Config
Log for #openttd on 17th October 2021:
Times are UTC Toggle Colours
01:27:28  *** snail_UES_ has joined #openttd
01:31:51  *** Wormnest has quit IRC
01:37:43  *** snail_UES_ has quit IRC
02:49:24  *** debdog has joined #openttd
02:52:46  *** D-HUND has quit IRC
03:00:07  *** snail_UES_ has joined #openttd
03:13:25  *** glx has quit IRC
03:18:30  *** Flygon has joined #openttd
03:40:46  *** snail_UES_ has quit IRC
05:16:28  *** WormnestAndroid has quit IRC
06:35:39  *** andythenorth has joined #openttd
06:45:47  <andythenorth> yo
07:08:28  <andythenorth> uuuf
07:08:36  <andythenorth> Iron Horse compile takes 45 seconds :(
07:08:41  * andythenorth should stop adding trains to it
07:13:40  *** Wolf01 has joined #openttd
08:11:01  <TrueBrain> https://github.com/OpenTTD/OpenTTD/pull/9603 <- ready for review, if anyone pleases
08:11:19  <TrueBrain> https://github.com/OpenTTD/website/pull/226 <- similar
08:15:37  <TrueBrain> fun stats: it takes the coordinator 180ms to send out all the servers in the serverlisting :D That could/should be optimized ;)
09:04:10  <DorpsGek> [OpenTTD/py-protocol] TrueBrain opened pull request #21: Fix: use add_trace_field() over add_context() for tracer https://git.io/JKaOC
09:05:20  *** Gustavo6046 has quit IRC
09:05:48  <DorpsGek> [OpenTTD/py-protocol] TrueBrain merged pull request #21: Fix: use add_trace_field() over add_context() for tracer https://git.io/JKaOC
09:06:01  <DorpsGek> [OpenTTD/py-protocol] TrueBrain created new tag: 1.4.1 https://git.io/JKa3D
09:07:27  <DorpsGek> [OpenTTD/game-coordinator] TrueBrain opened pull request #73: Fix: sample traces per command, not globally https://git.io/JKasS
09:09:38  <DorpsGek> [OpenTTD/game-coordinator] TrueBrain merged pull request #73: Fix: sample traces per command, not globally https://git.io/JKasS
09:11:03  *** frosch123 has joined #openttd
09:11:32  *** tokai has joined #openttd
09:11:32  *** ChanServ sets mode: +v tokai
09:17:59  <DorpsGek> [OpenTTD/game-coordinator] TrueBrain created new tag: 1.5.1 https://git.io/JKa8R
09:18:30  *** tokai|noir has quit IRC
09:18:37  <TrueBrain> right, lets see if this way of tracing gives a better overview :)
09:18:55  <TrueBrain> frosch123: I like to believe you joined to approve https://github.com/OpenTTD/OpenTTD/pull/9603 and https://github.com/OpenTTD/website/pull/226 ? :D
09:23:09  *** HerzogDeXtEr has joined #openttd
09:23:18  <frosch123> is that the original titlegame submission, or the updated one?
09:24:02  <frosch123> https://www.tt-forums.net/viewtopic.php?p=1248838#p1248838
09:24:42  <DorpsGek> [OpenTTD/website] frosch123 approved pull request #226: Add: 12.0 announcement https://git.io/JKauG
09:25:00  <TrueBrain> frosch123: as I wrote in the PR: "Used the latest from the forum-thread, as requested by the winner."
09:25:35  <TrueBrain> (so yes, the updated one :D)
09:25:39  <frosch123> aw, reading :p
09:26:11  <TrueBrain> :D :D :D
09:26:18  <DorpsGek> [OpenTTD/OpenTTD] frosch123 approved pull request #9603: Doc: Prepare for 12.0 release & title game https://git.io/JKazS
09:26:28  <frosch123> i assume the other things in the changelog were backported before :)
09:26:38  <TrueBrain> yeah
09:26:48  <TrueBrain> I even scripted it! :P
09:27:47  <TrueBrain> tnx frosch123  :) Lets kick 12.0 out of the door .. finally :P
09:28:01  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #9603: Doc: Prepare for 12.0 release & title game https://git.io/J2bDo
09:28:16  <frosch123> why does dorpsgek spam info@ so much recently :/
09:28:19  *** Samu has joined #openttd
09:30:24  <TrueBrain> DorpsGek? Or GitHub? :P
09:30:30  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain created new tag: 12.0 https://git.io/JKaVH
09:30:44  <TrueBrain> and because I fucked something up, and I still don't understand what :P
09:30:58  <DorpsGek> [OpenTTD/website] TrueBrain merged pull request #226: Add: 12.0 announcement https://git.io/J2NEm
09:30:59  <frosch123> you got fan mail btw :p
09:31:09  <TrueBrain> owh boy
09:31:29  <TrueBrain> the acceleration thing?
09:31:36  <TrueBrain> I read it, it made me giggle :)
09:32:04  <frosch123> the bottom part of the compliment sandwich is missing though
09:32:06  <TrueBrain> hmm .. if Cloudflare R2 hurries up, we can ditch the OVH VPSes .. :)
09:36:40  <TrueBrain> right, lets prepare the socials ..
09:40:26  <TrueBrain> 10 million unique impressions on Steam news
09:40:27  <TrueBrain> wow
09:40:40  <TrueBrain> 250k unique users
09:56:15  <TrueBrain> and of course it is now waiting for the Mac build ... lol
10:00:07  <DorpsGek> [OpenTTD/website] TrueBrain created new tag: 1.4.37 https://git.io/JKahU
10:12:22  <TrueBrain> @topic set 1 12.0
10:12:22  *** DorpsGek changes topic to "12.0 | Website: *.openttd.org (source: github, translator: translator, server list: servers, wiki: wiki) | Don't ask to ask, just ask | 'Latest' is not a valid version, 'Most recent' neither | English only"
10:13:19  <TrueBrain> okay, I think I did everything
10:13:21  <TrueBrain> I hope ..
10:14:16  <TrueBrain> GOG works, Steam works .. news posts, socials done ..
10:15:08  <TrueBrain> now we wait and see what the Game Coordinator does :P
10:20:39  <Eddi|zuHause> what's a game coordinator?
10:20:48  <TrueBrain> still can't believe 29 languages are translated in full .. :)
10:26:41  <frosch123> still, some common ones are missing, like italian
10:28:00  <frosch123> hmm, also, steam reviews are back to overwhelmingly positive
10:28:16  <frosch123> i expected it would drop, when all the fanboys were done with reviewing
10:29:12  <frosch123> oh dear, all the reviews are memes
10:33:37  *** virtualrandomnumber has joined #openttd
10:34:05  *** virtualrandomnumber has quit IRC
10:45:23  *** sla_ro|master has joined #openttd
10:51:13  <TrueBrain> Eddi|zuHause: https://github.com/OpenTTD/game-coordinator
10:51:22  <TrueBrain> frosch123: yeah, lot of reviews are .. lo
10:54:04  <TrueBrain> we will see if 12.0 removes a few of those negative reviews :P
10:54:17  <TrueBrain> next up is the tutorial part of the game, to get even better reviews :)
10:56:57  <frosch123> tutorials are a maintenance hell... some guy just made a 5 part yt series on signals, but at least the first two are already outdated, because 12.0 hides block signals for newbies
10:57:08  <TrueBrain> yeah
10:57:37  <TrueBrain> same for GS-assisted tutorials
10:58:02  <frosch123> and i don't think that is just bad luck. it's more like: tutorials are made about complicated guis, but complicated guis are also more likely to be changed...
10:58:32  <TrueBrain> so first, we need to revise our GUIs, you say? :)
10:59:07  <frosch123> haha, good excuse to never talk about tutorials again :)
10:59:55  <TrueBrain> :D
11:00:12  <frosch123> gui reviews are doomed though. whenever zorg enters the discussion, it's dead :p
11:01:04  <TrueBrain> I have a button that fixes that :P
11:04:10  <frosch123> i wanted to add the factorio train order gui to tyler's discussion, but when zorg joined, i dropped the effort :)
11:04:39  <TrueBrain> please do add to the discussion; I rather ban Zorg from GitHub ;)
11:09:30  <TrueBrain> can't believe 2TallTyler didn't add Factorio btw :P
11:11:20  *** gelignite has joined #openttd
11:35:51  <dP> is there some way to force turn connection to direct servers?
