IRC logs for #openttd on OFTC at 2022-12-14
⏴ go to previous day
00:06:32 *** gelignite has quit IRC (Quit: Stay safe!)
00:13:22 <glx[d]> could be enough indeed
00:19:29 <petern> But is that the best order?
00:20:34 <Pruple> roadtypes are likely to be more varied than either rail or tramtypes, imo.
00:20:52 <petern> glx[d]: Glad I mentioned it (kinda) in my comment 🙂
00:22:01 <glx[d]> that's why reviews exist 🙂
00:22:46 <glx[d]> but I should have tested more too
00:22:48 <petern> I'm impressed with myself, I only looked at the code, I didn't test it.
00:23:34 <petern> That also means it's understandable code too.
00:24:40 <glx[d]> I guess most existing NRT newgrf redefine tunnel entrance via actionA to make it wide enough to not glitch
00:25:21 <petern> Pruple: Purple is here deciding the spec, I like 🙂
00:26:05 <glx[d]> it seems everywhere road takes precedence over tram
00:27:52 <glx[d]> except for overlays, both are drawn with tram over road
00:29:49 <Pruple> it makes sense to me for road to always have priority, since generally there'll be a lot more different roadtypes, changing over years (dirt, gravel, paved, fancy, different markings over the years), each of which could conceivably have a different tunnel or boomgate style.
00:29:50 <petern> Only allow one NewGRF to touch road/tram/rail types, then there's no conflict.
00:30:21 <Pruple> but there's only so many different ways to draw tram tracks; one way, I'd say.
00:30:30 <Eddi|zuHause> only allow one newgrf active simultaneously. should solve all conflicts
00:31:30 <Pruple> it's not really a conflict either way. you might get clashing graphical styles but you do that anyway if you load multiple train sets etc.
00:35:47 <Pruple> should I do an issue? 😛
00:37:07 <reldred> I dunno, I can imagine a tram tunnel perhaps wanting to be wider than a road tunnel, perhaps even slightly taller. And if I think of something like Wally’s suspended monorail that will really want the tram tunnel to override a road tunnel.
00:37:27 <reldred> That said, I’m not developing a road set, pruple is 😛
00:37:47 <reldred> I’m willing to take a fat L on this one 😛
00:39:52 <Pruple> tram priority for tunnels, road priority for level crossings? 🙂
00:39:52 <Pruple> suspended monorail is a feature-abuse like pipelines or conveyor belts and should not be considered 😉
00:41:22 <reldred> Yeah I’d agree with that, the road graphics would be the more significant in a road+tram rail crossing.
00:42:47 <reldred> And I wouldn’t be inclined to go out of the way to block any particular feature, though I do think better methods to achieve those need to be explored (much like the eye candy tram overlays)
00:44:42 <Pruple> I'm not saying block it, just saying give most consideration to what the feature is actually intended to be used for.
02:52:40 *** Wormnest has quit IRC (Quit: Leaving)
03:55:56 *** debdog has quit IRC (Ping timeout: 480 seconds)
06:42:26 *** keikoz has quit IRC (Ping timeout: 480 seconds)
07:10:03 *** sla_ro|master has joined #openttd
09:18:07 <petern> Why did you open a discussion? o_O
09:31:27 *** gelignite has joined #openttd
09:47:04 *** D-HUND is now known as debdog
10:08:47 <Eddi|zuHause> my package from china arrived about 10 days earlier than they said it would
10:10:03 <Eddi|zuHause> now i need to figure out (again) how to open this monitor
10:25:24 <andythenorth> hold the box up 1 m in the air, and cut the tape at the bottom
10:25:28 <andythenorth> I bought a monitor
10:34:13 <petern> My heating has lost it.
10:36:34 <andythenorth> was it Yodel's fault?
10:44:08 <petern> I always love it when the docs for a project tell you how to build it, but not what it does or how to set it up.
12:01:36 <petern> Hmm, should I try Docker containers for deployment?
12:02:21 <petern> Seems a bit wasteful for each deployment to have its own dotnet runtime etc.
12:10:44 <andythenorth> /me watching Big Clive
12:10:57 <andythenorth> rona -> bed -> youtube
12:34:56 <petern> These trousers have shrunk, they are not longer enough :/
12:39:49 <andythenorth> have your legs adjusted to suit?
