IRC logs for #openttd on OFTC at 2021-08-26
⏴ go to previous day
00:41:59 *** peter1138 has quit IRC (Remote host closed the connection)
00:46:35 *** peter1138 has joined #openttd
00:46:35 *** ChanServ sets mode: +o peter1138
02:12:26 *** Wormnest has quit IRC (Quit: Leaving)
02:43:54 *** debdog has quit IRC (Ping timeout: 480 seconds)
02:44:46 *** ChanServ sets mode: +v tokai
02:51:36 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
05:23:38 *** sla_ro|master has joined #openttd
06:48:53 *** andythenorth has joined #openttd
07:03:02 *** jottyfan has joined #openttd
07:34:34 *** andythenorth has quit IRC (Quit: andythenorth)
07:39:31 *** andythenorth has joined #openttd
09:45:18 *** jottyfan has quit IRC (Quit: jottyfan)
09:47:49 <Timberwolf> I think I found a legitimate situation where block signals are better than path-based. It's when you have a track layout like ____/_/_/_/_/_/_/_/____ , and any of the slips is a valid option.
09:48:45 <Timberwolf> If the first few slips are blocked, the pathfinder gets called at a PBS entry signal and YAPF goes "well yeah, there's a green signal, but it's all the way over there so I'd prefer to do something closer"
09:50:53 <Timberwolf> Whereas if the entry is a block, the train keeps moving.
09:51:25 <Timberwolf> It is also a bit of a corner case, because why the hell would you build that :)
09:53:48 <dwfreed> You build that a lot in factorio
09:53:55 <dwfreed> not that openttd is factorio, but
09:56:42 <dwfreed> the biggest bottleneck is stations; if you need high throughput, that kind of layout leading into a station with equal number of platforms is great
09:57:11 <Timberwolf> Factorio only has block signals. I note that in any mid-late game multiplayer session, there's the inevitable "all of my trains have gone wrong, where's the OpenTTD player?"
09:58:51 <Timberwolf> Usually for us the problem is either "the oil depot fits 2 trains, we have 3, but production has dropped and now everything is blocked"
09:59:32 <Timberwolf> Or, "someone's megaproject starved the coal bus, the trains didn't get auto-refuelled, now one of them has stopped somewhere along the track and it's causing utter chaos"
09:59:46 <nielsm> programmable signals/route restrictions when?
10:00:30 <Timberwolf> I often build "safe" level crossings where walls go up if a train is approaching.
10:19:59 *** andythenorth has left #openttd
11:01:20 *** tokai|noir has joined #openttd
11:01:20 *** ChanServ sets mode: +v tokai|noir
11:06:40 *** roadt has quit IRC (Ping timeout: 480 seconds)
11:08:05 *** tokai has quit IRC (Ping timeout: 480 seconds)
11:38:55 *** jottyfan has joined #openttd
12:41:06 *** virtualrandomnumber has joined #openttd
12:41:21 *** virtualrandomnumber has quit IRC ()
13:02:02 *** sla_ro|master has quit IRC ()
13:59:25 *** jottyfan has quit IRC (Quit: jottyfan)
14:06:12 *** ChanServ sets mode: +v tokai
14:13:06 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
14:29:31 *** Eddi|zuHause has quit IRC ()
14:33:03 *** Eddi|zuHause has joined #openttd
14:33:51 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #8688: Feature: Hide block signal GUI by default https://git.io/JE0sM
15:00:04 *** jottyfan has joined #openttd
15:03:57 <TrueBrain> Timberwolf: I just use the LTN mod .. solves a lot of issues :D
15:04:06 <TrueBrain> Trains wait in depot till they are needed
15:20:05 <Timberwolf> Depots are BAD FEATURES.
15:20:14 <Timberwolf> We know this, because Chris Sawyer removed them from Locomotion :)
15:21:01 <Timberwolf> And no transfer tracks/storage in RCT, either.
15:24:05 <Timberwolf> One thing I found out the other week that amused me is one of the Disney parks has the train storage for Big Thunder Mountain on the other side of the tracks for the Disney Railroad without any grade separation, so when they want to add trains to the roller coaster they have to call up the railroad operators and go, "can we get our trains out please?"
15:33:49 *** Wormnest has joined #openttd
15:34:59 *** Samu has quit IRC (Ping timeout: 480 seconds)
15:43:13 *** Progman has joined #openttd
16:35:45 *** Gustavo6046 has joined #openttd
16:50:38 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
16:54:42 *** virtualrandomnumber has joined #openttd
17:17:42 *** jottyfan has joined #openttd
17:32:26 *** virtualrandomnumber has quit IRC (Quit: virtualrandomnumber)
17:41:56 *** jottyfan has quit IRC (Quit: jottyfan)
17:48:21 *** HerzogDeXtEr has joined #openttd
18:18:22 <TrueBrain> " which is much less obvious than the dev team thinks" <- I just love it if people put words in our mouths and stamp on it
18:18:27 <TrueBrain> I shouldn't read forums :D
18:28:37 *** Gustavo6046 has quit IRC (Ping timeout: 480 seconds)
18:31:02 *** Gustavo6046 has joined #openttd
18:35:23 <Beer> Hello! Dealing with C++, is it considered better to use [] operator or at() function to access a single char?
