IRC logs for #openttd on OFTC at 2020-02-18
⏴ go to previous day
00:01:30 *** Wormnest has joined #openttd
00:09:16 *** Groud[m] has joined #openttd
00:15:56 *** fiddeldibu[m] has joined #openttd
00:18:32 *** dude[m]1 has joined #openttd
00:23:41 *** glothit7ok[m] has joined #openttd
00:39:02 *** hylshols7qui[m] has joined #openttd
01:06:22 *** elliot[m] has joined #openttd
01:06:59 *** pina[m] has joined #openttd
01:07:53 *** tokai|noir has joined #openttd
01:07:53 *** ChanServ sets mode: +v tokai|noir
01:09:20 *** magdalena[m] has joined #openttd
01:10:55 *** yur3shmukcik[m] has joined #openttd
01:13:47 *** vanessa[m] has joined #openttd
01:13:52 *** ciet[m] has joined #openttd
01:15:05 *** paulus[m] has joined #openttd
01:16:46 *** natmac[m] has joined #openttd
01:17:53 *** iarp[m] has joined #openttd
01:29:48 *** buggeas40d[m] has joined #openttd
01:33:47 <Eddi|zuHause> weird, i could load the cities skylines save yesterday, but that exact same savegame now throws simulation errors on load
01:40:37 *** amal[m] has joined #openttd
01:40:58 *** fjodor[m] has joined #openttd
01:41:25 *** osvaldo[m] has joined #openttd
01:43:10 *** ircer[m] has joined #openttd
01:49:18 *** olivier[m] has joined #openttd
02:06:51 *** khavik[m] has joined #openttd
02:09:12 *** ist5shreawf[m] has joined #openttd
02:09:59 *** Corns[m] has joined #openttd
02:11:08 *** ad5twoknebor[m] has joined #openttd
02:11:13 *** twom[m] has joined #openttd
02:13:08 *** arron[m] has joined #openttd
02:14:51 *** pothyurf[m] has joined #openttd
02:15:43 *** Heiki[m] has joined #openttd
02:16:30 *** albert[m] has joined #openttd
02:18:56 *** cawal[m] has joined #openttd
02:21:03 *** robert[m]1 has joined #openttd
02:21:55 *** jact[m] has joined #openttd
02:26:32 *** natalie[m] has joined #openttd
02:27:03 *** nartir[m] has joined #openttd
02:33:54 *** lapav[m] has joined #openttd
02:34:00 *** jeeg[m] has joined #openttd
02:48:12 *** UncleCJ[m] has joined #openttd
02:53:59 *** patricia[m] has joined #openttd
03:02:44 *** labs[m] has joined #openttd
03:02:50 *** yoltid[m] has joined #openttd
03:06:24 *** nolep[m] has joined #openttd
03:07:31 *** karoline[m] has joined #openttd
04:03:28 *** Wormnest has joined #openttd
08:13:05 *** andythenorth has joined #openttd
08:21:04 <andythenorth> should I join reddit? :P
08:35:45 *** WormnestAndroid has quit IRC
08:36:01 *** WormnestAndroid has joined #openttd
09:15:30 <Eddi|zuHause> 5 years ago? maybe
09:16:21 <andythenorth> currently recommending disabling 90º turns to make pathfinding more reliable
09:16:27 <andythenorth> empirically, the opposite is true
09:17:29 <Eddi|zuHause> for ships it should probably be disabled
09:17:38 <andythenorth> no they get stuck
09:17:42 <andythenorth> unless we fixed that
09:17:45 <andythenorth> maybe we fixed that
09:18:30 <Eddi|zuHause> yeah, not awake yet, i meant the opposite
09:19:04 <andythenorth> with 90º disabled, occasionally a lost train will run into a terminus station that is too short for it
09:19:07 <andythenorth> then it can never leave
09:19:22 <andythenorth> with 90º enabled, it will turn on the crossing in front of the station
09:20:41 <Eddi|zuHause> but if 90° is enabled, it might prefer going over that crossing, instead of turning around on the platform
09:20:52 <Eddi|zuHause> which causes unnecessary traffic
09:20:59 <Eddi|zuHause> and potential deadlocks
09:21:26 *** andythenorth has joined #openttd
09:21:31 <Eddi|zuHause> which is, i believe, what the person meant
09:21:45 <andythenorth> swings, roundabouts
09:22:04 <Eddi|zuHause> roundabouts are terrible
09:22:15 *** adikt-- has joined #openttd
09:22:37 <Eddi|zuHause> no idea what you mean by that
09:23:28 <andythenorth> english colloquialism
09:24:50 <Eddi|zuHause> then i don't agree with you.
09:25:23 <andythenorth> you're favouring 90º disabled?
09:25:52 <Eddi|zuHause> typically, the game behaves better when there's less options
09:26:44 <andythenorth> so the conclusion is, I shouldn't join reddit
09:28:29 <Eddi|zuHause> the more options a train has to choose, the more likely it'll pick a disruptive choice, because something wasn't properly modeled as a pathfinder penalty
09:29:22 <Eddi|zuHause> like, if you have parallel tracks, don't put in X-crossings so "trains can bypass breakdowns"
09:30:28 <Eddi|zuHause> because a) the pathfinder cannot distinguish normal trains from broken down trains, and b) the pathfinder has no concept of blocking other incoming trains
09:38:05 *** andythenorth has joined #openttd
09:51:20 <DorpsGek_III> [OpenTTD/nml] nielsmh commented on issue #80: openttd_version: OpenTTD var 21/1A result has changed since OpenTTD moved to git https://git.io/JvchK
09:56:41 <DorpsGek_III> [OpenTTD/nml] andythenorth commented on issue #80: openttd_version: OpenTTD var 21/1A result has changed since OpenTTD moved to git https://git.io/JvchK
10:00:02 <DorpsGek_III> [OpenTTD/nml] andythenorth commented on issue #80: openttd_version: OpenTTD var 21/1A result has changed since OpenTTD moved to git https://git.io/JvchK
10:00:03 <DorpsGek_III> [OpenTTD/nml] andythenorth closed issue #80: openttd_version: OpenTTD var 21/1A result has changed since OpenTTD moved to git https://git.io/JvchK
10:15:17 <andythenorth> a [n% full load] order would ease ship design :P
10:15:42 <andythenorth> trying to provide a range of ideal capacities from 40t to 800t is an arse
10:15:52 <andythenorth> but % load has been rejected multiple times
10:25:28 *** ChanServ sets mode: +v tokai
10:26:01 *** andythenorth has joined #openttd
10:50:06 <Eddi|zuHause> i don't think %-load has been rejected per se, it's just embedded into the whole "order UI is horribly overloaded already" problem
11:16:57 *** tokai|noir has joined #openttd
11:16:57 *** ChanServ sets mode: +v tokai|noir
11:48:55 *** Wormnest has joined #openttd
11:57:27 *** mczapkie has joined #openttd
12:00:35 <mczapkie> Hello, I would like to notify, that hg@hg.openttdcoop.org doesn't respond...
