Ok you got me there, I forgot
21:57:41FromDiscord<ambient> But why Kotlin/Godot? šŸ¤”
21:57:42FromDiscord<enthus1ast> the godot team looked at nim a few years ago to adopt for godot (instead of gdscript)
21:57:55FromDiscord<enthus1ast> or in addition
21:57:59FromDiscord<scipio> In reply to @pyolyokh "No, go fuck yourself.": I honestly don;t know which nerve I struck. I don't care either. You do you.
21:58:00FromDiscord<ambient> In reply to @enthus1ast "the godot team looked": What did they think?
21:58:16FromDiscord<ShalokShalom> In reply to @ambient "But why Kotlin/Godot? šŸ¤”": Cause Kotlin is good supported
21:58:16FromDiscord<ShalokShalom> Godot 4 create a new API
21:58:17FromDiscord<ShalokShalom> Its either Haxe or Kotlin
21:58:26FromDiscord<pyolyokh> In reply to @scipio "I honestly don;t know": You're genuinely surprised that "But don't push me. Understood?" got a hostile response?
21:58:38FromDiscord<enthus1ast> @ambient\: no i guess
21:58:51FromDiscord<scipio> In reply to @pyolyokh "You're genuinely surprised that": Nope, not surprised. Yet you seem to be
21:59:24FromDiscord<auxym> In reply to @ambient "is there any language": rust has a lot of seemingly pretty good GUI options, eg. slint and druid
21:59:48FromDiscord<scipio> In reply to @pyolyokh "You're genuinely surprised that": I'm talking about the cause of your passive-agressive lead-up to that.
21:59:55FromDiscord<ambient> In reply to @auxym "rust has a lot": As far as I'm aware all of these are very immature and none were used to create any big & popular application
22:00:00FromDiscord<Elegantbeef> Children children it'll be ok
22:00:21FromDiscord<scipio> In reply to @auxym "rust has a lot": Druid development is pretty much absent, I think its founder is working on a new repo
22:00:27FromDiscord<ShalokShalom> Daddy solves this
22:01:14FromDiscord<pyolyokh> In reply to @scipio "I'm talking about the": ah, that's much more easily explained. People do this bit all the time. "I predict that if Nim has a first class [random thing I care about] library that people will pour in.", but they are less direct. They don't the prediction, but instead leave it unstated while talking up how a much more popular language (python) has the thing they want.
22:01:26FromDiscord<pyolyokh> (edit) "In reply to @scipio "I'm talking about the": ah, that's much more easily explained. People do this bit all the time. "I predict that if Nim has a first class [random thing I care about] library that people will pour in.", but they are less direct. They don't ... the" added "state"
22:02:20FromDiscord<sealmove> Is it possible to configure testament to use nimcache and not bloat project dir with compiled test modules?
22:02:31FromDiscord<pyolyokh> Someone honestly saying "I would like this thing", I like. Someone saying "Nim will jump five points on the TIOBE index if it has this thing." is at least a nice alternative. The norm gets old.
22:03:04FromDiscord<scipio> In reply to @pyolyokh "ah, that's much more": What you don't seem to be understanding, is that I am _not_ your enemy. I'd love Nim to succeed. It's a fact it doesnt have a job market right now, and it's been around for years. Why is there no demand? I'm asking those questions and looking for solutions. The questions I might ask via hypotheses which you seem to take personally. Don't. Won't end well, nor is it needed. I'd like Nim to succ
22:03:35FromDiscord<auxym> I'm not convinced data science and co is really the killer app compared to python. dynamic features and REPL are a real nice-to-have in that space IMO
22:04:47FromDiscord<pyolyokh> There are Nim kernels for Jupyter, which is the best thing in that space, and there's stuff like Arraymancer.
22:04:54FromDiscord<Elegantbeef> I mean python took how many years to take off
22:05:15FromDiscord<Elegantbeef> A language without massive corporate backing is in an uphill struggle
22:05:19FromDiscord<Elegantbeef> Look at Dlang
22:05:40FromDiscord<ambient> In reply to @pyolyokh "There are Nim kernels": The problem is that the entire notebook is compiled as a single file, so any change = full recompile
22:05:44FromDiscord<pyolyokh> and while python was eating perl's lunch, nobody knew ML would be the final rocket to strap to it. Early days had bioinformatics, maybe.
