IRC logs for #openttd on OFTC at 2024-06-19
⏴ go to previous day
00:08:04 <Eddi|zuHause> should the icons switch direction? :p
01:40:47 *** Wormnest has quit IRC (Quit: Leaving)
02:03:17 *** herms6 has quit IRC (Quit: bye)
02:11:14 *** debdog has quit IRC (Ping timeout: 480 seconds)
02:54:46 *** gnu_jj_ has joined #openttd
02:58:14 *** gnu_jj has quit IRC (Ping timeout: 480 seconds)
04:35:23 *** virtualrandomnumber has joined #openttd
04:35:47 *** virtualrandomnumber has quit IRC ()
04:39:24 *** johnfranklin has joined #openttd
04:42:17 <DorpsGek> - Update: Translations from eints (by translators)
04:59:36 *** johnfranklin has quit IRC (Ping timeout: 480 seconds)
05:02:04 *** johnfranklin has joined #openttd
05:06:49 *** Extrems has quit IRC (Remote host closed the connection)
05:06:53 *** Extrems has joined #openttd
05:15:04 *** keikoz has quit IRC (Ping timeout: 480 seconds)
05:30:04 *** Leopold_ has joined #openttd
05:46:56 *** YourOnlyOne has joined #openttd
05:53:41 *** johnfranklin has quit IRC (Ping timeout: 480 seconds)
05:57:09 <merni> rubidium42viaGitHub: Isn't it better to just delete the grfcodec download pages then? What's the point of having an outdated version there?
06:02:26 <pickpacket> What even is the grfcodec?
06:05:16 *** Leopold_ has quit IRC (Remote host closed the connection)
06:06:57 *** gwyd4016 has joined #openttd
06:06:57 <gwyd4016> It's how NFO is compiled
06:11:32 <merni> and how grfs are decompiled
06:29:06 <blathijs> merni: Rubidium: One reason to keep the old grfcodec pages is to keep the old tarballs, which are not (currently) published on github (there are tags, but not releases with artifacts). Also people might still link to the old place, so it should at least be replaced with a redirect if removed.
07:24:30 <LordAro> there's no particular reason why we couldn't upload new builds to the cdn
07:25:45 <peter1139> Someone™ just needs to fix it :)
07:33:45 <pickpacket> What kind of changes are made to the grfçodec?
07:34:04 <pickpacket> Nice weird c there. Thanks, android
07:34:46 <gwyd4016> Maybe I should update my version
07:35:44 <gwyd4016> The default container version only allows for tiny little sound files
07:47:01 *** mryakov has joined #openttd
07:47:01 <mryakov> Hello, I have a suggestion
07:47:01 <mryakov> In openttd we have waypoints, that have two order options, go via and go via non stop. If we choose go via, we can enable loading/unloading options to customise vehicle behaviour when it enter any of stations while go to waypoint
07:47:59 *** johnfranklin has joined #openttd
08:14:16 <wensimehrp> Probably they meant by implicit orders?
08:17:40 <Eddi|zuHause> i don't see a sane UI for that
08:19:38 <Eddi|zuHause> translation: "if an order is "go via, have load/unload etc. options apply to all implicit intermediate stops"
08:21:13 <mryakov> Eddi|zuHause: Exactly
08:22:17 <Eddi|zuHause> i'm against this, because you cannot clearly communicate this behaviour to the user.
08:22:59 <reldred> and implicit stops are a frequent fuckup for the path finder
08:23:10 <reldred> Bad Things(TM) happen.
08:24:20 <mryakov> Eddi|zuHause: Existing buttons for loading/unloading options, currently it disabled when user select waypoint in order list.
08:24:20 <mryakov> But if waypoint is go via, it simply enables this buttons
08:24:56 <Eddi|zuHause> mryakov: yes, but how do you tell a user what this button does?
08:25:24 <wensimehrp> I'm also against this
08:25:24 <wensimehrp> For any normal player it makes no sense to load/unload at a waypoint
08:25:24 <wensimehrp> And implicit orders oftentimes break routes
08:27:07 <Eddi|zuHause> mryakov: and this functionality must equally be available for "go to" orders, when they allow implicit stops
08:27:22 <Eddi|zuHause> then the button needs to do two separate things
08:31:36 <Eddi|zuHause> wensimehrp: that might be backwards. players often notice broken routes because it adds strange implicit orders. the reason for the broken route is usually something else (like depot placement, missing electrification, blocked signals)
08:32:45 <mryakov> Eddi|zuHause: Now for stations, an explanation is written in the list of orders about what the train will do at the station if alternative options are selected
08:32:45 <mryakov> the same thing will happen with the list of routes
08:32:45 <mryakov> For example "go via Foo (Full load any cargo if enter station)"
08:34:05 *** mindlesstux has joined #openttd
08:35:04 <wensimehrp> 🤔 iirc Transport Fever 2 has a route tool to check if the destination is accessible for the train.