12:05:24 *** Lejving_ has joined #openttd
12:23:53 *** andythenorth has joined #openttd
12:33:47 *** Smedles has joined #openttd
12:36:00 *** andythenorth has joined #openttd
13:34:29 *** Smedles has joined #openttd
13:48:17 *** ChanServ sets mode: +v tokai
14:55:59 <Samu> it adds a penalty when a ship is in the docking tile
14:56:55 <Samu> means that it's going to increase the number of node searchs needlessly
14:58:35 <Samu> the ship isn't even close to the destination when the pathfinder makes the search
14:59:07 <Samu> the penalty could be obsolete when the ship arrives closer
14:59:14 *** tokai|noir has joined #openttd
14:59:14 *** ChanServ sets mode: +v tokai|noir
15:00:35 <LordAro> Samu: i noticed that too - it definitely seemed to be doing the pathfinding more often than necessary
15:01:52 <Samu> i think the 90 degrees as penalties are also affecting it
15:03:30 <spnda> So uh, what is a spritelayout from NML in NFO? I can't find much if any reference to it... Also I saw that Airports use tilelayout. Do Objects, Industries, Houses or anything else use that?
15:07:06 *** andythenorth has joined #openttd
15:13:04 <Samu> 10222 (14119 with curve penalties) rounds from A to B, 14479 (22078 with curve penalties) rounds from B to A
15:13:14 <Samu> regardless, this ship would be lost
15:13:19 <Samu> with a limit of 10000 rounds
15:14:25 <spnda> andythenorth: yesterday you said "needs to follow industries and objects imho". Could you share a example of a multi tile/single tile object/industry in NML? I just want to look at something to get some ideas.
15:19:57 <Samu> when the ship pathfinds from B to the ship depot, it only needs 7856 rounds
15:20:37 <Samu> but when it pathfinds from B to A, skipping the ship depot, it needs more than 10000
15:22:56 <Samu> distance from depot to A is mere 5 tiles
15:23:22 <Samu> but the number of rounds searches goes up from 7856 to 22075, that's weird
15:24:46 <andythenorth> for single tile, general store, hardware store
15:24:57 <andythenorth> multi-tile: builders yard, recycling yard
15:24:59 <spnda> Oh I had no idea you were on github now
15:25:23 <andythenorth> you might just want the generated nml though
15:26:06 <andythenorth> the bigger FIRS industries sometimes have horrible complexity in the spritelayouts
15:26:20 <andythenorth> you might want to ignore those, lots of ternary checks for snow, land slope etc
15:26:28 <andythenorth> forest is a prime example of madness :P
15:27:03 *** D-HUND is now known as debdog
15:27:49 <spnda> You've got the generated NML anywhere?
15:28:02 <spnda> Oh I am blind. Found it
15:36:03 <Samu> there is a bug somewhere related to docking tiles
15:36:40 <Samu> if it is a dock it needs 22075 search nodes
15:36:57 <Samu> but if in place of a dock it is a ship depot, it only needs 8230 rounds
15:38:40 *** ChanServ sets mode: +v tokai
15:39:00 <Samu> gonna try with a buy instead of ship depot
15:39:12 *** supermop_work_ has joined #openttd
15:41:41 <Samu> i place a dock in front of the buoy and bam... now it requires 22075 nodes
15:42:08 <Samu> and the ship is still going to the buoy
15:42:38 <FLHerne> Samu: Are you putting the dock and buoy so they have exactly the same destination tile?
15:42:49 <spnda> andythenorth: Hmm, industries use industry tiles, objects don't... I kinda want to go the way objects are done, though. How do objects do it?
15:42:53 <FLHerne> (bearing in mind that docks now have more than one possibly)
15:44:13 <Samu> i know but, going from 8230 rounds to 22075 is a huge difference imo
15:44:26 <Samu> considering it's exactly the same destination
15:44:50 <spnda> FLHerne: Oh that looks quite helpful.
15:45:01 <spnda> Didn't know objects do it with a switch... interesting
15:45:46 <FLHerne> spnda: The switch is for the different orientations
15:46:44 <FLHerne> I don't know how you've implemented it with roadstops
15:47:17 <spnda> I haven't, that's the thing
15:47:35 <spnda> And I've only ever made trains and roadvehicles with NML, so I have no idea how others work
15:49:19 <FLHerne> I guess you want similar behaviour to rail stations, which don't actually exist in NML yet :/
15:51:25 <FLHerne> Hm, looks like that uses tilelayouts, which seems a bit pointless for road stops
15:51:46 <spnda> I thought of tilelayouts too
15:52:39 <FLHerne> They've only ever got one tile, unless you plan to change that?
15:53:03 <spnda> Well, it'd be an idea to maybe make multi tile stops
15:56:11 <spnda> But I am still not sure if that's actually a good idea or if it's necessary-
15:58:54 <planetmaker> right, ssh access to repos should work again
16:01:43 <FLHerne> spnda: I'm not convinced they're worth having, but could be wrong
16:02:06 <spnda> Not really, but it's been an idea that came up
16:14:38 <Eddi|zuHause> so i spend half an hour fiddling with CSUR, and immediately get to the limitations of it... i don't think this is the right thing for me...
