IRC logs for #openttd on OFTC at 2024-07-11
⏴ go to previous day
01:54:35 *** Wormnest has quit IRC (Quit: Leaving)
02:03:42 *** herms6 has quit IRC (Quit: bye)
02:43:59 *** debdog has quit IRC (Ping timeout: 480 seconds)
04:43:23 <DorpsGek> - Update: Translations from eints (by translators)
05:01:21 <johnfranklin> Temperature of feeling: 49 deg-C
07:17:47 *** reldred has quit IRC (Quit: User went offline on Discord a while ago)
08:10:13 <yiffgirl> i have an idea for making more organic looking tree groups that would obsolete half of this patch but that can wait
08:17:47 <andythenorth> Most tree patches get stuck
08:29:07 *** reldred has joined #openttd
08:29:09 <reldred> CONTROL TREES BY HEIGHT
08:29:16 <reldred> SNOW/DESERT AWARE TREES
08:29:31 <reldred> PLACEMENT RULES FOR AUTO GENERATING TREES
08:31:19 <truebrain> Your capslock is on
08:31:30 <yiffgirl> aren't trees already desert aware
08:31:30 <yiffgirl> or can you plant cacti anywhere
08:31:30 <yiffgirl> <- has never played tropical
08:31:36 <reldred> or I'm just very excited
08:31:49 <truebrain> Sorry, what do you say? I can't hear all that well, someone was yelling in my EAR
08:32:19 <LordAro> it's a shame bash.org died
08:32:44 <truebrain> shows how often I visited it 😛
08:32:53 <yiffgirl> been gone a long while now.
08:33:29 <yiffgirl> i could have sworn it's been down longer
08:34:05 <truebrain> The only reason bash.org was nice .. to quote these kind of things. Has been too long since I have seen that 🙂
08:34:23 <reldred> cruise control for cool
08:53:32 *** mindlesstux has joined #openttd
10:09:15 <peter1138> Bash.org stopped updating years ago
10:14:09 <truebrain> about the same time that IRC died? Owh wait ..... 😛
11:05:26 <peter1138> Bash.org stopped updating years ago
11:05:54 <peter1138> LordAro, 80 miles in, why have I done this...
11:11:42 *** mindlesstux has quit IRC (Ping timeout: 480 seconds)
11:15:49 *** mindlesstux has joined #openttd
11:50:20 <LordAro> peter1138: well you're nearly there then
12:33:53 <truebrain> hmm .. seems you can't create a Private Subnet on Oracle Cloud with IPv6 to talk to the Internet. You can on AWS. The small differences 😄
14:03:18 *** gelignite has joined #openttd
14:40:42 <goddess_ishtar> I miss bash.org
16:43:29 *** gelignite has quit IRC (Quit: Stay safe!)
17:42:46 <michi_cc> Apply /s were appropriate
17:48:20 <andythenorth> best PR for some time 😄
17:49:17 <andythenorth> michi_cc: oh so it's also 'realistic'? 🙂
17:50:44 <michi_cc> English WP page seems to be a bit more sparse, but default-green is basically like the german Selbstblock (which seems quite close to UK automatic block signalling).
17:51:18 <andythenorth> I started reading about it last year, and found some stuff about semaphores mechanically defaulting to 'danger'
17:51:35 <andythenorth> but then I stopped reading 😛
17:52:20 <michi_cc> Well, Selbstblock is not controlled by some other entity, but only based on the state of the track after the signal. If it was default-red, no train could ever run 🙂
17:54:04 <michi_cc> With modern electronic signal boxes, all signals are centrally controlled, so default state may vary to whatever. But who would be that retentive...? 😛
17:54:19 <andythenorth> green is faster 😛
18:09:39 <Rubidium> michi_cc: modern signal boxes is kinda oxymoronic
18:11:08 <goddess_ishtar> > It is well known that path signals are slower. After all, they show red all the time, and red is very much the colour for stop and not go in western cultures. Ergo, red signals are slow signals.
18:11:17 <Rubidium> modern "signalling" is actually signalless
18:15:17 <michi_cc> 1970's modern of course 😛
18:15:37 <andythenorth> it is much faster
18:16:20 *** Wormnest has joined #openttd
18:16:36 <truebrain> michi_cc: at the very least add a screenshot to the PR! 😛
18:17:07 <andythenorth> video would be better 😄
18:26:12 <Rubidium> andythenorth: no, it's not realistic... for the Netherlands at least
18:29:10 <andythenorth> wait, did you make a signal grf too? 😛
18:30:09 <Rubidium> if you have the non-one-way signals not going to a junction, going against them does not turn them all red. They'll be green/red based (okay, there's also yellow/amber/orange, but that's not what's added/changed here) on the actual occupied block, not on the track reservation. Signals at junctions will be turned to green based on reservations though ;)
18:30:09 <michi_cc> Oh, no, I had DBSetXL in the NewGRF list.
18:31:15 <truebrain> michi_cc: a lot of screenshots! 😛
18:31:36 <Rubidium> and some random sound in the background ;)
18:32:39 <michi_cc> Rubidium: Oops, I didn't even check because Windows have some warning about exclusive audio devices 🙂 I assumed it would have no sound 🙂
18:33:42 <andythenorth> "no good deed goes un-nitpicked" 🙂
18:34:11 <michi_cc> Also, I never claimed to have copied some real-world signal system exactly, so your opinion on red signals against the path may vary.