11:35:55  <dP> to check latency
11:38:34  *** EER has joined #openttd
12:13:43  <LordAro> \o/
12:13:46  <LordAro> well done all
12:14:03  <LordAro> first non-04/01 release in how many years? :)
12:15:19  <frosch123> 1.2.0-RC4 (2012-04-01)
12:15:31  <frosch123> less than 10 :p
12:15:48  <peter1138> Hah, title-save changes colours during start up... I guess the livery schemes flag is loaded later...
12:17:13  <frosch123> client settings affecting the titlegame are always weird :p
12:19:21  <LordAro> do we need to be more careful about restarting the GC now too?
12:20:19  <frosch123> it kicks all turn connections, doesn't it?
12:20:59  <frosch123> should not affect our 99.99% SLA
12:21:06  <LordAro> haha
12:21:29  <LordAro> @calc 0.01*365*24*60
12:21:29  <DorpsGek> LordAro: 5256
12:21:40  <LordAro> @calc 0.0001*365*24*60
12:21:40  <DorpsGek> LordAro: 52.56000000000001
12:23:28  <Wolf01> Yay, finally map scrolling in title screen
12:28:09  <LordAro> scrolling isn't super smooth, it has to be said
12:28:19  <LordAro> need that subpixel movement
12:30:31  <peter1138> timing innit
12:31:28  <dP> especially debug builds lag quite a lot in the main menu now
12:31:44  * andythenorth turned away for like 2 mins and there's loads of chat
12:31:49  <LordAro> ah yeah, debug build probably isn't helping
12:32:00  <andythenorth> oh it was more like 5 hours
12:32:08  <andythenorth> that's 5 hours of sorting out Lego parts :P
12:32:09  <andythenorth> oof
12:32:15  * andythenorth having a clearout
12:32:32  <Wolf01> <andythenorth> that's 5 hours of sorting out Lego parts :P <- I can relate
12:34:42  <andythenorth> got about 1kg to sell
12:34:54  <LordAro> "Return signal window how it was. Don't think everyone are fools."
12:35:20  <Wolf01> I think you'll regret it, build a home expansion instead :P
12:39:27  <LordAro> it is a bit strange that the signal window settings under company, rather than construction
12:39:34  <LordAro> was there any rationale for that?
12:40:25  <Wolf01> Also orders settings are there
12:40:50  <Wolf01> And service intervals
12:41:26  <Wolf01> Hmmm, the only setting which should belong to company is the company colour
12:41:57  <LordAro> hmm
12:42:00  <dP> only settings that should belong to company are company settings :p
12:42:14  <LordAro> orders/servicing doesn't really belong in construction either :p
12:42:17  <dP> meaning SF_PER_COMPANY
12:42:50  <andythenorth> related: Zorg (LC) has given feedback which caused GarryG to publicly abandon his Aus industries set https://www.tt-forums.net/viewtopic.php?p=1248866#p1248866
12:44:47  <Wolf01> Meh
12:45:24  <frosch123> was aus industries development active in the last 2 years?
12:45:43  <frosch123> garry is one of the top-scorers who have not migrated their bananas content
12:45:49  <frosch123> i.e. no upload since two years
12:46:59  <Wolf01> Hah! I can get profit with ships, strange
12:47:05  <peter1138> Wolf01, bug
12:48:14  <Wolf01> It's the first time I use ships in small maps (64x64), I didn't expect to make profit with a 20 tiles route
12:50:37  <LordAro> https://pasteboard.co/gRg6cK65u5ZP.png
12:50:38  <LordAro> uh.
12:51:30  <LordAro> https://pasteboard.co/x3BkQPg0XpA7.png also, some additional characters that should be ignored when sorting
12:51:37  <andythenorth> oh that again
12:51:45  <andythenorth> that came up in reddit too
12:52:35  <andythenorth> not massively a fan of attempted bullying of contribs
12:53:05  <andythenorth> oh fucking block signals
12:53:09  <andythenorth> still going on
12:53:16  * andythenorth can't be arsed with this community
12:53:18  *** andythenorth has quit IRC
12:53:30  <LordAro> ..
12:54:40  <frosch123> LordAro: are you using the sprite font? without cjk chars?
12:54:47  <Wolf01> Hmm, the newgrf settings didn't upgrade one grf (ISR-style docks 1.3 -> 1.4)
12:54:50  <LordAro> yeah
12:55:13  <frosch123> Wolf01: grfid changed?
12:55:14  <Wolf01> Also the ISR objects
12:55:32  <Wolf01> No, if I try to add it again it says "duplicate"
12:57:22  <Wolf01> City objects too
12:57:29  <Wolf01> Wtf
13:03:03  <Eddi|zuHause> did they properly update the a14 version?
13:16:30  *** ZDaemon is now known as Cyberdemon
13:26:37  *** glx has joined #openttd
13:26:37  *** ChanServ sets mode: +v glx
13:28:11  <glx> maybe version is changed on bananas only and not inside the newgrf (some GS/AI have the same issue)
13:41:46  *** WormnestAndroid has joined #openttd
14:00:24  *** EER has quit IRC
14:19:16  <TrueBrain> ... so that server entry .. what a wankers are some people .. guess it is time to find out how to IP ban :)
14:20:07  <TrueBrain> IPs are hidden ofc, lolz .. eeeuuhhh .. hmm
14:20:22  <dP> not that hidden but it's a dynamic ip
14:20:45  <TrueBrain> it is also funny to see that people do not understand invite-only vs public :D Haha .. the server-list will be very annoying soon enough ..
14:20:50  <TrueBrain> guess we need to give that some attention :)
14:23:49  <frosch123> lol, i think i found the servername you meant
14:23:54  <frosch123> is it about signal gui?
14:24:05  <TrueBrain> yeah .. server will be gone in the next few minutes :)
14:24:19  <TrueBrain> if you get personal, I get personal, it is that simple :)
14:26:36  <TrueBrain> and .. its gone
14:26:42  <LordAro> :)
14:27:04  <TrueBrain> really, so sad this is needed. People can be real shitheads
14:27:16  <LordAro> "Server address: +ABC" does seem a bit like of a bug
14:27:33  <LordAro> or are they really all STUN/TURN servers?
14:27:34  <TrueBrain> link?
14:27:48  <LordAro> TrueBrain: in the server list window
14:27:54  <TrueBrain> or you mean that all 12.0 servers are like that?
14:27:59  <TrueBrain> all 12+ servers are listed by invite-code
14:28:10  <TrueBrain> we cannot predict beforehand if a user can connect to the IP, even if we can :)
14:28:27  <LordAro> mm
14:28:45  <TrueBrain> also, this makes things a lot simpler, as everything uses the same flow
14:28:46  <LordAro> but we know whether it's an address or not though, could be "Invite code: +ABC.." instead?
14:29:11  <TrueBrain> good suggestion; please make a PR in https://github.com/OpenTTD/master-server-web :D
14:29:40  <TrueBrain> owh, wait, web does it correct
14:29:42  <LordAro> i'm looking at OTTD itself :p
14:29:42  <TrueBrain> guess the client you mean?
14:29:43  <TrueBrain> well, in that case you know where to find the code :D :D
14:31:13  <LordAro> adding new strings to releases is a pain though ;)
14:31:32  <TrueBrain> it is
14:31:33  <TrueBrain> and you should have tested the betas/RCs :P :P :P
14:31:39  <LordAro> :<
14:32:07  <TrueBrain> haha :)
14:32:59  <TrueBrain> "The moving background on the title screen is making me nauseous. Seriously. "
14:33:15  <TrueBrain> I do not disagree
14:34:13  <DorpsGek> [OpenTTD/OpenTTD] YouYouTheBoxx opened issue #9614: [Bug]:  https://git.io/JKokN
14:36:37  <frosch123> [16:26] <TrueBrain> and .. its gone <- how long does it cache?
14:36:56  <TrueBrain> frosch123: 5 minutes
14:37:12  <frosch123> it's still listed in-game for me
14:37:54  <TrueBrain> hmm, here too. That is odd .. let me think why this is the case ..
14:42:43  <TrueBrain> I think this message didn't propagate to all game coordinator instances or something .. bit odd :)
14:43:11  <TrueBrain> I simply blocked the IP, so the TCP connection times out after a while
14:43:30  <TrueBrain> after which the normal disconnect stuff should kick in
14:43:30  <TrueBrain> the entry is gone from the web .. so something did what it should be doing
14:44:27  <TrueBrain> frosch123: seems gone now?
