IRC logs for #openttd on OFTC at 2022-06-09
⏴ go to previous day
01:00:32 *** Smedles has joined #openttd
01:56:50 <wallabra> I wanna do highway interchanges!...
01:57:34 <wallabra> I think I'll have to resort to SimCity 4 with Network Addons Mod for the time being, to quench that profound and utter desire, fundamentally msaculine in its nature.
01:59:34 <wallabra> Mmm, unexpected termination of services.exe. This wine ages like milk!
02:26:34 *** Wormnest has quit IRC (Quit: Leaving)
02:50:22 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:59:59 *** D-HUND is now known as debdog
05:24:27 *** jpw has quit IRC (Ping timeout: 480 seconds)
06:02:41 <dwfreed> wallabra: Cities: Skylines
06:03:34 <dwfreed> You can totally sandbox some highways
06:20:40 <wallabra> it definitely nailed the openttd vibe
06:28:11 <wallabra> also if it says "too much traffic" just keep trying, it'll get through soon enough
06:28:15 <wallabra> then it takes like 2 mins to do your thing
06:54:49 <jpw> for a start, it actually works
07:26:44 *** HerzogDeXtEr has joined #openttd
07:43:39 <peter1138> Is there an RTX Quake yet?
08:37:46 *** jpw has quit IRC (Ping timeout: 480 seconds)
08:46:45 <FLHerne> wallabra: tbf, I have trouble interpreting the rest of your prompt as anything meaningful :p
08:46:55 <FLHerne> and yes, it's amazing it works at all
08:47:23 <wallabra> FLHerne, yeah, but to be fair, you're not an AI! AIs are much more imaginative than us!... :P
08:47:54 <wallabra> I joke, but it's always interesting to see how they deal with such absurdist, quasi-Monty-Pythonic aberrations of text.
08:48:18 <wallabra> peter1138, Quake 2 yes, not Q1
08:48:30 <wallabra> which is a pity because obviously Q1 had a lot more atmosphere and was, imho, a much better fit for such adaptation.
09:17:57 <Samu> i need to investigate why mine doesn't do any cache clears not requiring pathfinding
09:19:57 <Samu> guess i need AIs that do remove roads
09:20:22 <Samu> or maybe only dictatorai
09:20:32 <Samu> the only one that removes roads
09:29:31 *** wallabra has joined #openttd
09:36:24 *** WormnestAndroid has quit IRC (Remote host closed the connection)
09:40:00 *** wallabra has joined #openttd
09:42:03 <Samu> glx22 fix is interesting, it does seem to keep synchronized with master
09:43:16 <Samu> at least by looking at the profit graph with 14 dictator ais
09:43:44 *** wallabra has joined #openttd
09:59:35 *** lobstarooo has joined #openttd
10:06:03 *** lobster has quit IRC (Ping timeout: 480 seconds)
10:06:06 *** lobstarooo is now known as lobster
10:14:13 <Samu> let me test without that fix
10:25:19 <Samu> there's more to it than it looks
10:28:10 <peter1138> wallabra, yeah Q2 is disappointing though.
10:29:05 <peter1138> wallabra, there's also an RT Doom, but it's also a bit disappointing - overblown colours etc.
10:29:57 <peter1138> Okay, do I buy all the equipment to replace my gear cables or just pay someone else to do it... Hmm.
10:30:31 *** WormnestAndroid has joined #openttd
10:46:33 <LordAro> i go for the latter option
11:12:40 <peter1138> Might try Halfords. The nearest local bike shop are worse...
12:07:47 *** Etua has quit IRC (Remote host closed the connection)
13:34:41 *** andythenorth has joined #openttd
13:34:50 <andythenorth> did we ban cb 36 changing vehicle length in stations?
