IRC logs for #openttd on OFTC at 2021-05-14
⏴ go to previous day
00:32:48 <DorpsGek> [OpenTTD/OpenTTD] 2TallTyler commented on issue #9241: Some tree brushes do not mark tiles as a rainforest https://git.io/JsvZq
01:39:13 *** Wormnest has quit IRC (Quit: Leaving)
02:04:37 <DorpsGek> [OpenTTD/OpenTTD] JGRennison opened issue #9264: Incorrect autoreplace behaviour when temporary free-wagon chains exceed the train length limit https://git.io/JsYgM
02:24:50 <DorpsGek> [OpenTTD/OpenTTD] James103 commented on issue #9264: Incorrect autoreplace behaviour when temporary free-wagon chains exceed the train length limit https://git.io/JsYgM
02:50:23 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:06:41 <DorpsGek> [OpenTTD/OpenTTD] techgeeknz commented on pull request #8279: Port GUI rendering improvements from JGRPP https://git.io/JsYPu
03:53:00 *** Flygon_ has joined #openttd
03:56:19 *** Flygon has quit IRC (Ping timeout: 480 seconds)
04:15:28 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
04:16:30 *** WormnestAndroid has joined #openttd
04:36:02 *** snail_UES_ has quit IRC (Quit: snail_UES_)
04:43:44 *** WormnestAndroid has quit IRC (Remote host closed the connection)
04:43:51 *** WormnestAndroid has joined #openttd
04:45:21 *** lobstarooo has joined #openttd
04:52:21 *** lobster has quit IRC (Ping timeout: 480 seconds)
04:52:21 *** lobstarooo is now known as lobster
05:00:05 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
05:06:51 *** snail_UES_ has joined #openttd
05:30:14 *** WormnestAndroid has joined #openttd
06:10:32 *** sla_ro|master has joined #openttd
06:35:38 *** HerzogDeXtEr has joined #openttd
06:47:58 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
06:48:02 *** WormnestAndroid has joined #openttd
07:21:38 *** supermop_Home_ has quit IRC (Ping timeout: 480 seconds)
08:16:06 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 updated pull request #9263: Codechange: [Network] Use std::string for passwords or hashes thereof https://git.io/JsmxH
08:28:45 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 updated pull request #9263: Codechange: [Network] Use std::string for passwords or hashes thereof https://git.io/JsmxH
08:36:08 <Rubidium> I guess some people beg to differ with you on that ;)
08:42:02 *** andythenorth has joined #openttd
08:42:10 <andythenorth> does GS run when game is paused?
08:45:46 <andythenorth> various tedious gut medicines
08:46:13 <Rubidium> andythenorth: based on the code I would guess it is. It might be that performing a command halts it because it needs to wait until the command has been executed
08:46:31 * andythenorth wants to plant a 'town hall' industry in every town
08:46:40 <andythenorth> and the only way to do that might be a dedicated GS
08:46:51 <andythenorth> which pauses the game on start, then places industries until all done
08:46:52 <Rubidium> andythenorth: I reckon you can prove it by just printing some (different) text in a loop and then yielding for that tick
08:48:01 <Rubidium> andythenorth: though... during map creation there are a lot of ticks where the game script seems to be running
08:48:39 <andythenorth> this isn't something I could release, it's just to test a concept
08:48:57 <andythenorth> there's no way players will be able to download both a specific FIRS version and the correct version of a GS
08:49:03 <andythenorth> and get them both into the game
09:03:30 <FLHerne> ^ "Bug" is that "Unload all" doesn't imply "Unload and leave empty" by default
09:05:44 <TrueBrain> surprised me too :P
09:05:56 <LordAro> that window is a mess
09:06:23 <LordAro> well, those buttons specifically are a mess
09:17:29 <FLHerne> Yeah, tbh I think it might be better if it did
09:17:40 <TrueBrain> LordAro: it is .. but a good design to fix it is hard
09:17:44 <peter1138> I might look at group-orders at some point.
09:17:46 <TrueBrain> without ofc going in a totally other direction
09:18:04 <FLHerne> LordAro: Have you seen the JGR version?