16:19:08 <spnda> Actually, I think I'll just make this single tile, eventhough I am sure there'll be some developers who would make awesome multi tile stations.
16:21:13 *** supermop_work_ has quit IRC
16:31:49 <Samu> just being a docking tile it adds a 300 cost to the path
16:32:10 <Samu> enough to increase 8235 to 22075
16:32:57 <Eddi|zuHause> i'm definitely of the opinion that there should be multi-tile road stations, but they only make sense if there are also ... *cough* ... state machines
16:35:28 <spnda> Eddi|zuHause: what do you really mean with that?
16:36:23 <Eddi|zuHause> state machines are what guides airplanes on the airport (where can you land, where are the loading spots, how can you taxi, is the taxiway empty)
16:36:40 <Eddi|zuHause> they could also be used to guide road vehicles on the road stop
16:39:10 <planetmaker> is that really needed or can it not simply be like "this tile allows unloading in this direction and entry from that"
16:39:47 <planetmaker> which basically current ones do by orientation
16:40:07 <planetmaker> couldn't they be just strung together to bigger ones?
16:40:15 <spnda> Which was basically my idea. Leave their functionality as is, and just have custom sprites/sprite layouts/animations etc.
16:40:25 <Eddi|zuHause> planetmaker: i can't see that being useful for anything
16:40:45 <Eddi|zuHause> planetmaker: then you can just make individual tiles that you can place next to each other
16:40:48 <planetmaker> eh, not? Works also for stations
16:41:10 <planetmaker> you need no state machines and create nice multi-tile stuff
16:41:26 <planetmaker> special movement is not a requirement for that
16:42:01 <spnda> well, stations use waypoints
16:42:10 <Eddi|zuHause> planetmaker: i'm thinking about drive-through stations with two or more parallel loading stops (instead of two in opposite direction)
16:42:21 <spnda> that would be good, yeah
16:42:22 <planetmaker> there are waypoints, but rail stations are not rail waypoints
16:42:26 <Eddi|zuHause> and more sensible tram turning loops
16:42:45 <Eddi|zuHause> i'm not talking about waypoints at all.
16:42:47 <planetmaker> Eddi|zuHause, yes, I see and know what you mean. That definitely would open new possibilities (and requires state machines).
16:43:07 <planetmaker> But I see that as no way a requirement for multi-tile road stations. They simply could function like multi-tile train stations
16:44:05 *** Wormnest has joined #openttd
16:44:15 <spnda> Put two next to each other and the dev would need to check if there's neighbours and adapt.
16:44:42 <planetmaker> (and I do believe both approaches can co-exist)
16:46:03 *** tokai|noir has joined #openttd
16:46:03 *** ChanServ sets mode: +v tokai|noir
16:56:49 <Eddi|zuHause> planetmaker: i didn't say it's a requirement. just a very strong "would be nice to have"
17:18:31 <DorpsGek_III> [OpenTTD/OpenTTD] kiwitreekor opened pull request #8008: Feature: Double-click to rename vehicle https://git.io/Jv4id
17:24:34 *** ChanServ sets mode: +v tokai
17:26:26 <Samu> I have an idea to improve 8001
17:27:38 <spnda> So for single tile stop, should I just do "default: set_rs_default;" and do "spriteset (set_rs_default, "img.png") { 16_view_template(0, 0) }"?
17:27:44 <spnda> Or should I have some layouts or something
17:28:06 <Samu> seems to need the same treatment as road vehicles approaching their destination
17:28:17 <Samu> not sure how that one was exactly solved, but...
17:32:04 <Samu> i used a distance of less than 8 when ships are approaching
17:32:23 <Samu> to start counting the penalties
17:33:07 <Samu> now i'm getting 8300 rounds instead of 22000...
17:36:56 <DorpsGek_III> [OpenTTD/OpenTTD] ldpl commented on pull request #8008: Feature: Double-click to rename vehicle https://git.io/Jv4Pr
17:38:57 *** tokai|noir has joined #openttd
17:38:57 *** ChanServ sets mode: +v tokai|noir
17:53:42 <Samu> unsure of my approach, to be honest
18:06:22 <Samu> what's the command to force push?
18:06:32 <Samu> system forgot it post format
18:07:03 <Eddi|zuHause> spnda: you should still have spritelayouts, so you can compose graphics from multiple sprites (stop sign, shed, whatever)
18:11:53 <nielsm> instead of state machines for custom road stops, how about lane layouts, I think you can describe that with 8 bits for a tile
18:13:39 <nielsm> 2 bit for direction the lanes go in (N-S, E-W, NE-SW, NW-SE), 1 bit for each lane which direction it has, and 2 bits for each lane to decide whether it accepts cars, trams, both, or none
18:13:50 <nielsm> and then you can invoke pathfinding on that
18:14:08 <nielsm> actually maybe that's not the right thing to do now I think about it
18:15:03 <DorpsGek_III> [OpenTTD/OpenTTD] SamuXarick opened pull request #8009: Change #8001: Don't add docking tile cost when ships are still too far from their destination https://git.io/Jv41T
18:15:42 <nielsm> (it doesn't allow for internal junctions)
18:16:30 <planetmaker> Road stops are somewhat more complicated than train stations:
18:16:52 <planetmaker> * drive-through (which has 2 lanes with common orientation).
18:17:12 <planetmaker> * terminal which has an entry and place(s) where to stop with an orientation
18:17:21 *** ChanServ sets mode: +v tokai
18:18:27 <nielsm> yeah I suppose someone would want to make freight terminals where trucks back into an unloading bay
18:18:47 <Samu> who's gonna test 8009 LordAro ?
18:19:24 <Samu> im worried about one thing, about deadlock ships
18:19:49 <Samu> and that penalty making the ship turn around
18:20:09 <Samu> back and forth, but im unsure if that's going to happen, but it's what's worrying me
18:32:57 <spnda> Eddi|zuHause: Though I still haven't really understood how spritelayouts are handled by the game (or loaded). Any references?