13:34:58 <andythenorth> I thought we did because desyncs or something
13:35:10 <andythenorth> or because it breaks PBS reservations, or causes train crashes or something
13:38:09 <LordAro> certainly seems undesirable
13:38:10 *** Etua has quit IRC (Quit: Etua)
13:55:01 <andythenorth> afaict grfs are doing it
13:59:36 <FLHerne> I'm sure it did cause desyncs
14:05:04 *** Smedles has joined #openttd
14:28:47 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
14:29:51 <andythenorth> JP Set seems to change vehicle lengths
14:32:39 *** andythenorth has quit IRC (Quit: andythenorth)
15:01:33 <Samu> hi glx, i figured counting calls could help find some irregular rises to pathfinder calls
15:03:14 <Samu> think i found some odd ones worth investigating
15:06:03 <Samu> what i'm saying is that, there might be some more cases of possible performance degradation that can't be measured due to not being perceptible
15:12:06 <glx> with many send to depot at the same time
15:14:11 <glx> 3 of your versions seems to affect companies
15:17:35 <Samu> 3 are still outputting the same play
15:31:49 *** kstar892[m] has joined #openttd
15:51:07 <Samu> Assertion failed at line 968 of D:\OpenTTD\OpenTTD GitHub\OpenTTD\src\framerate_gui.cpp: newpoint.x <= lastpoint.x
15:51:16 <Samu> i hate it when im in the middle of something
15:53:34 <Samu> gonna try pintpoint to it
15:53:50 <Samu> but with the graph dude crashing me, i won't be able to save game
15:54:36 <peter1138> Do I go out on the bike or watch Ms Marvel?
15:54:55 <peter1138> I can't go with the group ride because the fast bike is the one that needs fixing...
15:55:03 <Samu> nevermind, emergency save will do :p
15:55:12 <LordAro> is your slow bike that slow?
15:55:48 <peter1138> Depends, I've managed one 20mph average ride on it, but normally 16-17.
15:57:03 <LordAro> sounds like excuses to me
16:00:54 <nielsm> Samu: when you get that assertion in the framerate graph window, has the game been frozen/hanging for a long time, or maybe been suspended in the debugger?
16:01:21 <Samu> i paused game, and immediately crashes
16:01:33 <nielsm> "paused" as clicking the in-game pause button?
16:01:48 <Samu> game was fast forwarding
16:01:57 <Samu> then i click pause and insta crash
16:11:24 <Samu> it displays a speed of 33 km/h, but it's actually stopped
16:21:27 *** gelignite has joined #openttd
16:34:27 *** Smedles has joined #openttd
16:49:52 <glx> indeed this vehicle is in a pathfinding loop
16:58:26 *** WormnestAndroid has joined #openttd
17:03:13 *** andythenorth has joined #openttd
17:08:56 <glx> ok so #124 has speed but is not moving, #28 copies #124 speed, #22 copies #28 and enters in pf loop
17:13:45 <andythenorth> needs a mind that can see round corners
17:16:25 *** wallabra has joined #openttd
17:21:50 *** Wormnest has joined #openttd
17:28:22 <glx> ok after being blocked for some time RoadVehFindCloseTo() returns nullptr even if there's a vehicle in front
17:56:54 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
18:22:14 <peter1138> Well that didn't work out too well.
18:23:11 <peter1138> That's 1978, not 9917.
18:30:50 <andythenorth> this will be something silly :P
18:38:58 <andythenorth> The best feature.
18:51:41 <LordAro> andythenorth: if the buy window has a link back to the depot, then it should definitely be able to know about the max speed
18:52:21 <andythenorth> the problem is the 'available vehicles' window iirc
18:52:28 <andythenorth> there's no railtype in that context
18:52:37 <andythenorth> frosch probably knows the real reason though
18:54:17 <Rubidium> the 'available vehicles' window is generic; from the vehicle list you can get to that window, and the "main" vehicle list does not have a specific tile/depot/railtype associated with it. Just all vehicles of that type
18:55:41 <andythenorth> we _could_ put something into extra callback info var to handle that case
18:55:55 <andythenorth> or otherwise specify a fallback
18:57:10 <Samu> just noticed master play is now different than 9917, looking at profit graphs, there are small differences
18:58:37 <Samu> seems to have happened on oct-dec 1976 period
19:06:22 <peter1138> LordAro, didn't have to worry about speed, nobody else turned up.
19:11:42 <Samu> current driver name is too large
19:15:13 <nielsm> if the windowing system layout code was better for "text that may break lines and shouldn't affect total width" it might look better
19:15:24 <nielsm> actually maybe I should see if it's possible to do in this case...
