IRC logs for #openttd on OFTC at 2024-07-23
⏴ go to previous day
01:52:00 *** Wormnest has quit IRC (Quit: Leaving)
02:58:00 *** debdog has quit IRC (Ping timeout: 480 seconds)
04:42:12 <DorpsGek> - Update: Translations from eints (by translators)
05:07:01 *** D-HUND is now known as debdog
06:32:43 <peter1139> Hmm, some NewGRFs without a cargo translation table manage to work fine with FIRS 5.
06:34:23 <peter1139> If they only use the cargo class properties, and no cargp-specific action 3, then there is no issue.
06:51:19 <johnfranklin> lol, I didn't wake up until it was morning in Europe
06:51:59 <peter1139> I still think that installing a default cargo table (as per 12646) is the best solution. Anything expecting the old behaviour isn't likely to be working properly anyaywa.
06:56:52 <andythenorth> Does 12646 have any downsides?
06:57:10 <andythenorth> Retconning a cargo table seems most appropriate to me
06:59:06 <peter1139> I have a second patch that allows installing a cargo table to the overridden NewGRF.
06:59:21 <peter1139> Which will allow any potential issue to be "solved".
07:22:21 *** reldred has joined #openttd
07:22:21 <reldred> winner winner chicken dinner it sounds to me
08:06:12 <peter1139> Oh actually, the patch also allows installing other translation tables, although that's not all that useful by itself.
08:26:31 *** Exec has quit IRC (Quit: File:///)
08:26:47 *** mindlesstux has joined #openttd
08:32:50 *** mindlesstux has joined #openttd
08:41:38 <gwyd4016> Is there a reason that when using double headed engines with lengths shorter than 8/8, the alignments stop working properly?
08:48:34 <peter1139> Probably because nobody reported an issue with it.
08:51:58 <peter1139> Multihead parts are reversed a different way to "depot flipping". Possibly that could be simplified.
08:56:18 <peter1139> No doubt somebody has seen the issue and then decided it was easier to do something else than find out if it's meant to be that way.
08:56:45 <peter1139> It's not like we need to care too much about TTDPatch compatibility at this point.
08:58:51 <reldred> Yeah I think those days are well and truly done
08:59:21 <reldred> I'm sure someone will cry and scream over it, but even as one of the old holdouts I've accepted our new OTTD overlords.
08:59:50 <peter1139> Multihead trailing engines use a different sprite number than leading engines, and that is used to select the rear sprite.
09:00:25 <gwyd4016> peter1139: Yeah I did that
09:01:27 <peter1139> When its a custom sprite, that causes it to adjust the sprite direction. But because it doesn't have the reverse direction flag, the offset-adjusting code is not applied.
09:06:14 <peter1139> Haha, the tailing engine logic is applied to road vehicles too.
09:06:53 <peter1139> I guess a NewGRF can specify spritenum 0xFE manually.
09:09:23 <gwyd4016> I think fixing this would probably break more sets that it would fix
09:09:49 <peter1139> How many sets use multihead with shorter engines?
09:10:10 <reldred> The JP+ folks probably do,
09:10:41 <reldred> otherwise Andy might be the best one to talk to, though I recall something he said once about sticking to 8/8 for some things
09:11:32 <peter1139> Andy doesn't count. :)
09:13:36 <reldred> yeah he just cheats and automates all his workarounds with python anyway
09:19:01 <peter1139> What's the workaround?
09:19:22 <peter1139> Is it feasible that nothing uses multihead and shorter than 8/8 engines because of the offset issue?
09:20:16 <peter1139> (I loaded some Japan stuff and the only multiheads I can find are 8/8)
09:21:45 <reldred> dunno, could cast the summoning ritual and see what they know
09:23:03 <reldred> Come to think of it, i think I ran into some issues with some of the multiheaded maglevs in the opengfx mars train set, that's got some shorter than 8/8 multiheads in it, I got frustrated and was just gonna skip those locos.
09:24:05 <peter1139> Don't think that's on Bananas.
