IRC logs for #openttd on OFTC at 2025-06-11
⏴ go to previous day
00:51:21 *** WormnestAndroid has quit IRC (Remote host closed the connection)
00:51:27 *** WormnestAndroid has joined #openttd
01:32:53 *** Wormnest has joined #openttd
02:02:06 *** herms has quit IRC (Quit: bye)
02:38:12 *** gnu_jj_ has joined #openttd
02:41:14 *** Wormnest has quit IRC (Quit: Leaving)
02:41:44 *** gnu_jj has quit IRC (Ping timeout: 480 seconds)
03:13:19 *** debdog has quit IRC (Ping timeout: 480 seconds)
04:42:37 <DorpsGek> - Update: Translations from eints (by translators)
04:58:34 *** SigHunter has joined #openttd
06:16:29 *** gelignite has joined #openttd
06:44:36 *** toktik is now known as Guest17933
06:52:51 *** Guest17933 has quit IRC (Ping timeout: 480 seconds)
07:20:13 *** cmcaine has joined #openttd
07:20:13 <cmcaine> Hello, contributing.md says I should post here if I'm thinking of contributing a feature to OpenTTD.
07:20:13 <cmcaine> I'd like to make the keyboard shortcuts visible in the UI. A [PR from 2022](https://github.com/OpenTTD/OpenTTD/pull/9732) by kuhnovic did most of the work by associating hotkeys with widgets and showing the hotkeys in tooltops, but it was not merged.
07:20:13 <cmcaine> If I developed a similar PR and addressed the translation issues, would it have a decent chance of being merged? What about an alternate proposal where the shortcuts for each button/widget are drawn on top of the button (perhaps when the user holds Alt, which matches some other applications)?
07:21:55 <cmcaine> Also, if someone else wants to just knock out a patch because it will be easy for them, then that's totally fine by me. I'm not precious about it.
07:23:15 <kuhnovic> If your PR contains a good solution that we all agree with then it will definitely be merged. I gave up on it because it was too deep of a rabbit hole and I had too little time for it. That doesn't mean it can't be done 🙂
08:06:08 <peter1138[d]> More dynamic interface scale, instead of going by what the OS says...? Lots of spacebars 🙂
08:07:21 <peter1138[d]> Sorry, flashing warning!
08:07:37 <peter1138[d]> (That's just resizing with SDL.)
08:08:31 <LordAro> i thought we double buffered to avoid that?
08:11:05 <peter1138[d]> It's the OpenGL stuff that was added on later.
08:12:41 <peter1138[d]> (Also, 19MB? Christ)
08:13:38 <peter1138[d]> Of course, this has a little "what's the native size of the game" issue...
08:15:49 <peter1138[d]> Comment from elsewher:
08:15:53 <peter1138[d]> > i changed it to 2x and it looks much better, i wonder why autodetect made it so tiny
08:16:05 <peter1138[d]> > Autodetect usually doesn't get it right
08:16:26 <peter1138[d]> Now, we know that there is no autodetect, we just go by what the OS tells us.
08:17:42 <peter1138[d]> (But it is called "Auto-detect size", so not issues on that assumption)
08:22:47 <LordAro> i have suddenly become unable to type any uppercase letters
08:23:06 <LordAro> shift has stopped working
08:23:43 <peter1138[d]> Have you tried the other shift?
08:23:54 <LordAro> yes, and sticky keys still activates, so it's not hardware
08:24:31 <peter1138[d]> Funny thing, I varely rarely use the right shift key, because the one on our BBC Micro broke early on...
08:25:33 <LordAro> on screen keyboard doesn't do anything weird
08:25:50 <LordAro> it's not that it just types lowercase, it doesn't type anything at all
08:26:16 <peter1138[d]> Try pressing and releasing all the other modifier keys.
08:27:04 <peter1138[d]> I had weirdness a few times when the OS hadn't registered a key being release for some reason.
08:27:17 <peter1138[d]> (The OS, or more likely Xorg.)
08:27:38 <LordAro> not a lot of xorg on windows :p
08:28:04 <peter1138[d]> That's your own fault 🙂
09:06:28 <wensimehrp> So some sort of ctrl+shift shortcut that triggers the shortcut layer?