18:36:30 <Eddi|zuHause> a) why should anyone care? b) why would you assume the kind of people that DO care came up with a consensus?
18:37:25 <TrueBrain> Beer: you mean like: 'char *a = "blabla"; a[3]'?
18:37:50 <TrueBrain> (your question is a bit random and lacking any sense of context :P)
18:38:59 <Beer> Sry, meant in a string object
18:39:39 <TrueBrain> an std::string I assume in that case?
18:39:40 <Beer> std::string a; a[0]; a.at(0);
18:39:54 <TrueBrain> yeah .. there is a huge difference between the two
18:39:59 <TrueBrain> latter does bound-checking
18:40:09 <TrueBrain> so "considered better" needs a lot more context to be answered
18:40:20 *** andythenorth has joined #openttd
18:40:27 <Beer> Bound checks are better, by default, I suppose?
18:40:36 <TrueBrain> why would that be true?
18:41:03 <andythenorth> TrueBrain don't read forums
18:41:11 <andythenorth> I have been reading them again, you don't need to :P
18:41:17 <Beer> Unless you know what you're doing, ensuring you access allocated memory is best?
18:41:18 <TrueBrain> andythenorth: my apologizes :P
18:41:24 * andythenorth had to do the NML release and read follow ups :P
18:41:32 <TrueBrain> Beer: I can only repeat what I wrote: "considered better" needs a lot more context to be answered
18:42:16 <TrueBrain> for (uint i = 0; i < a.length(); i++) a[i];
18:42:17 <Beer> In which standard case don't you want to do bound checks?
18:42:21 <TrueBrain> does that really need bound-checking?
18:42:48 <TrueBrain> is it really useful to waste CPU time on checking: is this index that is guaranteed to be in range in range?
18:43:00 <andythenorth> OpenTTD nearly dying
18:43:13 <Eddi|zuHause> andythenorth: i think that's an improvement?
18:43:21 <andythenorth> too early to tell
18:43:23 <TrueBrain> at least there is "nearly" in there :)
18:43:30 <andythenorth> TrueBrain that's really bad python
18:43:35 <andythenorth> you don't need all those parentheses
18:43:48 <TrueBrain> andythenorth: pretty sure .length() also fails :P
18:43:58 <andythenorth> uint- is that the new typing in 3.something
18:44:14 <TrueBrain> TrueBrain is talking, it must be Python :P
18:44:51 <Beer> TrueBrain: OK I see what you mean. Thx for the bound check piece of info!
18:45:06 <TrueBrain> for these questions, always look at the C++ references websites
18:45:13 <TrueBrain> they are pretty verbose about these things -most- of the time :)
18:45:28 <Beer> That's what I do. Missed the bound check thingy
18:45:49 <TrueBrain> std::string isn't as clear in it as std::vector: "This is in contrast with member operator[], that does not check against bounds."
18:45:52 <Beer> Thought it was a matter of practices/habits, and came here for code style guidelines :)
18:46:07 <TrueBrain> owh, 100 people 120 opinions on the matter
18:46:12 <TrueBrain> but none matter without more context :)
18:47:09 <andythenorth> ok reddit complete
18:48:08 <TrueBrain> lot of pictures on reddit
18:48:52 <andythenorth> that's all my fun over
18:50:39 <DorpsGek> - Update: Translations from eints (by translators)
18:51:20 <TrueBrain> I can do a dance for you instead?
18:55:35 <andythenorth> save it for next livestream :)
19:03:19 <TrueBrain> you don't pay my subscription!
19:11:17 <Samu_> nice, visual studio works again
19:13:56 <Samu_> or maybe not, stuck at "loading extensions"
19:18:49 *** sla_ro|master has joined #openttd
19:23:19 <TrueBrain> you know what that means ... cake for everyone!
19:37:38 <Samu_> I am inclined to making a pf setting to customize docking tile penalty, so that I can set it to 0 on my games :p
19:38:06 <glx> there are way too many pf settings
19:39:17 <Eddi|zuHause> the cake is still a lie?