14:45:31  <frosch123> yes
14:45:58  <TrueBrain> good enough for now :)
14:47:30  <TrueBrain> so far still ~15% are TURN connection .. so that seems to be working as expected
14:49:57  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #9614: [Bug]: refresh rate doesn't work with vsync https://git.io/JKokN
14:57:06  *** Wormnest has joined #openttd
15:03:48  *** Limyx826 has joined #openttd
15:24:04  <TrueBrain> what we could do next time when we make a change like disabling signals by default, is to make a popup when people start the game the first time after upgrading
15:24:05  <TrueBrain> asking what they want
15:24:13  <TrueBrain> that would have solved this whole blabla instantly :)
15:24:39  <TrueBrain> (I was thinking about how to replace the Order GUI, and realised that what-ever we do, people will bitch about it a lot .. realising a good solution might be to allow users to switch between the two)
15:25:40  <frosch123> i think you try too hard to please everyone :)
15:25:52  <TrueBrain> clearly I don't :P :P
15:26:21  *** esselfe has joined #openttd
15:26:33  <TrueBrain> I still think TrueTTD will be the way forward for changing the GUI :)
15:27:01  <LordAro> oh no
15:27:07  <FLHerne> Maybe the signal toolbar itself could have a 'show advanced signals' toggle button
15:27:11  <frosch123> anyway, game presets... i still love my 4 types of players theory. so on first start we could make an interview about the player personality and then select a default preset
15:27:24  <TrueBrain> that works too :)
15:27:38  <FLHerne> That would make it clear to new players that they're not required, without outright hiding them
15:28:15  <frosch123> but yes, gui settings are sometime better in the window they affect, than in some other settings window
15:28:18  <FLHerne> Problem with the order GUI is that then every new order feature will need adding to both of them
15:28:29  <FLHerne> and the old one will get increasingly more insane
15:28:40  <TrueBrain> yup
15:28:42  <FLHerne> of course, JGRPP players are quite happy with the insane version
15:28:56  <TrueBrain> but those are a very niche group of people :)
15:28:59  <FLHerne> (have you *seen* the JGR order window? It really is quite amazing)
15:29:01  <dP> I thought of doing similar stuff for cmclient, asking on a first run if they want vanilla-like config or full bananas
15:29:10  <dP> but in vanilla options aren't as clear
15:29:49  <TrueBrain> FLHerne: maybe in general this could work, that we move a lot of settings inside the window they impact
15:30:03  <TrueBrain> with a generic icon on every window that has settings
15:30:28  <frosch123> 3 vertical dots :)
15:31:20  <frosch123> 3 vertical dots that show a dropdown menu with rare settings
15:31:39  <frosch123> that would also work for "show hidden engines" in the purchase window
15:31:45  <frosch123> and other rare buttons that have to be somewehre
15:31:54  <TrueBrain> if we do the presets stuff, we can already move a lot of settings to presets
15:32:12  <TrueBrain> things like "max bridge length" etc could be in the construction toolbar I guess?
15:32:23  <frosch123> well, there is some advantage in having all settings in one window: the option to filter "different from default"
15:32:28  <TrueBrain> vehicle never expire in buy new vehicle window?
15:32:34  <frosch123> that helps in communicating your preferrences with others
15:32:48  <frosch123> TrueBrain: no, that is a game setting
15:32:55  <TrueBrain> I guess we could keep the settings window .. just make it less required :)
15:33:00  <frosch123> you have to decide it on game start, technically
15:33:07  <TrueBrain> ah, that would make more sense
15:33:21  <TrueBrain> currently you can change it in-game too :P
15:33:31  <dP> max bridge length is a game configuration setting, not construction/ui one
15:33:38  <dP> similarly with some others
15:33:40  <frosch123> yes, and for the longest time you had to open the console to type resetengines then :p
15:33:55  <TrueBrain> so guess first we need to do presets, and clear that mess up
15:34:05  <TrueBrain> so it is much more obvious which settings should be done BEFORE a game :P
15:35:02  <TrueBrain> lol, you can enable/disable inflation in-game?
15:35:06  <TrueBrain> I clearly never look at settings :D
15:36:14  <frosch123> what you can change in game is usually very technical, and not intuitive
15:36:39  <frosch123> there are plenty of settings which you can change in singleplayer, but not in multiplayer because we do not support reloading newgrf in the middle of multiplayer games
15:36:56  <TrueBrain> it also feels very niche, what you can change
15:37:09  <TrueBrain> like inflation .. the only reason I can think of to change it in-game, is if you forgot to disable it before
15:37:18  <TrueBrain> but you are not going to enable/disable it during the game multiple times
15:37:22  <TrueBrain> that just makes .. very little sense
15:37:33  <frosch123> it's pretty much: you can change everything, unless it is knows to break something
15:38:04  <frosch123> "if it doesn't make sense, just don't do it"
15:38:16  <TrueBrain> the 1995 way of doing this, I guess
15:38:43  <TrueBrain> I wonder what happens if we just revert it: you cannot change those things in-game, unless you do a magic dance first, or what-ever :P
15:38:43  <Wolf01> <TrueBrain> but you are not going to enable/disable it during the game multiple times <- like me
15:39:08  <TrueBrain> Wolf01: you enable/disable inflation multiple times during a game?
15:39:11  <TrueBrain> why? :)
15:39:30  <dP> when you have a server with players it's really handy that you can change most settings without kicking them all out
15:39:48  <dP> and resetting everything they done
15:40:16  <TrueBrain> one thing OpenTTD is a star in, confusing the niche use-cases for the generic use-cases, and make GUIs based on the first group, not the latter :)
15:40:46  <Wolf01> I start with it enabled, then after some years I disable it, then I enable it again when I have too much money
15:40:48  * TrueBrain looks at the Order GUI :P
15:41:07  <TrueBrain> Wolf01: lol, yeah, that is fair.
15:41:27  *** ioangogo_ has joined #openttd
15:42:08  <Wolf01> Or I'll keep it enabled directly, like breakdowns
15:42:20  <frosch123> yes, i also read about people enabling/disabling inflation. it's usually to somehow attach inflation to technological progression
15:42:36  <Wolf01> I disable inflation oonly when I plan to start like in year 2000
15:42:37  <TrueBrain> yeah, a broken functionality used to compensate for something else :)
15:43:03  <Wolf01> Because you can't start at all in year 2000 with inflation
15:43:14  <TrueBrain> haha, no, it is that broken for sure :)
15:43:48  <TrueBrain> but I guess that is mostly what makes the settings GUI annoying .. it is a mix of truly useful settings to change in-game, and those that you only want to change in rare occasions :)
15:44:47  *** ioangogo has quit IRC
15:45:11  <DorpsGek> [OpenTTD/OpenTTD] LordAro opened pull request #9615: Change: Don't use 'server address' string in server list when displaying an invite code https://git.io/JKoDT
15:45:22  <TrueBrain> :o he actually did it! :P
15:45:45  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #9615: Change: Don't use 'server address' string in server list when displaying an invite code https://git.io/JKoDB
15:45:54  <TrueBrain> I love how we can have those itches we need to scratch :D
15:45:55  <LordAro> fastest approval evar
15:46:12  *** Gustavo6046 has joined #openttd
15:46:53  <TrueBrain> it is just a good PR
15:47:11  <dP> oh, btw, with all that settings mess it's still missing gui for font selection and hotkeys xD
15:47:33  <TrueBrain> so why didn't you make a PR for that yet?
