00:12:00 | * | derpydoo joined #nim |
00:13:45 | FromDiscord | <threefour> In reply to @riku5543 "Honestly the only problem": The home key returns you to the top |
00:15:00 | FromDiscord | <riku5543> That's good to know lmao |
01:29:55 | FromDiscord | <double_spiral> uhm https://media.discordapp.net/attachments/371759389889003532/1250260760660283492/Untitled.png?ex=666a4b92&is=6668fa12&hm=a9553c2c8f13cdc73258a2b04f236a50513457d1b2af1ae512336c9b11b0aa9f& |
01:40:16 | * | rockcavera quit (Remote host closed the connection) |
01:40:40 | FromDiscord | <riku5543> If you're using vscode extensions it likes to spawn 1000 copies of nimsuggest |
01:43:02 | * | rockcavera joined #nim |
01:58:39 | FromDiscord | <kots> If you're using choosenim and nimlangserver, it's probably this: https://github.com/dom96/choosenim/issues/339 |
02:26:39 | FromDiscord | <albassort> When did we destroy tainted strings and why |
02:27:00 | FromDiscord | <ElegantBeef> Years ago, bad UX |
02:27:13 | FromDiscord | <albassort> I know we did years ago but I just don't know exactly when I stopped having to deal with them |
02:27:33 | FromDiscord | <albassort> I do agree though, more than two different kinds of strings is a bit silly because you never know the context |
02:30:11 | FromDiscord | <ElegantBeef> Well it was more of a noob trap |
02:33:17 | * | beholders_eye quit (Read error: Connection reset by peer) |
02:38:46 | * | beholders_eye joined #nim |
02:51:41 | * | beholders_eye quit (Read error: Connection reset by peer) |
03:00:59 | * | krux02 quit (Ping timeout: 268 seconds) |
04:46:49 | FromDiscord | <Phil> What a kind way to say that it'd absolutely screw me over |
04:46:56 | FromDiscord | <Phil> (edit) "What" => "That is" |
05:02:28 | FromDiscord | <albassort> internally they're refereed to as "Phil traps" |
05:33:09 | * | redj quit (Quit: No Ping reply in 180 seconds.) |
05:33:17 | * | ntat joined #nim |
05:35:02 | * | redj joined #nim |
05:57:41 | * | PMunch joined #nim |
05:58:31 | * | tiorock joined #nim |
05:58:31 | * | rockcavera is now known as Guest5394 |
05:58:31 | * | Guest5394 quit (Killed (iridium.libera.chat (Nickname regained by services))) |
05:58:31 | * | tiorock is now known as rockcavera |
06:02:09 | FromDiscord | <nnsee> In reply to @riku5543 "Honestly the only problem": 🥲 when did we stop teaching people what the buttons on keyboards do |
06:02:42 | FromDiscord | <nnsee> (no offense to you personally) |
06:03:21 | FromDiscord | <riku5543> Eh I mostly only use the mouse in a web browser so I didn't really know the shortcuts besides tab for navigation and space for selection. (And stuff like open tab, close tab, close browser) |
06:03:33 | FromDiscord | <nnsee> well it's not really a shortcut |
06:03:41 | FromDiscord | <nnsee> it's a dedicated button with a single purpose |
06:03:57 | FromDiscord | <nnsee> there's actually two of those buttons on your keyboard! (if you use a full size keyboard) |
06:04:15 | FromDiscord | <nnsee> numpad 7 with numlock off also acts as `home` |
06:04:20 | FromDiscord | <riku5543> Also the home and end keys are a little inaccessible on this laptop \:P They're on the keypad at least but it's unlabelled so I haven't developed the muscle memory for them yet |
06:04:31 | FromDiscord | <nnsee> that's fair |
06:05:44 | FromDiscord | <riku5543> (It's normally on the same row as the function keys which is atrocious) So I had to unlearn using them for programming and stuff but now I wanna get the muscle memory back using the keypad cause I really miss it haha. Oh and also, page up and page down are right next to the up arrow key 🥲 |
06:06:04 | FromDiscord | <riku5543> (((Really funny in a terminal))) |
06:07:51 | FromDiscord | <nnsee> i had a laptop with that layout, i low-key kind of preferred it actually |
06:16:26 | FromDiscord | <odexine> “Haha, low key” |
06:32:41 | FromDiscord | <nnsee> damn it Rika |
06:32:45 | FromDiscord | <nnsee> stop being funny |
06:36:52 | * | rockcavera quit (Remote host closed the connection) |
06:42:28 | * | jjido joined #nim |
07:09:53 | FromDiscord | <wannabelokesh> what are the possible values for `nim -d` and what exactly does it do? |
07:10:07 | * | ntat quit (Quit: Leaving) |
07:17:32 | FromDiscord | <riku5543> You can control specific nim features with it, like compiling the binary in release mode or including ssl support. I don't know every option for it though |
07:21:24 | FromDiscord | <riku5543> https://media.discordapp.net/attachments/371759389889003532/1250349218489434152/image.png?ex=666a9df4&is=66694c74&hm=3832b86d0644c06405166485af12ef518b2ea665f4a4eb4d1dbaec53f3341fd7& |
07:21:38 | FromDiscord | <riku5543> I found this table [here](https://nim-lang.org/docs/nimc.html) that shows them |
07:30:27 | * | derpydoo quit (Ping timeout: 252 seconds) |
07:33:32 | FromDiscord | <wannabelokesh> Thanks. |
07:40:08 | * | jjido quit (Quit: My laptop has gone to sleep. ZZZzzz…) |
07:46:18 | * | sadome joined #nim |
07:46:18 | * | sadome quit (Changing host) |
07:46:18 | * | sadome joined #nim |
07:46:19 | * | sadome quit (Excess Flood) |
07:50:02 | FromDiscord | <albassort> What happens if segfault while -d:noSignalHandler |
07:51:21 | FromDiscord | <albassort> In reply to @wannabelokesh "Thanks.": there are infinite options. -d: is not compiler implemented, so they can be arbitrary. You can detect them using if defined("") |
07:51:38 | FromDiscord | <albassort> (edit) "In reply to @wannabelokesh "Thanks.": there are infinite options. -d: is not compiler implemented, so they can be arbitrary. You can detect them using if defined("")" => "sent a code paste, see https://play.nim-lang.org/#pasty=GYcRjaFf" |
07:52:24 | FromDiscord | <albassort> sent a code paste, see https://play.nim-lang.org/#pasty=OoBbjPhK |