09:18:10 <FLHerne> It's just getting absurd
09:18:36 <FLHerne> (you can choose sets of cargo to station-refit to and so on)
09:21:11 <TrueBrain> did anyone ever make a patch that the order window tells you which stations you can reach?
09:21:17 <TrueBrain> so you can just click them, instead of scrolling to it?
09:22:45 <peter1138> can you click on a station in the station list?
09:23:22 <FLHerne> In general, OTTD UI seems quite good at being able to click on the thing
09:23:26 <peter1138> (I mean click on with goto order to add the order, not just click on it)
09:23:48 <andythenorth> it scrolls to the station
09:24:43 <TrueBrain> well, what happens to me a lot, that I clone a vehicle on an other part of the infra, and that I forget to change the orders
09:24:54 <TrueBrain> but it just leaves the depot and starts to do what-ever-the-fuck-it-wants
09:25:05 <TrueBrain> kinda miss the feedback of "station not reachable" a lot easier :D
09:25:15 <FLHerne> Yay, this time I got the crash-on-exit even when using the in-game Exit option
09:25:24 <andythenorth> something something simutrans
09:25:26 <FLHerne> (and it crashed my entire desktop again)
09:28:54 <peter1138> andythenorth, yes, group-orders
09:29:17 <andythenorth> and group consist templates :P
09:29:38 <andythenorth> group 2-way EOL settings
09:29:43 <peter1138> we have group colours, so why not group everything else ;)
09:29:55 <andythenorth> group colours the best
09:30:17 <peter1138> seems to be grand "fix" to everyone's routing issues...
09:30:24 <andythenorth> for times when you've built signals so complicated, for no obvious reason, that you exploit marginal hacks
09:30:35 <andythenorth> instead of using 1-line-per-train as intended
09:30:40 <peter1138> the penalty is too high so it makes it use a path anyway. or something? i dunno.
09:31:13 <peter1138> probably the penalty is high due to the massive chains of pointless presignals
09:31:45 <peter1138> Even devs don't get safe-waiting-points right though :)
09:36:57 <peter1138> Hmm, maybe I should run subtropic next time
09:38:15 <Rubidium> two buffers on an otherwise physically connected track
09:40:31 <peter1138> ^ Something to do with that.
09:40:36 *** frosch123 has joined #openttd
09:42:37 <andythenorth> that's an early frosch123 is it holiday?
09:43:13 <frosch123> yesterday was lazy christians day. today is hangover bridge day. but i didn't drink yesterday
09:45:23 <frosch123> but hey, i did my taxes yesterday. that counts for at least 2 workdays, doesn't it?
09:51:56 *** Samu has quit IRC (Quit: Leaving)
09:51:57 <_dp_> afaict the only reason to not have twoway eol is to make life easier for newbies that signal everything with twoway signals
09:53:05 <_dp_> which is even less meaningful considering the recent effort to make pbs a newbie signal
09:53:06 <andythenorth> something something nothing can ever change because 2 way EOL is the only way to make PBS work
09:53:42 <frosch123> _dp_: what about all the effors to ban pbs from servers to reduce cpu usage? :p
09:54:16 <andythenorth> are they the same servers that run big train grfs?
09:54:24 <frosch123> andythenorth: we changed default vehicles yesterday. is that enough change for this week?
09:54:32 <_dp_> frosch123, though at openttcoop level of complexity that may actually be true :p
09:55:33 <frosch123> i should host a server "hard game, no inflation, no infracost, no breakdowns, to town authority"
09:55:47 <_dp_> frosch123, also pbs are "slow" in a gameplay sense, not cpu usage
09:56:27 <frosch123> what do you mean? does it take longer to build them?
09:56:58 <frosch123> or do you mean lack of priority lines, so slowdown of mainlines?