15:47:34  <TrueBrain> tssk
15:47:36  <LordAro> font selection is Hard(tm)
15:47:39  <LordAro> hotkeys less so
15:47:47  <dP> TrueBrain, E_NOT_ENOUGH_TIME
15:47:53  <TrueBrain> hotkeys is just evil ... like I even didn't know you could remap them
15:48:25  <dP> hotkeys are kinda hard too actually
15:48:43  <dP> there are many nuances like remapping modifier keys
15:49:03  <dP> and keycode/charcode mess
15:50:32  <frosch123> the stupid part about hotkeys is that you cannot remap map scrolling :p
15:51:58  <frosch123> i had another idea about presets: we could make a survey website, you answer some questions about yourself (i played the original, i played ttdp, i play only on steam...) and your playstyle, and then you upload your openttd.cfg
15:52:22  <frosch123> and then we try to correlate the playstyles with the settings
15:52:56  <frosch123> though i suspect we only get random data, and can't conclude anything :p
15:53:49  <dP> I bet many don't even realize there are some setting they could've used
15:54:18  <dP> at least I'm still finding those from time to time xD
15:54:34  <Wolf01> Me too
15:55:47  <DorpsGek> [OpenTTD/OpenTTD] Kuhnovic opened pull request #9616: Feature: Railways can be created "through" bridges, tunnels, stations… https://git.io/JKo5u
15:56:26  <Wolf01> Makes sense
15:56:57  <Wolf01> I never noticed it wasn't like this before
15:57:24  <TrueBrain> Okay, that PR sounds cool
15:57:42  <LordAro> it does
15:57:50  *** Compu has joined #openttd
15:58:09  *** Compu has quit IRC
16:01:50  <glx> <frosch123> well, there is some advantage in having all settings in one window: the option to filter "different from default" <-- having the option to change a setting in the window it affects, doesn't force us to remove it from the settings window
16:03:08  <frosch123> yes, i agree. settings should be in multiple places
16:03:41  <dP> it sounds cool but I never noticed that to be a problem as well
16:04:07  <dP> though i guess I don't use autorail that much
16:06:42  <dP> btw, add preview for #9616
16:08:28  <frosch123> oh, i totally noticed it before. but i got trained to drag "correctly" and thus no longer encountered it
16:10:52  <glx> TrueBrain: preview is broken ;)
16:11:14  <TrueBrain> So fix it! :p
16:11:16  <dP> ideally autorail should just build bridges :p
16:12:44  <frosch123> just let an ai  play for you
16:12:46  <dP> also code is kinda meh recomputing highlight on draw instead of update
16:17:43  <dP> oh god, it's even worse, it recomputes whole selection for every tile just to check that tile is in it
16:21:16  <LordAro> "File tools/settings.py is read-only; trying to patch anyway"
16:21:18  <LordAro> curious.
16:24:09  <TrueBrain> Expecially as the same code runs on CI which works fine
16:26:24  <glx> options: -u 1001:121 <-- there's a difference between CI and preview
16:26:51  *** nielsm has joined #openttd
16:39:12  <TrueBrain> Nicely spotted! I missed that several lookovers.. so that makes the fix easy ;)
16:52:57  *** Flygon has quit IRC
17:01:13  <DorpsGek> [OpenTTD/OpenTTD] LordAro opened pull request #9617: Feature: Button to toggle showing advanced signal types https://git.io/JKKrm
17:01:18  <LordAro> well that was easy.
17:05:34  <frosch123> what happens if block signals are selected as active signal, and then you hide them?
17:06:09  <glx> same as changing the setting I guess
17:06:18  * LordAro hurriedly recompiles to check
17:08:06  <dP> I don't rly like the titles on toolbar windows in general though
17:08:35  <dP> I don't even know any other game that does that
17:09:31  *** Gustavo6046_ has joined #openttd
17:10:22  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9617: Feature: Button to toggle showing advanced signal types https://git.io/JKK1a
17:10:45  <LordAro> frosch123: answer: block signals are still selected
17:10:49  <LordAro> no other 'issues'
17:11:39  <DorpsGek> [OpenTTD/OpenTTD] LordAro updated pull request #9617: Feature: Button to toggle showing advanced signal types https://git.io/JKKrm
17:12:07  <LordAro> and as glx says, same as if you change the setting with a signal selected (though admittedly that was a bit harder to do)
17:12:21  <frosch123> hmm, i guess the widgets are just hidden. i was worried there would be a chance for nullptr deref, when a non-existing widget is lowered
17:13:05  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #9617: Feature: Button to toggle showing advanced signal types https://git.io/JKKDu
17:13:09  <LordAro> yeah, window is not that fancy
17:13:33  <TrueBrain> I don't like the icon doesn't change depending on setting, but not for now :p
17:13:43  *** Gustavo6046 has quit IRC
17:13:43  *** Gustavo6046_ is now known as Gustavo6046
17:13:46  <LordAro> TrueBrain: yeah, i was going to do that, but then i checked the finance window
17:13:50  <LordAro> it doesn't change either
17:13:53  <LordAro> so ¯\_(ツ)_/¯
17:14:13  <TrueBrain> Exactly why I say not for now ;)
17:14:25  <TrueBrain> Consistency before correctness :D
17:14:29  <DorpsGek> [OpenTTD/OpenTTD] LordAro merged pull request #9615: Change: Don't use 'server address' string in server list when displaying an invite code https://git.io/JKoDT
17:14:32  <LordAro> damn right
17:14:55  <frosch123> oh, i have a bad idea: ctrl+click that icon to hide the stupid semaphores :p
17:15:04  <LordAro> oho
17:15:11  <LordAro> bigger change
17:15:32  <frosch123> in case you wondered, that was no suggestion to add to that PR :)
17:17:37  <dP> btw, by the same logic as block signals shouldn't semaphores be hidden as well by default?
17:17:59  <LordAro> see the original PR
17:19:24  *** Gustavo6046 has quit IRC
17:21:21  *** Gustavo6046 has joined #openttd
17:25:30  <DorpsGek> [OpenTTD/OpenTTD] LordAro opened pull request #9618: Fix #9614: Refresh rate dropdown was still active when vsync was enabled https://git.io/JKKdA
17:25:59  <TrueBrain> LordAro: did you test if you give a lower Hz than your display
17:26:04  <TrueBrain> it still uses your display Hz?
17:26:17  <TrueBrain> (so refresh-rate on 30, vsync on, does that give 60Hz?)
17:26:39  <LordAro> i didn't
17:26:50  <TrueBrain> that I think is the more important part of that bug ticket
17:27:36  <TrueBrain> I like how our server-list went from somewhat sane server-names to a complete shitshow in a day :)
17:28:23  <LordAro> TrueBrain: it does indeed still do 30
17:28:24  <LordAro> bleh.
17:30:16  <LordAro> TrueBrain: by 'use your display Hz', do you mean in the framerate window, or elsewhere?
17:30:28  <TrueBrain> framerate window
17:30:34  <TrueBrain> as I think if you put refresh-rate on 30
17:30:36  <TrueBrain> and you vsync
17:30:39  <TrueBrain> you only generate 30 fps
17:30:45  <TrueBrain> which kinda works
17:30:50  <TrueBrain> but it means disabling refresh-rate isn't ideal
17:31:02  <TrueBrain> just going above your display Hz doesn't have any effect
17:31:06  <TrueBrain> but I didn't validate this
17:31:13  <TrueBrain> just how I remembered it working :)
17:31:16  <LordAro> hmm
17:31:44  <TrueBrain> I think "vsync" should imply "no refresh rate restrictions" in our code
17:31:47  <TrueBrain> and let the vsync pump it
17:31:55  <TrueBrain> but that also needs testing for non-OpenGL :)
17:31:58  <LordAro> yeah
17:34:05  <dP> is there any point in allowing passworded public servers?
17:34:37  <dP> kinda feel like that should be hidden deep in settings somewhere
17:35:20  <frosch123> public servers are the only ones, where password makes sense
17:35:39  <dP> well, yeah, I guess
17:35:40  <frosch123> why have a password for invite-only or even private?
17:35:42  <dP> so hide password in general
17:35:56  *** andythenorth has joined #openttd
17:36:07  <frosch123> TrueBrain: it's funny, of all the 12.0 servers, there are only 3 dedicated ones :p
17:36:14  <andythenorth> seriously considering using Honeycomb to trace Iron Horse compile
17:36:17  <andythenorth> is that mad?
17:36:22  <TrueBrain> frosch123: yup; fully what I expected btw :)
17:36:28  <TrueBrain> andythenorth: yes, but fun too :)
17:36:57  <andythenorth> I was thinking 'is there a python tracing equivalent, with a good UI'
17:37:01  <andythenorth> then my brain engaged
17:37:18  <TrueBrain> frosch123: means the new network stuff is working .. people now clearly easier jump in a game with friends :) Most have 2 players active, also what I would expect. So that is looking well :)
17:37:30  <dP> is there any public statistics on the number of invite-only servers btw?
17:41:35  <LordAro> "Graphics frame rate: 998.12 frames/s"
17:41:38  <LordAro> well, i definitely unlocked it
17:45:09  <DorpsGek> [OpenTTD/OpenTTD] LordAro updated pull request #9618: Fix #9614: Refresh rate dropdown was still active when vsync was enabled https://git.io/JKKdA
17:48:34  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #9618: Fix #9614: Refresh rate dropdown was still active when vsync was enabled https://git.io/JK6LG
17:49:10  <DorpsGek> [OpenTTD/OpenTTD] Kuhnovic updated pull request #9616: Feature: Railways can be created "through" bridges, tunnels, stations… https://git.io/JKo5u
17:55:36  <DorpsGek> [OpenTTD/OpenTTD] Kuhnovic updated pull request #9616: Feature: Railways can be created "through" bridges, tunnels, stations… https://git.io/JKo5u
17:59:29  * andythenorth hopes this PR is good
17:59:41  <andythenorth> it scratches an itch
18:01:40  <LordAro> "Service subsidy offered: First 34,856 of <invalid cargo> service from (undefined string) to  will attract a 0 year subsidy from the local authority!"