18:33:14 <Eddi|zuHause> spnda: sorry, not the right person to ask that
18:33:29 <spnda> ok, I'll search further
18:34:44 <Eddi|zuHause> nielsm: yeah, proper backing up instead of this weird spinny thing
18:34:55 <peter1138> Industry Tiles or House Tiles...
18:38:13 <FLHerne> spnda: I think spritelayouts correspond fairly directly to a varaction2
18:41:10 *** Progman has joined #openttd
18:42:21 <DorpsGek_III> [OpenTTD/OpenTTD] kiwitreekor commented on pull request #8008: Feature: Double-click to rename vehicle https://git.io/Jv4MJ
18:48:22 <DorpsGek_III> [OpenTTD/nml] andythenorth commented on pull request #85: Change: industry tiles and houses now support 16 accepted cargos https://git.io/Jv4Ms
18:48:39 <andythenorth> that needs a review though :)
18:48:48 *** sla_ro|master has joined #openttd
18:49:44 <peter1138> FLHerne, yeah. But the good thing is all the "hard work" there is already done.
19:00:38 *** frosch123 has joined #openttd
19:05:02 *** WormnestAndroid has quit IRC
19:05:11 *** WormnestAndroid has joined #openttd
19:23:57 <glx> #7998 is weird, I fail to see how it's possible
19:27:57 <LordAro> getting companyhq of nonexistent company?
19:28:36 <_dp_> but it checks for invalid id first
19:28:38 <glx> maybe, but ResolveCompanyID() should take care of that
19:28:53 <frosch123> i think the problem is the CompanyID enum
19:28:56 <glx> and it does the exact same test as the assert
19:28:56 <frosch123> it is defined twice
19:29:03 <frosch123> probably one is signed, the other unsigned
19:29:08 <frosch123> so the -1 is broken inbetween
19:29:32 <LordAro> frosch123: that is the case for all the script stuff :p
19:29:41 <LordAro> wish there was a better way than redefining everything..
19:29:46 <_dp_> also can fail if _current_company somehow got corrupted
19:30:09 <frosch123> TileIndex loc = ::Company::Get((CompanyID)company)->location_of_HQ; <- that line is nonsense
19:30:19 <frosch123> "company" is already a CompanyId, just in a different namespace
19:30:36 <frosch123> and ::CompanyId is defined as byte (probably new since 1.10)
19:30:45 <glx> yeah in this file sometimes it's ::CompanyID, sometimes CompanyID
19:31:11 <LordAro> yeah, that seems like a good bet to me
19:31:17 <frosch123> glx: can you check the disassembly in msvc?
19:31:24 <frosch123> whether there is a byte cast
19:32:15 <frosch123> also ScriptCompany::ResolveCompanyID does two different casts
19:32:16 <glx> anyway there are really some inconsistencies in casts in this file
19:32:25 <frosch123> the one to IsValidID is different to the one in return
19:32:39 <LordAro> sounds like a bit of a mess
19:33:50 <frosch123> most casts cast to ::Company, only the cast in GetCompanyHQ lacks the ::
19:34:35 <frosch123> GetPresidentGender as well
19:34:48 <glx> CompanyID is ScriptCompany::CompanyID for intellisense
19:35:51 <frosch123> well, half of the calls to ::Company::Get cast to ::CompanyID, the other half to CompanyID
19:36:44 *** supermop_work_ has joined #openttd
19:36:49 <frosch123> it is also full with casts to the same type :p
19:38:23 <frosch123> glx: LordAro: I am pretty sure you can reproduce it with AICompany::GetCompanyHQ(256)
19:38:48 <glx> that's easy to test indeed
19:38:54 <frosch123> ResolveCompanyID will cast the 256 to 0, think it is a valid company, and still return 256
19:39:35 <LordAro> now i'm looking at the squirrel generation
19:39:43 <LordAro> 500 line awk script is horrifying
19:39:56 <frosch123> just look at ResolveCompanyID, it's easy
19:40:09 <LordAro> that said, the fact that someone translated it to vbs is worse
19:40:09 <_dp_> ResolveCompanyID should probably return ::CompanyID
19:40:20 <_dp_> also -1 in enums doesn't seem to be a good idea :p
19:40:25 <frosch123> yes, and it should reject 256
19:40:33 <glx> LordAro: API generation ?
19:40:34 <LordAro> _dp_: definitely not ones defined to be bytes :p
19:41:10 <LordAro> i was wondering if some doc comments (or similar) could be used instead of just redefining all the functions and types
19:41:20 <_dp_> LordAro, well, technically here it's 0xFF in the one that defined as byte...
19:41:28 <glx> and in cmake I made it happen on build
19:41:37 <_dp_> LordAro, and that's exactly the problem xD
19:44:24 <frosch123> i am preparing a pr
19:44:47 *** andythenorth has joined #openttd
19:49:52 <andythenorth> what's this Github inbox thing?
19:50:02 * andythenorth standard reaction to anything new: what's this crap? :P
19:50:10 <glx> replacement of notifications
19:50:24 <Eddi|zuHause> i'm starting to get the hang of this CSUR thing, but there still are some problematic limitations/glitches
19:50:24 <glx> requires more micromanagement
19:50:46 <andythenorth> oof fake email crap
19:51:02 * andythenorth opts out of the beta
19:52:27 <andythenorth> double-click rename eh?
