IRC logs for #openttd on OFTC at 2026-01-25
⏴ go to previous day
00:06:52 <kaibaneddy> peter1138: Yep, already did 🙂
00:10:30 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
00:18:16 *** WormnestAndroid has joined #openttd
01:44:00 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
01:46:09 *** WormnestAndroid has joined #openttd
03:34:48 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
03:35:15 *** WormnestAndroid has joined #openttd
03:39:51 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
03:42:13 *** WormnestAndroid has joined #openttd
03:43:59 *** Wormnest has joined #openttd
04:25:12 *** Wormnest has quit IRC (Quit: Leaving)
04:37:22 *** WormnestAndroid has quit IRC (Remote host closed the connection)
04:37:26 *** WormnestAndroid has joined #openttd
04:44:09 *** Zathras_1 has joined #openttd
04:47:30 *** Zathras_4 has quit IRC (Ping timeout: 480 seconds)
04:51:16 <DorpsGek> - Update: Translations from eints (by translators)
06:01:39 *** reldred has joined #openttd
06:01:39 <reldred> god damn the intellisense terminal integration in vscode is like someone smacking my figners with a hammer
06:01:45 <reldred> glad I figured out how to turn it off
09:04:44 <peter1138> I think 10 minutes practicing on the piano probably isn't enough.
09:05:07 <peter1138> And by practice I mean "trying to play pieces that I can't play"
09:06:28 <andythenorth[d]> so shall I copy Pikka's "leading brake van" for Horse 👀
09:07:54 <peter1138> As methods go, that's probably the simplest and least awkward.
09:11:29 <_zephyris> Piano music set for OpenTTD when?
09:12:55 <peter1138> There's loads, but generally transcribed instead of recorded MIDI, so sound bad.
09:32:30 <_zephyris> Exactly what I mean 😉
09:32:55 <_zephyris> I've been tempted, got an ancient Clavinova, but haven't worked out how to convert recordings to midi
09:33:24 <peter1138> I just plug in a USB stick and hit record, straight to MIDI.
09:33:39 <peter1138> But I don't practice enough so will make mistakes.
09:35:05 <_zephyris> The (really) old ones use some proprietary format
09:40:37 <_zephyris> I'm done... sorry for the numerous updates...
09:42:29 <peter1138> Rubidium's comment is possibly pre-last update.
09:43:55 <_zephyris> I'm demonstrating severe lack of attention to detail 😉
09:45:06 <_zephyris> I blame toddler interference
09:46:56 <Rubidium> you need to train the toddler better. The toddler should have been asking why it looks weird for each of the sprites :D
09:47:35 <peter1138> I didn't see before.
09:49:18 *** gelignite has joined #openttd
09:50:27 <peter1138> Not expecting this to work but trying anyway :)
09:52:04 <reldred> we're all really excited for it peter, id honestly forgotten you'd raised a PR for it til one of the JP+ mafioso mentioned it. Last I thought it was still a theoretical.
09:52:38 <reldred> this solves a great many conundrums of both cross .grf and cross feature logic.
09:54:16 <peter1138> Bah, gotta document the parameters of every 'interface' (function prototype, heh) implementation.
09:54:40 <peter1138> Unless doxygen has a "see other definition" thing?
09:55:32 <peter1138> Almost makes me consider just using a giant switch block instead :o
09:58:24 <_zephyris> @Rubidium Excellent idea!
10:00:48 <peter1138> Okay, that seems to satisfy it.
10:01:56 <reldred> I wonder if this is something that can be abstracted by NML or already is, but I'd like to be able to just in one go test everything within 1 tile, or everything within 2 tiles, etc. And for houses it'd be nice if that abstraction takes into account that a 2x2 building where all tests happen from the northern tile that when I want to test within a 1 tile radius I want to test 1 tile south of the
10:01:56 <reldred> southern tiles for instance.
10:03:18 <reldred> but yes tile testing for badges is cool and nice
10:03:47 <peter1138> Well I did consider radius instead of offset.
10:04:19 <peter1138> But then do you want "is present" or "percentage covered"
10:04:24 <peter1138> And how expensive is that...
10:04:36 <reldred> Offset I think is still very very useful for some purposes. Not ones that i'm interested in but the folks who want to align houses to roads for instance care very deeply about orientation
10:05:11 <reldred> My uses would just be 'is a building with this badge within 1 tile radius of me'
10:05:29 <reldred> as it drastically simplifies stuff I was doing with tile classes
10:05:42 <reldred> anyways classes was a pain
10:06:20 <reldred> especially cross .grf
10:06:45 <reldred> the desire with badge testing is very much that we don't care if the item is from another .grf if it has the badge that's what we're interested in
10:06:54 <peter1138> Currently it uses the standard offset mechanism, so you provide a byte with 4 bits of x and 4 bits of y. Only allows checking a 16,16 area centred at 8,8.