18:01:44  <LordAro> i think i broke something
18:01:49  <TrueBrain> hahaha :D
18:01:57  <TrueBrain> its a feature!
18:02:11  <andythenorth> oof I need orange, but I use all the palette orange for magic graphics generation stuff
18:02:27  <dP> it's kinda hard to say what's the right way to do complex highlights like #9616 in vanilla actually
18:02:40  <dP> I created a whole new thing for stuff like that and it's still evolving
18:02:46  <dP> https://github.com/citymania-org/cmclient/blob/master/src/citymania/cm_highlight_type.hpp#L192
18:24:54  <DorpsGek> [OpenTTD/OpenTTD] LordAro merged pull request #9617: Feature: Button to toggle showing advanced signal types https://git.io/JKKrm
18:25:06  <DorpsGek> [OpenTTD/OpenTTD] LordAro merged pull request #9618: Fix #9614: Refresh rate dropdown was still active when vsync was enabled https://git.io/JKKdA
18:25:09  <DorpsGek> [OpenTTD/OpenTTD] LordAro closed issue #9614: [Bug]: refresh rate doesn't work with vsync https://git.io/JKokN
18:25:17  <TrueBrain> you go girl!
18:25:19  <DorpsGek> [OpenTTD/OpenTTD] Kuhnovic updated pull request #9616: Feature: Railways can be created "through" bridges, tunnels, stations… https://git.io/JKo5u
18:26:57  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9608: Change: Disable AI and Game Script when loading the title game https://git.io/JK6VU
18:28:33  <TrueBrain> I love how all the platforms are active about 12.0 .. tt-forums: NAH! I'm good ;)
18:29:31  <DorpsGek> [OpenTTD/OpenTTD] LordAro opened pull request #9619: Fix #9595: Always use plural forms of cargo for subsidy strings https://git.io/JK6wh
18:31:04  <andythenorth> forums is ...
18:31:07  <andythenorth> words
18:31:14  <LordAro> words and things
18:31:23  <andythenorth> 'not enjoyable like it used to be'
18:32:42  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9619: Fix #9595: Always use plural forms of cargo for subsidy strings https://git.io/JK6K4
18:33:04  <andythenorth> ha ha FML
18:33:06  <TrueBrain> LordAro: I have no clue what that PR actually does, so I cannot really review it :D
18:33:07  <andythenorth> the block signals
18:33:21  <TrueBrain> feels like a long time ago, someone had an itch?
18:33:21  <DorpsGek> [OpenTTD/OpenTTD] nielsmh commented on pull request #9608: Change: Disable AI and Game Script when loading the title game https://git.io/JK6KS
18:33:21  <andythenorth> apparently we are destroying servers
18:33:25  <DorpsGek> [OpenTTD/OpenTTD] nielsmh closed pull request #9608: Change: Disable AI and Game Script when loading the title game https://git.io/JVzdR
18:33:35  <andythenorth> PBS overloads servers, must use block
18:33:52  <TrueBrain> we should have removed block signals years ago
18:33:57  <TrueBrain> so many lies have been generated around them
18:34:09  <andythenorth> I am truly surprised by the level of hatred
18:34:21  <andythenorth> I have watched a 9 year old trying to learn the game signals
18:34:27  <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #9619: Fix #9595: Always use plural forms of cargo for subsidy strings https://git.io/JK66D
18:34:28  <andythenorth> it is fucking impossible
18:34:30  <LordAro> TrueBrain: my PR?
18:34:35  <TrueBrain> LordAro: yeah
18:34:48  <TrueBrain> andythenorth: it just shows how poorly informed many many people were
18:34:54  <TrueBrain> making it more clear that this was the right move :)
18:35:26  <nielsm> I think maybe it's the "PBS causes tiny delays in train movement because they don't turn green the same tick the path required becomes free, but only a little later" got shortened to "PBS has bad performance", and that got twisted into "PBS destroys server performance"
18:35:27  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #9619: Fix #9595: Always use plural forms of cargo for subsidy strings https://git.io/JK6iB
18:36:22  <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #9619: Fix #9595: Always use plural forms of cargo for subsidy strings https://git.io/JK6ih
18:36:31  <TrueBrain> andythenorth: where did you read that nonsense?
18:37:22  <andythenorth> https://www.reddit.com/r/openttd/comments/q9vyrz/comment/hh0dli4/?utm_source=share&utm_medium=web2x&context=3
18:37:24  <nielsm> "PBS destroys server performance" is a myth that keeps getting perpetuated
18:37:39  *** jottyfan has joined #openttd
18:37:42  <andythenorth> PBS is frigging annoying that it's always red
18:37:54  <andythenorth> I suspect that is actually the reason people like block, but they don't know how to explain it
18:38:00  <andythenorth> I don't mean advanced coop people
18:38:14  <andythenorth> I mean those people who do 'all block sections except at exits'
18:38:38  <nielsm> I still really realy want the signaling patches from JGR ported over
18:39:05  <andythenorth> the 'PBS appears to have latency' thing is quite real
18:39:09  <andythenorth> 'appears'
18:39:12  <nielsm> long reservations, routing restrictions, and braking updates
18:39:47  <TrueBrain> there, replied :P
18:39:59  *** jottyfan has quit IRC
18:40:11  <andythenorth> one chinese whisper at a time TB :P
18:40:43  <andythenorth> lol
18:40:54  * andythenorth back to drawing trains
18:42:33  <TrueBrain> lol, that reddit thread is funny .. some people make the weirdest comparisons :D
18:44:50  <LordAro> i kinda want to make a comment defending 2talltyler (they're just the ones that made the change, it was going to happen anyway), + advertising #9617
18:44:56  <LordAro> there's not really a good place to put it though
18:45:45  <andythenorth> it's irritating AF having to ship a crippled game because some self-declared elite players are too smart to find a setting
18:46:00  <andythenorth> and the idea that block is 'normal' is really lovely and retro
18:46:17  * andythenorth not very articulate, need more sleep
18:46:18  <TrueBrain> LordAro: seems others are fighting that "just made the change" already
18:51:40  <LordAro> i made a comment, buried deep in the downvoted comment chain where noone's looking anyway :)
18:52:25  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #9619: Fix #9595: Always use plural forms of cargo for subsidy strings https://git.io/JK67d
18:52:29  <TrueBrain> a real YOLO approval
18:52:51  <LordAro> no, a real YOLO approval would be approving the two_way_eol thing
18:52:53  <LordAro> :p
18:53:23  <TrueBrain> lol; especially as I know 0 about that part of the game
18:53:31  <TrueBrain> but from what I read from others, that is not a good idea :)
18:53:34  <LordAro> me neither
18:54:11  <LordAro> but afaict people that actually use those two way block signals in an advanced manner (and they are advanced signals now), seem to agree it's useful
18:56:51  <TrueBrain> some things in OpenTTD are settings, and I do not understand why :P
18:57:04  <dP> andythenorth, imo for coop players it's more about feeling that game moves farther from the playstyle they like it for than finding a setting
18:57:29  <andythenorth> I looked in my empathy cupboard and found it was bare
18:57:59  <andythenorth> :)
18:58:54  <TrueBrain> LordAro: maybe we should just revamp the whole signal / reverse / station / .. part of OpenTTD, as in, make it much more understandable what those settings etc mean
18:59:04  <TrueBrain> maybe a window with some visuals of what the settings change
18:59:13  <TrueBrain> with some text actually explaining what it implies
18:59:47  <TrueBrain> JGRPP has yellow signals too now, I believe ... no clue what they do. But that all mixes in with the same thing, I guess :)
19:00:21  <TrueBrain> similar for acceleration settings etc
19:00:26  <TrueBrain> just a small window with some track and a train
19:00:31  <TrueBrain> and when you change settings, you see how it impacts it
19:00:57  * andythenorth wonders if we could make PBS seem less broken
19:00:59  <dP> andythenorth, well having few attack assholes in the midst doesn't rly help their cause xD
19:01:14  <TrueBrain> andythenorth: the red-signal thing should be fixible
19:01:22  <andythenorth> dP civil discourse is always helpful
19:01:30  * andythenorth speaking as someone who sometimes fails at that
19:01:35  <andythenorth> :)
19:02:13  * andythenorth reading the hover help texts
19:02:28  <andythenorth> the PBS description ain't easy :)
19:03:07  <andythenorth> oh I like how the signal goes green when selected in the UI :)
19:03:17  * andythenorth hasn't seen the UI since we made it hideable
19:03:32  <andythenorth> is the signal UI hidden by default?