07:52:31 | FromDiscord | <albassort> (edit) "https://play.nim-lang.org/#pasty=OQeTFanG" => "https://play.nim-lang.org/#pasty=bkDTjEGq" |
07:52:39 | FromDiscord | <albassort> (edit) "https://play.nim-lang.org/#pasty=CHFUzAPX" => "https://play.nim-lang.org/#pasty=ziMSkyvd" |
07:53:37 | FromDiscord | <albassort> sorry it doesn't take a string variable |
07:53:46 | FromDiscord | <albassort> sent a code paste, see https://play.nim-lang.org/#pasty=ckKCOqgM |
07:53:52 | FromDiscord | <albassort> this is how you do it properly |
07:54:06 | FromDiscord | <albassort> and this would check if you compiled with↵-d:js |
07:54:46 | FromDiscord | <zumi.dxy> In reply to @albassort "What happens if segfault": well you get the default signal handler↵instead of `(Attempt to read from nil?)` you get `(Core dumped)` |
08:17:09 | * | spine-o-saurus joined #nim |
08:19:20 | FromDiscord | <pmunch> For some reason the IRC bridge sent your message to the #offtopic room @wannabelokesh, so I answered with some more details there |
08:21:04 | * | gst joined #nim |
08:21:27 | FromDiscord | <albassort> sent a long message, see https://pasty.ee/iePMoRKr |
08:28:31 | FromDiscord | <wannabelokesh> In reply to @pmunch "For some reason the": Hey, sorry my previous question was a little misguiding. I apologize, didn't mean to waste your time. ↵I just wanted to see if it doesn't do something global with the flag -d. ↵I'm not learning tooling thing. Before that (detailed in off-topic) I was only doing `compile -r` . |
08:28:44 | FromDiscord | <wannabelokesh> (edit) "thing." => "thing for now." |
08:29:23 | FromDiscord | <albassort> well, i figured. I just thought more context would help |
08:29:38 | FromDiscord | <wannabelokesh> In reply to @albassort "there are infinite options.": i don't understand "inifite options" thing |
08:30:03 | FromDiscord | <albassort> its arbitrary, you can put anything in -d: and you can check for it yourself |
08:30:15 | FromDiscord | <albassort> meaning, if your code calls for a compile time options, you can use -d: for it |
08:30:35 | FromDiscord | <albassort> so getting a table of options might appear useful, but its a case by case basis for each module |
08:31:27 | * | spine-o-saurus quit (Ping timeout: 264 seconds) |
08:32:40 | FromDiscord | <wannabelokesh> In reply to @albassort "its arbitrary, you can": is it something like ↵ifdef and ifndef in c? |
08:33:00 | FromDiscord | <albassort> I've never worked with large C projects so idk |
08:33:03 | FromDiscord | <albassort> i assume so |
08:33:16 | FromDiscord | <wannabelokesh> no it's basic c, frankly saying. |
08:33:42 | FromDiscord | <albassort> by which i meant, my c has only ever been a part of another language's code |
08:36:13 | FromDiscord | <nnsee> In reply to @wannabelokesh "is it something like": nim's ifdef is `when defined(whatever): echo "i'm defined!"` |
08:36:31 | FromDiscord | <nnsee> and yeah `-d` functions with that |
08:38:40 | FromDiscord | <wannabelokesh> sorry I'm personally a C lover, but I don't wanna write C yet. I am too much amazed by systems programming. That's why I wanted to try out Nim. Rust is overhyped, no, I'll definitely do rust, but not currently. I wanna independent kernel things but not in c/cpp, idk it can be stupid idea. |
08:39:01 | FromDiscord | <wannabelokesh> (edit) "sorry I'm personally a C lover, but I don't wanna write ... C" added "(professional)" |
08:39:29 | FromDiscord | <wannabelokesh> In reply to @nnsee "nim's ifdef is `when": okay, I guess I guessed right.. |
08:40:43 | FromDiscord | <albassort> In reply to @wannabelokesh "sorry I'm personally a": Hmm. Well. C will get you as far as your knowledge of engineering will take you, generally speaking. I do like C but I feel if you cost benefit the time you spend making sure everything is alright then it doesn't make much sense to use compared to Rust/Nim |
08:41:25 | FromDiscord | <albassort> C also I feel gives you the illusion of infinite support dating back 50 years but in reality code can be so archaic its not worth using. |
08:41:38 | FromDiscord | <albassort> (edit) "C also I feel gives you the illusion of infinite support dating back 50 years but in reality code can be so archaic ... its" added "/ dangerous (because C is very easy to mess up)" |
08:42:56 | FromDiscord | <albassort> I like C conceptually more than rust I think though. |
08:43:14 | * | spine-o-saurus joined #nim |
08:43:16 | FromDiscord | <wannabelokesh> with great powers, come great responsibilities (which not everyone can handle.. I guess) |
08:43:34 | FromDiscord | <wannabelokesh> In reply to @albassort "I like C conceptually": conceptually... |
08:43:48 | FromDiscord | <albassort> It's a matter of practicality and productivity and eventually this becomes important to you lol |
08:43:55 | FromDiscord | <albassort> I'm not sure when it happened though for me |
08:44:42 | FromDiscord | <albassort> It's why things like C++ beat out C in the professional world, even though it's a mess it's a less dangerous mess and slightly more productive |
10:05:25 | * | spine-o-saurus quit (Quit: snakeeyes!) |
10:10:41 | * | gst quit (Quit: Client closed) |
10:27:32 | NimEventer | New Nimble package! libcapstone - Futhark generated wrapper around libcapstone, see https://github.com/m4ul3r/libcapstone-nim |
10:28:18 | * | beholders_eye joined #nim |
10:31:09 | FromDiscord | <fabric.input_output> how do I set flags for nimble in a project? |
10:31:34 | FromDiscord | <fabric.input_output> doing `--define:"mingw"` in the `config.nims` doesn't make `nimble build` use mingw |
10:40:53 | FromDiscord | <Robyn [She/Her]> You could do `{.define: "mingw"}` at the top of your main file? |
10:44:09 | FromDiscord | <fabric.input_output> I guess I could |