22:08:32FromDiscord<Elegantbeef> Yea i mean julia didnt explode
22:08:52FromDiscord<Elegantbeef> It's ostensibly a 'better python'
22:10:05FromDiscord<Elegantbeef> But then again i dont see the value in a 'better python' cause python is generally shit
22:10:05FromDiscord<jos> is there a macro that can clean up something like this?
22:10:15FromDiscord<jos> sent a code paste, see https://play.nim-lang.org/#ix=4l1X
22:10:21FromDiscord<jos> i want something more like `B(a: A())`
22:10:29FromDiscord<jos> (edit) "i want something more like `B(a: ... A())`" added "addr"
22:10:36FromDiscord<jos> like u can do in C
22:10:53FromDiscord<jos> `B { a: &A() {} }` works in C and rust
22:11:02FromDiscord<jos> so declaring big literals of nested objects is nice
22:11:14FromDiscord<jos> but in nim A() doesn't have an address yet
22:12:23FromDiscord<Elegantbeef> I can see dangling pointers in your future
22:12:34FromDiscord<jos> its a c ffi
22:12:38FromDiscord<jos> with some really large structs
22:12:48FromDiscord<jos> i just need an easier way to edit them
22:12:53FromDiscord<pyolyokh> In reply to @jos "is there a macro": not a macro, but you can do `B(a: (ref A)())`
22:13:05FromDiscord<Elegantbeef> sent a code paste, see https://play.nim-lang.org/#ix=4l1Z
22:13:11FromDiscord<Elegantbeef> Works but again i dont know why you're using `addr a`
22:13:29FromDiscord<jos> it's because it's a c ffi, is there another way to pass a ptr?
22:13:46FromDiscord<pyolyokh> sent a code paste, see https://play.nim-lang.org/#ix=4l20
22:14:10FromDiscord<Elegantbeef> `ref T` is compatible with `T` assuming you understand the GC implications
22:14:20FromDiscord<Elegantbeef> But nope there is not another way to get a pointer to stack
22:14:34FromDiscord<Elegantbeef> I'm just weary of anyone storing a pointer to a stack value in Nim
22:15:14FromDiscord<jos> the c func won't hold any refs to it, it's a guarantee
22:15:25FromDiscord<jos> sent a code paste, see https://play.nim-lang.org/#ix=4l21
22:15:39FromDiscord<jos> In reply to @Elegantbeef "`ref T` is compatible": i actually can't get this to work for some reason
22:16:12FromDiscord<jos> https://media.discordapp.net/attachments/371759389889003532/1063219871971803228/image.png
22:16:33FromDiscord<deech> In reply to @scipio "What you don't seem": I really like GUI apps but I'm not sure where you see non-web GUI's being a blocker for adoption in this day and age.
22:16:34FromDiscord<Elegantbeef> Well i mean you need to `cast[ptr T](myRef)`
22:16:44FromDiscord<jos> oh i see
22:16:51FromDiscord<jos> well for me the whole point is to clean up the declarations
22:16:54FromDiscord<jos> so i can actually read them
22:17:01FromDiscord<jos> what it does under the hood doesn't really matter that much to me actually
22:17:04FromDiscord<deech> (edit) "In reply to @scipio "What you don't seem": I really like GUI apps but I'm not sure where you see ... non-webGUI" added "lack of" | "GUI's" => "GUI frameworks"
22:17:04FromDiscord<Elegantbeef> Well there is no way to clean this up really
22:17:22FromDiscord<jos> i was hoping there might be a macro or something already since i think this is pretty common for c ffis
22:17:39FromDiscord<Elegantbeef> There isnt any reason for a macro
22:17:52FromDiscord<jos> i have a lot of these and i edit them directly, i want them to be readable
22:17:58FromDiscord<jos> some of them are 20-30 lines long
22:18:00FromDiscord<scipio> In reply to @deech "I really like GUI": speed? with having an html/css/js frontend, the backend/midend message communication becomes a bottleneck. as well as no fullstack approach in one language
22:18:45FromDiscord<pyolyokh> I imagine the thought process isā†µ1. I could clean this up with a procā†µ2. but a proc would have its own stackā†µ3. I could use a macro to put that logic in the caller so that the objects will live on the caller's stack
22:18:47FromDiscord<deech> In reply to @scipio "speed? with having an": I mean which business domain actually wants this?