19:05:03  <LordAro> yes
19:05:19  <TrueBrain> over 1500 successful connections already today :)
19:09:54  <dP> also way more players on private servers than yesterday https://i.imgur.com/fThjLjt.png
19:10:01  <frosch123> TrueBrain: how is the "red" signal thing fixable?
19:10:14  <frosch123> what should green/red mean for pbs?
19:10:24  <nielsm> path signals can't show green when no path is set for them
19:10:41  <nielsm> because they can't know if a path is possible, except in some special cases
19:10:48  <nielsm> where they would be equivalent to a block signal
19:11:02  <TrueBrain> frosch123: if I understand it correctly (correct me if I don't), the train now breaks as it sees a red light, not? A tick after the PBS makes the reservation and goes green, right?
19:11:10  <dP> showing red when there is no path makes no more sense than green :p
19:12:08  <nielsm> if the path signals showed green by default and then turned red if a train tried reserving and failed, that would be signals turning red last-minute which is a huge no-no (don't make adversary signal changes during approach)
19:12:26  <andythenorth> I did try swapping to block in my current game
19:12:29  <andythenorth> it looks better
19:12:42  <frosch123> andythenorth: there is a grf which replaces the pbs signals to always show green
19:12:48  <andythenorth> ha
19:12:51  <frosch123> if you just dislike the color, use that
19:12:58  <andythenorth> I considered making one already :)
19:13:09  <frosch123> functionally the state of pbs pretty pointless, and means nothing
19:13:21  <andythenorth> PBS has absolutely no concept of 'block' yes/no?
19:13:29  <nielsm> another argument could be to have path signals being dark (no lights) when they have no path
19:13:32  <andythenorth> it's not a guard to the next section?
19:13:44  <nielsm> path signals don't have blocks, no
19:14:00  <nielsm> they allow trains wanting to pass to attempt to reserve a path to the next signal/end of line
19:14:16  <andythenorth> but we know how to detect trains in a block following a signal?
19:14:18  <nielsm> and if the train can reserve a path, then the signal turns green and lets the train pass
19:14:44  <nielsm> there are no blocks, there are track pieces, and track pieces can be reserved or not
19:14:45  <TrueBrain> frosch123: and if I understand that correctly, what could help for example is to poke the PBS signal a few ticks earlier
19:14:50  <frosch123> i guess pbs-green becomes meaningful with jgr yellow states
19:14:59  <andythenorth> but we know how to detect a block in the case of block signals?
19:15:08  <frosch123> TrueBrain: then go for the "yellow" state
19:15:13  <TrueBrain> sure
19:15:14  <andythenorth> I don't mean 'can PBS do it?' I mean we have the capability already
19:15:30  <TrueBrain> frosch123: as I said, should be fixable :)
19:15:42  <nielsm> yeah with JGR's long reservations (yellows) then trains reserve multiple signals ahead (if they need to, for their current speed)
19:15:49  <frosch123> yellow state mean (last time i checked, maybe i am completely wrong now), that paths are reserved over multiple pbs signals in advance
19:16:06  <frosch123> so the signals along the path which were not yet passed can show non-red
19:16:11  <nielsm> yellow signals are just one way to indicate it really...
19:16:18  <frosch123> no idea whether they should show green or yellow
19:16:31  <andythenorth> I guess reserving all the way to next junction / occupied block was tried and rejected?
19:16:49  <nielsm> in danish color light signals, the last signal before a stop would be showing a single green, and any signal with a green as next would be showing double green
19:16:49  <frosch123> andythenorth: jgr introduces "braking distance"
19:16:52  <TrueBrain> in general, there is a lot to improve in OpenTTD regarding signals :) At least to make it more clear to the player what is actually going on
19:17:00  <andythenorth> and look cooler :P
19:17:06  <andythenorth> more eye candy
19:17:29  <nielsm> yeah JGR has a trio of features there
19:17:46  <TrueBrain> I really like the idea of more realistic breaking :)
19:17:56  <nielsm> long reservations through PBS, braking distances, and routing restrictions programming on signals
19:17:59  <dP> how do yellow signals help to pack more trains onto the line? they won't be able to reserve more than one signal ahead anyway
19:18:22  <nielsm> the trains would be going slow too, then
19:18:24  <Rubidium> just get rid of signals and use ERTMS ;)
19:18:50  <TrueBrain> PR when? :P
19:19:48  <nielsm> braking distances feature in JGR's requires long reservations, because trains will no longer go faster than they'll be able to brake from in the distance available to them.
19:20:11  <nielsm> and long reservations somewhat require the routing restrictions programming of signals to have full features
19:20:13  <frosch123> nielsm: does that increase distance between trains?
19:20:27  <nielsm> yes, it means trains will need at least a full braking distance between them
19:20:38  <frosch123> in vanilla trains can follow each other with a gap of one signal block
19:20:47  <dP> that's way too much realism imo :P
19:20:49  <frosch123> does jgr enforce distance to be > braking disance?
19:20:49  <nielsm> yep, and brake on a dime
19:21:09  <TrueBrain> I like how this game slowly gets more and more realistic settings, mainly as more CPU becomes available .. can't imagine that being possible "back in the day" :P
19:21:18  <frosch123> haha, yeah, that's a problem :) it's really cool, but imagine the shitstorm :p
19:21:26  <TrueBrain> make it a setting
19:21:30  <TrueBrain> hide it somewhere
19:21:31  <nielsm> JGR makes trains brake if they are approaching a red signal, so they never go faster than they'll be able to brake from
19:21:34  <TrueBrain> change the default in 10 years
19:21:35  <TrueBrain> GO GO GO :)
19:21:56  <nielsm> so sure, you can have trains moving with two tiles between them
19:22:01  <TrueBrain> new disaster type: sleepy driver
19:22:01  <nielsm> they'll all be going 40 km/h
19:22:11  <michi_cc> TrueBrain: Trains should not brake if the path reservation can be successfully extended (if they do, then something got broken somewhere along the way). What does happen is that a train that waits at a path signal will not check for a good path each and every tick (because of exessive lag), in contrast to block signals that are refresh the tick a train leaves a block.
19:22:16  <frosch123> TrueBrain: we can make the settings-preset interview really offensive :p "i am an old fool, i want to play like in 95" :p
19:22:32  <TrueBrain> michi_cc: ah, tnx :)
19:22:54  <TrueBrain> so many cool projects to pick up with OpenTTD
19:23:00  <TrueBrain> as I really want to try a segment-based pathfinder :P
19:24:13  <michi_cc> You could for most simple track configurations probably trigger a path refresh on "signal update", but as path signals have no defined block boundaries, that will of course still only help on some situations.
19:25:02  <nielsm> one possible "fix" idea I have for the PBS latencies where trains waiting on a signal don't check every tick if they can get a path now, would be to have a PBS failing to reserve "poke" at the reserved tile they hit, somehow, and set a flag pointing at the signal or train, and then when that tile loses its reservation it can instantly signal back to the train that last poked the tile to check
19:25:03  <nielsm> again
19:25:10  <dP> I'm surprised it's still not segment-based
19:25:20  <michi_cc> But then as the average player dude is more like "I want to trains to use one station", it doesn't matter at all. It might matter for competitive goal servers, but again I'd guess you have some game knowledge if you play there.
19:25:46  <TrueBrain> nielsm: the hackish way, but that would work too :)
19:26:01  <TrueBrain> michi_cc: yeah, this is so far in the niche .. most players won't notice it either way :)
19:27:00  <nielsm> but anyway... the thing I keep going back to, the third feature that long reservations and braking distances also comes with: routing restrictions programming of signals
19:27:15  <TrueBrain> if you are such a fanboy, backport it already :P
19:27:17  <TrueBrain> :D :D :D
19:27:21  <nielsm> which might potentially obsolete all the (imo ugly) programming done with pre-signals
19:27:26  <michi_cc> So yes, path signals slow down trains if you build coop-like high performance, tick synchronized train networks. Which of course got shortened to: "path signals have bad performance".
19:28:10  <michi_cc> Just like the immortal myth of the path signals that are built into stations.
19:28:24  <TrueBrain> hahaha
19:28:31  <TrueBrain> we should make a myth-guide-to-OpenTTD :D
19:28:54  <nielsm> blog post?