09:10:30 <LordAro> fitting everything within the button without making a mess might be tricky
09:14:05 <_zephyris> My suggestion for a first solution is be automatic addition of the keyboard shortcut to button tooltips.
09:14:55 <wensimehrp> LordAro: It's impossible
09:15:48 <_zephyris> Text on larger icon buttons could work - for mocking up, use white text with a 1px (scaled with UI size) black drop shadow, because that's the OTTD style.
09:16:21 <_zephyris> (fonts doesn't make it impossible, but may make it weird for some configurations)
09:22:46 <peter1138[d]> For mocking up I generally just implement the feature, and then leave it lying around as a branch that no-one else ever sees...
09:23:50 *** reldred has joined #openttd
09:25:10 <reldred> I've been getting a lot of mileage out of that gif lately
09:26:46 <peter1138[d]> We do have a black outline in one place... for the depot cargo icon overlays.
09:27:02 <peter1138[d]> And yeah, it's inefficientlly drawing the cargo icons 5 times 🙂
09:36:39 <xarick> hi, I have a serious question
09:37:38 <xarick> peter can you revise that part is needed, since you worked on that recently
09:37:51 <xarick> i can't make it reachable
09:38:32 <peter1138[d]> Because I'm not going to look at something you've now posted at least 3 times.
09:39:46 <peter1138[d]> In fact, I'm just not going to.
09:42:33 <xarick> oh the last line is reachable, my bad
09:43:02 <xarick> but the other for loop... I can't for the best of me enter it
10:35:51 <peter1138[d]> Rebase from 4 years ago, or rewrite...
10:39:20 <_zephyris> peter1138[d]: Ugh, my graphic design mind dislikes that... That's inconsistent with original icon design and fonts
10:40:28 <_zephyris> It's OK, you don't need to care 😉
10:42:03 <peter1138[d]> But yes, it's a bit shoddy with non-1x sprites.
10:44:49 <peter1138[d]> Don't look now 😉
10:46:13 <peter1138[d]> The feet are the best.
10:47:49 <peter1138[d]> If we only had one blitter, I'd probably have tried to implement a nicer outline system.
10:47:54 <peter1138[d]> But as we have about 5 million...
10:48:24 *** ahyangyi has joined #openttd
10:48:24 <ahyangyi> Do we have usage stats on blitters?
10:48:29 <reldred> one blitter to rule them all? and in the darkness bind them?
10:50:08 <peter1138[d]> "But 8bpp support!"
10:51:13 <reldred> anything legitimately hardware bound to 8bpp is probably running short on RAM and CPU cycles to be running OpenTTD in 2025.
10:51:16 <ahyangyi> Haven't looked at this channel recently but the new toyland looks so pleasant
10:51:48 <reldred> Because thinking back to the old PC's I had, the Toshiba 110CT was 8bpp bound and it couldn't run the original TTD at a solid 60fps, it *bogged*
10:52:22 <reldred> The IBM 380 had enough guts to run TTDPatch smoothly, but it was also 16/24bit capable
10:52:39 <ahyangyi> (busy building my new pc, which should now have enough RAM and CPU to run OpenTTD in 2025)
10:54:13 <peter1138[d]> Nah we keep making it slower.
10:54:42 <xarick> make openttd multithreaded
10:55:11 <reldred> I'm OK with making it slower if it means y'all can get more audacious with the game, like fancy new map array, etc. 😛
10:56:14 <reldred> OpenGFX2 for openloco
10:56:32 <reldred> i can't stand how loco looks
10:56:41 <reldred> like a brown pixellated smear
10:57:06 <LordAro> mm, a bit too "British"
10:57:15 <LordAro> odd though, because obviously RCT is great
10:57:35 <reldred> yeah haven't been to the UK since '96 but what i remember was pretty much a brown pixelated smear
10:57:47 <reldred> yeah I know, RCT/RCT2 was lovely.
10:58:03 <LordAro> that's unfair, just a bit too "Britain in February"
10:59:47 <andythenorth> emojis in badges?
11:03:17 <_zephyris> reldred: realism != good
11:03:50 <_zephyris> I would quite enjoy cargo icons etc. in strings, factorio style
11:07:12 <peter1138[d]> That was a side-quest I thought about and then forgot about :p
11:11:48 <peter1138[d]> Needs to have a unicode range dedicated to cargo slots, as layouting/drawing strings is past the point of having parameters.
