<< 25-11-2025 >>
01:17:42FromDiscord<russ4806> What is the recommended way (or launch sequence) to get the nimlangserver working on linux? I've built it with `nimble install nimlangserver`, but I'm not sure if I'm supposed to spark it up myself when I want to use it or not. Weirdly, I did the same in Windows and it just started working beautifully/immediately with vscode.
01:18:31FromDiscord<russ4806> I suspect it might be because my vscode is running up in windows, and my linux is WSL (a guest VM)... but I thought I'd check if there was some trick I don't know before I root around too much.
01:26:39FromDiscord<russ4806> Never mind - I forced the path to it and it is working now. Unclear why it wasn't working before as it was definitely in my `$PATH`. šŸ¤”
01:28:38FromDiscord<Buckwheat> most likely a WSL scenario, it's not exactly perfect
01:28:52FromDiscord<Buckwheat> most likely a WSL-specific scenario, it's not exactly perfect
01:34:30*Jjp137 quit (Quit: Leaving)
01:37:58*Jjp137 joined #nim
05:04:50FromDiscord<kapendev> In reply to @nervecenter "Apologies if this isn't": C++ can be crazy, but it is really nice if you stay inside your way of doing things. No idea, random thoughts. I kinda avoid this videos anyway.
05:05:09FromDiscord<kapendev> (edit) "this" => "those"
05:05:34FromDiscord<Buckwheat> perfection isn't real anyways, everything will have flaws, even good things
05:59:43FromDiscord<nervecenter> sent a long message, see https://pasty.ee/DmnLerPW
05:59:54FromDiscord<nervecenter> (edit) "https://pasty.ee/rmgDvSBi" => "https://pasty.ee/vBcLUWiU"
06:00:00FromDiscord<nervecenter> (edit) "https://pasty.ee/kmiAkzlk" => "https://pasty.ee/qWCvhFne"
06:00:24*SchweinDeBurg quit (Quit: WeeChat 4.8.0-dev)
06:01:03FromDiscord<Buckwheat> eh, I just think C++ is feature bloat.
06:02:13FromDiscord<saint._._.> In reply to @Buckwheat "other than that, Futhark": Oh cool!
06:02:19FromDiscord<saint._._.> Does this apply to macos at all
06:03:34FromDiscord<Buckwheat> macOS already has a fix for it, and has for a while, the recent versions should workā†µ(@saint._._.)
06:03:47FromDiscord<Buckwheat> my patches just fixed build for FreeBSD and OpenBSD
06:07:27*SchweinDeBurg joined #nim
06:17:24FromDiscord<saint._._.> In reply to @Buckwheat "macOS already has a": awesome thanks!
06:21:14FromDiscord<Buckwheat> eh don't thank me, I'm not the Futhark maintainer, I just wrote some patches for it!ā†µ(@saint._._.)
06:21:43FromDiscord<Buckwheat> thank PMunch for creating it!
06:22:36FromDiscord<Buckwheat> I'm just happy to help with it wherever I have the skillset to do so, and I just happened to be able to fix it for those 2 platforms
06:31:25FromDiscord<gokr> Futhark is awesome, just note that it does not cover macros etc, so some libraries that offer you tons of macros that you are supposed to use - those you will need to take care of in some other way. But on the other hand, if the library uses macros to generate things (like say long enums) in the library Futhark will catch the end result. A bit hard to explain concisely but I would guess Futhark fits most cases the best
06:31:36FromDiscord<mr_rowboto> I feel like Futhark is just one of nim's greatest features
06:31:52FromDiscord<mr_rowboto> Like, a lot of people that come here, do so just because of Futhark
06:32:33FromDiscord<mr_rowboto> Having a low-to-mid level language _with a GC_ and C interop feels out of this world.
06:32:43FromDiscord<mr_rowboto> Feels like cheating
06:33:51FromDiscord<gokr> Using Futhark to wrap ORX game engine did lead to a few things I had to solve, but it still feels more robust than the previous c2nim version. Using Futhark to wrap NATS client lib ā€¦ well, Claude did that all on his own šŸ™‚
06:34:09FromDiscord<mr_rowboto> oof
06:34:16FromDiscord<mr_rowboto> yeah
06:34:30FromDiscord<mr_rowboto> despite whatever people say about AI
06:34:42FromDiscord<mr_rowboto> it can do some awesome tasks that I def not want to do
06:35:23FromDiscord<Buckwheat> can speed up some, but... should not be relied too heavily on, especially for large code snippetsā†µ(@mr_rowboto)
06:35:28FromDiscord<Buckwheat> it is not perfect
06:35:36FromDiscord<gokr> Indeed, it is ā€¦ mind blowing. I know lots of devs are skeptic but IMHO they are missing out.