19:52:29 <glx> of course I'm testing in 1.10 with a test AI expecting 1.11
19:52:34 * andythenorth increases the number of things to not give a fuck about :P
19:53:18 <DorpsGek_III> [OpenTTD/OpenTTD] frosch123 opened pull request #8010: Fix #7998: Crash when scripts tried to access companies with invalid IDs. https://git.io/Jv4y9
19:53:26 <glx> let's restart openttd after modifying the AI
19:56:47 <glx> frosch123: indeed AICompany.GetCompanyHQ(256); crashes
19:57:16 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro commented on pull request #8010: Fix #7998: Crash when scripts tried to access companies with invalid IDs. https://git.io/Jv4yN
19:57:25 <glx> I'll test your PR after dinner
19:58:20 * andythenorth hasn't checked logs
19:58:32 <andythenorth> someone talked about it last year
19:58:55 <DorpsGek_III> [OpenTTD/OpenTTD] frosch123 commented on pull request #8010: Fix #7998: Crash when scripts tried to access companies with invalid IDs. https://git.io/Jv4yA
19:59:50 <frosch123> edited comment with some addition, if you already read it
20:03:01 * andythenorth wonders about half-size ships in buy menu
20:03:23 <frosch123> i thought you already did that
20:03:29 <frosch123> or was it pikka with planes?
20:03:36 <andythenorth> I think pikka has done it somewhere
20:03:46 <andythenorth> we do it with some airports also
20:04:04 <frosch123> well, with your gui zoom they are normal size now :p
20:04:25 *** WormnestAndroid has quit IRC
20:04:25 *** ad5twoknebor[m] has quit IRC
20:04:25 <frosch123> you should consider adding the original sprites as 2x zoom sprites
20:04:33 <andythenorth> hmm, I could literally make half-size a parameter option :P
20:04:49 *** adikt-- has joined #openttd
20:05:30 *** WormnestAndroid has joined #openttd
20:05:30 *** ad5twoknebor[m] has joined #openttd
20:05:30 *** rubenwardy has joined #openttd
20:05:30 *** dwfreed has joined #openttd
20:05:32 <andythenorth> "Action 7/9 condition 0F..12" what even means? :)
20:06:50 *** WormnestAndroid has quit IRC
20:06:50 *** ad5twoknebor[m] has quit IRC
20:07:47 *** WormnestAndroid has joined #openttd
20:07:47 *** ad5twoknebor[m] has joined #openttd
20:07:47 *** rubenwardy has joined #openttd
20:07:47 *** dwfreed has joined #openttd
20:08:37 <andythenorth> checks for labels defined?
20:09:33 <frosch123> glx: btw. it also crashes/not crashes as expected before/after fix on linux
20:09:49 <frosch123> andythenorth: the intention is probably the same as for railtypes
20:10:29 <frosch123> vehicle grf checks whether a road/tramtype is defined, to pick which type to assign to its vehicles
20:10:51 <frosch123> though it's probably redundant by now, there are so many compatible/equivalent propertiers, that noone has to check single labels
20:11:25 <andythenorth> but we need it for consistency?
20:12:08 <frosch123> railtype authors are weirdos, if you need different tracks for 25/50/75 kmh, you also need that test probably :p
20:12:19 * andythenorth trying to find how nml does it for railtypes
20:13:16 <andythenorth> railtype_available
20:13:42 <andythenorth> builtin_roadtype_available seems to be there
20:14:05 *** WormnestAndroid has quit IRC
20:14:05 *** ad5twoknebor[m] has quit IRC
20:14:11 <andythenorth> builtin_tramtype_available checks 0x11
20:14:39 <frosch123> it only needs one of the pairs, the other one is the inverse, which nml can do itself
20:15:28 *** WormnestAndroid has joined #openttd
20:15:28 *** ad5twoknebor[m] has joined #openttd
20:15:28 *** rubenwardy has joined #openttd
20:15:28 *** dwfreed has joined #openttd
20:15:40 * andythenorth looking for the openttd implementation
20:15:44 <andythenorth> I have NFI how this stuff works :)
20:16:20 <andythenorth> ok they come in pairs I see
20:16:26 *** WormnestAndroid has quit IRC
20:16:26 *** ad5twoknebor[m] has quit IRC
20:16:33 <frosch123> i fixed them last month or so, they were broken in ottd :)
20:17:54 *** WormnestAndroid has joined #openttd
20:17:54 *** ad5twoknebor[m] has joined #openttd
20:17:54 *** rubenwardy has joined #openttd
20:17:54 *** dwfreed has joined #openttd
20:20:24 <andythenorth> oops, JGR isn't the correct place to look
20:20:35 * andythenorth too many openttd checkouts
20:24:03 * andythenorth wonders whether they are railtypes, rail types, or rail_types or track_types
20:24:11 <andythenorth> probably consistency doesn't matter much
20:24:17 <andythenorth> just confusing having to pick one every time :P
20:25:21 <frosch123> i stated using "track type" when referring to all three rail/road/track types
20:25:33 *** WormnestAndroid has quit IRC
20:25:33 *** ad5twoknebor[m] has quit IRC
20:25:42 <frosch123> i am probably not consistent with space/underscore/nospace
20:25:46 *** WormnestAndroid has joined #openttd
20:25:46 *** ad5twoknebor[m] has joined #openttd
20:25:46 *** rubenwardy has joined #openttd
20:25:46 *** dwfreed has joined #openttd
20:26:27 <frosch123> i see why it's named "synthon", not "synchron"
20:27:59 *** WormnestAndroid has quit IRC
20:27:59 *** ad5twoknebor[m] has quit IRC
20:28:19 <andythenorth> I wonder about refactoring nml to track_type in more places
20:28:44 <andythenorth> I hate that there are two symbols resolving to 0x4A
20:28:58 *** WormnestAndroid has joined #openttd
20:28:58 *** ad5twoknebor[m] has joined #openttd
20:28:58 *** rubenwardy has joined #openttd
20:28:58 *** dwfreed has joined #openttd
20:32:08 <frosch123> there should be three
20:33:19 <_dp_> Samu, yeah, I've seen it
20:33:39 <_dp_> gonna pretend it's jgr's bug until proven otherwise :P
20:34:29 <andythenorth> current_railtype, current_roadtype, current_tramtype
20:34:53 <andythenorth> docs say "Please note that while all variables are available for all vehicles types, some of them only make sense for one or more vehicle types. For example checking current_railtype for a non-rail vehicle doesn't make sense at all."