11:15:52 <peter1138[d]> Hmm, actually dynamically allocated may be a nice approach, so a codepoint is allocated when the string code for "cargo icon" with parameter "passengers" is seen, and a handler installed to deal with it when layouting/drawing (similar to the sprite fallbacks)
11:16:43 *** Flygon has quit IRC (Read error: Connection reset by peer)
11:17:02 <peter1138[d]> Badge icons embedded in strings?
11:22:21 *** Emru has quit IRC (Quit: Gateway shutdown)
11:22:38 <peter1138[d]> U+F0000-U+FFFFD allows 65534 custom glyphs, probably enough.
11:25:54 <mnhebi> peter1138[d]: 256bpp blitter when?
11:27:44 <mnhebi> :widdle_goblin:: it was a joke
11:28:09 <reldred> getting more mileage on those gifs
11:31:10 <xarick> here I come with my pointless bench runs
11:31:43 <xarick> performance seems all over the place... doesn't meet my expectations
11:34:14 *** foodliker has joined #openttd
11:34:14 <foodliker> reldred: still is now
11:34:44 <mnhebi> Pointless? If you doing it for the sake of some objective, it is not pointless.
11:35:09 <mnhebi> As long as it has value to you, it is not pointless.
12:04:22 *** WormnestAndroid has quit IRC (Remote host closed the connection)
12:04:35 *** WormnestAndroid has joined #openttd
12:42:25 <LordAro> mnhebi: you're new here
12:53:49 <mnhebi> I have always been here.
12:55:05 <reldred> You’ve been a bit quiet for a while
13:11:49 *** keikoz has quit IRC (Ping timeout: 480 seconds)
13:42:49 *** keikoz has quit IRC (Ping timeout: 480 seconds)
13:44:24 *** masteroktagon has joined #openttd
13:44:24 <masteroktagon> Hello, I want to try to implement one-way trams in OpenTTD. Is this more a JGR-scope or "Default-scope" patch?
13:44:48 <talltyler> I would welcome one-way trams in vanilla if they're implemented properly. 🙂
13:45:17 <talltyler> (meaning properly coded, intuitive to users, etc.)
13:48:47 <masteroktagon> xarick: No one-way-one-lane but asymmetric to match the roads
13:52:32 <masteroktagon> would of course need a lot of new graphics but I am going to do the coding first
13:54:09 *** keoz has quit IRC (Ping timeout: 480 seconds)
13:54:20 <talltyler> Luckily the graphics are pretty simple, you'd just need to copy the existing sprites and erase the rails that aren't there 🙂
13:56:54 <talltyler> Might be worth layering junction sprites to avoid combinatorial explosion, similar to how rail junctions work (GRFs can provide specific junction sprites, not sure I've ever seen anyone use them though)
13:59:50 <masteroktagon> yes sounds like a good idea
14:00:23 *** D-HUND is now known as debdog
14:02:11 <masteroktagon> Also the question is how do we handle tram stops? Only draw half of them? Requires new sprites and may be a problem on normal roads. Only allow double track stations? Very unintuitive, but the first I will try to achieve because its easier.
14:05:49 <talltyler> I'd draw the usual double track sprites, because they also serve as bus stops. GRFs can provide one-way stations (and already do).
14:06:42 <talltyler> Automatic one-way stations would be an interesting stretch goal, but I think this could be a later PR to not do too much at once (both for you, and for reviewers).
14:07:05 <talltyler> (Automatic one-way station sprites would also need to look at roads, which is a harder challenge 🙂 )
14:12:54 <peter1138[d]> Yeah, I don't think one-way trams needs any new graphics, just functionality. NewGRFs can already provide custom graphcs for one-way sprites.
14:17:22 <talltyler> Not even one-way tram tracks? Tram tracks are already provided by OpenTTD (for TTD or incomplete base sets) and you'd have to make them separate sprites that base sets can replace anyway...
14:18:36 <peter1138[d]> "one-way-ness" only applies to the tile with the one-way signjs.
14:20:09 <peter1138[d]> So you could perhaps draw that tile with only one side.