12:42:51 <Pruple> it's not the whole suit, just the trousers
14:57:39 *** ag has quit IRC (Quit: User went offline on Discord a while ago)
15:35:20 <Eddi|zuHause> well... package didn't contain a bobcat.
15:48:55 *** matematysek has joined #openttd
15:48:55 <matematysek> Eddi|zuHause: bobcat? you mean animal or construction machine?
17:13:11 *** supermop_toil has quit IRC (Remote host closed the connection)
17:13:21 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
17:13:29 <pickpacket> I was hoping all the specs for the base industries would be available as nml files in one of the repos. Are they?
17:15:16 <petern> I messed up my pipeline. I forgot about the .scss files that need to be compiled. I don't know how 😦
17:20:53 <pickpacket> petern: do you know where I can find the specs for the base industries?
17:21:51 *** birdjj has quit IRC (Ping timeout: 480 seconds)
17:22:51 <LordAro> though i'm not actually sure where those specs are... i'd expect them to be in the baseset grf... which would mean they should be in ogfx somewhere
17:23:31 <LordAro> but perhaps the actual logic is within the game itself
17:23:32 <pickpacket> LordAro: well, in any format would be better than not at all :D
17:23:36 <Pruple> the baseset grf is just realsprites
17:23:53 <pickpacket> yeah, I've already checked there
17:25:34 <LordAro> in which case i am also curious about where they're defined
17:30:59 <LordAro> that is some horrifying code
17:46:30 <glx[d]> imagine without the macro
17:50:50 <andythenorth> it's kind of beautiful
18:02:33 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
18:03:20 *** WormnestAndroid has joined #openttd
18:05:50 *** HerzogDeXtEr has joined #openttd
18:45:31 <DorpsGek> - Update: Translations from eints (by translators)
19:11:24 <andythenorth> I've been in bed for 2 days
19:11:35 <andythenorth> is it PR approval and merging time yet?
19:12:05 <andythenorth> think of the bandwidth reduction from cutting out all those Horse action 1s
19:13:09 <petern> That would require my change to work properly first 🙂
19:16:12 <andythenorth> well, Horse has at least 10 users, so it would be as much as 180 MB saved
19:16:23 <andythenorth> the reductions in the AWS bill will be huge
19:22:55 <pickpacket> the AWS bill? What's running there?
19:33:06 <pickpacket> I still don't get it. Are build agents running there?
19:34:13 <andythenorth> binary distribution, website, in-game content, multiplayer discovery service
19:34:27 <andythenorth> builds are github actions far as I understand
19:34:33 <pickpacket> oh shit. That is a lot
19:34:40 <pickpacket> who foots the bill and how big is it?
19:35:27 <andythenorth> (1) historical community donations built up an operating fund, also there were some AWS open source credits (2) no idea
19:36:19 <petern> > Sorry we had to remove zBase/aBase due to costs
19:37:13 * pickpacket doesn't know what zBase/aBase is, nor what Horse refers to
19:37:52 <petern> They're the hideous 32bpp extra zoom basesets that new players seem to love.
19:40:55 <pickpacket> kill them. Kill them with fire!
19:44:39 <dP> btw, newgrf format could use some better image compression
19:45:41 <dP> jpeg is lossy and I'm not sure about licensing
19:45:59 <petern> Lots of tiny little high-entropy sprites are not great for compression.
19:46:05 <LordAro> is webp noticably different to png?
19:46:05 <dP> though there seems to be something called jpeg xl nowadays
19:47:18 <dP> LordAro: seems to be slightly better
19:47:22 <petern> The one that Chrome has removed?
19:47:35 <dP> but I guess it needs testing on real sprites
19:49:37 <petern> Might help for 32bpp EZ but not much on standard sprites.
19:53:39 <andythenorth> petern: I thought that PNG would have compressed as much it can, but 36 MB of Horse grf compresses to 4 MB of zip
19:56:45 <LordAro> what are we using libpng for then?
19:56:46 <andythenorth> I don't know much about compression formats, but .tar seems to be just an archive wrapper, compression is minimal
19:56:59 <LordAro> that is exactly what a tape-archive is :)
19:57:00 <andythenorth> (Horse uses .tar which is 36 MB)
19:57:15 <Eddi|zuHause> there's lots of redundance in the pseudosprites, which aren't compressed at all
19:58:12 <Eddi|zuHause> and .tar isn't compression either, it's just "put these files together"
19:58:22 <andythenorth> so I have learnt 🙂
19:58:33 <petern> PNG is your source files 🙂
19:58:52 <Eddi|zuHause> which is why you usually combine tar with a compressor that takes a single file, like gz, xz, bzip2, ...