06:36:40FromDiscord<Buckwheat> I think the problem is less of skepticism and more of "is this code I can actually use without violating copyright or copyleft of any sort?"
06:36:55FromDiscord<gokr> Seeing Claude make advanced macros in Nim was ā€¦ well, di not expect it to pull that off.
06:37:28FromDiscord<Buckwheat> or if it's code that a human can even maintain... cause good luck trying to vibe code the bugs out of your code heh, you're gonna have to get hands on with things like that
06:37:55FromDiscord<Buckwheat> throwbacks to when someone vibecoded a SaaS solution and then it got hacked multiple times
06:40:40FromDiscord<mr_rowboto> In reply to @Buckwheat "it is not perfect": perfection isn't real anyways, everything will have flaws, even good things
06:40:43FromDiscord<gokr> I have been ā€œvibingā€ large parts of Nim BUT ā€¦ I also refactor and make Claude (primarily Sonnet 4.5) simplify, clean etc so that the code looks fine to me. I argue such a way of working with it works great. The end result is as clean as any
06:40:46FromDiscord<mr_rowboto> šŸ˜›
06:41:10FromDiscord<mr_rowboto> Don't worry, I don't use AI for big stuff. For me is just glorified autocomplete.
06:41:25FromDiscord<Buckwheat> that's true, but I still think a human's gonna have a much sharper eye. at most, I think AI should just be a slightly smarter debugger.ā†µ(@mr_rowboto)
06:41:33FromDiscord<mr_rowboto> I wouldn't ever ship a single line of AI code that:ā†µ1- I don't understandā†µ2- I haven't read
06:42:11FromDiscord<Buckwheat> what aboutā†µ3- is code under a license that I have to ensure I respect?ā†µ(@mr_rowboto)
06:42:28FromDiscord<mr_rowboto> ~~I don't believe in licencing~~
06:42:38FromDiscord<Buckwheat> le gasp!
06:42:40FromDiscord<gokr> I urge you to try Claude Code if you havenā€™t yet. If you think 20usd/month is fine of course. There are cheaper options too.
06:42:43FromDiscord<mr_rowboto> haha
06:42:53FromDiscord<Buckwheat> eh, I personally write all my code AGPL 3.0 anyways
06:43:16FromDiscord<mr_rowboto> I usually write ISC
06:43:45FromDiscord<Buckwheat> I used to do BSD 2-Clause in the past but I kinda like the AGPL for my stuff now because I think it's funny when people use my garbage code
06:43:48FromDiscord<mr_rowboto> sometimes unlicensed even
06:44:09FromDiscord<Buckwheat> I definitely don't bother with a license for proof-of-concepts
06:44:43FromDiscord<Buckwheat> maybe I should start saying to get me a cup of coffee in return for the pinky strain the project required
06:45:04FromDiscord<gokr> Regarding the license part - to be honest I think that train has left the station. In the sense that AI is so deeply used now that even if one could discern some pattern or something, it wouldnā€™t be considered as copied. Just my guess.
06:47:11FromDiscord<Buckwheat> nah I'm totally doing it, the Get Buckwheat A Coffee license\:ā†µyou can use this code as long as it remains open source at all costs (as well as if you make it SaaS or use it over a network or something)ā†µand if you like it, you buy Buckwheat a coffee because he uses Emacs and that's tough on the pinky
06:50:50FromDiscord<nnsee> seems like the beerware license:ā†µ> <[email protected]> wrote this file. As long as you retain this notice you can do whatever you want with this stuff. If we meet some day, and you think this stuff is worth it, you can buy me a beer in return Poul-Henning Kamp
06:51:45FromDiscord<Buckwheat> yeah pretty much the beer license, but I don't drink (hence coffee) and a lil bit of... uh what's the word... well whatever, trying to make you feel bad for my poor pinkyā†µ(@nnsee)
06:52:08FromDiscord<Buckwheat> (actually I play the Guitar and Piano so I don't get Emacs pinky thankfully)
06:57:54FromDiscord<nnsee> i just use vim šŸ˜Ž
06:58:35FromDiscord<Buckwheat> that used to be me for a few years... til I had a dreamā†µ(@nnsee)
06:59:26FromDiscord<Buckwheat> ...I know that sounds like a joke, but I'm serious, I had a dream that involved Emacs and the question of "and if you're not using Emacs, why?" and when I woke up I was like "...it is my destiny to use Emacs" so I learned it and never went back
06:59:38FromDiscord<Buckwheat> can YOU say that YOUR editor choice came to you in a dream though?