10:54:15 | * | beholders_eye quit (Ping timeout: 264 seconds) |
10:57:24 | FromDiscord | <albassort> @Robyn [She/Her] how u |
11:05:33 | Amun-Ra | why not just do it a proper way: nimble -d=mingw build |
11:08:08 | FromDiscord | <Robyn [She/Her]> In reply to @albassort "<@524288464422830095> how u": Dead inside xD |
11:08:15 | FromDiscord | <Robyn [She/Her]> Sick with a horrid migraine, hbu? |
11:08:26 | Amun-Ra | hmm, define in nim.cfg works |
11:08:27 | FromDiscord | <fabric.input_output> In reply to @Amun-Ra "why not just do": that's how I have been doing it until now |
11:08:32 | FromDiscord | <albassort> somewhat productive, figuring out how to pass error-codes between rust and nim |
11:08:40 | FromDiscord | <albassort> im settling on a enum passed via pointer |
11:09:06 | Amun-Ra | fabric: try nim.cfg |
11:09:43 | FromDiscord | <fabric.input_output> I tried, it used `gcc` and linker errors occurred. I think it overrides the default `nim.cfg` or something |
11:09:46 | Amun-Ra | I redefine mingw compilers in my local setup, perhaps that's why it works but it may work also for global settings |
11:09:47 | FromDiscord | <albassort> in nim you can passl and passc |
11:09:52 | FromDiscord | <albassort> why cant you passd in a .nims |
11:10:11 | FromDiscord | <albassort> although, im not really sure the use case for -d:mingw? |
11:10:15 | FromDiscord | <albassort> (edit) "-d:mingw?" => "-d:mingw in the .nims?" |
11:10:43 | Amun-Ra | I'm compiling my nim projects via make, I don't have to manually add defines |
11:12:29 | FromDiscord | <albassort> i... well... ok then |
11:13:22 | FromDiscord | <fabric.input_output> the name of the `.nims` file is `config.nims`? or is it `$project.nims`? |
11:13:42 | FromDiscord | <Robyn [She/Her]> In reply to @albassort "somewhat productive, figuring out": Oooh |
11:13:48 | Amun-Ra | config.nims |
11:14:03 | FromDiscord | <Robyn [She/Her]> In reply to @albassort "im settling on a": Why not just return an int? It's basically free to copy |
11:14:21 | FromDiscord | <albassort> well I'd rather use that for the intended return type |
11:14:47 | FromDiscord | <albassort> i've seen c libraries doing it both ways |
11:15:07 | Amun-Ra | albassort: -d=mingw is used for crosscompiling for windows |
11:15:29 | FromDiscord | <albassort> yeah but you'd never want it defined in your .nims |
11:15:34 | FromDiscord | <fabric.input_output> In reply to @albassort "why cant you passd": passd doesn't exist, I think you mean `--define:` |
11:15:48 | FromDiscord | <albassort> and that hasn't worked? |
11:15:53 | FromDiscord | <fabric.input_output> nope |
11:16:48 | FromDiscord | <Robyn [She/Her]> In reply to @albassort "i've seen c libraries": Fair enough |
11:28:30 | * | acidsys quit (Ping timeout: 268 seconds) |
11:44:04 | * | ntat joined #nim |
11:46:02 | * | acidsys joined #nim |
11:49:28 | * | krux02 joined #nim |
11:50:17 | FromDiscord | <sOkam! 🫐> what does the latest version of `{.pure.}` for enums do exactly? |
12:02:07 | FromDiscord | <odexine> nothing iirc |
12:03:01 | FromDiscord | <Robyn [She/Her]> In reply to @albassort "well I'd rather use": I don't get it :derp: |
12:03:09 | FromDiscord | <Robyn [She/Her]> In reply to @heysokam "what does the latest": Pure does nothing |
12:03:28 | FromDiscord | <Robyn [She/Her]> But iirc it used to enforce accessing the enum via a dot expr |
12:03:47 | FromDiscord | <Robyn [She/Her]> (So `Foo.Bar` to access `Bar`) |
12:04:00 | FromDiscord | <sOkam! 🫐> it did do that yeah |
12:04:14 | FromDiscord | <sOkam! 🫐> that's why I was asking for the new version's behavior |
12:04:47 | FromDiscord | <sOkam! 🫐> well, that's dumb. because it was a useful feature 😦 |
12:13:11 | FromDiscord | <sOkam! 🫐> does anyone know what `nnkEnumTy()[0]` is used for?↵The macros manual only lists them as `nnkEmpty` with no other examples |
12:13:17 | FromDiscord | <Robyn [She/Her]> In reply to @heysokam "well, that's dumb. because": Was it? How so? |
12:13:42 | FromDiscord | <Robyn [She/Her]> In reply to @heysokam "does anyone know what": I can't think of any reason tbh |
12:14:00 | FromDiscord | <sOkam! 🫐> In reply to @chronos.vitaqua "Was it? How so?": same way `var one: Type = 1` is useful. explicitness↵there is also some edge cases I don't remember, that were very very useful for binding generation |
12:14:07 | FromDiscord | <Robyn [She/Her]> In reply to @chronos.vitaqua "Was it? How so?": You can still fully qualify your access it |
12:14:22 | FromDiscord | <Robyn [She/Her]> (edit) "In reply to @chronos.vitaqua "Was it? How so?": You can still fully qualify your access ... it" added "to" |
12:14:26 | FromDiscord | <sOkam! 🫐> But you don't get name collisions if you use Pure |
12:14:58 | FromDiscord | <Robyn [She/Her]> If the names collide I'm pretty sure you're forced to fully qualify access anyway |
12:14:58 | FromDiscord | <sOkam! 🫐> literally the field was not listed, but they were still found correctly when implying the type in an argument, etc |
12:15:13 | FromDiscord | <sOkam! 🫐> In reply to @chronos.vitaqua "If the names collide": you were not |
12:15:15 | FromDiscord | <sOkam! 🫐> now you are |
12:15:29 | FromDiscord | <Robyn [She/Her]> Yeah I'm talking about now |
12:15:47 | FromDiscord | <Robyn [She/Her]> I'd say that's a fine reason to remove functionality from `pure` |
12:16:11 | FromDiscord | <sOkam! 🫐> i don't see it as a good reason |
12:16:21 | FromDiscord | <sOkam! 🫐> what damage was it causing? |
12:17:19 | FromDiscord | <Robyn [She/Her]> The loss of a developer's choice to write code how they want to? 🤷♀️ |
12:17:44 | FromDiscord | <Robyn [She/Her]> I don't see any harm for pure existing or not, but also I'm a newer Nim user than most active Nimians here |