09:24:20 <reldred> I think it is, I'd have to check
09:24:34 <emperorjake> Some of the earliest JP+ MUs have 7/8 dualheaded
09:26:25 <peter1139> What model is that (and what NewGRF?)
09:26:26 <emperorjake> Come to think of it, Vactrain Set has an even shorter dualheaded vehicle
09:28:00 <emperorjake> peter1139: JP+ Multiple Units, 6340 series of 1914
09:29:02 <emperorjake> 2CC TrainsInNML has heaps of shorter dual headed MUs as well
09:29:03 <peter1139> So yes, making it work automatically is not possible without breaking these sets.
09:29:11 <peter1139> Unless some extra flag is used.
09:30:00 <reldred> are those sets doing anything to fix it though? or is it a widespread pain-in-ass that folks have just been putting up with?
09:30:53 <peter1139> Yes, they will be doing the same trick that used to be needed for depot flipping shorter vehicles.
09:31:48 <emperorjake> It's as simple as having a separate template for flipped vehicles
09:35:54 <peter1139> gwyd4016, so there you go. The reason is that NewGRFs expect alignments to be wrong, so it can't be fixed.
09:37:05 <peter1139> If the NewGRFs tested for spritenum == 0xFE then I could have a solution, but they don't.
09:38:09 <emperorjake> Dualheaded engines also have that weirdness with the purchase sprite, requiring a special template to counteract the offset
09:38:56 <reldred> sounds like a lot of fucking around that could be avoided for future generations
09:39:00 <reldred> someone think of the children
09:39:01 <peter1139> If you think "the engine is multihead, lets draw both parts automatically" is weird, sure.
09:40:25 <peter1139> I'll agree it's inconvenient for some sets due to their interpretation of what the purchase list should look like. But it's not weird per-se.
09:42:42 <LordAro> my computer has crashed twice this morning
09:49:06 <peter1139> Today's task is reimplementing code that we removed because we didn't need it...
09:51:30 <peter1139> Other things have been redesigned since which means a straight revert isn't going to cut it.
10:36:26 <andythenorth> gwyd4016: Sometimes Horse code has notes, I’ll look
10:38:12 <andythenorth> I suspect all Horse dual-head uses 8/8
10:55:01 <gwyd4016> peter1139: Yeah I've used a test for "vehicle_is_flipped" and I assume that this has implications beyond being flipped for dual headedness
10:55:24 <gwyd4016> And then just used a sprite with a different alignment based on whether or not the vehicle is flipped
11:04:26 <peter1139> I'm not sure "vehicle is flipped" works for trailing multihead parts.
11:04:36 <peter1139> (I mean, I am sure it does not.
11:05:15 <peter1139> If that flag was used, then the offsets would working already.
11:06:07 <peter1139> Hmm, unless "vehicle_is_flipped" actually tests property 0x80 + 0x48 being 0xFE.
11:15:53 <peter1139> I should check things before thinking I know how it works.
11:18:00 <andythenorth> More fun to discover by trial and error 😛
11:18:09 <peter1139> It does not test for 0xFE though, it sets that bit 1 is set.
11:20:40 <peter1139> Oh dear, neighbour having an argument over glass broken by a football...
11:21:04 <peter1139> The football I chucked over the fence yesterday when I found it on my lawn.
11:21:14 <peter1139> Hope it wasn't me :p
11:21:29 <yiffgirl> i'm going away on holiday for a couple of weeks, so i might not be able to respond to reviews of my prs.
12:21:40 *** HerzogDeXtEr has joined #openttd
13:01:34 *** asymptotically2 has quit IRC (Remote host closed the connection)
13:09:27 *** asymptotically2 has joined #openttd
13:20:28 <peter1139> Crap, my rechargable battery is not recharged :(
13:42:07 <peter1139> Hmm, should I get a Garmin watch...