20:35:42 <andythenorth> it bothers me that author can use 'current_roadtype' on tram and get lies back
20:35:47 <andythenorth> I don't like APIs that lie
20:35:51 <andythenorth> maybe that's just me
20:36:15 <andythenorth> maybe I know too much about the implementation, and that tram isn't a real thing
20:36:37 *** WormnestAndroid has quit IRC
20:36:37 *** ad5twoknebor[m] has quit IRC
20:36:58 *** WormnestAndroid has joined #openttd
20:36:58 *** ad5twoknebor[m] has joined #openttd
20:36:58 *** dwfreed has joined #openttd
20:41:15 *** WormnestAndroid has quit IRC
20:41:16 *** ad5twoknebor[m] has quit IRC
20:42:22 *** dwfreed has joined #openttd
20:42:34 *** WormnestAndroid has joined #openttd
20:42:34 *** ad5twoknebor[m] has joined #openttd
20:44:46 <spnda> Ahhh I can't get myRoadStopResolverObject to like what I want to do... Just crashing all the time on ResolverObject::Resolve
21:07:41 <glx> [20:09:35] <frosch123> glx: btw. it also crashes/not crashes as expected before/after fix on linux <-- yup same here with MSVC
21:09:58 * andythenorth will write some nml docs and back away
21:10:05 <andythenorth> done > perfect, right? :P
21:10:40 <Samu> i'm skeptical about "amount *= best_rating + 1;"
21:10:52 <Samu> what is that + 1 supposed to be doing
21:12:55 <Samu> what if it doesn't exist
21:13:07 <glx> safety if best_rating is 0
21:14:17 <nielsm> remember that means amount = amount * (best_rating + 1)
21:14:58 <_dp_> it was like that already, I didn't change rating calculatien
21:15:26 <Samu> yes, but i suspect the assert triggered it because of that
21:16:29 <_dp_> Samu, i suspect you are wrong :p
21:16:49 <Samu> if there are 2 stations, one with 255 rating, other with 0, and i move 1 cargo
21:17:04 <Samu> i will be getting 256 > 255
21:18:27 <Samu> but still, perfect round numbers and there's still an extra :(
21:20:13 <andythenorth> can we never ever please change the newgrf spec ever again? :)
21:20:24 <andythenorth> it's now 100% awesome
21:22:50 <Samu> UpdateStationWating has amount += ge.amount_fract;
21:23:58 <Samu> if ge.amount_fract is 0, amount becomes 256, GB(256, 0, 8) =
21:24:34 <Samu> there has to be a reason that +1 exists
21:28:09 <frosch123> sounds like 1.10 gets delayed forever
21:28:54 <Samu> moves 1 "real" cargo and adds a fraction of 0
21:29:17 <Samu> so now I see what that + 1 is supposed to do
21:30:55 <andythenorth> so we're not changing current_roadtype to current_tracktype? o_O
21:30:57 <Samu> if i move 1 unit, i multiply it by (rating + 1)
21:31:05 <andythenorth> that would be a PITA for existing sets
21:31:47 <Samu> maybe it should compute with (rating + 1) in the cases where we're moving cargo?, what about a rating of 0
21:32:19 <frosch123> you could add current_tracktype, and make nml print "deprecated" for current_railtype
21:32:36 <frosch123> (and remove road/tram versions)
21:33:00 <frosch123> i still wonder whether we can readd some compatibility code
21:33:02 <Samu> stations with rating 0 don't get cargo, it's a condition
21:33:45 <frosch123> nml is in a python2/3 situation, many things are incompatible, and there is no proper way to have both versions
21:37:19 <andythenorth> also do I have to keep adding these {{ttdp|no}} things
21:37:22 <andythenorth> or is ttdp dead?
21:37:36 <andythenorth> when do we add a {{jgrpp} } formatter?
21:38:14 <frosch123> why {{ttdp}}? always {{ottdp}}
21:38:26 <frosch123> the ottd version is always handy, and the ttdp icon comes at no cost then
21:39:29 <frosch123> what is the jgrpp logo? a patch over the ottd logo?
21:40:58 <frosch123> unicorn is already assigned to nuts
21:41:07 <frosch123> iron pony also already exists
21:41:50 <frosch123> also, do you remember fonsinchen avatar?
21:48:19 <Eddi|zuHause> anyone remember the unicorn cake?
21:48:52 <frosch123> didn't everyone bring one?
21:49:11 <andythenorth> or I can, then we can learn how wrong I was
21:52:54 <andythenorth> urgh 0x4A in nfo spec
21:55:13 <andythenorth> the next byte with powered info is trains only?
21:55:42 <frosch123> the output of two bytes as \wx is a bit weird, but i guess it was easier
21:55:58 <andythenorth> unpowered RVs would be broken?
21:56:09 <Eddi|zuHause> are we sure we know what's going on in the regression change?
21:56:27 <frosch123> andythenorth: if someone was smart, that bit is always set for roadtypes
21:58:22 <frosch123> Eddi|zuHause: is your nfo rusty?
21:58:30 <andythenorth> that's ORing 0x100 ?
21:58:38 <andythenorth> which gives me 1 bit set?
21:58:57 <andythenorth> usually compiles fine
21:59:12 * andythenorth wonders when it will be reddit/r/jgrpp
21:59:14 <Eddi|zuHause> frosch123: i haven't tried to understand it yet, just seemed a bit much at first glance
22:02:33 <frosch123> just rewind irc to 20:10 utc+1
22:02:46 <frosch123> it's almost a matching answer
22:03:33 <Samu> jesus christ, that save provided has 87 newgrfs + missing ones, how do you even test this
22:04:00 <Samu> can't load, missing newgrfs
22:07:17 <peter1138> Hmm £160 a month to lease a Nissan Leaf for 2 years...
22:07:18 <andythenorth> the main thing is to increase the newgrf limit
22:07:22 <andythenorth> peter1138: do it!
22:07:30 <peter1138> So £6000 to use a car... for 2 years... and get nothing back. Hmm!
22:07:44 <andythenorth> how much is a normal car?