09:57:29 <_dp_> you can cram more trains into a line with block signals of the same interval
09:57:38 <_dp_> figuratively speaking pbs are slow to switch
09:59:23 <andythenorth> trying to repro my result where a single Iron Horse train cuts 50% of the ffwd rate compared to similar base set train
09:59:39 <andythenorth> not getting stable results, the ffwd speed swings around wildly
09:59:57 <peter1138> The argument about performance was originally throughput, IIRC. It just got morphed into performance, and then CPU performance...
10:00:07 <frosch123> andythenorth: easy, play on 64x64 with no industries
10:00:32 <frosch123> when a train is the only thing in a game, it's effect on cpu load is at maximum
10:00:51 <peter1138> (Also, block signals do path finding to work out what state they need to be, so "in theory" should cost more in CPU time ;)
10:00:55 <_dp_> terrible name for the bug btw
10:01:03 <andythenorth> I need to turn off autosave and stuff
10:01:13 <peter1138> (Well, not path finding, but track following)
10:01:17 <andythenorth> configuring my game for ffwd-based speed profiling has too many confounding factors :P
10:01:47 <andythenorth> anyway, I think Horse kills performance
10:02:21 <TrueBrain> horse power .. always ruining everything
10:02:42 <andythenorth> chasing rainbows
10:03:24 <peter1138> I performance checked my tree tick change on 4096x4096 and noticed the standard game loop tick is ~ 11ms already.
10:03:39 <peter1138> So I decided I didn't care.
10:04:07 <frosch123> i wondered about measuring cpu time of newgrf
10:04:17 <frosch123> tic/toc is too expensive to do at every newgrf call
10:04:28 <frosch123> so i wondered about a statistical approach
10:04:48 <frosch123> store in some global what newgrf thing we are doing, and then have some sampling thread
10:05:24 <andythenorth> we only need some indication of 'makes a difference, or not' ?
10:05:32 <andythenorth> actual precise timings don't matter?
10:05:45 <peter1138> Nice, eat a whole CPU core monitoring what is eating performance :D
10:06:01 <peter1138> Could be a useful debug tool.
10:06:03 <frosch123> we have enough cores for that by now
10:06:14 <andythenorth> I mean...I can only use one of the cores before the mac dumps the thermals
10:06:33 <andythenorth> 65 i9 poor buying decision
10:07:06 <peter1138> Cargo destinations eh?
10:07:18 <andythenorth> cargodist is now out of fashion
10:07:27 <andythenorth> pikka has persuaded me it's net worse to play with
10:08:06 <frosch123> "because it plays the game for you" or some other reason?
10:08:46 <andythenorth> because it chokes passenger routes
10:08:57 <andythenorth> and because it doesn't work with FIRS supplies unless tricks are used
10:09:03 <Timberwolf> Passenger generation isn't scaled for it, imo.
10:09:33 <peter1138> It requires existing established routes is its main issue for me.
10:09:36 <andythenorth> and because it prevents multiple destination pickups from one station
10:09:38 <frosch123> ah, so "because it doesn't do what i want'" :)
10:09:47 <Timberwolf> I found it much nicer in daylength games, where small villages actually end up with a small branch line.
10:09:51 <andythenorth> 'out of fashion' not broken
10:10:13 <andythenorth> as an automated transfer system cdist is really clever
10:11:24 <andythenorth> Timberwolf but it would need to scale passenger generation _down_ which is weird
10:11:46 <andythenorth> and also transfer payments are a shitshow with big cdist pax network
10:13:00 <FLHerne> Not really, you just ignore them :p
10:14:19 <peter1138> Because cdist works on existing routes, it means that point 2 point still works.
10:14:25 <Timberwolf> I think the correct thing is Simutrans-style for things like passengers.
10:14:58 <Timberwolf> They have a destination in mind, if you have a station near the destination they will generate, if not they won't.
10:15:38 <Timberwolf> Maybe a tolerance on travel time (like Transport Fever).
10:15:57 <peter1138> You need something to tell you where things want to go.
10:16:39 <Timberwolf> There is also the opposite problem, which is it being impossible to start with passengers.