07:44:53yetiThere is no emacs locked in syndrome! I'm open to try every editor that gives me eshell, gnus, orgmode and tramp!
07:49:52FromDiscord<Buckwheat> eh, I'll just make a plugin that locks me into Emacs in my Emacs
08:13:49*Jjp137 quit (Read error: Connection reset by peer)
08:14:10*Jjp137 joined #nim
08:28:18FromDiscord<aethrvmn> The power of emacs is elisp right? The power of elisp is macros right? We could make a better, easier configurable (that doesnt make me have to learn a lisp dialect exclusive to this) editor as the modern emacs like nvim to vim, or a post modern emacs like helix to vim
08:29:17FromDiscord<aethrvmn> The power of emacs is elisp right? The power of elisp is macros right? We could make a better, easier configurable (that doesnt make me have to learn a lisp dialect exclusive to this) editor as the modern emacs like nvim to vim, or a post modern emacs like helix to vim. Written in nimā„¢ļø
09:31:32yetithere are lots of failed emacs competitors. doit. create one whose name starts with n(im)...
09:32:46yetijed still is nice, peppy (python based) was a nice idea, a fac of emacsalike editors is periodically posted in usenet, the list is looooooong.
09:34:44yetidespite liking lisp and forth because small is beautiful, I often curse about their incredible diversity. but when I want to do complex text stuff like a mix of jupyter notebook and literate programming, I see no alternative to accepting the weitd lisp dialect of emacs.
09:34:53yetiweird
10:11:23FromDiscord<ray1_hatfield94401> sent a long message, see https://pasty.ee/ebUCtNQb
11:21:34*beholders_eye joined #nim
12:15:05FromDiscord<nocturn9x> I gotta thank araq for mentioning deepwiki in the nimony README
12:15:07FromDiscord<nocturn9x> pretty cool
13:39:19*beholders_eye quit (Ping timeout: 264 seconds)
13:40:48*beholders_eye joined #nim
14:12:40*thunder joined #nim
14:15:55*beholders_eye quit (Ping timeout: 240 seconds)
14:18:02*beholders_eye joined #nim
14:46:05*maytham joined #nim
14:49:15*beholders_eye quit (Ping timeout: 240 seconds)
14:50:16*beholders_eye joined #nim
14:52:56*thunder quit (Quit: Leaving)
14:55:40*maytham quit (Ping timeout: 256 seconds)
15:24:39FromDiscord<tauruuuuuus> After he made me discover it I've been using it and pitching it quite a bit
15:24:48FromDiscord<tauruuuuuus> Favourite AI tool yet
16:12:36FromDiscord<Buckwheat> sounds like a better use of AI than vibecoding some slop you can't maintain
16:32:23FromDiscord<tauruuuuuus> sent a long message, see https://pasty.ee/foHuZtMe
16:32:37FromDiscord<tauruuuuuus> BTW claude is not bad at all with nim!
16:40:56FromDiscord<Buckwheat> even then I still think that unless you get human developers to write code snippets to train AI on, I will not support vibecoding, you have no clue where those snippets came from and who knows what licenses you're violating... or maybe it's rather HOW MANY licenses you're violating
16:41:06FromDiscord<Buckwheat> even then I still think that unless you get human developers to write code snippets themselves to train AI on, I will not support vibecoding, you have no clue where those snippets came from and who knows what licenses you're violating... or maybe it's rather HOW MANY licenses you're violating
16:41:29FromDiscord<Buckwheat> but also, at that point... are YOU the developer or is the AI the developer?
16:41:43FromDiscord<Buckwheat> food for thought
16:44:09FromDiscord<nnsee> i vibe-code sometimes. no, i'm not the developer, i'm more the senior sorta-developer-project-lead, telling the new hire what to code and fixing their code when they fuck up or do something that i don't quite agree with
16:45:24FromDiscord<Buckwheat> ah, so my job will be harder now because I have to rewrite that code on top of adding new code!