10:07:06 <reldred> Yeah that would be fine
10:07:18 <reldred> I'd like radius, but I won't throw my toys out of the pram if I don't get it
10:07:22 <peter1138> That's "probably" fine, and at least means existing nearby stuff can be/is used.
10:07:39 <peter1138> What's the size of a parameter...
10:07:52 <reldred> Be easy to retrofit existing logic to use badges instead of classes for instance
10:08:38 <reldred> I think for a lot of authors the sorta checks we'd want to do are within limited radiuses as well but I can't speak for everyone ofcourse.
10:08:48 <peter1138> Yeah, so we could stuff more bits AND keep the offset logic.
10:09:01 <peter1138> e.g. enough 4 bits for radius.
10:09:38 <peter1138> Or maybe less bits, because checking lots of tiles every time something is drawn will suck.
10:10:09 <reldred> It's usually only neighboring tile we're after. I've done 2 tile checks before with classes so a cluster of industrial buildings for instance can cross a road.
10:10:54 <reldred> Direction is useful, an example quoted to me was a house checking the neighboring road type to select an appropriate ground sprite.
10:12:22 <reldred> road types badge that is
10:12:40 <reldred> id presume they'd code a badge for certain common surface textures used between their family of grfs
10:13:42 <reldred> Personally I'm mostly interested in placement logic for houses that's compatible between .grf's, since I maintain a few different sets I'd like something looking for a commercial building to find it across multiple sets of mine.
10:15:59 <andythenorth[d]> could just do a big tile loop in nml
10:16:00 <_zephyris> Standard offset and radius search are good for other bits of newgrf, no complaints, so just emulate...
10:16:43 <andythenorth[d]> hmm mapping a function across ranges in grf? 👀
10:17:16 <reldred> andythenorth[d]: that's what I was wondering whether it was more appropriate as an abstraction inside NML than an underlying behavior in the spec.
10:18:05 <reldred> I mean it needs to get iterated through tile by tile one way or another, just wondering if it's computationally more expensive in NML or in engine to do that iteration
10:34:19 <peter1138> Well, if it's done by the author then NML or OpenTTD is not hiding the complexity.
10:34:35 <peter1138> That's been a problem with scripting as well.
10:34:49 *** Zathras_1 has quit IRC (Quit: Connection reset by beer!)
10:35:38 *** Zathras_7 has joined #openttd
10:41:49 <reldred> well either way, stoked
10:42:42 <reldred> cross grf cross feature tile checking is going to be a massive boon
10:46:03 <andythenorth[d]> " this variable is likely to be expensive to evaluate"
10:46:13 <andythenorth[d]> FIRS animation FPS impact, all over again?
10:46:34 <andythenorth[d]> I can't imagine this **not** being used in the graphics chain
10:46:35 <reldred> mobile ottd players btfo anyway
10:46:59 <reldred> i'd mostly use it for construction checks so not tooooo bad
10:47:29 <andythenorth[d]> wonder if a tile could have subscribers, and trigger a callback to them when it changes
10:47:40 <reldred> and most graphjics checks will be a 1 tile radius
10:48:02 <reldred> or very specifically checking the two bottom faces of a tile
11:14:25 <_zephyris> Yup. In my (houses) experience most useful by far is testing offset of one or two in cardinal directions, followed by a radial search for nearest up to a radius of about 4.
11:29:13 <peter1138> TBH for houses checking roads, the badges don't seem that useful,
11:30:10 <reldred> there's interest for matching ground sprite to the sidewalk surface
11:30:23 <reldred> Not something I'd do but the JP+ folks are interested in it.
11:42:42 <_zephyris> Give vanilla infrastructure vanilla badges?
11:43:47 <_zephyris> (well, I mean give road tiles a basic badge like road)
11:48:08 <ahyangyi> I also wish vanilla tiles check a fixed set badges to decide whether to draw the foundation cliffs
11:48:09 <Rubidium> just give everything a badge badge :D
11:48:46 <reldred> participation award badge
11:48:56 <ahyangyi> Right now it's impossible to make a sloped terrain as a custom object, and ensure nearby vanilla tiles still draw foundation cliff correctly
11:49:28 <ahyangyi> Because the vanilla tiles cannot tell between slopes and foundations
11:51:36 <ahyangyi> This is what I was talking about: the vanilla rail fails to draw the NW cliff because it does not know the custom object there is a slope.