10:17:23 <Timberwolf> I always found that with the Simutrans model, it gets a bit, "cool, you've built a nice network between these towns, shame all the passengers want to go to this other place halfway across the map"
10:19:01 <andythenorth> I considered splitting the cargo type to local and long distance pax
10:19:07 <andythenorth> but they can't get on the same bus :P
10:19:12 <andythenorth> actual subtypes anyone?
10:21:35 <andythenorth> I read those entire threads recently for $some reason
10:21:47 <andythenorth> and the cargo_age_period
10:22:03 <andythenorth> maybe _that_ is why I am having a forums break
10:22:28 <andythenorth> 100 pages of watching people talk past each and other and make bad decisions due to weird nitpick social politics
10:22:47 <andythenorth> 'life is better without tt-forums'
10:23:19 * andythenorth loves tt-forum really
10:23:20 *** WormnestAndroid has quit IRC (Read error: Connection reset by peer)
10:23:55 *** WormnestAndroid has joined #openttd
10:43:59 <Timberwolf> The hard thing is a lot of the people who gave good feedback or contributed a lot have wandered off over the years, making the nitpicking and the not-really-reading proportionately larger.
10:45:48 <Timberwolf> If I go back to my original Road Vehicles dev thread there's *way* more advice about technical issues, on-topic answers to questions and people responding to my monologues on how I'm thinking of approaching certain issues.
10:51:31 <Timberwolf> This is a general problem IMO, I don't think the modern platforms have the same kind of depth of discussion that the old forums had.
10:52:00 <Timberwolf> I find this a lot with reddits, people don't really discuss stuff, they share screenshots and photos of things.
11:01:46 <Timberwolf> Discord can be good, it's very dependent on the first few people to reply to set the tone.
11:29:13 <andythenorth> the early grfs I made were absolutely dependent on forums
11:29:23 <andythenorth> for getting the shape of an idea, that sort of discussion was great
11:57:54 <Timberwolf> I do miss those days, when you could post some musings on vehicle scales or gameplay ideas and someone would go, "I tried that with my ___ set, it sort of works but I found... there's also ___ if you haven't tried it"
12:09:47 *** urdh has quit IRC (Quit: Boom!)
12:19:43 *** m1cr0m4n has quit IRC (Quit: G'luck)
12:19:44 <Timberwolf> Yeah, it's really good for that if you get the time right.
12:20:20 <Timberwolf> If you get it wrong it quickly falls to "uhhh but it doesnt fit with the landscape tho" and never recovers.
12:20:45 <peter1138> There's always some troll on there talking about removing block signals...
12:23:04 *** m1cr0man has joined #openttd
12:23:05 <andythenorth> funny that only happens when you're there :D
12:23:19 * andythenorth didn't get a memo
12:29:03 <TrueBrain> I really like how the new translator signup gave a bunch more translators
12:29:09 <TrueBrain> like .. an awful lot more :P
12:29:19 <TrueBrain> people like clicking buttons more than sending emails, I can conclude?
12:31:03 <Rubidium> or maybe a lot of those emails were never delivered due to some unknown reason?
12:33:54 <glx> it's one of the fastest approval :)
12:36:16 <FLHerne> TrueBrain: Has there been a corresponding increase in the amount of stuff translated?
12:36:53 <TrueBrain> FLHerne: that is a question for frosch123 ; he likes to do stats :P
12:49:32 <glx> highly depends on language I guess
13:15:09 *** snail_UES_ has joined #openttd
13:33:19 <frosch123> FLHerne: i checked early april. most translators were either active in the first days after sign-up, or shortly before release
13:33:34 <frosch123> very few are active constantly. quite some have never been active
13:34:03 <frosch123> the latter is also backed up by the percentage of people who never accept the invitation
13:34:43 <frosch123> anyway. i concluded it makes no sense to make any stats without waiting at least a year. so see you in october
13:35:40 <frosch123> oh, and then there were a few, who did not translate anything since sign-up on github, but who translated 70% of a language 10 years ago (or someone with the same nickname) :p
13:37:18 <frosch123> FLHerne: oh another stat: some teams use the team chat a lot, some not at all. of those who do, some discuss in native, some in english
13:39:17 <frosch123> the people who sign-up are also very different. some sign-up to github only for us. some have been active in other projects for years, some claim to be professional translators, some are obviously underaged kids, who do stuff like p1sim
13:39:36 <frosch123> (p1sim is only an example for equivalent efforts)
13:43:14 <TrueBrain> haha, all the links from there are no longer available or "draft"
13:44:22 <LordAro> mm, seems some sort of Drama is occurring/about to occur
13:45:01 <frosch123> what is the economical potential of hosting a irc network?