16:45:29FromDiscord<Buckwheat> semi-joking
16:46:03FromDiscord<nnsee> i obviously wouldn't use it for large projects or codebases, more single-shot stuff that i'll use and probably discard later on
16:46:19FromDiscord<nnsee> or tiny, self-contained stuff like react components or such
16:46:23FromDiscord<Buckwheat> I think that if you really are gonna use such a tool, probably the only real use case
16:46:24FromDiscord<nnsee> where it's easy to review
16:46:31FromDiscord<Buckwheat> is just some sort of proof-of-concept
16:47:09FromDiscord<Buckwheat> and then trying to deep down figure out what was done so you know how to meaningfully use what you've learned
16:48:29FromDiscord<Buckwheat> not like you have to memorize literally everything ever about programming, nobody can memorize EVERY library. more about building those core concepts and lexicon-agnostic techniques that you can visualize and apply regardless of implementation language.
16:49:04FromDiscord<Buckwheat> even I'll have like 5, 6, maybe more tabs open in my browser of documentation anytime I program
16:49:43FromDiscord<Buckwheat> ...sometimes I even just have the source code of whatever it is I'm working with open
16:49:57FromDiscord<Buckwheat> ...sometimes I even just have the source code of whatever library I'm working with open
16:51:16FromDiscord<tauruuuuuus> In the end, the most problematic issue I see is an educational one, and that's not limited to programming
16:52:10FromDiscord<tauruuuuuus> As in, how do we evaluate people skills and how do we teach stuff, especially when not in person...programming wise, I think I am for banning AI for junior devs LOL
16:52:16FromDiscord<nervecenter> LLMs suck hard at domain-specific scenarios. I've found this repeatedly. They're good for get examples of usage from documentation, but that's about it. And 99% of my code is domain specific, where an LLM will have no idea how to properly architect data and logic for my use case. It's a boilerplate generator for webdev.
16:52:27FromDiscord<nervecenter> (edit) "get" => "getting"
16:53:00FromDiscord<Buckwheat> even with my proof-of-concepts ironically, I manage to always pull off my idea better just by myself
16:53:01FromDiscord<nervecenter> Happy to be proven wrong, but I've found "vibecoders" aren't doing anything that's very particular
16:53:13FromDiscord<Buckwheat> I'll second you on that
16:53:52FromDiscord<Buckwheat> I have grown a very particular style of writing code and even structuring how my code works that I don't think anyone except myself can really replicate it
16:54:10FromDiscord<Buckwheat> an LLM just, does not know how to write like I do, and thus it won't ever nicely slot into what I do
16:54:27FromDiscord<nnsee> In reply to @nervecenter "LLMs suck hard at": i think the way i currently use LLMs is very much a "slightly more sophisticated boilerplate generator"
16:54:40FromDiscord<tauruuuuuus> In reply to @nnsee "i think the way": That's exactly it
16:55:00FromDiscord<Buckwheat> ehe, kinda like how for me it's nothing more than a plugin for a debugger that I can curse atā†µ(@nnsee)
16:56:27FromDiscord<nnsee> i recently noticed that claude seemed to curse at me _a lot_, unprompted
16:56:57FromDiscord<Buckwheat> pfff... the clankers are taking over!
16:57:20FromDiscord<nnsee> was a bit confused by that until i realized my "response style preference" prompt had a "it's fine to curse" sentence in it that it interpreted as "curse at every occasion whenever possible"
16:58:20FromDiscord<xtrayambak> I personally use LLMs to be a sycophant for my work whenever I feel demotivated :P
16:58:31FromDiscord<Buckwheat> also probably not the most exciting thing in the world or anything extremely novel... but I managed to use Futhark to call Assembly procedures from Nim granted they had a C header file with a function declaration
16:58:41FromDiscord<Buckwheat> useless probably, but it was really cool to pull off
16:59:07FromDiscord<nnsee> that preference also basically says to try to be as unsycophantic as possible, and that it's fine to disagree with me or suggest alternative solutions in order to prevent the XY problem
16:59:22FromDiscord<nnsee> which i've found actually does increase the quality of the responses
16:59:22FromDiscord<xtrayambak> In reply to @nnsee "that preference also basically": Aw, that's no fun though
16:59:55FromDiscord<Buckwheat> honestly, I think whenever one is demotivated, probably not the best time to be doing work, and also a great way to introduce bugs
17:00:13FromDiscord<Buckwheat> if I don't feel like programming, I just don't, I know I'll have a jolt of energy later
17:00:55FromDiscord<Buckwheat> but at the same time, I guess programming is different for me in that it's like a puzzle game for me, and it really stimulates my brain
17:01:27FromDiscord<xtrayambak> In reply to @Buckwheat "but at the same": I like programming with faster "rewards" or where I can see visible results
17:01:46FromDiscord<Buckwheat> but is it really a reward if you didn't earn it?ā†µ(@xtrayambak)
17:01:51FromDiscord<xtrayambak> which is a huge pain in of itself, albeit I think we should discuss this in #offtopic
17:01:56FromDiscord<Buckwheat> or is it just you cheating yourself to get some quick dopamine?
