IRC logs for #openttd on OFTC at 2022-07-11
⏴ go to previous day
00:08:01 *** Soni has quit IRC (Ping timeout: 480 seconds)
01:13:06 *** Wormnest has quit IRC (Quit: Leaving)
01:34:25 *** Montana_ has quit IRC (Remote host closed the connection)
02:13:21 *** debdog has quit IRC (Ping timeout: 480 seconds)
02:24:28 *** D-HUND is now known as debdog
05:23:43 *** WormnestAndroid has quit IRC (Read error: No route to host)
05:23:45 *** WormnestAndroid has joined #openttd
05:45:33 *** andythenorth has joined #openttd
06:04:34 *** Smedles has joined #openttd
06:07:36 *** andythenorth has quit IRC (Quit: andythenorth)
06:20:57 *** andythenorth has joined #openttd
06:22:07 *** sla_ro|master has joined #openttd
07:12:50 *** andythenorth has quit IRC (Quit: andythenorth)
07:51:34 *** andythenorth has joined #openttd
08:04:25 *** andythenorth has quit IRC (Quit: andythenorth)
08:08:28 *** andythenorth has joined #openttd
08:50:10 *** rafalk42 has joined #openttd
09:20:44 *** rafalk42 has quit IRC (Remote host closed the connection)
09:47:35 *** WormnestAndroid has quit IRC (Remote host closed the connection)
10:24:46 *** sla_ro|master has quit IRC ()
10:35:12 *** supermop_work has quit IRC (Read error: Connection reset by peer)
10:40:50 *** WormnestAndroid has joined #openttd
11:09:35 *** andythenorth has quit IRC (Quit: andythenorth)
11:17:37 *** andythenorth has joined #openttd
11:21:45 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
11:26:36 *** WormnestAndroid has joined #openttd
11:30:30 *** tokai|noir has joined #openttd
11:30:30 *** ChanServ sets mode: +v tokai|noir
11:37:21 *** tokai has quit IRC (Ping timeout: 480 seconds)
12:43:30 <glx> with all these loops it really looks like a broken pathfinder
12:44:17 <LordAro> are you sure you're not maximising the path cost, rather than minimising it?
12:44:34 <Samu> just tried 90 degrees and it found no path, very strange
12:46:58 <Samu> if i forbid 90 degrees by design, it can find a path
12:47:13 <Samu> if i put a cost to 90 degrees turn equal to max cost, it can't find
12:47:38 <Samu> possibly something to do with the queue
12:54:12 <Samu> the more i think i know, the less I do
13:00:47 <Samu> what I think goes wrong: directions again
13:02:35 <Samu> what happens when a neighbour (a 90 degrees one) fails to pass the cost check
13:02:49 <Samu> it doesn't go into the Open
13:03:16 <Samu> only those that pass 90 degrees go into open, and can then be put on the closed list
13:03:42 <Samu> nah, i can't visualize it
13:05:19 <Samu> maybe the goal segment isn't being detected?
13:05:39 <Samu> but that would be even more stranger, cus it is already working for the other cases
13:23:36 *** wallabra_ has joined #openttd
13:28:26 *** wallabra has quit IRC (Ping timeout: 480 seconds)
13:51:09 <Samu> very strange, i put coast cost = 0
13:51:17 <Samu> and it now can find a path for 90 degrees
13:56:43 <Samu> why the heck is it now building
13:56:57 <Samu> lol... nice that it builds, but i really wanted to understand why
14:01:45 <Samu> I added the node ID check
14:01:55 <Samu> and apparently it was needed
14:03:09 <Samu> but by doing this, I open a precedent for whenever I need to build bridges/tunnels
14:03:48 <Samu> how am I going to relly on nodeID for collisions
14:25:01 <Samu> i think im storing too much information about a tile
14:32:11 <FLHerne> Samu: without reading the last month of scrollback, what is this supposed to be doing?
14:32:26 <FLHerne> standard A*-like pathfinder for rails?
14:32:49 <FLHerne> are you trying to find both rails in parallel at the same time?