22:07:51 * andythenorth has never bought one
22:08:02 <peter1138> Last car cost me £4500... and lasted 17 years.
22:08:19 <peter1138> But that seems impossible now.
22:08:21 <andythenorth> my vehicle owning history consists of 1) Indian tuk-tuk goods carrier 2) Ford Transit
22:08:37 <peter1138> I did just buy a new bike though.
22:08:46 <andythenorth> bit rainier though
22:09:01 <DorpsGek_III> [OpenTTD/nml] frosch123 commented on pull request #85: Change: industry tiles and houses now support 16 accepted cargos https://git.io/Jv4QB
22:09:59 <andythenorth> peter1138: get a Transit :P
22:10:18 <peter1138> Vans are not comfortable drives :/
22:10:34 <Eddi|zuHause> i wish there was a car that i actually fit in...
22:11:07 <andythenorth> Transit fits everything
22:11:16 <Eddi|zuHause> not in the driver seat
22:11:20 <andythenorth> hmm, didn't fit assembled children's bunk beds
22:11:21 <peter1138> I fit in a modern Peugeot 308.
22:11:41 <peter1138> They did cheat a bit by making the steering wheel smaller than normal. Sportier, I think they call it.
22:11:46 <Eddi|zuHause> if you're above 1,90m, you fit in no car
22:11:55 <DorpsGek_III> [OpenTTD/nml] glx22 commented on pull request #85: Change: industry tiles and houses now support 16 accepted cargos https://git.io/Jv4Qr
22:12:28 <peter1138> It does make finding a car harder, but there are plenty that I can fit in, at 198 cm.
22:12:56 <Eddi|zuHause> maybe i haven't looked hard enough
22:13:28 <glx> [21:55:44] <frosch123> the output of two bytes as \wx is a bit weird, but i guess it was easier <-- if it's about #85 it was easier to not use \wx but I think the NFO is more readable with it :)
22:13:33 <peter1138> I was looking a Citroen C4 Grand Picasso... I actually had to have the seat forward a bit, there was so much room...
22:14:22 <frosch123> glx: i had to swap the bytes to check their order :)
22:15:02 <andythenorth> Eddi|zuHause: it's good marketing? :P
22:15:12 <DorpsGek_III> [OpenTTD/nml] frosch123 commented on pull request #85: Change: industry tiles and houses now support 16 accepted cargos https://git.io/Jv4QD
22:15:19 <DorpsGek_III> [OpenTTD/nml] frosch123 approved pull request #85: Change: industry tiles and houses now support 16 accepted cargos https://git.io/Jv4QS
22:16:53 <glx> yeah cargo amount reads as \wxamountcargo
22:17:44 <peter1138> Eddi|zuHause, nah, the steering wheel is on the wrong side :D
22:17:46 <glx> but with \wx the pair is easy to spot if you have many cargos
22:18:15 <peter1138> Eddi|zuHause, I guess it's a concept that doesn't actually exist, right?
22:18:33 <glx> drive on the right side then ;)
22:18:46 <Eddi|zuHause> peter1138: i seem to remember reading that they had a prototype to test-drive it
22:21:08 <Eddi|zuHause> in the news report that i read a few months back it said that production starts end of 2020, they seemed to have pushed that back a year
22:28:32 <frosch123> never heard of it, just read the wiki page
22:29:22 *** WormnestAndroid has quit IRC
22:29:22 <frosch123> they ran a successful fundraiser for 50 M€, while they have only 10k orders
22:29:42 <frosch123> those numbers do not add up
22:30:03 *** WormnestAndroid has joined #openttd
22:30:24 *** Wormnest has joined #openttd
22:30:37 <Eddi|zuHause> 10k orders with 500€ minimum payment upfront? my maths is not the greatest, but that comes out at 5M€?
22:30:59 <andythenorth> that's not enough for car development
22:31:11 <andythenorth> that's barely enough to build a basic web app
22:31:23 <andythenorth> any other finance?
22:31:37 <frosch123> it wasn't their first fundraiser
22:32:10 <Eddi|zuHause> i haven't really looked into their finances
22:32:28 <frosch123> but well, i do not know the resource cost for such a car
22:32:30 <Eddi|zuHause> it was just a thing i once read in the news, and i thought "hey, that sounds neat"
22:32:33 <andythenorth> frosch123: merging #85? :)
22:33:26 <frosch123> 25k€ per car (of which 9k are for the battery only)
22:33:31 <Eddi|zuHause> frosch123: company wiki pages doesn't really emit the most unbiased vibes
22:34:59 <frosch123> Eddi|zuHause: yes, but they should give you one side of the interval
22:35:20 <frosch123> and if even the best case sounds weird, i don't know
22:35:52 <andythenorth> glx: I shall update the docs then? How does it work now? Just lists?
22:36:17 <glx> same as before, just 16 instead of 3
22:36:25 <frosch123> Eddi|zuHause: but well, many other hyped companies also never made any profit
22:37:17 <frosch123> andythenorth: i thought you can merge in nml as well?
22:37:42 <andythenorth> GH suggests rebase and merge?
22:38:00 <glx> yes that's the usual way :)
22:38:47 <frosch123> oh, we also disabled squash? misclick or intention?
22:41:16 <DorpsGek_III> [OpenTTD/nml] andythenorth merged pull request #85: Change: industry tiles and houses now support 16 accepted cargos https://git.io/Jv4tu
22:42:32 <glx> I though squash was allowed
22:43:17 <frosch123> it usually is, but not for nml
22:43:30 <frosch123> that's why i ask :)
22:43:32 <LordAro> if only there were an org owner around to fix it...
22:44:12 <frosch123> i am only around 1% of the time, so i do not know what other people did intentionally
22:44:39 <frosch123> github has no admin log, as far as i know
22:44:53 *** WormnestAndroid has quit IRC
22:45:27 *** WormnestAndroid has joined #openttd
22:45:54 <andythenorth> thanks everybody :D
22:46:37 <Eddi|zuHause> great, now we can finally release 1.9 :p
22:46:49 <andythenorth> just in time for christmas
22:47:38 <andythenorth> "Global property 16" looks done in nml, anyone confirm that?