18:34:12 <andythenorth> the single yellow aspects kind of make sense for how it's working
18:34:14 <Rubidium> though to be realistic (for the NL), the rightmost signal on the bottom "row" should remain green when the train reverts at ~4 seconds and only turn red at ~7 seconds (or yellow at ~6 seconds)
18:34:37 <andythenorth> wall-clock based signalling grf spec? 😛
18:34:48 *** HerzogDeXtEr has joined #openttd
18:35:01 <Rubidium> michi_cc: I know you didn't, it was andythenorth suggesting it to be realistic
18:35:12 <michi_cc> But really, I'm totally open on whether backward passed signals should turn red or remain green. Comments in the PR welcome.
18:36:20 <gwyd4016> Back passable signals should default red, if realism is the flavour of the day, a signal shouldn't be able switch from green to red in front of a train
18:36:22 <goddess_ishtar> wait those are path signals?
18:36:27 <Rubidium> IMHO less logic is better, and to be honest realistic isn't what people would expect
18:36:37 <andythenorth> 'faster' is the goal 🙂
18:36:41 <goddess_ishtar> why are they yellow
18:36:55 <gwyd4016> Some grfs just make them look like that
18:37:30 <michi_cc> goddess_ishtar: Signal graphics of DBSetXL, which I had in my config randomly
18:39:06 <goddess_ishtar> maybe not the best choice to represent differences from vanilla
18:39:19 <Rubidium> gwyd4016: in the real world those signals that automatically turn green and then might turn red when a train is coming from the other side, won't be encountered by the train as the path onto there cannot be made by a higher level interlocking system. Actually the "automatic green" signals are technically extremely simple
18:42:20 <Rubidium> ... and primitive. In the Netherlands that's literally 50s/60s technology. A current is passed through an isolated block of the track, if that current goes away (short circuit by train axles, loose wire, power outage, ...) then it'll automatically fall into the red/danger state (by our friend gravity)
18:42:37 <michi_cc> Rubidium: The backwards signals are actually more complex than it might seem. Green path signals with a reservation can't be turned red automatically, because then another train that enters a block might turn it back to red if timed right, and the original train gets stuck.
18:43:33 <michi_cc> And as the map doesn't store the direction of the reservation, you can't really decide if it is okay to change the signal or not. Or the signal code would have to make an additional vehicle search every time.
18:44:20 <michi_cc> It's actually easier to just turn them red when the reservation is made. And I really dislike leaving them green all the time even if the block really is occupied.
18:46:22 <Rubidium> michi_cc: that's why I'm saying simpler is better than realistic, especially as realistic (for the NL) isn't quite what you would expect if you don't know how it works physically
18:47:12 <michi_cc> Oh, I thought you meant the code with the signal state reverting on a failed reservation.
19:02:44 *** kuhnovic has joined #openttd
19:02:44 <kuhnovic> andythenorthviaGitHu: We're doing clickbait PR titles now? 😆
19:03:00 <kuhnovic> For the record: it did work
19:34:08 <truebrain> _jgr_: It always annoyed me how slow it could be, but never bothered to figure out why 😄 Tnx a lot 🙂
19:35:56 <_jgr_> Not to worry. I'd been using it on some big savegames and scrolling past somf of the chunks was getting a bit painful 😛
19:36:20 <truebrain> 😄 If you have more fixes or improvements, they are welcome!
19:36:26 <truebrain> (context: OpenTTD savegame reader)
19:36:34 <truebrain> I realised it doesn't broadcast in this channel for PRs 😄
19:52:36 *** Wormnest has quit IRC (Quit: Leaving)
19:54:48 *** Wormnest has joined #openttd
20:18:03 *** Leopold_ has joined #openttd
20:21:51 *** Leopold_ has quit IRC (Remote host closed the connection)
20:28:54 *** nielsm has quit IRC (Ping timeout: 480 seconds)
20:48:49 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
21:09:14 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
21:34:55 <andythenorth> was it merged yet? 😛
21:38:59 *** Flygon has quit IRC (Read error: Connection reset by peer)
21:43:47 *** ChanServ sets mode: +v tokai
21:45:36 <reldred> Not that fast then eh
21:46:40 <reldred> Maybe openttdcoop will start using them after this
21:50:43 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
23:01:43 *** Wormnest has quit IRC (Quit: Leaving)
23:25:09 <locosage> reldred: only if they actually get fixed :p
23:25:43 <reldred> Tf is wrong with path signals anyway!
23:26:06 <locosage> slow switching and blocking bug
23:27:26 <reldred> sounds like fake news to me
23:27:48 <locosage> there is a setting and a bug for that so 🤷♂️
23:29:36 <reldred> There’s a setting to add fake news? nice
23:29:49 <locosage> there is a setting for switching speed
23:29:59 <locosage> problem is on lowest it's still slower than block
23:33:32 <_jgr_> It's really a polling interval. At the minimum value it's every tick, that ought to be fast enough.
23:34:06 <locosage> same thing, more often they poll the faster they switch
23:34:21 <locosage> at least in theory xD
23:35:38 <locosage> and I tested it with 1 tick interval, it's still noticeably slower than block signals
continue to next day ⏵