08:35:04 <wensimehrp> Maybe we should have something like this, to check if there is a valid route between places.
08:35:29 <wensimehrp> And lines could replace shared orders I think?
08:36:46 <Eddi|zuHause> the "lines" tool of transport fever is nice, but much more limited than our current orders
08:37:24 <Eddi|zuHause> because it doesn't allow for dynamic rerouting of trains based on track occupancy
08:38:30 <Eddi|zuHause> and that makes it difficult to "check for valid lines", because the search space is much bigger when you include all the potential detours.
08:39:27 <Eddi|zuHause> and you won't catch cases where there exists a route, but it is unreasonably long.
08:40:16 <mryakov> Eddi|zuHause: I am not sure about this, its only about waypoints
08:40:16 <mryakov> Its will be extremely usefull if you build srnw networks or similar stuff.
08:40:16 <mryakov> But have same behavior on station orders, I cant imagine network configuration, where you need that propagation
08:42:25 <Eddi|zuHause> mryakov: that's a problem with your suggestion. you can't imagine a situation where that's needed, but leaving it out makes the potential use cases much more narrow, tailored only to the use case you have originally in mind
08:44:09 <mryakov> Eddi|zuHause: But I did not suggest doing this for stations.
08:44:09 <mryakov> Waypoints and stations are different orders, it is normal that their behavior will be different
08:45:34 <Eddi|zuHause> but on a technical level, your suggestion isn't really dependent on it being a waypoint order. it's about the "non-stop" part of the order, which can be part of station orders as well
08:46:34 <Eddi|zuHause> so you have to implement it for all orders, and then deliberately take it out of station orders
08:47:31 <Eddi|zuHause> for which there isn't a justification.
08:57:56 <mryakov> Can you pls point code location, which handle it
09:07:47 <Eddi|zuHause> probably order_cmd.cpp and order_gui.cpp
09:08:09 <mryakov> Eddi|zuHause: For me its looks like additional if in Vehicle::BeginLoading, where we check if waypoint order have additional info and add it in vehicle order bits.
09:08:12 <Eddi|zuHause> _cmd for the internal behaviour, and _gui for the buttons
09:08:56 *** johnfranklin has quit IRC (Quit: johnfranklin)
09:48:11 <peter1139> I think most people would just say remove non-non-stop orders.
09:50:27 <Eddi|zuHause> and sometimes it's silly to listen to "most people" :p
09:51:48 <mryakov> peter1139: And that broke any type of srnw network
10:04:31 <mryakov> mryakov: Any feedback ?
10:25:46 *** andtos12 has joined #openttd
10:26:20 *** tabytac has joined #openttd
10:40:19 *** mindlesstux has quit IRC (Ping timeout: 480 seconds)
10:45:08 *** mindlesstux has joined #openttd
10:46:26 <kuhnovic> This discord is about lunch
11:02:27 <truebrain> en bikes; don't forget the bikes!
11:08:43 <pickpacket> what's the openttd/newgrf directory equivalent on Windows?
11:09:58 <pickpacket> i.e. the equivalent of the ~/.local/share/openttd/newgrf/ on linux
11:16:21 *** techmagus has joined #openttd
11:17:57 *** YourOnlyOne has quit IRC (Ping timeout: 480 seconds)
11:19:39 <pickpacket> _glx_: Documents\... something?
11:22:24 <talltyler> Mine is `C:\Users\tyler\Documents\OpenTTD\newgrf\`
11:37:29 <peter1139> Not e bikes but en bikes.
11:44:57 *** techmagus has quit IRC (Quit: WeeChat 4.4.0-dev)
12:16:19 <peter1139> Fixing AIs spamming roads by... making roads automatically disappear if they're not used?
12:16:25 <peter1139> Seems like a weird solution :)
12:19:24 <LordAro> there was an patch that aged rails from the dawn of time
12:22:31 <ahyangyi> The other side of the question might be "when is it acceptable when the AI's road vehicles (ab)use other companies' roads"
12:23:35 <ahyangyi> Assuming this is a question that can be answered in AI script parameters, I wonder if I can play with weird setups such as A can use B's road, B can use C's road and C can use A's road, but not the other way around
12:29:52 <truebrain> `But, as with any QoL feature, it was ignored.`. I like how the forum keeps finding new rules for us to follow
12:30:03 <truebrain> very frustrating, the list is getting rather long and often contradicts itself
12:30:31 <peter1139> Not just the forum.