11:52:23 <ahyangyi> Something that I thought was unfixable before, but with badges I realize it might be fixable now 🙂
11:53:10 <reldred> Im honestly not much a fan of those backfaces anyway, they've given me some grief building certain things before.
11:53:46 <ahyangyi> I am neither, but as long as they exist, it helps if they behave consistently.
11:54:56 <ahyangyi> Anyways, perhaps it could be either a badge or a new prop, whatever. I think I should file an issue so the dicussion is not lost.
11:55:29 <ahyangyi> Uh, not sure if it's a bug or a feature request though. But feature requests do not go to Github issues, so bug it is then.
11:57:00 <kaibaneddy> maybe it should just draw the foundation on the neighbouring tiles if the object/house/teva has the "do not draw foundations" flag set? 🤷♀️
11:57:35 <ahyangyi> Probably, hence I realize I should open an issue.
12:07:13 <ahyangyi> kaibaneddy: Would you mind if I copy your comment to the issue? I think it's probably the right fix.
12:21:42 <andythenorth[d]> _zephyris: this
12:34:57 <Rubidium> yes, how dare we release 15.1 without fixing #15166 :D
12:41:18 <peter1138> > Obvious limitation is that there (likely) is some NewGRF that depends on this bug existing.
12:43:06 * Rubidium wonders what original features were removed in 15.1.
12:43:25 <LordAro> well there is a notable "feature" that was removed...
12:43:44 <Rubidium> LordAro: RCE you mean?
13:36:50 <talltyler> "My suggestions for 10,000 vehicles, merging all the climates into one, and improving rivers in my favorite baseset of zBase haven't been implemented, 2/10 stars"
13:47:12 <_zephyris> talltyler: Ah, I see. I was going to dig through github history, but couldn't be bothered in the end...
13:47:37 <talltyler> They're all in Discussions 🙂
13:49:53 <_zephyris> To be fair, a higher single player RV limit probably makes sense. Let people build RVs till their CPU melts.
13:52:12 <talltyler> Sure, but I have zero patience for someone who makes unsolicited suggestions, and then two weeks later, whines that they haven't been implemented.
13:52:45 <talltyler> Even if we were a commercial game, nobody is entitled to anything beyond what is advertised at the time of purchase.
13:53:03 <_glx_> they're free to implement it themselves
13:53:03 <talltyler> (Some of my frustration might be bleeding over from similar issues with my day job, apologies 🙂 )
13:56:12 <_zephyris> Oh totally, I was very close to just closing the discussion. Tried to be restrained and educate on how to give feedback instead.
14:23:42 *** gelignite is now known as Guest286
14:23:46 *** gelignite has joined #openttd
14:30:11 *** Guest286 has quit IRC (Ping timeout: 480 seconds)
15:11:42 <andythenorth[d]> I was just going to close #15167
15:17:59 *** MinchinWeb[m] has quit IRC (Read error: Connection reset by peer)
15:18:12 *** MinchinWeb[m] has joined #openttd
15:25:36 <peter1138> It's your own wagons that in the wrong place.
15:26:05 <peter1138> But in general NewGRF vehicle sets like to use different offsets from vanilla.
15:27:49 <peter1138> I suppose OpenGFX 2 could use the y adjustment that NewGRFs use.
15:29:52 <kaibaneddy> I dunno, the loco looks very low relative to the tracks...
15:31:37 <ahyangyi> Hmm, disucssion is also forwarded here.
15:33:40 <pickpacket> peter1138: my own wagons are maybe a little to high on the y, but the floss looks too low imho
15:37:12 <peter1138> It's not the floss.
15:37:23 <peter1138> It's just how original TTD placed trains.
15:38:00 <peter1138> i.e. put other engines or wagons in your train, and it's your wagons that are mismatched.
15:42:29 <peter1138> > It's just how original TTD placed trains.
15:44:21 *** Wormnest has joined #openttd
15:48:44 <pickpacket> well, is it more important to look like TTD or to look *right*?
15:49:08 <andythenorth[d]> let me know when we've decided 😛
15:49:21 <andythenorth[d]> Horse has been changed maybe 3 times, due to feedback 😛
15:50:44 *** MinchinWeb[m] has quit IRC (Ping timeout: 480 seconds)
15:50:50 *** MinchinWeb[m] has joined #openttd
16:21:29 <ahyangyi> pickpacket: At least, when using TTD graphics, it is important to look like TTD I guess
16:34:31 <_zephyris> That's not the problem. Base set graphics have to use the same sprites/offset for the purchase menu and the horizontal track view. So it's not that it's trying to look like TTD, it's to make the purchase menu look correct.