22:48:15 <andythenorth> people have made grfs and stuff, probably fine? :P
22:51:09 <frosch123> for nrt you may need the reverse approach. nml may have stuff implemented which does not exist in ottd
22:51:55 <frosch123> esp all those special flags changed N times
22:52:26 <glx> I unchecked "include administrators" so it's like openttd
22:52:53 <frosch123> glx: we checked that for all repos but openttd
22:53:01 <frosch123> it's only unchecked for openttd, because eints needs it
22:53:32 <glx> I just wanted to see the merge options available
22:53:52 <glx> and I can see "squash and merge" and "rebase and merge"
22:54:06 <andythenorth> frosch123: specific to nrt, when the docs are done, I'll write a test grf in nml examples dir
22:54:07 <frosch123> i already enabled "squash", but got distracted by gh asking my password again, so i forgot to mention it :p
22:54:39 <glx> the only disabled stuff is "create a merge commit"
22:55:31 <andythenorth> oof the prejudice against merge commits :P
22:55:36 <frosch123> also, should we make lordaro owner, so he can push the dorpsgek thingie to all repos?
22:56:18 <TrueBrain> Only if he enabled 2FA :p
22:56:40 <frosch123> good point, he didnt
22:56:46 <TrueBrain> GitHub has audit logs btw, under the organisation
22:57:18 <TrueBrain> But asking is easier :D
22:57:35 <TrueBrain> It is 1800 here, still weird ... timezones suck
22:57:54 <andythenorth> come home TrueBrain!
22:58:08 <TrueBrain> DorpsGek thingies can be done via PRs btw, no need to be owner
22:58:08 <frosch123> there was harsh whether in nl today
22:58:23 <TrueBrain> I fixed the rest already ;p
22:58:23 *** nielsm is now known as Guest16766
22:58:38 <frosch123> i could barely understand my coworker, when the rain was hitting the window
22:58:57 <TrueBrain> It was raining like crazy here too ... I am around the equator ...
22:59:11 <TrueBrain> "Always blue skies" my ass
22:59:13 <frosch123> anyway, where is utc-4? greenland?
22:59:41 <TrueBrain> Further south and west :)
22:59:41 <andythenorth> TrueBrain: you in India?
23:00:08 <TrueBrain> UTC-4, as mentioned
23:00:16 <LordAro> is the US east coast -4 right now?
23:00:20 <TrueBrain> I am further west than New York, silly enough
23:00:28 <TrueBrain> And a lot more south
23:01:01 <andythenorth> oh the Netherlands
23:01:10 <TrueBrain> Was in curacao last week
23:01:11 <andythenorth> how's the weather in Kingdom of Netherlands?
23:01:27 <TrueBrain> I just told you .. raining
23:01:47 <andythenorth> silly andythenorth
23:01:50 <TrueBrain> Lucky enough had to work all day, so didn't care
23:01:57 <andythenorth> funny countries that still have empires
23:02:01 <frosch123> it doesn't look more west than ny on the map...
23:02:11 <andythenorth> England still has a bit of an Empire
23:02:25 <andythenorth> but Scotland and some of N. Ireland might want out
23:02:28 <TrueBrain> It was more west according to then plane :p
23:02:46 <TrueBrain> I might be complete off
23:03:34 <TrueBrain> Such a hard life ... being send to here while my boss pays ... sitting in the sea with a beer ... such hard life ..
23:03:51 <frosch123> oh, you are there for work?
23:04:21 <TrueBrain> My work can be fun :)
23:04:54 <TrueBrain> But home soon again, and my own bed is going to be lovely
23:05:22 * andythenorth might go to bed soon
23:05:32 <TrueBrain> It is 1800 silly ... :p :p
23:05:48 <andythenorth> I shall stay up another 5hours
23:06:15 <TrueBrain> Off to meet client. Bye!
23:09:47 <andythenorth> maybe tomorrow, more newgrf spec fun :P
23:09:56 <andythenorth> can I start new Horse yet? :P
23:10:19 <frosch123> isn't it fish's turn?
23:10:42 <frosch123> oh wait, you were there earlier today
23:10:53 <andythenorth> somehow trains are more fun
23:10:58 <andythenorth> ships are just tedious :P
23:11:04 <andythenorth> road hog depends on NRT
23:11:10 <andythenorth> and we seem to think that's broken? :P
23:11:36 <frosch123> ah right :) so, chips?
23:11:50 <andythenorth> no chips is nfo stations only, it's mad
23:11:54 <andythenorth> waiting for yexo
23:11:55 <andythenorth> I'll do a test grf for NRT in NML, then I'll do some Hog
23:12:03 <DorpsGek> LordAro: Yexo was last seen in #openttd 7 years, 11 weeks, 4 days, 8 hours, 55 minutes, and 31 seconds ago: <Yexo> <NGC3982> The station glitches are CHIPS related. I adressed it to Andy the other day, and it seems like it's b0rked in some way. <- it's still in the issue tracker for CHIPS, but I haven't had time to look at it yet
23:12:13 <andythenorth> trains are mostly symmetrical which is very helpful for drawing
23:12:34 <frosch123> LordAro: you can stalk yexo, foobar and others :)
23:13:12 <LordAro> frosch123: Google, wasn't it?
23:13:37 <frosch123> yes, google to linkedin or xing
23:15:12 <frosch123> you can't stalk me anymore, it worked for some years, but now there is some young footballer with the same name
23:15:30 <frosch123> it's always footballer who break it
23:21:42 <frosch123> probably, iirc it was eddi's idea to add variables which noone needs
23:21:57 <frosch123> i skipped them intentionally for 1.9
23:24:51 *** andythenorth has left #openttd
23:30:17 *** WormnestAndroid has quit IRC
23:30:52 *** WormnestAndroid has joined #openttd
23:38:57 *** WormnestAndroid has quit IRC
23:39:01 *** WormnestAndroid has joined #openttd
continue to next day ⏵