12:31:12 <truebrain> could I write an LLM to detect if a PR is a QoL change, and auto-close if it improves it?
12:32:52 <peter1139> Oh, my mistake, it was just the forum.
12:34:38 <peter1139> Now we can't make any changes if a fork is affected?
12:36:26 <truebrain> that sounds like: I did not upstream my work, and now I am mad upstream implemented a similar functionality differently?
12:36:32 <truebrain> or am I reading this wrong?
12:36:43 <truebrain> in general I kinda never understand these kind of passive aggressive remarks, I have to admit
12:41:23 <peter1139> Hmm, what are we supposed to do with Windows events that we can't process right now, but can do later?
12:45:03 <peter1139> truebrain, something like that, I think.
12:45:30 <truebrain> peter1139: Not have the `case ..:` entry? 😄
12:46:06 <truebrain> why did I write `:` there? No clue .... feels weird.
12:53:43 <peter1139> That would be not handling it at all.
12:59:31 <truebrain> Vague question vague answer, ofc 😄
13:03:27 <pickpacket> > "Fixing AIs spamming roads by... making roads automatically disappear if they're not used?" -- It sounds to me like it's the AI devs that should fix it.
13:05:12 <peter1139> truebrain, we 'fixed' the OneDrive/Dropbox or whatever it was file-saving issue by using a different Windows API call.
13:05:20 <peter1139> But I don't think that's a fix here.
13:05:40 <peter1139> Drop Windows support?
13:11:52 <efessel> Speaking of windows - if I save a game in windows with NewGRF's loaded, should that save game also be able to load in linux (thinking of search path issues) - provided the same GRF's are in the proper locations on the linux server?
13:12:29 <LordAro> saves are completely OS agnostic
13:13:13 <efessel> so if a path is mangled in a windows config, that won't matter in the savegame since only the ID is used
13:17:53 <ahyangyi> pickpacket: My question (as posted above) is whether "spamming roads" is always *undesired*.
13:17:53 <ahyangyi> Considering that the same player that complain about bots spamming roads might also complain when the bots don't spam roads and instead reuse the player-built roads...
13:20:01 <peter1139> There, I have no longer ignored this suggestion.
13:24:05 <peter1139> Why am I being set on fire?
13:26:28 <truebrain> peter1139: lol, now that is a solid reply 🙂
13:38:50 <andythenorth> truebrain: One doesn’t write an LLM, silly 🙂
13:39:07 <andythenorth> One asks an LLM to write an LLM
13:42:50 <talltyler> I replied too with the other classic answer about opening a PR instead of shouting into the void on a forum
13:48:21 <Eddi|zuHause> andythenorth: isn't "self-replicating" one of the signs of life?
13:56:14 <_glx_> ahyangyi: roads are owned but not private, so you can't really prevent road usage
13:57:35 <truebrain> Eddi|zuHause: the odd thing about any definition of what "signs of life" are, it is defined by us, humans. So what is the meaning of defining this, really? 🙂 😛
13:58:40 <Eddi|zuHause> are we channeling descartes or something?
14:15:47 <ahyangyi> _glx_: Agreed - so that's more of an AI setting, since there's nothing to do in the game itself
14:19:11 *** amal[m] has quit IRC (Quit: Client limit exceeded: 20000)
14:23:37 <peter1139> AIs could 1) be aware of the cost of building and maintaining roads (seems simple) vs reusing roads, 2) be aware of traffic levels/rail crossings to avoid congestion (not sure how they can do this)
14:24:14 <LordAro> iirc the default AI pathfinder does account for existing roads
14:24:34 <LordAro> obviously the default cost for them is too high
14:25:09 <LordAro> accounting for rail crossings should be relatively doable, at least on construction
14:25:19 <LordAro> traffic levels is a fair bit harder, yeah
14:52:23 <LordAro> """Having to even think about the $^S variable in your exception handlers is simply wrong. $SIG{__DIE__} as currently implemented invites grievous and difficult to track down errors. Avoid it and use an END{} or CORE::GLOBAL::die override instead."""
14:52:28 <LordAro> I've reached a good place in the perl documentation
15:04:32 <peter1139> The "why is perl still a thing in 2024?" place?