16:35:39 <peter1138> It could use the vertical pitch thing that NewGRFs use, I guess?
16:36:26 <peter1138> > Y-Offset for train sprites
16:37:01 <peter1138> Hmm, or not, setting that in the additional NewGRF won't affect the main data, will it.
16:37:32 <_zephyris> Presumably not, I think I tested it at one point.
16:39:21 <peter1138> Put a hack in so that if it's the system baseset NewGRF it applies to baseset trains? Hmm.
17:38:12 <talltyler> Now that 15.1 is out, let's get people testing this 🙂
17:38:36 <talltyler> mmtunligit: Now you can PR randomization 😄
17:39:44 <mmtunligit> gotta do final cleanup on the house stuff before I PR it
17:40:44 <talltyler> Yeah, no rush, of course. But now you're not blocked by waiting on this PR to be merged 🙂
17:40:52 <mmtunligit> Then I want to PR a generalized endian buffer pair for vectors so I can just not have to think about the weird ass errors I was getting with roadstops
17:41:09 *** Smedles has joined #openttd
19:00:08 <_zephyris> How's the controller PR looking? That'd be another good candidate to merge and get tested
19:15:04 <rito12_51026> _zephyris: Needs rebase
19:17:51 <_zephyris> I'd forgotten that I'd already left an encouraging message
20:12:36 <_zephyris> Is this just a documentation problem?
20:13:57 <ahyangyi> It does say "For other 80+x variables confer the TTD town structure", though the link destination looks a bit ancient
20:14:30 <andythenorth[d]> frosch usually warns against 80+ vars as a general principle
20:15:56 <ahyangyi> Also the link contains an invalid anchor `#_TownArray`, probably referring something that existed before the last update of the page
20:16:14 <ahyangyi> *Last changed on 2006-02-27*
20:16:36 <ahyangyi> Apprently the 2006 doc is too new for our purpose
20:37:37 <andythenorth[d]> some of town array was deleted
20:37:51 <andythenorth[d]> there was a partially implemented proposal to provide newgrf town control
20:37:57 <andythenorth[d]> which was eliminated in favour of GS
20:38:12 <andythenorth[d]> (iirc - memory is fallible)
20:51:19 <_zephyris> andythenorth[d]: I'll do some blame, I can kinda see that now you mention it
20:52:27 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
20:52:28 *** WormnestAndroid has joined #openttd
20:58:23 <rito12_51026> Why do I need graphics set to start dedicated server?
21:04:38 <Rubidium> it used to be map generation
21:05:07 <Rubidium> but also for things like screenshots
21:05:55 <Rubidium> ofcourse you can add a lot of special casing for it, but that's likely a more brittle solution than requiring a graphics set
21:06:38 <rito12_51026> I can take screenshots on headless server?
21:07:40 <andythenorth[d]> we have to get rid of Toyland
21:07:42 <Rubidium> yup, just type screenshot in the OS console, or use screenshot with rcon
21:08:11 <andythenorth[d]> a forum / discord joke between about 20 people is ruining the game for other players
21:08:25 <andythenorth[d]> one of the arts of democracy is knowing when to not go with the vote
21:08:32 <rito12_51026> Oh no it failed 🙁
21:14:47 <peter1138> andythenorth[d], kindly, fuck right off.
21:16:37 <andythenorth[d]> I just think it's a pointless own goal
21:17:24 <peter1138> I get it, you also don't like the new title game, and are using confused player reviews as an excuse to complain about it.
21:17:35 <andythenorth[d]> the title game doesn't affect me
21:18:01 <peter1138> It clearly does, you've moaned about it several times.
21:18:03 <andythenorth[d]> for reasons I didn't even explore, my self-compiled OpenTTD doesn't use it
21:18:28 <ahyangyi> Andy is probably just sensitive about this kind of negative feedback
21:18:34 <peter1138> If you compile 15.0 or 15.1, it will.
21:18:45 <andythenorth[d]> I'm sensitive about how much effort goes into OpenTTD
21:18:53 <andythenorth[d]> and how much 15.0 advanced the game
21:19:07 <andythenorth[d]> and how much needless driving away of players a discord joke causes
21:19:42 <andythenorth[d]> that most of the people complaining can barely string 4 coherent words together and ARE JUST ANGRY TYPING doesn't make them wrong
21:20:15 <peter1138> It wasn't a joke. _zephyris has done a good job on Toyland.
