IRC logs for #openttd on OFTC at 2025-01-06
⏴ go to previous day
00:28:48 <LordAro> duckfullstop: libxdg-basedir (optional) usage was removed in part because debian was removing the library
00:30:34 <LordAro> good grief that's a lot longer ago than i thought
01:11:55 <peter1139> Right. Well, it's not going to get lost now.
01:17:45 <peter1139> /home/runner/work/_actions/OpenTTD/OpenTTD-git-hooks/main/hooks/check-message.py:6: SyntaxWarning: invalid escape sequence '\d'
01:25:04 * dwfreed hands peter1139 an 'r'
01:26:25 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
01:32:00 <dwfreed> or "#\\d+" but r is generally preferred
01:33:01 <peter1139> Oh. Okay. Well, it's not anything I'm touching here.
01:33:24 <peter1139> Nice though, lots of warnings, my code is clang-safe only :S
01:40:46 <peter1139> Okay, a couple of sillies, but now compiling with gcc.
02:05:09 <peter1138> auto cni = std::make_unique<CompanyNewsInformation>(STR_NEWS_COMPANY_LAUNCH_TITLE, c);
02:05:09 <peter1138> AddNewsItem(GetEncodedString(STR_NEWS_COMPANY_LAUNCH_DESCRIPTION, cni->company_name, t->index),
02:05:09 <peter1138> NT_COMPANY_INFO, NF_COMPANY, NR_TILE, c->last_build_coordinate.base(), NR_NONE, UINT32_MAX, std::move(cni));
02:06:02 <peter1138> auto cni = std::make_unique<CompanyNewsInformation>(STR_NEWS_COMPANY_LAUNCH_TITLE, c);
02:06:02 <peter1138> EncodedString str = GetEncodedString(STR_NEWS_COMPANY_LAUNCH_DESCRIPTION, cni->company_name, t->index);
02:06:02 <peter1138> AddNewsItem(std::move(str),
02:06:02 <peter1138> NT_COMPANY_INFO, NF_COMPANY, NR_TILE, c->last_build_coordinate.base(), NR_NONE, UINT32_MAX, std::move(cni));
02:06:13 <peter1139> First one breaks. Second one works.
02:11:22 *** gelignite has quit IRC (Quit: Stay safe!)
02:16:44 <peter1139> Sorry for brutalising the CI.
02:22:48 <peter1139> GAH, I need whatever plugin in VS Code that's doing that to not do that :(
02:48:12 *** Flygon has quit IRC (Read error: Connection reset by peer)
03:10:13 *** Wormnest has quit IRC (Quit: Leaving)
03:19:41 *** gnu_jj_ has quit IRC (Ping timeout: 480 seconds)
03:21:59 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:27:19 *** D-HUND is now known as debdog
04:43:20 <DorpsGek> - Update: Translations from eints (by translators)
06:08:29 *** keikoz has quit IRC (Ping timeout: 480 seconds)
07:16:13 <DorpsGek> - Add: summary for week 01 of 2025 (by OpenTTD Survey)
09:13:37 <peter1139> I open up my browser, go to the website address. It loads and displays all my past conversation.
09:13:50 <peter1139> It then also needs that I need to login.
09:17:40 *** kuka_lie has joined #openttd
09:34:02 <andythenorth> £400 for an extra 24 GB
09:34:17 <andythenorth> I guess I won't be running many LLMs locally
09:41:15 <LordAro> your productivity will suffer immeasurably
09:41:41 <peter1139> kuhnovic, $work says no.
10:07:52 <peter1139> I wonder how to add #include checks into the commit checker.
10:08:45 <peter1139> Just to check that stdafx.h is first.
10:15:43 <LordAro> clang-format with everything else disabled?
10:16:10 <peter1138> Not sure clang-format knows that stdafx.h needs to be first.
10:17:19 <LordAro> should be able to enforce the disallowed functions header last too
10:17:36 <LordAro> the trick would probably be "with everything else disabled"
10:40:24 <peter1139> Do not eat all the Timtams at once challenge.
10:42:30 *** gelignite has joined #openttd
11:13:10 <_zephyris> What would opengfx2 as default actually require?
11:14:02 <_zephyris> As far as I can see, the OTTD code just points at a url... Though a cdn.openttd.org one
11:15:57 <peter1139> Well, step 1 is do you consider it ready?