14:20:22 <peter1138[d]> But it would connect to tiles with both sides.
14:20:40 <talltyler> That is for masteroktagon to solve, I guess. 🙂
14:20:40 <talltyler> JGRPP detects one-way-ness by looking at the entire road segment between junctions, and vehicles on a segment detected as one-way due to one-way markers use both lanes.
14:23:23 *** kuka_lie has joined #openttd
14:24:12 <masteroktagon> I thought of giving roads (specifically trams) for bits more (maybe m8) in the map array to have an extra bit for in/out directions (the new bits prob. would have to be negated to allow backwards compabiliy)
14:24:52 *** gelignite has joined #openttd
14:28:03 <talltyler> I haven’t looked at the map array docs but one-way level crossings are very likely to be a problem, I think they’re close to full 🙂
14:30:11 <masteroktagon> there are some bits left ex. the four bits FEDC of m8
14:41:22 <peter1138[d]> Hypothetically, if a town has a straight road with no junctions and a company places twon road tiles with one-way markers at each end, is the entire road one-way?
14:42:10 <andythenorth> bigger map array 😛
14:45:16 <peter1138[d]> And then if the town adds a junction in the middle?
14:45:27 <talltyler> Then it’s no longer one-way
14:58:45 *** Wormnest has joined #openttd
14:59:20 <masteroktagon> Not sure if this is the place to ask, but I am getting a compiler error like this. I suspect I might have forgotten to configure something. Does anyone know where this comes from? I have not changed something at this point.
14:59:20 <masteroktagon> openttd/src/settingsgen/../core/format.hpp:33:13: error: redefinition of ‘struct fmt::v11::formatter<T, Char>’
14:59:20 <masteroktagon> 33 | struct fmt::formatter<T, Char> : fmt::formatter<typename T::BaseType> {
14:59:20 <masteroktagon> | ^~~~~~~~~~~~~~~~~~
14:59:22 <masteroktagon> /home/oskar/Programmierprojekte/cpp/openttd/src/settingsgen/../core/format.hpp:17:13: note: previous definition of ‘struct fmt::v11::formatter<T, Char>’
14:59:22 <masteroktagon> 17 | struct fmt::formatter<E, Char> : fmt::formatter<typename std::underlying_type_t<E>> {
14:59:24 <masteroktagon> | ^~~~~~~~~~~~~~~~~~
14:59:45 <LordAro> you're lucky this came up yesterday :p
15:12:55 <masteroktagon> is g++ 12 new enough?
15:18:11 <peter1138[d]> Hmm, I appear to have a dodgy "behind the knee" muscle.
15:20:34 <andythenorth> is the pain in the left or right of the knee
15:20:47 <andythenorth> I have no idea what I'm talking about I just happened to research it about 2 hours ago
15:21:09 <LordAro> mine is probably tendon related
15:21:19 <andythenorth> there were a lot of words like `anterior`, so I stopped
15:21:37 <peter1138[d]> Mine is probably 100 miles at the weekend related.
15:21:39 <andythenorth> but I am going to massage gun my hamstrings, quads and calf muscle
15:21:47 <andythenorth> that usually works
15:21:53 <dwfreed> andythenorth: anterior means front :P
15:22:06 <peter1138[d]> Can't find the cable for my massage gun 😦
15:22:22 <andythenorth> what about Anteromedial thigh?
15:22:55 <andythenorth> "popliteal fossa"
15:23:07 <andythenorth> hmm yeah internet doctoring, unwise
15:25:02 <peter1138[d]> Neat, I found it.
15:25:28 <andythenorth> guidance varies on what it's safe to massage gun
15:25:34 <andythenorth> but 'back of the knee' seems to be nope
15:25:58 <peter1138[d]> If it hurts a bit (can't easily bend) should I avoid riding the bike later?
15:26:12 <peter1138[d]> "Doctor it hurts when I do this"
15:26:21 <masteroktagon> sorry to interrupt you again, I am still having problems compiling (this time it seem like it does not choose the right libc++ version and thus missing std::source_location). Anyone having an Idea how to fix this?