21:20:49 <peter1138> And you, of course, know that when 16.0 comes around, it will have its own title game.
21:21:11 <andythenorth[d]> it's not a written rule though, 15.2 could have its own title game
21:21:40 <ahyangyi> 16.0 doesn't have to be released near the end of 2026 though.
21:22:47 <belajalilija> release 16,0 on 1st of april
21:22:51 <belajalilija> with a new toyland map
21:22:51 <_zephyris> New title game game competition can start whenever
21:23:03 <andythenorth[d]> if Toyland wins again...then I will see the joke
21:23:23 <andythenorth[d]> but please, not with that sound effect featured
21:23:34 <_zephyris> "Join the competition" is a very positive thing to respond with when people are annoyed
21:25:28 <andythenorth[d]> "have another vote"
21:27:36 <ahyangyi> belajalilija: Speaking of which, what about making the toyland title game a permanent and undemocratic feature
21:27:42 <ahyangyi> But only for April the first
21:28:32 <belajalilija> invert that, have normal maps only appear on leap days
21:29:02 <belajalilija> and only if the moon, mars and the sun align correctly
21:29:13 <rito12_51026> Why not to change title game depending on the landscape chosen in new game window?
21:29:25 <andythenorth[d]> because that would require a PR 🙂
21:30:07 <locosage> pick title game depending on outdoor temperature
21:30:19 <rito12_51026> Then we could have 4 polls for each release
21:30:24 <_zephyris> pick title game based on blood sugar level
21:30:37 <belajalilija> _zephyris: actuallys thats a good idea
21:30:37 <_zephyris> loading a new game when UI option changes doesn't seem viable
21:30:41 <belajalilija> would encourage healthy eating
21:31:53 <_zephyris> Lol, me and peter both replied at about the same time
21:32:09 <locosage> _zephyris: when you see toyland it's time to go to a doctor? 🤭
21:34:41 <peter1138> Well, 8 minutes apart.
21:36:12 <andythenorth[d]> all black title game
21:36:13 <ahyangyi> peter1138: Seems I haven't looked at Toyland since this patch
21:36:13 <ahyangyi> It looks so much better now 🙂
21:36:53 <peter1138> Yeah, that change was probably the one that prompted the "joke" to use a toyland title game.
21:39:38 <belajalilija> here's some title game proposals
21:39:59 <belajalilija> to really bring the game back to its roots
21:43:22 <_zephyris> Wow, that isometric on isometric really messes with my head!
21:43:25 <belajalilija> here's some for all those toyland haters
21:43:54 <andythenorth[d]> the cubes one is phenomenal
21:43:55 <belajalilija> _zephyris: it does mine too, feels like when people post weird compressed screenshots
21:44:51 <_zephyris> My brain is really noping out on that candyfloss forest
21:45:35 <belajalilija> its not better like this either
21:46:39 <belajalilija> this one feels almost like a heightmap
21:54:05 *** Zathras_4 has joined #openttd
21:57:20 <ahyangyi> The cube feels like this
21:57:35 <ahyangyi> industries and forests pegged on the board
22:01:02 <belajalilija> weird seeing stuff like this
22:07:48 <LordAro> no particular evidence that steam post is complaining about toyland. could just be ogfx2 ;)
22:08:37 <andythenorth[d]> I made sweeping assumptions
22:13:57 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:16:04 <peter1138> If it's not that, then... the request is bizarre because the option to switch graphics set has been in the game for a very long time.
22:17:48 <andythenorth[d]> many players may not understand what a base set is?
22:18:00 <andythenorth[d]> we changed quite a lot at once - base set, main menu
22:18:13 <andythenorth[d]> was the graphics options window rework 15?
22:18:18 <andythenorth[d]> or was that in 14 already?
22:18:40 <andythenorth[d]> change is good, just freaks some people out
22:50:53 <belajalilija> im sure that will solve it
22:51:47 <andythenorth[d]> change is good, just freaks some people out
22:51:53 <andythenorth[d]> was there an echo?
22:54:17 <belajalilija> people seem more freaked out this time than last
22:54:29 <belajalilija> but there's also just a lot more people now
22:55:45 <locosage> well, invite system creates more issues than solves imo
23:02:55 <peter1138> They thing developers read tt-forums, cute.
23:43:00 <_zephyris> Hmm, no debug button for towns?
23:46:07 *** jfkuayue has joined #openttd
23:46:07 <jfkuayue> is "one company for all players open access" approach disabled in vanilla 15?
continue to next day ⏵