19:29:09  <TrueBrain> would be the perfect moment
19:29:20  <TrueBrain> if any of you would be up for it, posting that in a few days would be ideal
19:29:43  <nielsm> I don't feel confident enough in the details of the problem to write a coherent blog about it
19:29:53  <TrueBrain> I have the same issue :P
19:30:46  <TrueBrain> can't help but to kinda laugh about the fact that this is the first release post on reddit with more than 20 replies :P
19:31:14  <michi_cc> I'm available in two weeks, but not before :) (Trip starting Thursday and a heap-load to do beforehand).
19:31:30  <dP> also most of those replies are hidden in one negative comment xD
19:31:49  <TrueBrain> michi_cc: well, a blog after that would still be interesting, just lost its momentum .. but I would still be all up for it :)
19:34:27  <michi_cc> Meanwhile, I have been writing some "nice" C++ that you'll "love", but that will have to wait till then as well.
19:34:47  <TrueBrain> .... oh-oh
19:34:54  <TrueBrain> what have you been up to .... :P
19:37:31  <michi_cc> Preview: "DoCommand<CMD_BUILD_SINGLE_RAIL>::Post(STR_ERROR_CAN_T_BUILD_RAILROAD_TRACK, CcPlaySound_CONSTRUCTION_RAIL, tile, _cur_railtype, TRACK_X, _settings_client.gui.auto_remove_signals);"
19:37:42  <michi_cc> Only makes sense if you know the current CommandProc definition :)
19:37:56  <michi_cc> Will get "interesting" for the network part, that is still open :p
19:38:02  <TrueBrain> oooooeeeehhhhhhh
19:38:24  <michi_cc> I.e. type-safe, none bit-stuffed command calls.
19:38:35  <TrueBrain> that would improve a lot of things :)
19:39:12  <michi_cc> And lot's of 'if constexpr' and args + ... + 0 :)
19:39:19  <dP> hell yeah xD
19:39:37  <TrueBrain> hopefully network would be easy too .. depends how you approached this, ofc :)
19:39:44  <TrueBrain> will be completely incompatible, but that is fine :)
19:39:57  <michi_cc> Actually, the most annyoing part will by desync debugging and replay.
19:40:21  <michi_cc> I know how I want to do the network part, I just have to figure out the proper magic c++ incantations :)
19:40:52  <TrueBrain> ghehe
19:41:03  <TrueBrain> can't wait to see how it looks
19:41:10  <TrueBrain> most likely I will sit in horror in a corner for a while
19:42:09  <nielsm> but will it let commands carry more data, perhaps in a more structured way?
19:42:14  <michi_cc> TrueBrain: The current worst part: https://gist.github.com/michicc/86e6b312c96b1d6566c3f556b90a1d53
19:42:16  <dP> guess I should wait with my advance into untackling that command mess: https://pastebin.com/RnLJY7sg
19:42:36  <michi_cc> nielsm: More data, absolutely, more structured, depends on the coding.
19:43:18  <nielsm> if it avoids some of all that terrible bitstuffing going on right now it's a huge win IMO
19:43:25  <andythenorth> wat :o
19:43:33  <andythenorth> the pathfinder is not segment based? :o
19:43:41  <andythenorth> ha ha andythenorth knows nothing :P
19:43:49  <andythenorth> is it tile based, not nodes and edges?
19:44:34  <frosch123> TrueBrain: cppcon starts in two weeks, want to visit? :p
19:44:38  <frosch123> *in one week
19:45:14  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #9620: Fix 3046a6ce: [Preview] building preview failed to patch LZMA https://git.io/JKi0z
19:45:43  <LordAro> isn't the bitstuffing at least partially for CPU cache improvements?
19:45:46  <TrueBrain> michi_cc: my C++ knowledge fails on that snippet completely :P
19:46:11  <michi_cc> It will make OTTD firmly C++17 (hello XP :D)
19:46:13  <LordAro> TrueBrain: so does mine
19:46:19  <TrueBrain> andythenorth: it is "node" based, which is for trains Tile + Trackdir
19:46:19  <dP> how does cpu cache matter for commands?
19:46:24  <TrueBrain> frosch123: if you pay, sure :)
19:46:29  <frosch123> LordAro: for commands? how many commands do you expect per second?
19:46:33  <LordAro> idk
19:46:39  <LordAro> just something i'd always assumed
19:46:53  <LordAro> but maybe it just dates back to the original TTD architecture?
19:47:01  <LordAro> bit stuffing via assembly might make more sense there :p
19:47:11  <dP> I'm more worried about full game recording I do but compression should take care of that
19:47:55  <frosch123> michi_cc: did you consider nlohmann::json for command packets, or some other standard serialisation format?
19:48:23  <frosch123> or is the fun part to make the metaclasses for serialisation?
19:48:32  <dP> json is kinda meh for commands, too much junk
19:48:39  <TrueBrain> LordAro: it wouldn't even surprise me if p1/p2 got added for network transmission :P
19:50:44  <nielsm> we already have several other data-to-code compilation steps in the build (languages, settings, scripting), what would adding something to compile some command structs and functions to transmit/work on them be?
19:51:53  <nielsm> as a bonus, if all commands are represented as some kind of data, you could compile it to a different language too (not sure if that's a good idea)
19:52:12  <TrueBrain> Java-port? :D
19:52:18  <dP> oh, btw, if they are binary packed it would be very nice to have some schema that can be parsed from, say, python for tools
19:52:27  <nielsm> yep
19:52:29  <frosch123> nielsm: do you want an openapi description of the commands? :)
19:52:41  <TrueBrain> Swagger them!
19:52:43  <dP> that's kind of what I started to do on that link I posted
19:53:15  <nielsm> really mostly something along the lines of protobuf, but maybe less full featured
19:53:43  <nielsm> (no need for versioning or nested types, probably)
19:54:28  <frosch123> binary/text serialisation is potato/potato to me. the interesting part is whether it is self-descriptive like json/msgpack/... or non-self-desriptive like boost::serialize/ottd-map-array/...
19:55:14  <frosch123> self-descriptive probably increases command packet size a lot :p
19:55:36  <TrueBrain> pretty sure size isn't a real issue, given currently a game consumes 600 bytes/s :P
19:55:40  <dP> self-descriptive is just a waste of bytes in this case imo
19:55:41  <frosch123> currently it's 4 ints (cmd, tile, p1, p2)
19:55:46  <frosch123> with descriptions it's a lot more :p
19:55:47  <TrueBrain> of which most is the sync-packet :)
19:56:30  <nielsm> yeah self-descriptive isn't meaningful when it's data in transit only, never at rest
19:56:49  <dP> btw, every command can have like one sync-byte for desync debugging
19:57:12  <nielsm> you have already established that you have two compatible versions of the game talking, they know the protocol they get
19:57:20  <dP> not only will it desync faster easing up debugging but may even eliminate the need for sync packet altogether
19:57:56  <DorpsGek> [OpenTTD/OpenTTD] glx22 approved pull request #9620: Fix 3046a6ce: [Preview] building preview failed to patch LZMA https://git.io/JKiwl
19:58:13  <TrueBrain> glx: I also can't remember why I added it
19:58:29  <TrueBrain> the CI variant doesn't use it .. I can only thing that it will fail sending the preview to AWS :P
19:58:45  <TrueBrain> wow, that was bad english
19:58:52  <TrueBrain> lets say ... thing = think
19:58:54  <TrueBrain> and it is fine :P
19:59:54  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #9620: Fix 3046a6ce: [Preview] building preview failed to patch LZMA https://git.io/JKi0z
20:00:56  <TrueBrain> 1928 successful connections made ... I wonder how high it goes today
20:00:56  <TrueBrain> just shy above 2000 I think
20:00:58  <TrueBrain> 2200 or something
20:01:15  <TrueBrain> average session consumes 2MB of bandwidth :P
20:01:26  <TrueBrain> (and is 23 minutes long)
20:01:44  <TrueBrain> we had 270 TURN sessions today, new record :)
20:02:40  <TrueBrain> https://openttd.grafana.net/dashboard/snapshot/nZFtYSW8Rk9cMs2ngkevho7CR0ru40Tq but for some reason the TURN graphs are broken
20:03:23  <nielsm> how much additional work would it be to extend the TURN support to proxy through websockets to the server hosting an emscripten build?
20:05:12  <dP> emscripten client I kinda get but why emscripten server? xD
20:05:48  <nielsm> because browser security doesn't allow the app to connect to anything else but the origin server
20:06:37  <TrueBrain> nielsm: very little; I just haven't gotten to doing it :)
20:07:12  <dP> nielsm, how will emscripten server connect to turn server then?