19:39:38 <Eddi|zuHause> or maybe "still alive"? :p
19:39:58 <Eddi|zuHause> ... this was a triumph... *sing*
19:42:08 <Samu_> looking at it for road vehicles I see there's a penalty for going through drive through and drive through occupancy
19:42:22 <Samu_> but there's none for ship
19:44:01 <Samu_> road_stop_penalty, road_stop_occupied_penalty, road_stop_bay_occupied_penalty, npf_road_drive_through_penalty, npf_road_dt_occupied_penalty, npf_road_bay_occupied_penalty
19:44:38 <Samu_> to be complete, ships would need 4 pf settings, 2 for npf, 2 for yapf
19:45:19 <Samu_> something like ship_docking_penalty, ship_docking_occuppied_penalty * 2
19:45:44 <Samu_> but ok, no more settings
19:46:06 *** WormnestAndroid has quit IRC (Remote host closed the connection)
19:47:48 *** WormnestAndroid has joined #openttd
19:52:06 <Samu_> the way it's currently set up, if I have 1000 ships in the docking_tile, the penalty just for that dock would be 1001 * 3 * 100
19:53:01 <Samu_> feels wrong, there should be a cap
19:53:39 <Samu_> for npf, the penalty can be at times 100 or 70
19:54:00 <Samu_> actually 1001 * 3 * (100 or 70)
19:56:02 <Samu_> I am so compelled to do a PR about this :(
20:03:24 <andythenorth> the 3 GS I used most are all pure goal scripts
20:03:35 <andythenorth> they don't modify anything in game apart from goal window or storybook
20:08:24 <peter1138> Probably a non-configurable cap makes sense
20:17:11 <Samu_> count starts as 1 just for being a docking tile
20:17:44 <Samu_> count then adds the number of ships in the tile atm
20:18:26 <glx> just use number of ships but with a cap to number of ships
20:19:13 <glx> like min(number of ships, some value)
20:20:42 <Samu_> the higher the penalty, the more likely for the pathfinder to spew ship is lost
20:20:49 <glx> if you have access to number of docking tiles that could be a good minimum
20:21:17 <Samu_> it's very sensitive because it's on open water
20:22:24 *** FLHerne_ has joined #openttd
20:23:28 *** FLHerne has quit IRC (Read error: Connection reset by peer)
20:24:17 <andythenorth> goes it a "I am just a goalscript flag"?
20:27:57 <Samu_> im revisiting #8001 savegames
20:31:03 <Samu_> dbg: [sl] ¯üåGame Load Failed
20:31:03 <Samu_> Internal error: Loader for 'zlib' is not available.
20:31:15 <Samu_> visual studio failing me
20:36:07 *** Speeder has quit IRC (Read error: Connection reset by peer)
20:36:31 *** Speeder has joined #openttd
20:55:20 *** Gustavo6046_ has joined #openttd
21:00:05 *** Gustavo6046 has quit IRC (Ping timeout: 480 seconds)
21:00:05 *** Gustavo6046_ is now known as Gustavo6046
21:06:41 *** sla_ro|master has quit IRC ()
21:07:28 <Beer> Are changes to networking documented? It seems there is no advertising to the master server?
21:07:43 <Beer> New configuration directives required?
21:08:47 <Beer> Testing 12.0-beta2 on a previousy working server configuration
21:16:46 <Samu_> cap of 1 ship is enough, due to pf max search nodes ship is lost
21:17:02 <Samu_> and yet that's still not good enough
21:17:15 <Samu_> at least for the savegames in #8001
21:22:41 <Beer> OK found network.server_game_type
21:24:01 *** Wuzzy has quit IRC (Remote host closed the connection)
21:28:44 *** Samu_ has quit IRC (Ping timeout: 480 seconds)
21:43:38 <andythenorth> lol testing random things
21:44:47 <glx> Beer: a conversion from old settings to new ones is supposed to happen, but somehow it doesn't work very well
21:50:40 <glx> hmm looking at the source, there's no conversion at all
21:53:53 <Beer> Wouldn't have helped me anyway, as I load read-only configurations, but that might be an interesting idea for adoption.
21:55:01 <Beer> I suppose documentation might be the 1st step, as expecting people to skim through code or install/launch client to check settings is not very accomodating
21:55:38 <Beer> Breaking all the dedicated servers around is gonna raise issues :P
21:56:44 <glx> main issue is most dedicated servers don't try beta
21:58:44 <glx> crazy there are still 1.3.0 servers
22:19:14 *** tokai|noir has joined #openttd
22:19:14 *** ChanServ sets mode: +v tokai|noir
22:25:56 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:26:09 *** tokai has quit IRC (Ping timeout: 480 seconds)
22:29:22 *** Gustavo6046 has joined #openttd
22:35:22 *** andythenorth has quit IRC (Quit: andythenorth)
22:45:02 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
22:46:54 *** Progman has quit IRC (Remote host closed the connection)
22:57:24 *** nielsm has quit IRC (Ping timeout: 480 seconds)
23:04:09 *** virtualrandomnumber has joined #openttd
23:04:22 *** virtualrandomnumber has quit IRC ()
23:22:35 *** FLHerne_ is now known as FLHerne
23:25:12 *** ChanServ sets mode: +v tokai
23:27:36 *** Etua has quit IRC (Quit: Etua)
23:32:06 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
continue to next day ⏵