22:19:26FromDiscord<jos> mmmm i can actually just do it with a proc i think
22:19:32FromDiscord<jos> and a closure
22:19:56FromDiscord<Elegantbeef> Let's make something simple, much more complex
22:19:57FromDiscord<deech> (edit) "this?" => "this enough that would justify the effort."
22:20:05FromDiscord<Elegantbeef> What's your entire code
22:21:05FromDiscord<Elegantbeef> if you share an image, i swear i'm sending a rocket to your house
22:21:26FromDiscord<scipio> In reply to @deech "I mean which business": let me flip that question: why would Google spend so much effort on developing and branding Flutter/Dart?
22:22:08FromDiscord<scipio> Flutter 3 isn't just targeting mobile but also Desktop
22:22:15FromDiscord<scipio> macOS, Windows
22:23:41FromDiscord<scipio> Firebase monetization? By deliberately not focusing on Dart low level?
22:24:13FromDiscord<scipio> Streaming UI as a service?
22:24:41FromDiscord<deech> In reply to @scipio "let me flip that": So, a unified API for mobile+web+desktop is a huge selling point but it takes Google resources to pull it off. No niche language community can hope to pull that off, it's hard enough to maintain even bindings.
22:26:25FromDiscord<deech> (edit) "pull that off," => "compete,"
22:27:25FromDiscord<scipio> @deech off-topic, but when Google was still called Backrub I was on the losing end in web search competing with it. Stanford vs TU/Eindhoven
22:27:30FromDiscord<ShalokShalom> In reply to @scipio "let me flip that": Market dominance
22:27:57FromDiscord<!!sharpcdf!!> how can i use `walkDirRec` to copy files, following and copying folder too?
22:27:58FromDiscord<ShalokShalom> https://blog.bencope.land/flutter-state-in-f-with-fable-and-redux/
22:28:11FromDiscord<ShalokShalom> In case you love to build a Flutter app
22:28:24FromDiscord<scipio> Well it's not that off-topic giving it some more thought.
22:28:26FromDiscord<!!sharpcdf!!> (edit) "how can i use `walkDirRec` to copy files, following and copying folder too? ... " added "ive tried using `copyFileToDir` but it just copies them all to the same dir"
22:28:47FromDiscord<Elegantbeef> if you dont need a filter `copyDir` existsā†µ(@!!sharpcdf!!)
22:28:53FromDiscord<scipio> If something is novel enough, like chatGPT or stable diffusion the world can be taken by storm
22:29:14FromDiscord<ShalokShalom> You overthink this
22:29:21FromDiscord<ShalokShalom> Look how many employees they have
22:29:30FromDiscord<Elegantbeef> What are you after sharp
22:29:36FromDiscord<scipio> Back then it was just Brin and Page vs us two
22:29:40FromDiscord<ShalokShalom> Companies like Google allocate their resources different
22:29:41FromDiscord<scipio> I'm talking 1996
22:29:50FromDiscord<ShalokShalom> They dont care about a small team like Flutter
22:29:59FromDiscord<ShalokShalom> For them, its an experiment
22:30:33FromDiscord<Elegantbeef> sent a code paste, see https://play.nim-lang.org/#ix=4l26
22:30:36FromDiscord<ShalokShalom> That's petty cash to them
22:30:41FromDiscord<deech> @scipio When my thing is in better shape I'll announce it and tag you here if you're still around. If you feel that strongly about GUI I wouldn't mind the help.