17:02:12FromDiscord<Buckwheat> don't have that, I'm on Matrixā†µ(@xtrayambak)
17:02:39FromDiscord<xtrayambak> In reply to @Buckwheat "or is it just": I can't cheat myself to get quick dopamine, because I simply don't feel accomplished while doing my typical work and I'm too demotivated to branch out
17:02:56FromDiscord<xtrayambak> I can't get dopamine while programming per se nowadays
17:03:11FromDiscord<Buckwheat> maybe you just need to find a more challenging task?
17:03:20FromDiscord<Buckwheat> push yourself a bit
17:04:06FromDiscord<xtrayambak> sent a long message, see https://pasty.ee/yYMkrTtf
17:04:24FromDiscord<Buckwheat> eugh... I already see the problem
17:04:34FromDiscord<Buckwheat> programming for Wayland period is absolutely horrid
17:04:47FromDiscord<xtrayambak> In reply to @Buckwheat "programming for Wayland period": it is, but I absolutely love wayland and its design
17:05:26FromDiscord<Buckwheat> can't share the same sentiments, I like the concept but I can't stand its execution, it feels like a full-on violation of UNIX philosophy
17:06:13FromDiscord<Buckwheat> but at the same time, I think Wayland is still very young and needs time to mature, if that makes sense... does not help with its current development team though
17:06:35FromDiscord<Buckwheat> what I'd truly love is if someone was enough of a psychopath to port Xenocara from OpenBSD to other UNIX-likes
17:06:39FromDiscord<xtrayambak> In reply to @Buckwheat "but at the same": I've been daily driving Wayland since around ~2021 and it's improved significantly
17:06:53FromDiscord<xtrayambak> In reply to @Buckwheat "what I'd truly love": I personally think X11 should just die apart from certain niches now
17:07:01FromDiscord<xtrayambak> and legacy systems, obviously
17:07:18FromDiscord<Buckwheat> I hardcore disagree, freedom of choice is what FOSS lives onā†µ(@xtrayambak)
17:07:52FromDiscord<Buckwheat> if you want to use Wayland, be my guest, but don't try to stop me from using X11, I'm perfectly happy with it and I could care less for Wayland after having used it for a while
17:07:54FromDiscord<xtrayambak> In reply to @Buckwheat "I hardcore disagree, freedom": fair, but I'd say it should only be maintained by hobbyists now and no amount of real manpower from teams should be spent trying to fix it
17:08:38FromDiscord<Buckwheat> I also hardcore disagree, I think if anything that Xenocara should be the X11 standard and that of course is maintained by the OpenBSD team who are very diligent about doing maintenanceā†µ(@xtrayambak)
17:08:49FromDiscord<xtrayambak> I'm currently also writing libwayland-client wrappers for Nim and I'll use that with the pre-existing, decent xcb and win32 bindings for Nim to make a winit-like library
17:09:19FromDiscord<nnsee> In reply to @Buckwheat "I hardcore disagree, freedom": i would generally agree, but there's no real benefit for clients (applications) and DE developers etc to have to have the burden of supporting two different protocols simultaneously
17:09:52FromDiscord<xtrayambak> In reply to @Buckwheat "I also hardcore disagree,": I have no clue what Xenocara is like, but I think Xorg should go die in a ditch. I feel very negatively about Xorg, not so much about the X11 protocol itself.