13:45:22 <frosch123> just some hedgefond who thought it was a teams alternative?
13:46:37 <glx> I have a warning free MinGW \o/
13:46:47 <TrueBrain> "GitHub Actions has encountered an internal error when running your job." <- ugh :P
13:47:29 <TrueBrain> that is just plain hilarious :D
13:49:05 <frosch123> let's have a team-building event where we draft fake-letters of our resignment, to learn what we like about our job.
13:49:53 <frosch123> i have no other ideas, why you would "draft" something like that
13:50:14 <TrueBrain> well, before you release it, it is a draft, ofc
13:50:24 <frosch123> "Fake logs, you can't have newlines in IRC messages!" <- i like that one :)
13:50:25 <TrueBrain> but then saying: no no, that content is not valid yet, as it is a draft, now that is weird
13:51:58 <glx> "we'll communicate when we can"
14:09:30 <milek7> but maybe they didn't want to publish until their replacement is running (libera.chat)
14:09:59 <LordAro> there are some gag orders going on
14:11:00 <LordAro> it's at least a bit serious
14:16:02 <_dp_> what is irc network anyway and how does that even matter in 2021?
14:18:41 <_dp_> it's "network" part that's a mystery to me, not "irc" :p
14:19:56 <_dp_> nobody calls jabber a "network" or discord
14:21:16 <SpComb> org or group of people that run some servers and configure them to form a network
14:21:51 <FLHerne> IRC's a federated protocol with multiple servers per network, often run by different parties
14:21:53 <SpComb> IRCNet/EFNet are the traditional networks, they're organized on a more national level
14:22:33 <FLHerne> _dp_: I'm connected to helix.oftc.net in Sweden, and you're connected to coulomb.oftc.net in London
14:22:48 <SpComb> then you have those like freenode that are run by a single org and organized more around software languages/projects
14:23:15 <glx> yeah freenode is very strict about channel names
14:23:33 <SpComb> and they are very much different networks with different policies and communities etc
14:24:18 <SpComb> a channel only exists on a specific network
14:25:17 <SpComb> you can't connect to an IRCNet server and join #openttd - or you can, but it's a different channel with different users. Probably either empty or very quiet
14:26:25 <glx> ahah ##openttd still exists on freenode, but it's empty now
14:26:39 <glx> "This unofficial channel is available as needed. Please consult the project website for its official IRC location. Thank you for using freenode!"
14:31:48 <frosch123> TrueBrain: how did the meeting with epic go? are they buying ottddistltd as planned?
14:32:23 *** jottyfan has joined #openttd
14:32:25 <TrueBrain> frosch123: wrong channel ffs
14:32:47 <andythenorth> use the secret channel
14:33:11 <TrueBrain> well, his text was only a draft, so we are fine
14:35:05 <TrueBrain> GitHub Actions really doesnt want to start on TrueWiki .. that is annoying
15:29:09 *** Gustavo6046_ has joined #openttd
15:30:31 *** Gustavo6046 has quit IRC (Ping timeout: 480 seconds)
15:30:31 *** Gustavo6046_ is now known as Gustavo6046
15:36:47 *** Wormnest has joined #openttd
15:54:15 *** Progman has joined #openttd
16:34:55 <andythenorth> goes it a patch?