20:07:12  <TrueBrain> basically, game-coordinator doesn't care which side can't use direct-ip / STUN
20:07:26  <TrueBrain> so as soon as I add websocket support to the GC, it will magically work for both client and server :)
20:07:42  <TrueBrain> I already did this for the bananas-server
20:07:46  <TrueBrain> so it is just repeating a known trick
20:08:23  <TrueBrain> main reason I haven't yet, is because I first want to make sure we can pay the bill for TURN :P Before I add additional load :)
20:09:52  <TrueBrain> but my goal is still to add a "Play Now" button on openttd.org, where you can just jump into the game and .. play :) From your browser, ofc!
20:12:15  <dP> I'm not sure "play now" is such a good idea
20:12:28  <dP> for testing sure, but for actually playing browser version kinda sucks
20:14:21  <LordAro> ^
20:15:07  *** gelignite has quit IRC
20:16:32  <glx> mouse capture is annoying
20:16:52  <nielsm> being able to use the system cursor instead would be a huge win usability wise there
20:20:09  <TrueBrain> Chrome allows this with Stadia
20:20:15  <TrueBrain> not sure it can be extended to other games
20:20:31  *** jottyfan has joined #openttd
20:20:47  <TrueBrain> but there are more things to fix before it becomes workable
20:20:50  <TrueBrain> still, it is a good goal to have :)
20:24:21  <glx> ok so preview fails to build if a user is specified, but it fails to publish if a user is not specified
20:24:57  <TrueBrain> yeah, fixing it ...
20:25:59  <glx> at least now we know why there was a user ;)
20:27:13  <Xaroth> dP: Installing an application is tedious, people don't want to do it if they can avoid it. Playing from the browser solves that, and as a result, leads to more potential users.
20:28:31  <dP> or they can try it in browser, find out that it sucks and never even try the real game :p
20:28:48  <glx> oh of course microsoft store still doesn't have 12.0
20:29:04  <Xaroth> but they at least would have tried it.
20:29:11  <TrueBrain> owh, right, orudge ! MSStore needs an update :D
20:29:16  <TrueBrain> we still have to automate that :P
20:30:47  <dP> Xaroth, do you want more people to try or more people to play? :p
20:31:07  <dP> it's kind of like android version though I guess there are no stats on how useful it is
20:31:38  <Xaroth> You counter your own argument :P
20:31:45  <frosch123> going by reddit screenshots, the android port is very popular
20:32:01  <Xaroth> If they don't like the game, there will be 0 difference between people downloading it, and people trying it in the browser.
20:32:28  <Xaroth> if they do like the game, being able to play it in the browser, as well as downloading it, will have a larger amount of users than just with downloading.
20:32:37  *** iSoSyS has joined #openttd
20:33:15  <TrueBrain> its funny how this conversation is a replica of a reddit thread linked earlier :P
20:33:37  *** iSoSyS has quit IRC
20:34:59  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #9621: Fix 01603d9e: [Preview] publishing preview failed due to recent changes https://git.io/JKihx
20:35:06  <TrueBrain> right, part 2 .. hopefully that is the end of it glx :P
20:36:40  <TrueBrain> hmm .. the memleak in the Game Coordinator is worse now .. so I guess it is related to activity :P
20:36:41  <DorpsGek> [OpenTTD/OpenTTD] glx22 approved pull request #9621: Fix 01603d9e: [Preview] publishing preview failed due to recent changes https://git.io/JKPek
20:37:12  <TrueBrain> for now, lets scale up the amount of memory assigned to those instances .. will try to trace tomorrow what is actually going on :)
20:38:25  <dP> android version has 1,000,000+ installs according to google play
20:38:49  <dP> more than steam, yet unlike steam users all these players don't rly show up in any stats
20:40:03  <TrueBrain> https://cdn.discordapp.com/attachments/337701432230805505/899396291140923392/unknown.png <- anyone any clue what happened this morning? :D
20:40:43  <TrueBrain> finally some CPU usage .. those machines are finally doing something :P
20:42:23  <Xaroth> about time! :P
20:42:33  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #9621: Fix 01603d9e: [Preview] publishing preview failed due to recent changes https://git.io/JKihx
20:42:56  *** nielsm has quit IRC
20:43:34  *** jottyfan has quit IRC
20:44:56  <frosch123> TrueBrain: you made a new tag at noon, so it restarted?
20:45:15  <frosch123> well, not noon, like 10am utc
20:46:13  <TrueBrain> that too
20:46:16  <TrueBrain> two tags even :)
20:46:32  <frosch123> what are the colors?
20:46:45  <TrueBrain> max, min, avg
20:46:55  <TrueBrain> there are two instances running, so basically blue and green are instance #1 and #2
20:47:04  <TrueBrain> (for memory)
20:47:09  <TrueBrain> for CPU it is a bit more complicated
20:47:46  <glx> nice stairs in memory usage
20:47:54  <TrueBrain> but one could say we have plenty of room to grow ;)
20:47:55  <glx> but they climb fast
20:48:00  <TrueBrain> just the memory issue I have to find/fix :)
20:48:17  <TrueBrain> which is very difficult :(
20:50:32  <TrueBrain> the most expensive functions in the GC are CLIENT_LISTING .. and this is mainly because I send 1 server at the time :P
20:50:35  <TrueBrain> so I can fix that :)
20:50:41  <TrueBrain> it takes, on average, 180ms
20:50:50  <TrueBrain> after that is SERVER_UPDATE, with 18ms
20:50:59  <TrueBrain> the rest is <5ms to process
20:51:05  <TrueBrain> pretty happy with that :)
21:00:30  <TrueBrain> glx: IT WORKS!!! :D
21:00:39  <glx> \o/
21:01:07  <TrueBrain> one less problem to worry about :)
21:08:18  <andythenorth> :)
21:23:26  *** Samu has quit IRC
21:24:13  <andythenorth> oh this improves my mood :) https://www.reddit.com/r/openttd/comments/q52o3i/only_have_access_to_two_types_of_signals/
21:27:38  <DorpsGek> [OpenTTD/OpenTTD] Laszlo65 opened issue #9622: OpenTTD 1.11 crashes on startup[Crash]:  https://git.io/JKPVe
21:29:50  *** sla_ro|master has quit IRC
21:32:12  <DorpsGek> [OpenTTD/OpenTTD] James103 commented on issue #9622: OpenTTD 12.0 crashes on startup[Crash]:  https://git.io/JKPVe
21:36:16  *** frosch123 has quit IRC
21:41:32  <dP> same line as #9593
21:47:52  *** snail_UES_ has joined #openttd
21:54:49  *** andythenorth has quit IRC
22:10:09  <glx> but no missing newgrf it seems
22:19:10  <DorpsGek> [OpenTTD/OpenTTD] glx22 commented on issue #9622: OpenTTD 12.0 crashes on startup[Crash]:  https://git.io/JKPVe
22:21:48  <TrueBrain> That assert and backtrace are not in agreement :D
22:22:00  <dP> was James missing newgrfs? xD
22:22:30  <glx> original gamelog looks clean
22:23:17  <glx> james one says emergency savegame
22:23:32  <glx> but nothing about newgrf
22:31:33  <DorpsGek> [OpenTTD/OpenTTD] glx22 commented on issue #9622: OpenTTD 12.0 crashes on startup[Crash]:  https://git.io/JKPVe
22:39:09  <glx> let's try to load the save (I can't get useful data from the log)
22:40:56  <glx> but a "broken" tile (or similar) may explain both, as without assert it will try to draw garbage
22:41:39  *** Wolf01 has quit IRC
22:54:28  <glx> hmm company name in #9622 crash save is title 1.12
22:56:04  <glx> and of course I miss some newgrf (more than 1300 and still missing some)
23:04:15  <DorpsGek> [OpenTTD/OpenTTD] Rau117 commented on pull request #9617: Feature: Button to toggle showing advanced signal types https://git.io/JKXlA
23:16:34  <glx> oh but the crash happens in the intro game, there's openttd.exe!SelectGameWindow::OnRealtimeTick(unsigned int delta_ms) Ligne 235    C++ in the trace
23:17:33  <glx> that explains why crash.sav has issues, because it adds newgrfs
23:23:19  <DorpsGek> [OpenTTD/OpenTTD] glx22 commented on issue #9622: OpenTTD 12.0 crashes on startup[Crash]:  https://git.io/JKPVe
23:40:08  *** HerzogDeXtEr has quit IRC

Powered by YARRSTE version: svn-trunk