17:09:57FromDiscord<Buckwheat> I would rather work on something mature than something that still feels like an alpha thoughā†µ(@nnsee)
17:10:17FromDiscord<xtrayambak> It served its purpose well for decades but it's time to modernize (for most people - if that's not you, I can't force you)
17:10:18FromDiscord<nnsee> have you worked with wayland recently? it in no way feels like an alpha and is much more pleasant to work with than X
17:10:21FromDiscord<Buckwheat> Xenocara is a security-hardened X11 fork used in OpenBSD, and it's a much cleaner version of X11ā†µ(@xtrayambak)
17:10:29FromDiscord<Buckwheat> yes, and I was not a fan of itā†µ(@nnsee)
17:10:35FromDiscord<xtrayambak> In reply to @nnsee "have you worked with": It is, as long as you're not working with libwayland
17:10:38FromDiscord<nnsee> to each their own i suppose
17:10:42FromDiscord<xtrayambak> which you aren't in most cases
17:10:58FromDiscord<xtrayambak> In reply to @Buckwheat "Xenocara is a security-hardened": Does it prevent screen loggers and key loggers?
17:11:17FromDiscord<Buckwheat> if that's why you prefer Wayland, I don't think you understand most display serversā†µ(@xtrayambak)
17:11:35FromDiscord<xtrayambak> In reply to @Buckwheat "if that's why you": That's one of the reasons, but not the paramount reason.
17:11:41FromDiscord<Buckwheat> I can tell you for certain if I was to write a keylogger for a UNIX-like, I would not be using the display server, I'd be reading from evdev
17:11:47FromDiscord<Buckwheat> and that's display server-agnostic
17:11:59FromDiscord<nnsee> how do you read from evdev without `input` privs?
17:12:11FromDiscord<xtrayambak> In reply to @Buckwheat "I can tell you": And that requires root privileges on most sane NIXes.
17:12:26FromDiscord<xtrayambak> So does the "`LD_PRELOAD` into your compositor" gotchas
17:12:32FromDiscord<Buckwheat> you wouldn't, but that's why you would use social engineering techniques or attempt to exploit a known vulnerability that allows privescā†µ(@nnsee)
17:12:47FromDiscord<nnsee> In reply to @xtrayambak "Does it prevent screen": it doesn't. and most of the "security" functionality was that it allowed you to run X as non-root, which upstream X has had for ages now as well
17:13:02FromDiscord<xtrayambak> In reply to @Buckwheat "you wouldn't, but that's": Social engineering is where your keylogger turns from deterministic to non-deterministic.
17:13:10FromDiscord<nnsee> In reply to @Buckwheat "you wouldn't, but that's": so an entirely different threat vector
17:13:21FromDiscord<Buckwheat> then clearly you also aren't too security-aware, as that's how a lot of modern malware worksā†µ(@xtrayambak)
17:13:32FromDiscord<nnsee> :^)
17:13:49FromDiscord<Buckwheat> I also don't think MOST UNIX-like malware is concerned about keylogging, not enough of a userbase on UNIX-likes to warrant it
17:14:00FromDiscord<nnsee> maybe i should specify that i work as a security researcher, penetration tester and redteamer IRL
17:14:04FromDiscord<Buckwheat> most of the time, it's rootkits or RATs i.e. Mirai or Katana
17:14:10FromDiscord<xtrayambak> In reply to @Buckwheat "then clearly you also": Nobody who isn't a lobotomite would look at a pkexec or sudo prompt that appeared for no reason and say "oh nice, let me enter my password!"
17:14:19FromDiscord<nnsee> just in case i get called "not security-aware" as well :')
17:14:47FromDiscord<nnsee> oh also
17:14:51FromDiscord<nnsee> can we move this to #offtopic
17:14:55FromDiscord<xtrayambak> Let me specify that I've been working with Wayland for a good while, just in case I get not-display-protocol-aware again :3
17:15:03FromDiscord<xtrayambak> In reply to @nnsee "can we move this": They do not have access to it.
17:15:06FromDiscord<xtrayambak> They use Matrix.
17:15:21FromDiscord<nnsee> i'm fairly sure there's a matrix bridge to #offtopic as well?
17:15:21FromDiscord<xtrayambak> (edit) "Let me specify that I've been working with Wayland for a good while, just in case I get ... not-display-protocol-aware" added "called"
17:15:27FromDiscord<xtrayambak> Let me check
17:15:51FromDiscord<Trayambak> Nope, if there is one, I cannot see it.