16:35:19 * andythenorth wants to circumvent industry count rules
17:02:28 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 opened pull request #9266: Codechange: use std::string in network chat https://git.io/Jssu9
17:07:32 <DorpsGek> [OpenTTD/OpenTTD] JGRennison opened issue #9267: Heap use after free in Squirrel delayed memory deallocation when AI deallocated https://git.io/JsszD
17:09:41 *** jottyfan has quit IRC (Quit: jottyfan)
17:12:31 *** Flygon_ has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
17:58:10 <glx> ok after upgrading my mingw install, release build with mingw32 stops failing to link, but it warns
18:00:32 *** andythenorth has quit IRC (Quit: andythenorth)
18:00:59 *** andythenorth has joined #openttd
18:01:36 *** andythenorth has quit IRC ()
18:26:39 *** gelignite has joined #openttd
18:29:39 *** jottyfan has joined #openttd
18:39:53 *** Gustavo6046 has quit IRC (Remote host closed the connection)
18:41:11 *** Gustavo6046 has joined #openttd
18:53:56 *** Gustavo6046 has joined #openttd
18:53:56 *** Gustavo6046 has quit IRC ()
18:53:59 *** Gustavo6046 has joined #openttd
19:38:29 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 commented on issue #9267: Heap use after free in Squirrel delayed memory deallocation when AI deallocated https://git.io/JsszD
19:44:30 *** gelignite has quit IRC (Quit: Stay safe!)
19:46:15 *** andythenorth has joined #openttd
19:50:50 *** andythenorth has quit IRC (Remote host closed the connection)
19:51:21 *** andythenorth has joined #openttd
19:54:43 *** andythenorth has quit IRC (Remote host closed the connection)
19:55:14 *** andythenorth has joined #openttd
19:59:11 <andythenorth> does the livery refit callback have limited vars available?
19:59:20 <andythenorth> trying to help someone on discord
20:03:42 *** andythenorth has quit IRC (Remote host closed the connection)
20:04:13 *** andythenorth has joined #openttd
20:04:49 <andythenorth> nice stable wifi network
20:06:16 *** luaduck has quit IRC (Ping timeout: 480 seconds)
20:13:38 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
20:14:08 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
20:14:35 *** WormnestAndroid has joined #openttd
20:18:09 <peter1138> Might chuck out some of these old laptops.
20:24:21 <andythenorth> is one a dead mac :P
20:28:59 <peter1138> Chunk bevels with non-integer scaling, eh?
20:40:16 *** andythenorth_ has joined #openttd
20:40:26 *** andythenorth has quit IRC (Ping timeout: 480 seconds)
20:43:08 <peter1138> 3 pixel wide bevels, yes.
20:43:18 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
20:47:56 <peter1138> andythenorth_, get me some non-integer sprite scaling
20:48:09 <andythenorth_> sprite scaling :o
20:48:28 <andythenorth_> non integer extra zoom?
20:48:35 <andythenorth_> goes it sub-pixel anti-aliasing?
20:50:18 *** WormnestAndroid has joined #openttd
20:54:22 *** jottyfan has quit IRC (Quit: jottyfan)
20:58:34 *** Wormnest has joined #openttd
20:58:38 <DorpsGek> [OpenTTD/OpenTTD] glx22 opened pull request #9268: [MinGW] Fix MinGW32 build failure and almost all MinGW32/MinGW64 warnings https://git.io/JsGOr
21:09:57 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #9263: Codechange: [Network] Use std::string for passwords or hashes thereof https://git.io/JsG3h
21:11:20 *** frosch123 has quit IRC (Quit: be yourself, except: if you have the opportunity to be a unicorn, then be a unicorn)
21:15:16 *** andythenorth_ has quit IRC (Quit: andythenorth_)
21:22:08 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 merged pull request #9263: Codechange: [Network] Use std::string for passwords or hashes thereof https://git.io/JsmxH
21:51:08 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
22:07:59 *** Gustavo6046 has quit IRC (Ping timeout: 480 seconds)
22:18:00 *** Gustavo6046 has joined #openttd
22:41:07 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:55:42 *** sla_ro|master has quit IRC ()
22:57:53 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
22:58:49 *** WormnestAndroid has joined #openttd
23:13:01 *** Progman has quit IRC (Remote host closed the connection)
continue to next day ⏵