15:26:37 <LordAro> "can't easily bend" seems like more than "a bit"
15:26:41 <andythenorth> I have started NOT climbing hard on my injured shoulder
15:27:04 <andythenorth> seems when it's injured it doesn't go away by using it more
15:27:08 <andythenorth> except when it does
15:27:16 <LordAro> masteroktagon: not immediately obvious. definitely choosing the right compiler? you reconfigured after installing the new version, etc?
15:28:13 <masteroktagon> yes installed g++-12 and used ln to link it to g++ (same with gcc). Then installed libc++-19-dev
15:28:38 <masteroktagon> and removed all contents from the build directory
15:28:55 <dwfreed> gcc would be using libstdc++, not libc++, generally, unless specifically built that way
15:28:58 <LordAro> libc++ is specific to clang (unless you're really trying)
15:29:13 <LordAro> i'm not sure i'd recommend using ln to override compilers like that either
15:29:48 <masteroktagon> dwfreed: well I have libstdc++-12-dev which seems to be the newest in apt
15:30:27 <_jgr_> A cmake command along the lines of `CC=/usr/bin/gcc-12 CXX=/usr/bin/g++-12 cmake .. -DCMAKE_BUILD_TYPE=Release` should be sufficient
15:30:45 <locosage> and make sure you do a clean build
15:34:42 <peter1138[d]> I fibbed, I can easily bend. It's just just a bit tender (not particularly ouch)
15:35:35 <LordAro> peter1138[d]: well, be careful with it
15:35:39 <LordAro> you don't want to end up like me :p
15:39:13 <peter1138[d]> What, fit and ready to do 250 mile rides?
15:39:28 <masteroktagon> _jgr_: seems to have fixed it (still building right now, but no errors). Thanks everyone!
15:40:05 <peter1138[d]> I like that VS Code (or Codium) can handle all that stuff for me.
15:40:23 <peter1138[d]> Or least, the cmake extension can.
15:41:49 <masteroktagon> well mine can't
15:42:42 <masteroktagon> And clangd recently decided to turn on clang-tidy by default and I have to watch out so it doesn't format all the code wrongly
15:51:16 *** akimoto has joined #openttd
15:58:43 *** gelignite has joined #openttd
16:05:29 <mnhebi> reldred: Yes...I got so many projects going on that I've only been able to work on weekends. At least now with my laptop I can work on them more when its quiet at work.
16:50:36 <peter1138[d]> Uhm. Hmm. Stupid utf-16 surrogate pairs.
16:53:20 <peter1138[d]> Fortunately, `Utf16DecodeChar` exists.
17:41:50 *** alpapilus has joined #openttd
17:41:50 <alpapilus> peter1138[d]: Senham central🫃
17:48:14 <talltyler> Assuming that’s typed by the player and not added based on cargo types at the station, looks like a useful feature to label platforms/waypoints/etc by players who like separating those out 🙂
17:48:56 <peter1138[d]> It's typed in my me but shouldn't actually be possible.
17:50:22 <peter1138[d]> And the bad offset is just OpenGFX 2 being wrong.
17:51:00 <talltyler> How do you type them?
17:57:28 <peter1138[d]> They're dynamically mapped a unicode codepoint.
17:58:06 <peter1138[d]> Because it's dynamically mapped, it's not reliable 🙂
18:15:23 <_jgr_> This seems like a bit of footgun for script authors, existing code using Valuate without a wrapper loop in the new scheme may or may not work as expected
18:19:39 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
18:19:50 <_jgr_> A partially valuated list, which doesn't indicate which elements have actually been done yet does not seem like something that should be exposed to external scripts at all
18:20:56 <_jgr_> If the script did something other than immediately call Valuate again, there's potential for resume_iter to be left dangling as well, which could be quite awkward
18:24:53 <_zephyris> Hmm, there's all kinds of complex newgrf cargo icon interaction isn't there...
18:26:51 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
18:27:08 *** WormnestAndroid has joined #openttd
18:29:06 <peter1138[d]> I don't think so?
18:30:04 <peter1138[d]> They're up to 10 pixels high, and can aligned within that 10 pixels. Easiest way to get correct positioning with replacement sprites is to just make it 10 pixels with a y-offset of zero.
18:32:15 <_zephyris> Oh, I mean for use in strings
18:32:24 <peter1138[d]> (The original sprites follow TTD's tradition of not wasting time on blank pixels, so they are various sizes with offsets to make up for it.