11:16:53 <peter1139> (Which is not the same as perfected :))
11:18:12 <_zephyris> Pretty much. There's a couple of minor bugs to squash.
11:18:22 <_zephyris> Feedback is good, and I've dogfooded for 6 months
11:19:40 <_zephyris> .... but the build/make process is somewhere between eccentric and bad
11:20:01 <_zephyris> Good product, bad code.
11:22:44 <_glx_> IIRC there's nothing to change in openttd itself, it's all done from content server
11:30:58 <LordAro> i can imagine we'd want to set up CI/uploads/infra "properly"
11:31:20 <peter1139> There's some references to OpenGFX, for tests and release builds (store releases do bundle it)
11:37:06 *** gelignite has quit IRC (Quit: Stay safe!)
12:26:32 <peter1139> Hmm, Timer triggers are a bit weird.
12:28:08 <johnfranklin> I misread as “timtam triggers”
12:36:42 <peter1139> 92% of hours played don't have more minutes per year than normal.
12:37:30 <peter1139> Hmm, stats for 14 and vanilla are quite different. 84% for vanilla.
12:37:40 <peter1139> 7200 seems popular. That's a lot.
13:02:59 *** gnu_jj_ has joined #openttd
13:04:51 *** gnu_jj has quit IRC (Ping timeout: 480 seconds)
13:14:08 <peter1139> Localised time-of-day?
13:14:53 <xarick> mfw 4 minutes into a video i realize its being narrated by an AI 😦
13:15:10 <andythenorth> welcome to the future
13:15:16 <peter1139> “9:00 a.m., 1st Jan 1950”
13:15:20 <xarick> im starting to believe the dead internet theory
13:15:35 <peter1139> Usually takes about 1-2 seconds to spot it's AI.
13:16:15 <xarick> what got me was the way the AI was pronouncing DRAM
13:16:56 <peter1139> They be rolling out AI generated voice translations too which is pissing off the creators.
13:18:26 <_glx_> Auto translating titles and descriptions is so annoying too
13:19:19 <peter1139> Hmm, we don't have date display localisation already. Hmm.
13:19:53 <peter1139> It's language-based, so kinda.
13:20:41 <andythenorth> lol @ dead internet theory
13:20:48 <peter1139> Normal players get D MMM YYYY, weirdos get MMM D, YYYY.
13:21:38 <belajalilija> The best people got yyyy mm dd
13:30:43 <peter1139> I dunno, maybe someone wants to address it.
13:32:38 <LordAro> one assumes there must be a workaround possible
13:42:12 <_zephyris> LordAro: Wouldn't be too hard to split sprite generation (the bad python) from actually making the grfs
13:42:34 <peter1139> Feel free to dive into all the APIs to find it.
13:43:13 <peter1139> If sprite generation is part of the build, then no reason to.
13:49:48 <peter1139> Actually I've got a solution for SDL2 now. Hah.
13:53:25 <_zephyris> Fair enough... Time to learn make I guess!
13:53:58 <peter1139> Or if the whole thing is buildable from python, do that...
13:55:26 <_zephyris> Err, yeah, should be 100% python.
14:00:47 <LordAro> can nmlc be used as a python module?
14:12:57 <_zephyris> Python 3.10.12 (main, Nov 6 2024, 20:22:13) [GCC 11.4.0] on linux
14:12:57 <_zephyris> Type "help", "copyright", "credits" or "license" for more information.
14:12:59 <_zephyris> ['__builtins__', '__cached__', '__doc__', '__file__', '__license__', '__loader__', '__name__', '__package__', '__path__', '__spec__']
14:33:56 <_glx_> _zephyris: you can manually import what you want using `from nml import <stuff>`
14:34:16 <peter1139> Is it wise to though?
14:36:11 <_glx_> well FIRS and IronHorse could theorically generate the AST directly instead of generating a to be parsed nml 🙂
14:37:49 <_glx_> but it's a pain to create the AST by hand
14:38:45 <_glx_> grf-py is better on this aspect
14:39:10 <_zephyris> Dunno about wise, but:```
14:39:10 <_zephyris> >>> from nml import main
14:39:27 <_glx_> yeah `nmlc` does that 🙂
14:40:12 <_glx_> #! /usr/bin/env python3
14:40:12 <_glx_> if __name__ == "__main__":
14:40:32 <peter1139> A build system would be able to determine which bits need to be built though.