17:15:55FromDiscord<Buckwheat> does this work on Matrix?ā†µ(@nnsee)
17:16:06FromDiscord<nnsee> In reply to @Trayambak "Nope, if there is": hm, strange, let me check
17:17:52FromDiscord<nnsee> https://matrix.to/#/!eqphGKnWIutLvVlDqB:envs.net?via=matrix.org&via=t2bot.io&via=envs.net
17:18:10FromDiscord<Buckwheat> your Mirais and Katanas and that sort
17:18:11FromDiscord<Buckwheat> maybe I got a bit brash when saying that... I'm not a professional but I'm currently a student that tinkers quite a bit, and from what I've personally seen, keylogging is not much of a UNIX-like thing and most of the time, threat actors want to instead get RATs on UNIX-likesā†µ(@nnsee)
17:18:17FromDiscord<Buckwheat> keylogging seems more to be something you worry about with Windows
17:19:00FromDiscord<xtrayambak> Mirai was a malware that used IoT devices with _no or weak password protection_
17:19:21FromDiscord<xtrayambak> No social engineering or zero days were involved in the original version, though newer ones do use stuff like Log4j iirc
17:21:06FromDiscord<Buckwheat> but even at that... in general malware for desktop UNIX-likes also doesn't really seem to be much of a thing besides that one instance of AUR getting a supply chain attack
17:29:38*beholders_eye quit (Ping timeout: 244 seconds)
17:34:38FromDiscord<d1_trainstation_01.bsp> hi server
17:35:46FromDiscord<nnsee> hello there
19:03:03FromDiscord<d1_trainstation_01.bsp> are there nim packages for fedora?
19:53:55*beholders_eye joined #nim
20:03:19FromDiscord<planetis_m> In reply to @d1_trainstation_01.bsp "are there nim packages": yes: https://copr.fedorainfracloud.org/coprs/planetis/nim/builds/ although not 2.2.6 yet
20:13:49FromDiscord<.tokyovigilante> Are there any known issues with code using both `asyncdispatch` and threads/channels when built with MSVC? I'm getting a weird race condition on startup /my logging thread seems to die and stall my app, which doesnt happen when using mingw.
20:15:27FromDiscord<nnsee> In reply to @.tokyovigilante "Are there any known": not quite what you asked for, but any reason you're not using chronos?
20:16:33FromDiscord<.tokyovigilante> asyncdispatch is part of stdlib and chronos isn't I guess, no particular reason though. Code works fine on Linux and with mingw on Windows, so didn't have a specific reason to change.
20:17:19FromDiscord<nnsee> personally i've found std/asyncdispatch to be quite ass with edge cases (not dissimilar to what you're describing) that just aren't a problem on chronos
20:20:06FromDiscord<.tokyovigilante> Fair, I can always give it a roll. The API is the same yeah?
20:20:09FromDiscord<Buckwheat> @.tokyovigilante woah, super unrelated... but it's a throwback seeing The Cheat in your pfp!
20:20:23FromDiscord<nnsee> In reply to @.tokyovigilante "Fair, I can always": not in all cases, but mostly yeah
20:20:46FromDiscord<nnsee> migrating from asyncdispatch to chronos should be fairly straightforward
20:21:05FromDiscord<.tokyovigilante> cool, I'll sic the AIs on it and see if it works
20:35:34FromDiscord<nnsee> In reply to @.tokyovigilante "cool, I'll sic the": let me know
20:46:37FromDiscord<d1_trainstation_01.bsp> In reply to @planetis_m "yes: https://copr.fedorainfracloud.org/coprs/planet": I donā€™t see it listed when I search for it on dnf
20:51:02*beholders_eye quit (Quit: WeeChat 4.6.3)
21:04:10FromDiscord<afonso.lima> sent a code paste, see https://play.nim-lang.org/#pasty=obOaYNrM
21:10:03FromDiscord<.tokyovigilante> sent a code paste, see https://play.nim-lang.org/#pasty=NveADTFY
21:10:28FromDiscord<.tokyovigilante> (edit) "https://play.nim-lang.org/#pasty=rCtXcitf" => "https://play.nim-lang.org/#pasty=JCLJJXpU"
21:10:50FromDiscord<.tokyovigilante> (edit) "https://play.nim-lang.org/#pasty=CkYeHLER" => "https://play.nim-lang.org/#pasty=qKMkeeNg"
21:11:08FromDiscord<.tokyovigilante> (edit) "https://play.nim-lang.org/#pasty=iIaGuWNq" => "https://play.nim-lang.org/#pasty=gJaxweZT"
21:11:33FromDiscord<.tokyovigilante> (edit) "https://paste.rs/usKga" => "https://paste.rs/4aMyh"
21:11:56FromDiscord<.tokyovigilante> (edit) "https://play.nim-lang.org/#pasty=emOvNiOW" => "https://play.nim-lang.org/#pasty=SSmweNzz"
21:17:02*beholders_eye joined #nim
21:48:46FromDiscord<.tokyovigilante> In reply to @nnsee "personally i've found std/asyncdispatch": Ok have run this test with both mingw and MSVC, definitely only affecting MSVC
21:49:31FromDiscord<nnsee> In reply to @.tokyovigilante "Ok have run this": think you have to break out the debugger and see where exactly it's stalling
21:51:03FromDiscord<nnsee> could very well be an edge case/bug in msvc
21:51:11FromDiscord<nnsee> it's not know as... the least buggiest compiler
21:53:03FromDiscord<Buckwheat> MSVC, my beloved...