12:18:13 | FromDiscord | <sOkam! 🫐> well that's what I was doing before, now its not allowed to do `Thing.one` and `one = 1` because both `one` are considered a valid symbol that collides and must be name-qualified |
12:18:44 | FromDiscord | <sOkam! 🫐> which wasn't happening when you marked the enum as pure before |
12:19:58 | FromDiscord | <sOkam! 🫐> with pure before, enum values didn't exist in the namespace at all, unless qualified↵if you wanted them to exist without qualifying, you could just not mark them as pure |
12:20:31 | FromDiscord | <sOkam! 🫐> now both Pure and non-pure both exist in the namespace and collide 🤦 which basically makes no sense to me why pure even exists now |
12:21:02 | FromDiscord | <Robyn [She/Her]> Yeah but other developers besides you using your libraries may prefer not to qualify enum access :P |
12:21:12 | FromDiscord | <sOkam! 🫐> overloadable enums i understand. removing pure? i don't 🤷♂️ |
12:21:12 | FromDiscord | <Robyn [She/Her]> In reply to @heysokam "now both Pure and": Pure exists purely for backwards compatibility, I'd imagine |
12:21:46 | FromDiscord | <sOkam! 🫐> In reply to @chronos.vitaqua "Yeah but other developers": then you would have used type inference |
12:21:52 | FromDiscord | <sOkam! 🫐> which also worked for pure enums |
12:21:57 | FromDiscord | <Robyn [She/Her]> Wdym? |
12:22:38 | FromDiscord | <sOkam! 🫐> afair, `one` would resolve to `Thing.one` if you were assigning to a `Thing` argument or variable |
12:23:02 | FromDiscord | <sOkam! 🫐> (in the context of `Thing {.pure.}` of course |
12:23:05 | FromDiscord | <sOkam! 🫐> (edit) "course" => "course)" |
12:24:04 | FromDiscord | <sOkam! 🫐> now you are -forced- to qualify, which you were not before, when multiple enums have the same field name for whatever reason, even if the enums are pure 😔 |
12:24:18 | FromDiscord | <sOkam! 🫐> so the "more options" point I don't see at all.. if anything I see less options |
12:24:24 | FromDiscord | <nnsee> i'm constantly being forced to qualify :( |
12:25:00 | FromDiscord | <Robyn [She/Her]> In reply to @heysokam "afair, `one` would resolve": Ah |
12:25:16 | FromDiscord | <Robyn [She/Her]> Yeah I don't know why the change was made then |
12:25:42 | FromDiscord | <sOkam! 🫐> exactly why I was asking how it works, in order to maybe understand the why 🙈 |
12:26:10 | FromDiscord | <sOkam! 🫐> it just makes no sense to me. feels like a "lost in overloadable enums implemenation" issue, tbh 🤷 |
12:26:52 | FromDiscord | <Robyn [She/Her]> What are overloadable enums anyway? |
12:28:17 | FromDiscord | <sOkam! 🫐> In reply to @chronos.vitaqua "What are overloadable enums": https://nim-lang.org/blog/2023/08/01/nim-v20-released.html |
12:28:26 | FromDiscord | <sOkam! 🫐> search for `overload` in there |
12:29:33 | FromDiscord | <sOkam! 🫐> there was a better example than that one, but can't find it atm |
12:29:59 | FromDiscord | <Robyn [She/Her]> Aaah okay thanks! |
12:53:59 | FromDiscord | <mratsim> In reply to @albassort "somewhat productive, figuring out": - https://github.com/mratsim/constantine/blob/master/include/constantine/core/serialization.h#L18-L35↵- https://github.com/mratsim/constantine/blob/master/scripts/gen_rust_bindings.sh#L21↵- https://github.com/mratsim/constantine/blob/master/constantine-rust/constantine-sys/src/bindings64.rs#L6-L22 |
13:03:23 | * | ntat quit (Remote host closed the connection) |
13:04:36 | * | ntat joined #nim |
13:32:01 | * | PMunch quit (Remote host closed the connection) |
13:57:23 | * | PMunch joined #nim |
14:01:10 | * | beholders_eye joined #nim |
14:09:01 | * | lucasta joined #nim |
14:24:03 | * | rockcavera joined #nim |
14:38:00 | * | PMunch quit (Quit: Leaving) |
15:01:10 | * | ntat quit (Quit: Leaving) |
15:03:40 | FromDiscord | <fabric.input_output> I don't think you have to fully qualify? |
15:07:30 | FromDiscord | <la_wea> What is this cursed thing, https://media.discordapp.net/attachments/371759389889003532/1250466514038034503/image.png?ex=666b0b31&is=6669b9b1&hm=a73bc3c300f4cda5886a8e5d0a7551ecd108d6c58945dfc0f9fc221ee1c9a97a& |
15:08:05 | NimEventer | New thread by basow53698: Nimqt compile error, see https://forum.nim-lang.org/t/11751 |
15:08:29 | FromDiscord | <Phil> Looks like a bug in the generated c code |
15:15:45 | * | lucasta quit (Ping timeout: 255 seconds) |
15:16:10 | FromDiscord | <la_wea> This is the thing: https://media.discordapp.net/attachments/371759389889003532/1250468695550201967/image.png?ex=666b0d39&is=6669bbb9&hm=7d3605c35041c9f3436099f6e361c5d066f8438e88a437e36cc060f2ca0fd4a5& |
15:17:11 | FromDiscord | <la_wea> Where arr is a `openArray[T]`. |
15:19:22 | FromDiscord | <odexine> likely a bug because you used gcc 14 |
15:19:32 | FromDiscord | <odexine> think you need to update nim |
15:21:40 | FromDiscord | <la_wea> https://media.discordapp.net/attachments/371759389889003532/1250470078361899193/image.png?ex=666b0e83&is=6669bd03&hm=278491beff71f967a9495b46b84e94755081bab11c4bb125ed325636380b5d2b& |
15:22:15 | * | ntat joined #nim |
15:25:39 | FromDiscord | <la_wea> And Mr.McAfee treats choosenim bin as a virus. |
15:27:38 | FromDiscord | <albassort> im using nimAES and using decryptCBC and the first 16 bytes are getting corrupted |
15:27:57 | FromDiscord | <albassort> is this expected behavior? |
15:30:11 | FromDiscord | <albassort> im padding it to the nearest 16 bytes, but its loosing the first block? |
15:32:23 | FromDiscord | <albassort> sent a code paste, see https://play.nim-lang.org/#pasty=AqUgVuxe |
15:50:27 | FromDiscord | <solitudesf> aint reading that, but `collect(for x in 0 .. add-1: "X").join("")` goes crazy |
15:50:56 | FromDiscord | <albassort> Well, I just wanted to create padding and I didn't want to look up how to create a string of a fixed length because I'm very tired |