19:59:03 <andythenorth> ok so 36 MB -> 4.4 MB is 87%? Remind me how to do basic school level maths
19:59:25 <DorpsGek> LordAro: 0.12222222222222223
19:59:26 <petern> So we already do .tar, can we do compression on top?
19:59:31 <petern> At least for transfer.
20:00:39 <pickpacket> petern: how are they transferred? Are we talking about when clients download them? I’m guessing the game uses libcurl?
20:01:33 <pickpacket> a web server can easily be configured to compress before transmit so that side is easy at least
20:02:11 <petern> These could be compressed as rest to be honest.
20:02:38 <LordAro> pickpacket: raw TCP, generally
20:02:41 <andythenorth> if I could upload as zip, they would go up the wire faster also 😛
20:02:55 <LordAro> i'm not sure if downloads are compressed, i think they are...
20:02:59 <dP> btw, I feel like openttd is somehow downloading stuff much slower than it could
20:03:48 <andythenorth> hmm the zip I have doesn't work with bananas
20:04:01 <andythenorth> it uncompresses the individual files, but reports they're incompatible
20:04:32 *** garlic_bread42 has quit IRC (Quit: User went offline on Discord a while ago)
20:05:30 <andythenorth> I would have been super shocked if TB hadn't looked at compression stuff 😛
20:06:32 * andythenorth wonders what the best image input format for grfs would be
20:06:39 <andythenorth> the one that compiles fastest 😛
20:09:08 <petern> The one with half the number of sprites due to not duplicating act1s?
20:09:33 <andythenorth> I wonder if one giant spritesheet would be better than many
20:10:13 <andythenorth> the patch I wrote is just leaving a hand grenade behind
20:10:41 <andythenorth> actually the whole automated palette guessing shenanigans in NMLC are the hand grenade
20:10:46 <andythenorth> I just pulled the pin
20:11:01 <LordAro> there's a lot of changes in there that don't seem relevant to the diff
20:11:12 <LordAro> is your black out of date, perhaps?
20:11:47 <andythenorth> I never quite figured that out
20:12:05 <andythenorth> that xkcd about time saved suddenly applied 😛
20:12:54 <LordAro> only if you only apply the calculation to yourself :p
20:13:05 <Eddi|zuHause> src/fileio.cpp: Debug(misc, 1, "Scanning for tars");
20:13:07 <Eddi|zuHause> src/fileio.cpp: Debug(misc, 3, "{} found as config directory", config_dir);
20:13:07 <LordAro> definitely not too new, mine shows no changes
20:13:17 <Eddi|zuHause> the debuglevels on these two might be flipped
20:13:25 <Eddi|zuHause> it's awfully spammy on level 1
20:13:37 <LordAro> you know where the PR button is :p
20:14:19 <andythenorth> yeah running black locally over the nml repo produces a huge diff
20:15:35 <andythenorth> rona brain is not wanting to figure that out 😛
20:16:24 <andythenorth> ignore ^ just a rebase to master
20:16:33 <andythenorth> thought we'd renamed master to main, but eh, maybe I misread
21:00:10 <glx[d]> andythenorth: only some repos
21:09:17 *** Samu has quit IRC (Read error: Connection reset by peer)
21:27:28 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
21:27:44 *** WormnestAndroid has joined #openttd
21:48:59 <andythenorth> hmm more variants for Horse then?
21:51:32 *** keikoz has quit IRC (Ping timeout: 480 seconds)
22:02:53 <andythenorth> maybe not all of them
22:02:58 <andythenorth> "most of the variants"
22:06:06 <petern> Randomize the variant parents.
22:08:12 <andythenorth> at compile time? or run time? 😛
22:08:23 <andythenorth> make it a callback not a property 😛
22:41:05 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:45:42 *** nielsm has quit IRC (Ping timeout: 480 seconds)
23:11:52 *** Wormnest has joined #openttd
23:13:01 *** sla_ro|master has quit IRC ()
23:23:08 *** gelignite has quit IRC (Quit: Stay safe!)
23:38:48 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
23:40:38 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
23:40:38 *** WormnestAndroid has joined #openttd
23:42:34 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
23:57:01 <petern> Do I need 24 thin-client computers?
23:57:19 <petern> Set up some kind of cluster... ;p
continue to next day ⏵