14:34:02 <LordAro> trains & choochoo are the only ones i'm aware of that explicitly do dual rails
14:34:03 <FLHerne> most of the dual-track AIs build one, then try to build the other track parallel, and sometimes fail in an ugly way
14:34:53 <FLHerne> I'm sure there was a second one before trAIns
14:35:33 <Samu> what does trAIns actually do?
14:35:49 <Samu> whatever it does, it looks good
14:36:18 <LordAro> pretty sure it was the topic of the author's master's thesis
14:36:22 <LordAro> there's a pdf somewhere...
14:36:48 <Samu> I was trying to figure on my own how to do it, but...
14:37:27 <FLHerne> so the answer is yes, it does do pretty much what you're trying to
14:38:18 <Samu> oh, wow, he has pictures
14:57:38 <Samu> wow, i need to look at trAIns A* implementation
14:57:49 <Samu> he switched some ordering of stuff
14:58:00 <Samu> maybe i need to do something similar?
15:10:18 *** sla_ro|master has joined #openttd
15:27:32 *** virtualrandomnumber has joined #openttd
15:27:37 *** virtualrandomnumber has quit IRC ()
15:33:19 <Samu> 22 with tunnel and bridge
15:33:27 <Samu> i haven't though of them yet
15:34:18 <Samu> I won't create junctions, i wont need depot parts
15:35:04 <Samu> he uses a single base tile
15:36:17 <Samu> he calls them DoubleTrackParts, I call them SegmentDir
15:37:57 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
15:39:02 *** WormnestAndroid has joined #openttd
15:39:04 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
15:39:05 *** WormnestAndroid has joined #openttd
15:40:37 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
15:40:38 *** WormnestAndroid has joined #openttd
15:56:28 *** HerzogDeXtEr has joined #openttd
16:10:41 *** sla_ro|master has quit IRC (Ping timeout: 480 seconds)
16:33:24 *** Smedles_ has joined #openttd
16:33:38 *** Smedles has quit IRC (Ping timeout: 480 seconds)
16:35:39 *** Smedles has joined #openttd
17:13:27 <Samu> alright, the only part I'm interested in is how he handled the so called "directions", to check if it is already in the closed list
17:16:57 <Samu> relatively simple compared to my approach :(
17:21:04 <Samu> i already use 3 different int64 for bitmasking
17:52:57 *** virtualrandomnumber has joined #openttd
17:53:06 *** virtualrandomnumber has quit IRC ()
18:00:32 *** Wormnest has joined #openttd
18:17:06 *** agneli has quit IRC (Remote host closed the connection)
18:17:49 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
19:13:44 *** Smedles has joined #openttd
19:37:29 <Samu> I was pretty convinced splits wouldn't occur again, I was wrong
19:48:39 <Samu> is it tracks merging or tracks splitting?
20:05:06 <Samu> I don't know what to do, I'm gonna give up
20:28:58 <Samu> how do you feel when you have to give up doing something you wanted to make but fail
20:29:21 <Samu> it feels like i'm so close to doing it
20:29:30 <Samu> but there's always something amiss
20:39:12 <Samu> thx for the help anyway, this is a too big of a task for my limited experience
20:47:11 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:14:56 *** andythenorth_ has joined #openttd
21:17:51 *** andythenorth has quit IRC (Ping timeout: 480 seconds)
21:26:46 *** jpw has quit IRC (Ping timeout: 480 seconds)
21:27:44 *** andythenorth_ has quit IRC (Quit: andythenorth_)
21:41:45 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
21:45:15 *** andythenorth has joined #openttd
21:50:41 *** andythenorth has quit IRC (Quit: andythenorth)
21:55:26 <Samu> what's the hotkey for fast forward
21:56:45 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
21:56:52 <Samu> not ideal for my testing
21:57:07 *** Samu has quit IRC (Quit: Leaving)
23:03:11 *** rafalk42 has joined #openttd
23:26:23 *** agneli has quit IRC (charon.oftc.net dacia.oftc.net)
23:41:23 *** agneli has quit IRC (charon.oftc.net dacia.oftc.net)
continue to next day ⏵