15:51:07 | FromDiscord | <albassort> I also needed to see it so it couldn't be random |
15:58:13 | * | beholders_eye quit (Ping timeout: 256 seconds) |
15:58:40 | FromDiscord | <la_wea> I think I better learn Rust and stop hopping between langs. |
15:58:48 | FromDiscord | <la_wea> (edit) "I" => "I'd" |
15:59:49 | FromDiscord | <nocturn9x> has anyone ever tried statically compiling musl with nim using gcc? |
16:00:03 | FromDiscord | <nocturn9x> I'd like to have portable binaries that don't depend on glibc and don't do dynamic loading |
16:07:14 | * | beholders_eye joined #nim |
16:07:44 | FromDiscord | <nocturn9x> although that seems incompatible with nimpy |
16:08:08 | Amun-Ra | you mean nim project with musl-gcc? |
16:08:22 | FromDiscord | <nocturn9x> I'd prefer clang if possible |
16:08:28 | FromDiscord | <nocturn9x> but musl-gcc is fine too |
16:08:40 | FromDiscord | <nocturn9x> sent a code paste, see https://play.nim-lang.org/#pasty=TPOCoPIl |
16:09:29 | FromDiscord | <nocturn9x> and `-fuse-ld=lld` seems to fix that |
16:10:31 | FromDiscord | <nervecenter> In reply to @nocturn9x "has anyone ever tried": Yes, I do this daily and have a set process for it |
16:10:45 | FromDiscord | <nervecenter> Let me double check some of my compiler flags |
16:10:49 | FromDiscord | <nocturn9x> thanks! |
16:14:35 | FromDiscord | <nervecenter> sent a long message, see https://pasty.ee/ePWYJKZw |
16:15:11 | FromDiscord | <nocturn9x> the only thing I include is (optionally) mimalloc |
16:15:15 | FromDiscord | <nocturn9x> do I need to compile that manually? |
16:15:28 | FromDiscord | <nervecenter> Is it a header or a compiled library? |
16:15:48 | FromDiscord | <nocturn9x> I'm not entirely sure actually |
16:15:53 | FromDiscord | <nervecenter> If it's a header you just need to place it sensibly and use the correct calling convention in your FFI proc. |
16:15:54 | FromDiscord | <nocturn9x> I installed it from the AUR, could be anything really |
16:16:02 | FromDiscord | <nocturn9x> In reply to @nervecenter "If it's a header": oh I don't use it directly |
16:16:06 | FromDiscord | <nocturn9x> it's a drop-in malloc replacement |
16:16:11 | FromDiscord | <nocturn9x> so Nim would be the one doing the work |
16:17:15 | FromDiscord | <nervecenter> Ah. So long as it's either a header or a static lib, you can include it. Either way I think you'd need to use a `--passL:` flag to make sure the Nim compiler gives it to `musl-gcc`. If it's dynamic in your AUR package, you may need to visit the project's repo and compile it statically yourself or just pull out the source files. |
16:18:01 | FromDiscord | <nervecenter> ALSO: |
16:18:33 | FromDiscord | <nervecenter> Any libraries you compile into `.a` files NEED to be compiled with `musl-gcc`, so if you fetch a project online to compile it yourself, you'll need to fiddle with the makefiles. |
16:18:45 | FromDiscord | <nocturn9x> sent a code paste, see https://play.nim-lang.org/#pasty=fsFJWgfb |
16:19:11 | FromDiscord | <nocturn9x> In reply to @nervecenter "Any libraries you compile": if the makefile isn't garbage, all I really need is `CC=musl-gcc` |
16:19:15 | FromDiscord | <nocturn9x> _hopefully_ |
16:19:49 | FromDiscord | <nervecenter> yeah that has worked for me with some pretty big projects so it's thankfully pretty simple in most cases |
16:20:01 | FromDiscord | <nocturn9x> I'm getting some weird errors |
16:20:05 | FromDiscord | <nocturn9x> which I think are `nimpy` related |
16:20:39 | FromDiscord | <nervecenter> Most likely, god I can't imagine linking into Python with a musl binary, you're probably gonna hit so many errors |
16:20:55 | FromDiscord | <nervecenter> Might need a musl-compiled Python |
16:21:01 | FromDiscord | <nocturn9x> O_o |
16:21:13 | FromDiscord | <nocturn9x> also holy shit is musl slow as heck |
16:21:17 | FromDiscord | <nervecenter> That's spitballing, I don't know that but it's my hunch |
16:21:22 | FromDiscord | <nervecenter> Not for me |
16:21:49 | FromDiscord | <nocturn9x> the generated code compared to clang is like |
16:21:51 | FromDiscord | <nocturn9x> a good 2x slower |
16:21:58 | FromDiscord | <nocturn9x> fuck that lol |
16:22:17 | FromDiscord | <nervecenter> And you're using `danger`? That doesn't seem right |
16:22:23 | FromDiscord | <nervecenter> Musl shouldn't be inherently slower |
16:22:32 | FromDiscord | <nervecenter> Try `release` |
16:23:42 | FromDiscord | <nocturn9x> I can _probably_ separate the nimpy part out of the main codebase |
16:23:44 | FromDiscord | <nocturn9x> let me try |
16:24:58 | FromDiscord | <nocturn9x> ahh nvm there's too many dependencies |
16:24:59 | FromDiscord | <nocturn9x> rip |
16:25:21 | FromDiscord | <nervecenter> Yeah plugging into Python may as well be pluggin into nodejs |
16:25:28 | FromDiscord | <nervecenter> (edit) "pluggin" => "plugging" |
16:25:33 | FromDiscord | <nocturn9x> that's what my tuner script is written in |
16:25:34 | FromDiscord | <nocturn9x> cuz pytorch |
16:25:35 | FromDiscord | <nervecenter> That's gonna have a wild dependency tree |
16:25:47 | FromDiscord | <nocturn9x> also it's still slow as heck |
16:26:24 | FromDiscord | <nervecenter> Are you translating back and forth between Nim objects/dataframes and Numpy matrices? Could be all the copying work |
16:26:37 | FromDiscord | <nocturn9x> oh no the python part is not running at all right now |
16:26:37 | FromDiscord | <nocturn9x> it's all nim |
16:26:47 | FromDiscord | <nocturn9x> release is much slower than danger too |
16:26:57 | FromDiscord | <nocturn9x> from ~31 to almost 60 to now 71 seconds |
16:27:12 | FromDiscord | <nervecenter> Make a compile option for regular gcc and glibc just to be sure |