18:32:30 <_zephyris> Haven't checked the baseset offsets
18:32:47 <peter1138[d]> What interaction is there in strings?
18:33:05 <peter1138[d]> (Considering they don't exist in any strings... except in this branch I cobbled together just now.)
18:36:48 <_zephyris> Just that, takes more than a cobble
18:37:16 <peter1138[d]> _glx_: can you hide the while() bit so that's done in a way that's internal to the system, but also suspndable?
18:37:26 <_zephyris> They should be 10x10px then?
18:37:45 <_zephyris> It's one of those things that's not obvioous when back-engineering from the original graphics
18:37:45 <peter1138[d]> 10px high. Width is whatever you want.
18:37:58 <peter1138[d]> Within reason, of course.
18:38:06 <peter1138[d]> They're a bit like badges, tbh 😄
18:38:52 <_zephyris> I've standardised them to the same width, makes sense for proportional display of cargo quantity in the station views
18:39:11 <peter1138[d]> Most sets that do them do, yeah.
18:39:52 <peter1138[d]> The cargo waiting view where they are all side by side uses a fixed spacing of 10px, regardless of the sprite width.
18:41:26 <peter1138[d]> Other places generally allocate enough space for the largest icon, and then centre horizontally within that space. But they're not centred vertically.
18:41:40 <peter1138[d]> (Because that's bad with the original sprites)
18:43:07 <peter1138[d]> Anyway, now I need to make sure the list of allocations is in sync with cached string layouts.
18:43:18 <peter1138[d]> I think that means just clearing it at the same time.
18:46:42 <peter1138[d]> Oh, of course, wrong data member.
18:50:35 <peter1138[d]> And, well... is this the way to go?
18:57:27 <peter1138[d]> Benefit: The game, and GS, could put cargo / badge icons anywhere in text.
18:57:58 <peter1138[d]> Things get a bit awkward if the font size is smaller than normal.
19:10:40 <_zephyris> Somewhat like the bus/truck/train/ship/plane icons in text
19:12:46 <peter1138[d]> They were always glyphs though 🙂
19:13:56 <_zephyris> Anyway, standardised 10x10 cargo icons
19:53:33 *** Emru has quit IRC (Quit: Gateway shutdown)
20:29:47 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
20:30:54 *** WormnestAndroid has joined #openttd
20:33:40 <peter1138[d]> Hmm, this hack would probably be a lot cleaner witht he side-by-side stuff... that doesn't yet work on Windows or mac OS 😮
20:47:15 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
21:05:23 *** akimoto has quit IRC (Remote host closed the connection)
21:09:39 <peter1138[d]> Spamming random unicode characters and seeing what gets loaded...
21:51:19 *** keikoz has quit IRC (Ping timeout: 480 seconds)
21:59:34 *** Wormnest has joined #openttd
22:01:24 *** ChanServ sets mode: +v tokai
22:07:51 <cmcaine> wensimehrp: I was thinking that the hotkeys are displayed shown as hints over their corresponding buttons when you hold Alt, similar to how holding Alt in some applications causes a top menu to be displayed with an underline under the letter in each item's name that is it's hotkey.
22:07:51 <cmcaine> Generally, when you hold Alt it only shows you the shortcuts that you access by pressing Alt+Whatever, but it's still a kinda similar idea.
22:08:25 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
22:12:53 <cmcaine> LordAro: Yes, I did a mock up where long key combinations are shown in a separate grey box near their button, but I think it would probably be possible to choose small-enough icons for modifiers that they are still legible; or it can just skip drawing long combinations.
22:12:53 <cmcaine> It's not really a good hotkey if it could be in the basic emacs config, anyway.
22:14:09 <cmcaine> _zephyris: Cool, I'll try the drop shadow, tho I think it's probably not as readable as an outline.
23:13:50 *** kuka_lie has quit IRC (Remote host closed the connection)
23:26:04 <wensimehrp> cmcaine: I think using the sprite font might work since most keys can be represented in letters in the English alphabet.
23:26:50 <wensimehrp> And for keys like ctrl alt shift you'll have to draw new sprites
continue to next day ⏵