22:30:47FromDiscord<!!sharpcdf!!> In reply to @Elegantbeef "if you dont need": oh snap i never saw that
22:30:49FromDiscord<!!sharpcdf!!> thanks
22:31:03FromDiscord<scipio> In reply to @ShalokShalom "They dont care about": I don't think Flutter is merely an experiment to Google in terms of strategy
22:31:09FromDiscord<!!sharpcdf!!> but can i copy the directory with a new name?
22:31:11FromDiscord<Elegantbeef> Deech it's not the shape that matters it's what it can do
22:31:25FromDiscord<!!sharpcdf!!> oh yea i can, thanks
22:31:32FromDiscord<ShalokShalom> In reply to @scipio "I don't think Flutter": I can tell you how those things are at Microsoft
22:31:40FromDiscord<ShalokShalom> And I doubt Google is any different
22:32:00FromDiscord<ShalokShalom> They love some market share and popularity among the programming crowd
22:32:09*rockcavera quit (Remote host closed the connection)
22:32:10FromDiscord<ShalokShalom> You know why they do this?
22:32:11FromDiscord<ShalokShalom> Jobs
22:32:12FromDiscord<Elegantbeef> Fuschia isnt an experiment either amirite
22:32:21FromDiscord<planetis> like people who complain ever lift a finger
22:32:33FromDiscord<ShalokShalom> For Google, everything is an experiment
22:32:43FromDiscord<ShalokShalom> Unless it brings really money
22:32:53FromDiscord<ShalokShalom> In reply to @planetis "like people who complain": Who complains?
22:33:04FromDiscord<Elegantbeef> Hey they have to move their keys to type their strongly worded messagesā†µ(@planetis)
22:33:21FromDiscord<Elegantbeef> Shalok is like "Shit are they on to me"
22:33:27FromDiscord<planetis> In reply to @ShalokShalom "Who complains?": not you
22:33:54FromDiscord<planetis> jeez wrong fish took the bait
22:34:07FromDiscord<pyolyokh> In reply to @planetis "like people who complain": if there's a boulder sitting in the middle of the road, people feel better about joining a group that's pushing, than going to push it themselves. Time isn't free, and if it's only one person the job might not even be worth doing.
22:34:45FromDiscord<Elegantbeef> pyolyokh explaining the bystander effect
22:34:54FromDiscord<planetis> no you missed the other group that shits in the side judging their form while they push
22:35:04FromDiscord<pyolyokh> I'm just saying it doesn't hurt to know who to ping, when you want help on your project that's in line with their interests
22:35:35FromDiscord<Elegantbeef> Planetis is now calling me out
22:37:02FromDiscord<planetis> don't be harsh on yourself beef, youre always helpful to others
22:37:36FromDiscord<Elegantbeef> Hey i'll use self deprecating humour if i want mom!
22:38:52FromDiscord<planetis> deprecate and rewrite away!
22:41:59FromDiscord<scipio> In reply to @pyolyokh "I'm just saying it": But serious question @pyolyokh : if you were to write an app/service like cryptowatc.ch desktop entirely in Nim, which (current) UI solution would you implement?
22:42:58FromDiscord<scipio> (edit) "cryptowatc.ch" => "cryptowat.ch"
22:43:42FromDiscord<scipio> (it being written in Rust right now, including development of Iced)
22:44:11FromDiscord<pyolyokh> sent a long message, see https://paste.rs/zQW
22:44:24FromDiscord<pyolyokh> (edit) "http://ix.io/4l2a" => "https://paste.rs/aoG"
22:45:32FromDiscord<scipio> https://cryptowat.ch/apps/desktop
22:46:05FromDiscord<scipio> it's being deployed currently as a tool to model your own dashboards, no coding required
22:47:59FromDiscord<Elegantbeef> Had to check, and oh owlkettle does have GL area so custom widgets like a graph could be made
22:53:48FromDiscord<Elegantbeef> But i guess in that case good bye mac support
22:54:10FromDiscord<pyolyokh> In reply to @scipio "it's being deployed currently": it looks good, and for something really slick there's probably a lot of performance that you can gain by having something more oriented around trading data than grafana just pairing queries with blobs of javascript to render the numbers the query gets back.