16:27:55 | FromDiscord | <nervecenter> I won't be able to help you with nimble or Atlas, I just use Makefiles lol |
16:28:49 | FromDiscord | <nocturn9x> ah okay it seems like clang is just slow |
16:28:54 | FromDiscord | <nocturn9x> (edit) "slow" => "faster" |
16:28:59 | FromDiscord | <nocturn9x> how is gcc so slow wtf |
16:29:12 | FromDiscord | <nocturn9x> does `-Ofast` somehow mean something different in gcc vs clang? |
16:29:38 | FromDiscord | <nocturn9x> ye it's gcc that's generating slow code |
16:29:50 | FromDiscord | <nervecenter> In gcc that's `-O2` or `-O3` |
16:30:01 | FromDiscord | <nocturn9x> how is clang so much faster wtf |
16:30:06 | Amun-Ra | there is no well defined optimization standard in C |
16:31:22 | FromDiscord | <nocturn9x> like 27 vs 54 seconds |
16:31:25 | FromDiscord | <nocturn9x> is no laughing matter |
16:31:29 | FromDiscord | <nocturn9x> is llvm just better now? lol |
16:31:42 | FromDiscord | <nervecenter> Did you try `-O2` or `-O3`? |
16:31:54 | FromDiscord | <nocturn9x> doesn't `-d:danger` already do that |
16:32:01 | Amun-Ra | no |
16:32:11 | FromDiscord | <nocturn9x> wat |
16:32:14 | FromDiscord | <nocturn9x> seriously |
16:32:20 | FromDiscord | <nocturn9x> it certainly does for clang |
16:32:24 | FromDiscord | <nervecenter> I thought `-d:release` used `-O2` |
16:32:26 | FromDiscord | <nocturn9x> or maybe it's -Ofast |
16:32:37 | Amun-Ra | wait, I'm not so sure now |
16:32:54 | FromDiscord | <nocturn9x> yeah no -O3 makes no difference |
16:33:16 | FromDiscord | <nervecenter> `-O3` is not guaranteed to be better than `-O2` but it's always worth a shot |
16:33:22 | Amun-Ra | but the 27 vs 54 second difference might be a result of something else |
16:33:23 | FromDiscord | <nervecenter> Most people use `-O2` everywhere |
16:34:10 | FromDiscord | <nocturn9x> like what |
16:34:45 | Amun-Ra | add --listcmd and check what compiler flags are used in both cases |
16:35:57 | FromDiscord | <nocturn9x> lol it's dupliating all options |
16:35:58 | FromDiscord | <nocturn9x> `-pthread -pthread -lm -lm -lrt -lm -lm -flto -ldl` |
16:37:31 | FromDiscord | <nocturn9x> `gcc -c -w -fmax-errors=3 -pthread -flto -Ofast -march=native -mtune=native -mbmi2 -mpopcnt -O3 -fno-strict-aliasing -fno-ident -fno-math-errno` |
16:37:33 | FromDiscord | <nocturn9x> this is the command |
16:38:16 | Amun-Ra | and clang? |
16:38:31 | FromDiscord | <nocturn9x> it doesn't print anything, just the final link command |
16:38:36 | Amun-Ra | hmm |
16:38:47 | FromDiscord | <nocturn9x> ah nvm |
16:39:09 | FromDiscord | <nocturn9x> `clang -c -w -ferror-limit=3 -pthread -flto -Ofast -march=native -mtune=native -mbmi2 -mpopcnt -O3 ` |
16:39:23 | FromDiscord | <nocturn9x> so literally the exact same |
16:40:54 | Amun-Ra | -O3 in both cases |
16:41:51 | FromDiscord | <nocturn9x> yes |
16:56:13 | rockcavera | clang has presented some better optimizations than gcc |
16:57:20 | rockcavera | but everything is a matter of testing and more testing. Remembering that compiling ocm -march=native and -mpopcnt may make the binary not very portable. |
16:58:00 | rockcavera | but if it's for your own use, there's no problem. |
17:23:12 | NimEventer | New Nimble package! libunicorn - Futhark generated wrapper around unicorn-engine, see https://github.com/m4ul3r/libunicorn-nim |
17:51:59 | * | tiorock joined #nim |
17:51:59 | * | rockcavera quit (Killed (zinc.libera.chat (Nickname regained by services))) |
17:51:59 | * | tiorock is now known as rockcavera |
18:13:42 | * | oddish quit (Ping timeout: 256 seconds) |
18:13:54 | * | oddish joined #nim |
18:20:56 | * | tiorock joined #nim |
18:20:56 | * | rockcavera quit (Killed (zirconium.libera.chat (Nickname regained by services))) |
18:20:56 | * | tiorock is now known as rockcavera |
18:38:26 | * | rockcavera quit (Killed (erbium.libera.chat (Nickname regained by services))) |
18:41:25 | * | xaltsc joined #nim |
18:52:13 | FromDiscord | <fabric.input_output> if a function has overloads, how can I select the one with the signature I want, besides doing `let fn: signature = functionName` |
18:58:40 | FromDiscord | <Robyn [She/Her]> In reply to @fabric.input_output "if a function has": How about (signature)(functionName)`? :P |
18:58:48 | FromDiscord | <Robyn [She/Her]> Think you'll need to specify the signature either way |
19:01:34 | Amun-Ra | 'https://play.nim-lang.org/#pasty=dwulfHAD' |
19:01:53 | Amun-Ra | (my terminal is adding that apostrophes) |
19:01:57 | Amun-Ra | these* |
19:03:13 | FromDiscord | <Robyn [She/Her]> In reply to @Amun-Ra "'https://play.nim-lang.org/#pasty=dwulfHAD'": Fabrico said besides that |
19:03:27 | Amun-Ra | ah, I've missed it |
19:03:56 | FromDiscord | <Robyn [She/Her]> No sweat |
19:04:00 | FromDiscord | <fabric.input_output> In reply to @chronos.vitaqua "How about (signature)(functionName)`? :P": yeah that works thanks |
19:04:28 | FromDiscord | <Robyn [She/Her]> No problem o7 |
19:04:51 | Amun-Ra | https://play.nim-lang.org/#pasty=OUpnmBQV |
19:04:54 | Amun-Ra | that works,too |
19:05:02 | FromDiscord | <Robyn [She/Her]> In reply to @chronos.vitaqua "How about (signature)(functionName)`? :P": I do like how Nim usually just works how you expect it to |
19:05:44 | FromDiscord | <Robyn [She/Her]> Amun-Ra, could I interest you in a simple Nim program I made for uploading piped content to the Nim playground via Pasty's API? :> |
19:06:25 | Amun-Ra | Robyn: I can look at it :) |
19:07:19 | FromDiscord | <Robyn [She/Her]> https://play.nim-lang.org/#pasty=CDXLMdNT ✨ |
19:07:45 | FromDiscord | <Robyn [She/Her]> Pretty simple ✨ |
19:08:49 | Amun-Ra | nice |
19:10:05 | FromDiscord | <Robyn [She/Her]> Indeed :] |
19:11:07 | Amun-Ra | it's very concise |
19:19:49 | FromDiscord | <Robyn [She/Her]> I didn't see a reason for it to be more complex aha |