14:40:40 <peter1139> If you do it in python you have to reimplement that yourself.
14:41:00 <peter1139> Also generating sprites concurrentlky.
14:47:31 <_glx_> ah ok it does it for all logs
14:51:45 <_zephyris> peter1139: Yup. Mostly already does, a simple source vs. output modification time check
14:52:13 <_zephyris> What do you mean by generating concurrently?
15:03:00 <locosage> I use nml as a module in grf-py
15:03:10 <locosage> but nml api isn't well-designed to be used like that
15:05:34 <locosage> grf-py works pretty well as build system and everything
15:05:44 <locosage> especially for basesets as I've done bonky with it
15:06:57 <locosage> don't think it does concurrent generation though atm
15:07:05 <locosage> but it's not hard to add
15:07:18 <locosage> just didn't feel the need atm
15:07:23 <locosage> it's pretty fast as it is
15:12:31 <_zephyris> I suspect good enough for now then...
15:12:50 <_glx_> locosage: isn't well-designed at all 😉
15:16:04 <_zephyris> Sounds like a clean python-only build is probably the key milestone for an OpenGFX2 1.0 then
15:20:01 *** gnu_jj_ has quit IRC (Ping timeout: 480 seconds)
15:39:04 *** wubbianalt has quit IRC (Quit: User went offline on Discord a while ago)
15:45:44 <peter1139> Nice, doing remote system updates and my VPN gave up for a bit.
15:46:39 <peter1139> Hmm, how does sound work with NML.
15:49:03 <peter1139> Looking at action11, it seems it doesn't check for a .wav file at all?
15:49:23 <peter1139> (That would make sense, as people have inadvertantly included stereo sounds before now.)
15:54:43 <peter1139> Okay, so looking at the code and not actually trying it, I think NML already supports sounds in Ogg Opus format.
15:54:49 <_glx_> seems it's all done via `sound` and `import_sound` builtins and indeed the only check is file existence and not being 0 bytes
15:57:28 <_glx_> also no check on the number of sounds
16:28:58 *** Wormnest has joined #openttd
17:00:47 <LordAro> possibly. no one really understands those newgrf makefiles that PM created
17:01:21 <peter1139> I think I have a revamped one somewhere.
17:22:24 <_glx_> they are overcomplicated because they are meant to be generic
17:24:20 <peter1139> They don't set up targets properly.
17:26:56 <_zephyris> Sounds a bit beyond me - I won't worry!
17:40:43 <peter1139> Let's see how long for it to fail. All my mac OS changes are blind.
17:42:23 <LordAro> would that fix #3966 too?
17:42:41 <LordAro> oh wait, it's not a known issue, is it?
17:43:51 <peter1139> #3966 is for Windows, so it doesn't touch that.
17:45:16 <peter1139> Technically that's SDL1 ;-) (excuses)
17:45:27 <LordAro> haven't we removed that driver yet?
17:45:55 <LordAro> regardless, possibly an entry in known-bugs that could be removed
17:56:01 <peter1139> Well yes, I was wondering.
17:56:39 <peter1139> As long as the issue still affects Windows then there's no much point removing the known issue.
17:57:23 <peter1139> mac OS compiles now though that doesn't mean it works.
17:59:36 <LordAro> you should fix windows too :p
18:04:35 <peter1139> I should also go out offroading on the bike tonight, but that's not going to happen either.
18:24:45 <truebrain> Basically, Wayland made things worse for Linux 😛
18:27:09 <truebrain> For context, if that helps
18:27:15 <truebrain> I still have nightmares 😛
18:29:17 <peter1139> Maybe I can simplify this to rely on auto-capture.
18:34:27 <peter1139> And yes, I am of course aware of Wayland being shit.
18:35:42 <peter1139> "Ready for prime time"
18:36:32 <andythenorth> "year of the Linux desktop"?
18:38:08 <peter1139> I can't test this on Wayland because I stick with Xorg.