13:47:06 <audigex> gwyd4016: ... I thought I was just going insane, gave up, and re-drew it 8/8
13:47:06 <audigex> Thankyou for restoring my faith in my sanity
13:50:44 <gwyd4016> ```template template_std_6(x, y) {
13:50:44 <gwyd4016> [x, y, 8, 29, -3, -20]
13:50:44 <gwyd4016> [x+13, y, 26, 24, -12, -17]
13:50:44 <gwyd4016> [x+44, y, 33, 16, -8, -12]
13:50:44 <gwyd4016> [x+83, y, 26, 24, 0, -15]
13:50:45 <gwyd4016> [x+114, y, 8, 29, -3, -16]
13:50:45 <gwyd4016> [x+127, y, 26, 24, -16, -15]
13:50:47 <gwyd4016> [x+158, y, 33, 16, -16, -12]
13:50:47 <gwyd4016> [x+197, y, 26, 24, -4, -17]
13:50:49 <gwyd4016> template template_std_6_flipped_align(x, y) {
13:50:51 <gwyd4016> [x, y, 8, 29, -3, -16]
13:50:51 <gwyd4016> [x+13, y, 26, 24, -16, -15]
13:50:53 <gwyd4016> [x+44, y, 33, 16, -16, -12]
13:50:53 <gwyd4016> [x+83, y, 26, 24, -4, -17]
13:50:55 <gwyd4016> [x+114, y, 8, 29, -3, -20]
13:50:55 <gwyd4016> [x+127, y, 26, 24, -12, -17]
13:50:57 <gwyd4016> [x+158, y, 33, 16, -8, -12]
13:50:57 <gwyd4016> [x+197, y, 26, 24, 0, -15]
13:51:01 <gwyd4016> I just swapped the offsets for view 1-4 with offsets for view 5-8
13:54:08 <peter1139> grf v846693 will change the offset to point to the centre of the vehicle, but also have separate bogies that follow the track.
13:56:28 <peter1139> andythenorth will use LLM tools to create 360 degree rotations of engines in Iron Horse 73
14:07:13 <gwyd4016> (Iron Ibex will still not be finished)
14:26:17 *** Flygon_ has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
14:40:04 <andythenorth> IronIbexGPT will be v0.8
14:40:33 <andythenorth> peter1139: You mean like “3d models in meshes”?
15:06:19 *** HerzogDeXtEr1 has joined #openttd
15:09:55 <ahyangyi> or in gaussian splatting, or in radiance field, or in SDF
15:10:12 <ahyangyi> meshes are just one of the many possibilities
15:12:34 *** HerzogDeXtEr has quit IRC (Ping timeout: 480 seconds)
15:27:03 *** HerzogDeXtEr has joined #openttd
15:31:53 <peter1139> 3D models automatically aliased to look like 8bpp 1x zoom sprites, eh?
15:33:49 *** HerzogDeXtEr1 has quit IRC (Ping timeout: 480 seconds)
15:43:48 <ahyangyi> With some magical renderers/shaders, perhaps
15:43:52 <goddess_ishtar> by that time all graphics will be obsolete
15:44:14 <goddess_ishtar> OpenTTD will beam the gamestate directly into your cybernetic implant
15:49:13 *** gelignite has joined #openttd
16:03:27 *** Wormnest has joined #openttd
17:21:02 *** gelignite has quit IRC (Quit: Stay safe!)
17:56:09 *** gelignite has joined #openttd
20:00:38 *** gelignite has quit IRC (Quit: Stay safe!)
20:35:49 *** nielsm has quit IRC (Remote host closed the connection)
21:07:40 <peter1139> Do we have anything similar to TileIndexDiffCByDiagDir but for sub-tile positions?
21:09:54 *** keikoz has quit IRC (Ping timeout: 480 seconds)
21:44:22 *** ChanServ sets mode: +v tokai
21:51:13 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
22:46:35 <wensimehrp> Is it fine to add a station variable 6C for station ID detection that omits nearby tile grf id restrictions?
22:47:59 <wensimehrp> This could be useful for intergrf communication I guess
continue to next day ⏵