19:20:41 | FromDiscord | <Robyn [She/Her]> Also random question for anyone here, but is there any alternative to the httpclient library in Nim that supports streaming HTTP responses? |
19:21:33 | * | rockcavera joined #nim |
19:22:28 | * | Batzy quit (Quit: https://quassel-irc.org - Chat comfortably. Anywhere.) |
19:22:52 | * | Batzy joined #nim |
19:26:00 | FromDiscord | <Robyn [She/Her]> And also if anyone knows how I'd handle chunks of JSON returned from a request at a time, that'd be useful too |
19:28:16 | * | Batzy quit (Ping timeout: 268 seconds) |
19:31:05 | FromDiscord | <fabric.input_output> In reply to @chronos.vitaqua "And also if anyone": streaming json? |
19:34:39 | FromDiscord | <Robyn [She/Her]> In reply to @fabric.input_output "streaming json?": Yeah, I'm interacting with Ollama's APIs and they have a streaming API which I'd like to support for my own bot |
19:42:13 | * | ntat quit (Quit: Leaving) |
19:44:35 | FromDiscord | <Robyn [She/Her]> Also wondering whether it's worth prefixing all my types with `Ollama`, I was doing it before but it seems a tad redundant now |
19:55:28 | FromDiscord | <fabric.input_output> maybe just `o` or `ol` |
19:57:37 | FromDiscord | <ElegantBeef> I personally try to avoid prefixing any type and just use modules in case of ambiguity |
19:58:15 | FromDiscord | <Robyn [She/Her]> Fair enough |
20:15:21 | FromDiscord | <bosinski2023> sent a code paste, see https://play.nim-lang.org/#pasty=kCjzjyuU |
20:30:15 | FromDiscord | <ElegantBeef> What's the QF macro and QFi Macor? |
20:30:18 | FromDiscord | <ElegantBeef> (edit) "Macor?" => "Macro?" |
20:31:42 | * | Batzy joined #nim |
20:39:05 | FromDiscord | <bosinski2023> sent a code paste, see https://play.nim-lang.org/#pasty=SQQYYSDv |
20:39:32 | FromDiscord | <ElegantBeef> Ah they're structs |
20:39:33 | FromDiscord | <bosinski2023> (edit) "https://play.nim-lang.org/#pasty=dJuNrvod" => "https://play.nim-lang.org/#pasty=gDxvaePc" |
20:39:40 | FromDiscord | <ElegantBeef> QF is a terrible name for a struct, down with the aliases |
20:40:46 | FromDiscord | <Robyn [She/Her]> Hey Beef, do you have any idea on any other HTTP libraries that support streaming responses? |
20:40:58 | FromDiscord | <ElegantBeef> stdlib? 😄 |
20:41:46 | FromDiscord | <Robyn [She/Her]> Besides the stdlib |
20:43:21 | FromDiscord | <ElegantBeef> In reply to @bosinski2023 "from a example when": Yea this API makes 0 sense withwhat is given |
20:43:53 | FromDiscord | <ElegantBeef> `QF` is an empty struct but they do `qf_malloc(filter.addr, ...)` |
20:44:51 | * | xaltsc quit (Quit: WeeChat 4.1.3) |
20:46:43 | FromDiscord | <bosinski2023> In reply to @elegantbeef "`QF` is an empty": agree - `gqf_wrapper.h` is not too helpful and makes little sense. In init they malloc the QF-thing, but don't return it ? |
20:46:55 | FromDiscord | <ElegantBeef> Well they're initing a global variable it seems |
20:47:59 | FromDiscord | <bosinski2023> In reply to @elegantbeef "Well they're initing a": so, then i just grab that global QF-thing, work with it and call `destroy()` when i'm done ? |
20:48:15 | FromDiscord | <ElegantBeef> Seems t obe |
20:48:18 | FromDiscord | <ElegantBeef> (edit) "t obe" => "to be" |
20:51:57 | FromDiscord | <bosinski2023> In reply to @elegantbeef "Seems to be": thx, regarding the `uint128_t` in `insert( uint128_t val)` can i use `stint.nim` for that ? |
20:58:54 | FromDiscord | <ElegantBeef> Only if it wraps the system's uint128 |
21:00:06 | FromDiscord | <ElegantBeef> You likely want https://github.com/rockcavera/nim-nint128 instead |
21:04:05 | FromDiscord | <nocturn9x> is it possible to only link certain libraries statically? |
21:04:10 | FromDiscord | <nocturn9x> like for example mimalloc |
21:04:33 | FromDiscord | <nocturn9x> would like to be my engine as dependency free as possible |
21:04:50 | FromDiscord | <nocturn9x> and glibc is already a pretty undesirable burden, but one I can't currently get rid of |
21:04:53 | FromDiscord | <nocturn9x> (edit) "and ... glibc" added "runtime" |
21:05:02 | FromDiscord | <nocturn9x> (edit) "and runtime glibc is already a pretty undesirable burden, but one I can't currently get rid of ... " added "cuz I need dlopen for nimpy" |
21:05:39 | FromDiscord | <ElegantBeef> You should be able to link specific libraries statically just by doing `-static -lname` |
21:07:10 | Amun-Ra | --passl=/path/to/libname.a |
21:07:33 | Amun-Ra | with eventual dynliboveride |
21:08:12 | FromDiscord | <nocturn9x> sent a code paste, see https://play.nim-lang.org/#pasty=SfmhMlod |
21:08:18 | FromDiscord | <nocturn9x> I only have .so files |
21:08:29 | FromDiscord | <ElegantBeef> Well you cannot statically link .so |
21:08:33 | FromDiscord | <ElegantBeef> So build the static libraries |
21:08:37 | FromDiscord | <nocturn9x> hm |
21:08:43 | FromDiscord | <ElegantBeef> Using the same toolchain as you want to compile your project with |
21:08:45 | FromDiscord | <nocturn9x> gonna have to clone the repo for mimalloc ig |
21:08:56 | FromDiscord | <ElegantBeef> Also doesn't `ldconfig` only search for .so and not .a |
21:09:50 | Amun-Ra | 'https://github.com/microsoft/mimalloc?tab=readme-ov-file#static-override' |
21:10:09 | Amun-Ra | build an object and pass it to the linker |
21:10:19 | FromDiscord | <ElegantBeef> Also I highly suggest using pkg-config for generally linking needs |
21:10:36 | FromDiscord | <ElegantBeef> sent a code paste, see https://play.nim-lang.org/#pasty=HnQhoWpS |
21:10:45 | FromDiscord | <nocturn9x> In reply to @Amun-Ra "build an object and": hm, so like |
21:11:09 | FromDiscord | <nocturn9x> `clang -c mimalloc` or something? |
21:11:28 | Amun-Ra | yes, you can do that with pragma |
21:11:54 | FromDiscord | <nocturn9x> uh, how? |