22:59:14*rockcavera joined #nim
23:00:02FromDiscord<scipio> In reply to @Elegantbeef "But i guess in": right. But CryptoWatch is also accepted in/ being distributed via the Mac AppStore
23:00:23FromDiscord<Elegantbeef> Nice, tell apple to add opengl support
23:01:37FromDiscord<Elegantbeef> Though i guess they still support opengl4.1 but idk
23:03:51*jjido quit (Quit: My laptop has gone to sleep. ZZZzzzā€¦)
23:08:37FromDiscord<ambient> Isn't OGL 3.3 already decent enough if you don't want something like compute shaders
23:08:44FromDiscord<Elegantbeef> Yea
23:08:56FromDiscord<Elegantbeef> I didnt know if modern m1s supported opengl 4.1 aswell
23:10:21FromDiscord<Elegantbeef> So in closing gtk's gl area is sufficient to make graphs if someone really wanted to
23:11:30FromDiscord<jos> do you even need hardware acceleration for a graph nowadays
23:11:57FromDiscord<ambient> you never did
23:12:03FromDiscord<Elegantbeef> Well you'd want to draw text on the graph on hover and the like
23:12:17FromDiscord<Elegantbeef> So i imagine it's more intelligent to use glarea than using a drawing area
23:12:31FromDiscord<jos> u could use the os primitives too
23:12:44FromDiscord<jos> i think stuff like cairo is good
23:12:46FromDiscord<Elegantbeef> We're talking about gtk
23:12:57FromDiscord<Elegantbeef> But yea drawing area or glarea would be sufficient
23:13:46FromDiscord<T0lk1en> Hey how do you get a binary value from a string
23:14:09FromDiscord<T0lk1en> If possible^
23:14:59FromDiscord<enthus1ast> @T0lk1en\: https://nim-lang.org/docs/strutils.html#toBin%2CBiggestInt%2CPositive
23:16:27FromDiscord<enthus1ast> sent a code paste, see https://paste.rs/zI4
23:17:07FromDiscord<T0lk1en> That seems to only work for numbers
23:17:30FromDiscord<T0lk1en> I want to convert chars into a value I can do math on
23:17:35FromDiscord<T0lk1en> Thatā€™s my goal
23:17:53FromDiscord<enthus1ast> then`'a'.int`
23:18:02FromDiscord<T0lk1en> Btw why does so much code have goo and bar in it
23:18:02FromDiscord<enthus1ast> 'a'.uint8
23:18:06FromDiscord<pyolyokh> In reply to @T0lk1en "Hey how do you": as a rule, loop over the bytes of the string and build new data according to the format. For a completely random binary format you might need custom code, maybe with some help from std/endians.ā†µIf you have some specific binary format there might be a serialization module for it in the nimble directory, and if you want to store data to/from binary strings in general then "nimble search serial"
23:18:39FromDiscord<T0lk1en> In reply to @pyolyokh "as a rule, loop": Ty
23:18:40FromDiscord<pyolyokh> In reply to @T0lk1en "Btw why does so": it's just tradition: <https://en.wikipedia.org/wiki/Metasyntactic_variable>
23:19:06FromDiscord<T0lk1en> Funny
23:19:32FromDiscord<T0lk1en> Iā€™m just trying to turn a char into a number I can do math on and also turn it back. Is there anything like that out there
23:19:39FromDiscord<T0lk1en> Doesnā€™t have to be binary
23:20:09FromDiscord<pyolyokh> for that a cast is enough, `int(c)`
23:20:17FromDiscord<T0lk1en> What is cast
23:20:23FromDiscord<T0lk1en> Wait thatā€™s it?
23:20:28FromDiscord<pyolyokh> yep
23:20:34FromDiscord<Elegantbeef> That's a type conversion
23:20:39FromDiscord<Elegantbeef> Dont call it a cast
23:20:50FromDiscord<enthus1ast> i know it beef
23:20:58FromDiscord<enthus1ast> itched your fingers
23:21:00FromDiscord<ambient> int(c) doesn't return just the ordinal?