21:53:09FromDiscord<Buckwheat> (I do not like it)
21:59:00FromDiscord<.tokyovigilante> In reply to @nnsee "could very well be": looks like its blocking trying to acquire the lock for `recv()`
21:59:15FromDiscord<nnsee> šŸ¤”
21:59:57FromDiscord<.tokyovigilante> I don't really mind but was hoping to get a MSVC build running because my mingw build is being flagged by Windows Defender at work, and I don't really want to fork out for an EV certificate (or even an OV certificate)
22:00:59FromDiscord<.tokyovigilante> sent a code paste, see https://play.nim-lang.org/#pasty=zPqDXYey
22:02:03FromDiscord<.tokyovigilante> well, on the call to `interlockedCompareExchange64` in that function anyway.
22:04:51FromDiscord<nnsee> hmmmmmmmm
22:05:10FromDiscord<nnsee> this maaayyybeeee seems line an alignment issue with FreeCell in nim
22:05:30FromDiscord<nnsee> are you willing to try a custom nim compiler with patches? :p
22:06:55FromDiscord<nnsee> if you are, long shot: https://github.com/nim-lang/Nim/blob/devel/lib/system/alloc.nim
22:07:57FromDiscord<nnsee> sent a code paste, see https://play.nim-lang.org/#pasty=uROsaUrx
22:08:05FromDiscord<nnsee> not really sure this will do much, but worth a shot
22:09:09FromDiscord<nnsee> (or was it `align(8)`?)
22:11:17FromDiscord<nnsee> in any case, this is worth posting in #internals, maybe araq will have some more insightful... insight
22:12:12FromDiscord<.tokyovigilante> Thanks for looking. I have built the compiler a bunch on linux but not windows yet. Will give that a try and if not may just accept mingw.
22:12:16FromDiscord<.tokyovigilante> ok, will do
22:18:37FromDiscord<.tokyovigilante> `C:\Users\ryan\Develop\Nim\lib\system\alloc.nim(100, 32) Error: invalid pragma: align: 8`
22:18:40FromDiscord<.tokyovigilante> `C:\Users\ryan\Develop\Nim\lib\system\alloc.nim(100, 32) Error: invalid pragma: align(8)`
22:18:43FromDiscord<.tokyovigilante> COMPUTER SAYS NO
22:18:47FromDiscord<.tokyovigilante> šŸ™‚
22:18:54FromDiscord<nnsee> wtf wait
22:18:58FromDiscord<nnsee> am i misremembering it
22:20:08FromDiscord<nnsee> oh, maybe you can't apply to objects https://nim-docs.readthedocs.io/en/latest/manual/pragmas/specific_pragmas/#align-pragma
22:20:22FromDiscord<nnsee> i'm sorry :')
22:20:35*cm_ joined #nim
22:20:51FromDiscord<.tokyovigilante> ah right, makes sense. I would have thought an object should be aligned to a pointer boundary anyway
22:21:01*cm quit (Ping timeout: 246 seconds)
22:21:02*cm_ is now known as cm
22:21:03FromDiscord<nnsee> you _would_ think so, wouldn't you
22:21:17FromDiscord<nnsee> let me look at some old notes
22:21:19FromDiscord<.tokyovigilante> will see what the boffins in #internal come back with
22:21:28FromDiscord<.tokyovigilante> (edit) "#internal" => "#internals"
22:21:31FromDiscord<.tokyovigilante> thanks
22:54:32*rockcavera joined #nim
23:27:59*tiorock joined #nim
23:27:59*rockcavera is now known as Guest2240
23:28:00*Guest2240 quit (Killed (calcium.libera.chat (Nickname regained by services)))
23:28:00*tiorock is now known as rockcavera