19:16:19 <nielsm> since there are plenty of other widgets in the window that ensure its width
19:41:43 <Samu> the issue happening in year 1976 is caused by a lot of vehicles in a long queue along a long road, similar as that in the crashed savegame i uploaded
19:41:58 <Samu> vehicles are blocked but not at 0 km/h
19:42:39 *** lobstarooo has joined #openttd
19:43:31 <glx> yes because one of the blocked vehicle is blocked for so long that the detection fails
19:44:10 <glx> so its speed is not set to 0 and vehicles behind it are affected by that
19:45:10 <glx> vehicle #124 blocked_ctr is above 1480
19:45:43 <glx> so it's blocked but doesn't see anything in front of it
19:50:18 *** lobster has quit IRC (Ping timeout: 480 seconds)
19:50:20 *** lobstarooo is now known as lobster
19:51:22 *** Etua has quit IRC (Quit: Etua)
19:54:17 <nielsm> never mind, I remember now what the issue with breaking text lines is, the window sizing calculation is single-pass, so the text area can't know how much width is minimum available from all the other controls, so it can't know how the width it has to line break at, so it can't determine its required height
19:54:44 <nielsm> it's still the issue of having to rework the window sizing algorithm to be two-pass
20:02:21 <Samu> i still can't figure out why it became out of sync with master, where exactly triggered this
20:04:07 *** gelignite has quit IRC (Quit: Stay safe!)
20:04:41 <Samu> not important but just curious
20:05:02 <Samu> mine also isn't synced with master
20:07:58 *** lobstarooo has joined #openttd
20:09:54 *** lobstarooo has quit IRC (Read error: Connection reset by peer)
20:10:23 <Samu> bottom left is still synced with bottom right
20:10:37 <Samu> but none are synced with master any more
20:12:10 *** lobstarooo has joined #openttd
20:12:28 <peter1138> Out of sync with master? Is that not expected if you invalidating caches?
20:13:29 <Samu> but glx was trying to keep it synced
20:14:06 <glx> I was just trying to prevent useless pf runs
20:14:27 <glx> but the blocage detection is broken
20:14:49 <peter1138> Have you fallen down a rabbit hole?
20:15:06 *** lobster has quit IRC (Ping timeout: 480 seconds)
20:15:11 *** lobstarooo is now known as lobster
20:16:09 <glx> all blocage tests rely on the result of this function, and at some point it stops returning a vehicle
20:21:18 <glx> hmm maybe I should just remove the speed check
21:02:42 <glx> crash save is still laggy, but it's not because useless pathfinding
21:05:14 <Samu> trackdir in that line should be new_td
21:05:48 <Samu> it's not really a bug because SlopeCost doesn't make use of trackdir
21:06:21 <Samu> but should one day use it, then yes, it's using the wrong one
21:07:38 <glx> trackdir is new_td in next iteration
21:09:42 <Samu> ah yes, further down below
21:09:50 <glx> new_td is for next tile not current tile
21:11:52 <andythenorth> does nmlc need to check palette of source images when output is nfo?
21:14:17 <Samu> so that means i have my code wrong
21:15:30 <andythenorth> seems I could have about 1.5s saving by not checking image palette
21:15:46 <Samu> my SlopeCost is... "different"
21:15:48 <andythenorth> that's about 10% in the case I'm looking at
21:21:27 <Samu> there are situations where tile is the same as tile_next, even in that loop
21:26:50 <Samu> damn it, it doesn't get the height of bridges correctly?
21:30:07 <Samu> nvm, i better head to bed, no time to test this today
21:33:31 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:34:04 <Samu> why is it getting the slope pixel z of water tile, when it's actually a tunnelbridge tile? ...
21:39:15 <Samu> okay, this is putting me on a water tile, must investigate tomorrow how this function works TileVirtXY(x, y);
21:39:29 *** Samu has quit IRC (Quit: Leaving)
21:39:51 *** andythenorth has quit IRC (Quit: andythenorth)
23:21:18 *** Smedles has joined #openttd
continue to next day ⏵