23:21:04FromDiscord<pyolyokh> yeah, to be clear `cast[char](300)` will be silently accepted, but `char(300)` will throw an error. type conversions are safer
23:21:14FromDiscord<Elegantbeef> It does `ord(char)` and `int(char)` should be the same
23:21:33FromDiscord<ambient> but if I want "0" to be 0, then that's not the ordinal
23:21:39FromDiscord<T0lk1en> Wtf is a cast
23:21:45FromDiscord<T0lk1en> What is ordinal
23:21:59FromDiscord<Elegantbeef> Cast in Nim reinterprets the binary data as a different type
23:22:11FromDiscord<T0lk1en> Okay
23:22:11FromDiscord<ambient> echo $("234".parseInt)
23:22:38FromDiscord<Elegantbeef> `int` and `float` are the same size on 64bit OS but `cast[int](32.2) != int(32.2)`
23:22:49FromDiscord<ambient> int is 64-bit on Mac?
23:22:52FromDiscord<pyolyokh> In reply to @T0lk1en "Wtf is a cast": programmer jargon: <https://en.wikipedia.org/wiki/Type_conversion> - a bad name for the previous conversation. The Nim manual has the terminology that Nim uses.ā†µordinal is math jargon: <https://en.wiktionary.org/wiki/ordinal>
23:23:04FromDiscord<Elegantbeef> On 64 bit osā†µ(@ambient)
23:23:22FromDiscord<ambient> Strange, because my code breaks on mac with int, instead of int64
23:23:29FromDiscord<pyolyokh> int is the size of the pointer on your architecture: <https://nim-lang.org/docs/manual.html#types-preminusdefined-integer-types>
23:23:32FromDiscord<ambient> On Windows it works fine with just int
23:24:19FromDiscord<T0lk1en> Wait Iā€™m not trying to turn a valid number that is stored as a chat to an int
23:24:32FromDiscord<T0lk1en> I want to turn a letter into an int
23:24:44FromDiscord<T0lk1en> Char^
23:24:51FromDiscord<pyolyokh> sent a code paste, see https://play.nim-lang.org/#ix=4l2n
23:25:22FromDiscord<T0lk1en> Iā€™ll be damned
23:25:23FromDiscord<T0lk1en> Ty
23:40:33FromDiscord<T0lk1en> Hey is there a way to tell a sequence how long it should be
23:41:13FromDiscord<T0lk1en> Iā€™m trying to run an empty sequence through a for loop howveee because it is empty I get out of bounds error
23:41:45FromDiscord<T0lk1en> This would be solved if I could tell a sequence how long it should be without filling it. However the length also changes so I canā€™t use an array
23:44:36FromDiscord<pyolyokh> it sounds like you want `newSeqOfCap`, but that's not going to save you from an out of bounds error. Maybe you want an initial and then a final `setLen`, initially for space to work, and finally after you've decided on a length
23:45:32FromDiscord<T0lk1en> How do you do setlen
23:46:49FromDiscord<pyolyokh> sent a code paste, see https://play.nim-lang.org/#ix=4l2q
23:48:12FromDiscord<T0lk1en> Ty
23:48:23FromDiscord<T0lk1en> You guys in this discord have helped sm
23:50:26FromDiscord<!!sharpcdf!!> how can i clear the screen of an illwill tui when it isnt fullscreen?
23:54:16FromDiscord<pyolyokh> like, to clear it on startup even if you're only going to write to part of the screen? Manually, I guess, with `echo "\e[2J"`
23:54:56FromDiscord<pyolyokh> why aren't you going fullscreen?
23:56:16FromDiscord<enthus1ast> @!!sharpcdf!!\: setting the cursor to 0 , 0 then overwrite stuff should work
23:56:31FromDiscord<pyolyokh> when writing to a terminal buffer the size of the screen, `illwillInit(fullscreen=true)` or =false didn't seem to make a difference.