18:43:35 <andythenorth> is that pronounced 'Zorg"
18:44:20 <johnfranklin> "cha-ou-ar-jee" or "Zorg" for me
18:47:21 *** tokai|noir has joined #openttd
18:47:21 *** ChanServ sets mode: +v tokai|noir
18:54:03 *** tokai has quit IRC (Ping timeout: 480 seconds)
19:01:10 <xarick> viewport scrolling with cursor at the edge?
19:12:35 <peter1139> Github doesn't give me a reply box to last comment :o
19:15:30 <Rubidium> but it does to the first comment, which is where I commented on
19:16:45 <peter1139> Weird and annoying.
19:30:23 <peter1139> The fact my StringParameter constructors completely fail to work with MSVC is a bit annoying...
19:31:59 <LordAro> do we really need Windows support though?
19:38:10 <xarick> viewport scrolling should ignore the toolbars
19:38:21 <xarick> they block window scrolling
19:38:34 <peter1139> You're talking about edge-scrolling which nobody uses.
19:38:49 <xarick> nobody uses because it doesn't work
19:40:06 <Heiki> I do, because it (mostly) works
19:49:20 *** firmat has quit IRC (Ping timeout: 480 seconds)
19:50:00 <peter1139> (It's about 95% turned off)
19:59:30 *** firmat has quit IRC (Ping timeout: 480 seconds)
20:08:55 <peter1139> Might need to do something else anyway though, probably doesn't play nice with std::move (maybe the existing code doesn't either)
20:21:54 <Rubidium> looks like MSVC just has problems with the braced initializer 'constructor' thinking something gets narrowed. It's definitely not the first one ;(
20:22:47 <_glx_> wensimehrp: can you attach an error message or `-s` trace to the issue ? (I'm lazy)
20:23:56 <_glx_> edge scroll is annoying anyway, often triggers when you don't want it
20:25:17 <peter1139> Which is precisely why it's designed to not scroll when you hit the toolbar.
20:26:58 <peter1139> Argh, fecking clangd.
20:28:25 <peter1139> Hopefully that'll fix the dedicated build.
20:39:25 <andythenorth> We’re supporting “achievements” yes/no?
20:51:09 <LordAro> peter1139: external game?
20:56:47 <truebrain> I doubt Steam sees that way of launching of OpenTTD as an "Official OpenTTD" version 🙂
20:57:41 <talltyler> Some people put JGRPP in their Steam folder to track their hours
20:57:51 <talltyler> A true case of spacebar heating 😛
20:59:05 *** gelignite has joined #openttd
20:59:38 <truebrain> I am sure it is a valid case to run OpenTTD 🙂 I just doubt it is one that Steam supports 🙂
21:00:02 <talltyler> triq55: Discord Moderator
21:07:10 <_zephyris> Did a quick dumb test. Steam plugin runs successfully when openttd is launched from steam, not when the executable installed through steam is run directly.
21:07:34 <truebrain> The latter can never work, as the SteamID of OpenTTD is only injected when Steam launches OpenTTD
21:07:45 <truebrain> so that is completely expected behaviour 🙂
21:08:00 *** kuka_lie has quit IRC (Quit: Lost terminal)
21:08:01 <truebrain> I just don't know if you manually add a game, if Steam is clever enough to figure things out
21:08:03 <truebrain> I would assume it is not
21:08:12 <truebrain> as it would be very tricky to do that 😄
21:11:10 <_glx_> yup you can see the injection in debug window
21:14:17 <xarick> is there a quick way to see which PRs have conflicts currently?
21:20:43 <_zephyris> truebrain: Hehe. Anyway, I did the dumb test - it does indeed not work.
21:21:17 <truebrain> 😄 Doing dumb things is my favorite mode of them all 😄
21:21:34 <truebrain> I am also VERY good at that btw
21:21:57 <xarick> Rubidium gives me hope
21:30:28 <_glx_> funny how nobody triggered the issue before waypoint `\FF` thing
22:05:31 <peter1139> I think a lot of authors think they can be only ascii characters anyway.
22:07:34 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:10:39 *** nielsm has quit IRC (Ping timeout: 480 seconds)
22:35:44 *** keikoz has quit IRC (Ping timeout: 480 seconds)
23:17:52 <_glx_> I could approve #344, but most of it is my code 🙂
23:21:52 <peter1139> Ah the one where pointing out an error was met with confusion.
continue to next day ⏵