21:12:33 | Amun-Ra | https://nim-lang.org/docs/manual.html#implementation-specific-pragmas-compile-pragma |
21:13:36 | FromDiscord | <nocturn9x> so I can do like |
21:13:51 | FromDiscord | <nocturn9x> sent a code paste, see https://play.nim-lang.org/#pasty=tmZBslkl |
21:13:57 | FromDiscord | <nocturn9x> or should that rather be in a nim script file? |
21:14:22 | FromDiscord | <nocturn9x> (edit) "https://play.nim-lang.org/#pasty=zgvhatas" => "https://play.nim-lang.org/#pasty=nQJUuRNz" |
21:14:44 | FromDiscord | <ElegantBeef> That's fine |
21:17:53 | * | xaltsc joined #nim |
21:17:53 | FromDiscord | <nocturn9x> is there a pragma for dyn lib override as well? |
21:18:00 | FromDiscord | <nocturn9x> ldd shows it's still linked dynamically |
21:20:33 | Amun-Ra | you don't need it in this case, mimalloc is not linked by default during the build process |
21:20:39 | FromDiscord | <nocturn9x> hm |
21:20:46 | FromDiscord | <nocturn9x> why is it still linked dynamically then |
21:22:00 | Amun-Ra | what's the outputof --listcmd? |
21:23:14 | FromDiscord | <nocturn9x> `clang -o /home/nocturn9x/heimdall/bin/heimdall /home/nocturn9x/heimdall/mimalloc.o [...]` |
21:23:35 | Amun-Ra | that |
21:23:40 | Amun-Ra | that's static linking |
21:23:48 | FromDiscord | <nocturn9x> it's not finding it tho |
21:23:58 | FromDiscord | <nocturn9x> the linker |
21:24:15 | FromDiscord | <nocturn9x> sent a code paste, see https://play.nim-lang.org/#pasty=kOcyCMqY |
21:24:30 | FromDiscord | <nocturn9x> sent a code paste, see https://play.nim-lang.org/#pasty=poiaPKaj |
21:24:34 | FromDiscord | <nocturn9x> my nim.cfg |
21:24:43 | FromDiscord | <bousilleur> hello, does anyone know what happened to yardanico, i was trying to find an old library he had for reference but his github account vanished |
21:24:58 | Amun-Ra | nocturn9x: remove "-lmimalloc" |
21:25:21 | Amun-Ra | -l is used for dynamic linking only |
21:25:22 | FromDiscord | <nocturn9x> ding! |
21:25:23 | FromDiscord | <nocturn9x> works |
21:25:24 | FromDiscord | <nocturn9x> thx |
21:26:34 | Amun-Ra | :) |
21:35:16 | FromDiscord | <bousilleur> is there a reason as to why packedjson no longer works? |
21:43:24 | FromDiscord | <ElegantBeef> -l is not only used for dynamic linking |
21:44:22 | FromDiscord | <ElegantBeef> https://play.nim-lang.org/#pasty=OSSiSRRD just to prove it |
21:44:39 | FromDiscord | <ElegantBeef> Shit I forgot to cat test.nim |
21:44:41 | FromDiscord | <ElegantBeef> I'm a failure |
21:46:11 | Amun-Ra | now do that without -static ;) |
21:47:35 | FromDiscord | <ElegantBeef> Why? |
21:48:58 | Amun-Ra | -static links whole program statically |
21:49:10 | FromDiscord | <ElegantBeef> Right but you can also do `-Bstatic` I believe |
21:50:59 | Amun-Ra | didn't -B just add dir to search path? |
21:51:03 | FromDiscord | <ElegantBeef> Ah seems like you can actually do `--push-state -static -lname --pop-state` |
21:51:34 | Amun-Ra | hmm |
21:52:11 | FromDiscord | <ElegantBeef> https://sourceware.org/binutils/docs/ld/Options.html hmm maybe not |
21:55:07 | FromDiscord | <ElegantBeef> `nim c --passL: "-Bstatic -L/tmp -lexpose -Bdynamic $(pkg-config --libs gtk4)" --verbosity:0 ./test.nim` does compile and link gtk4 |
21:58:38 | * | beholders_eye quit (Read error: Connection reset by peer) |
22:00:02 | Amun-Ra | hmm, so "-Bstatic -L/tmp -lexpose -Bdynamic" == "/tmp/libexpose.a" |
22:00:11 | FromDiscord | <ElegantBeef> Actually that's just cause it searches for the single one |
22:00:21 | FromDiscord | <ElegantBeef> since `libexpose.a` is the only existent library |
22:01:15 | Amun-Ra | right, bc of static |
22:01:38 | FromDiscord | <ElegantBeef> No |
22:01:47 | FromDiscord | <ElegantBeef> Cause ld searches for `.so` then `.a` |
22:01:57 | FromDiscord | <ElegantBeef> Removing `-Bstatic` it still gets the static |
22:02:02 | Amun-Ra | hmm |
22:02:03 | FromDiscord | <ElegantBeef> Adding in a `.so` means it cannot find the `.a` |
22:02:29 | Amun-Ra | I've to test it, interesting |
22:03:08 | Amun-Ra | I wonder whether it worked always that way or not |
22:03:44 | FromDiscord | <ElegantBeef> It is documented to work that way, I cannot figure out how to toggle on static preference then disable it |
22:04:17 | * | beholders_eye joined #nim |
22:10:11 | FromDiscord | <ElegantBeef> https://stackoverflow.com/a/6578558 does seem to indicate you can do them in lists, but no avail here |
22:11:46 | FromDiscord | <ElegantBeef> Though I am trying `--passL` with Nim so that could be the issue |
22:17:15 | * | beholders_eye quit (Read error: Connection reset by peer) |
22:21:06 | strogon14 | Why not use the -l:libexpose.a syntax shown in the first answer to the SO question? |
22:22:49 | * | beholders_eye joined #nim |
22:24:50 | FromDiscord | <ElegantBeef> Cause that relies on the library being at a specific location and is not portable |
22:27:05 | FromDiscord | <ElegantBeef> Oh wait am I daft, it still searches the path 😄 |
22:27:51 | FromDiscord | <ElegantBeef> But yea it does not compose with the king of linking `pkg-config` |
22:41:00 | * | beholders_eye quit (Read error: Connection reset by peer) |
22:41:26 | * | beholders_eye joined #nim |
22:41:30 | FromDiscord | <Robyn [She/Her]> When storing an LLM's messages in a database, how would guys handle sessions and then memory recall? |
22:42:52 | FromDiscord | <ElegantBeef> Me personally? I'd nuke the LLM and not require a database any further |
22:47:27 | FromDiscord | <Robyn [She/Her]> Lmao |
22:47:45 | FromDiscord | <Robyn [She/Her]> I'm making a personal assistant so I can't nuke the LLM :P |
22:57:49 | FromDiscord | <sOkam! 🫐> nuke the project, llm+db gone :smart: |
22:58:49 | FromDiscord | <Robyn [She/Her]> I want a personal assistant though :P |
23:57:23 | * | xet7 joined #nim |