15:05:08 <LordAro> but also "this thing was written in ~2010 and we have to still support it"
15:05:57 <peter1139> Apparently I'm allowed to finish early tomorrow. Not sure why.
15:07:07 <peter1139> Probably some sportsball thing?
15:07:31 <peter1139> Ah yes. Roads might be quiet then.
15:08:44 <andythenorth> time off for cycling?
15:09:15 <peter1139> Sounds like it to me.
15:21:06 <peter1139> Hmm, XML validation...
15:28:30 *** gelignite has joined #openttd
16:31:18 *** HerzogDeXtEr has joined #openttd
17:13:30 *** Wormnest has joined #openttd
17:40:20 <peter1139> Hmm, 100km or 100miles?
17:43:52 <Rubidium> 100 Scandinavian miles! ;)
17:51:44 <peter1139> LordAro could do that, not me so much.
17:52:19 <LordAro> would take a few days though
17:56:56 <peter1139> You ask that as if there is anything else.
17:57:29 <efessel> I live in the US though
17:57:53 <efessel> Lots of roadies in here?
18:27:58 <Eddi|zuHause> scandinavian mile is 10km, or so?
18:30:17 <efessel> Have to account for all the climbing
18:30:52 <Eddi|zuHause> 1 german mile is 7,5km
18:31:05 <Eddi|zuHause> but nobody really knows this anymore
18:32:52 <Eddi|zuHause> it isn't helped that on old signs distances are usually written in "hours" rather than "miles" (where 1 mile = 2 hours)
18:33:41 <Eddi|zuHause> that results in a walking speed of just under 4km/h
18:47:16 <Timberwolf> Rural UK has the occasional instance of very old signposts for "place x: 3 1/2 mi", you drive 4 miles and get a "place x: 2 3/4 mi"
18:49:08 <rutoks> What do you think about adding an HTTP API method to Bananas API to download actual archive with the data?
18:49:08 <rutoks> Currently if I want to install something to my game via a script, the best way I found was starting the game via console and using `content` console command via stdin/stdout.
18:54:53 <LordAro> BaNaNaS licensing agreement only allows us to distribute content via OTTD
18:55:57 <rutoks> Maybe you should close it with this info as well?
18:56:24 <LordAro> i could be wrong about the above, of course
18:56:42 <LordAro> TB wouldn't have raised that issue if that were the cast
18:57:02 <LordAro> there was definitely something about distribution rights somewhere
18:59:45 <Eddi|zuHause> rutoks: it was difficult enough to convince some of the grf creators to agree to the current terms of automatic delivery. and some of them never agreed, making some especcially very old content weirdly difficult to access
19:00:51 <Eddi|zuHause> i think the biggest problem is delivery of older versions to people that don't have a savegame that demands it.
19:06:45 <Timberwolf> I have some frustrating old savegames from the early ttdpatch era where the grf config didn't get baked into the savegame.
19:07:15 <Timberwolf> In theory I could load them if I could work out what combination of DBSet/George's/whatever else I used when I made them...
19:07:29 *** Flygon has quit IRC (Read error: Connection reset by peer)
19:36:47 <truebrain> LordAro: You are not wrong, but not completely right. The ToS allows us to distribute the latest version of the content (as in, via any channel; we never mention how that distribution happens). But any older content can only be distributed for "savegame compatibility", which basically means only via in-game 🙂
19:37:15 <truebrain> so all content via in-game: yes. content via a HTTP .. partially 🙂
19:50:48 <truebrain> mostly because authors didn't want their "old broken content to still be available", which I can understand
19:51:17 <truebrain> also the reason no http download exists (yet); it is hard to make a system that only allows for new content without indexers keeping old links till the end of times to the older content
19:53:59 <Timberwolf> Too right, I want people using my new broken content!
20:03:34 *** andriydohniak has quit IRC (Quit: User went offline on Discord a while ago)
20:17:34 *** nielsm has quit IRC (Ping timeout: 480 seconds)
20:30:14 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
20:32:47 <Eddi|zuHause> newer is always better.
21:09:34 <reldred> In the early days people uploaded our old broken shit they scraped off the forums to BaNaNaS and it caused us a tonne of grief having to tell them over and over again it was already fixed
21:10:05 <reldred> Or, in those early days, was broken because OpenTTD didn’t yet have a full newgrf implementation.
21:33:27 *** keikoz has quit IRC (Ping timeout: 480 seconds)
21:50:31 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
21:56:47 <Eddi|zuHause> i don't think bananas existed yet, because we were very active to remove any stuff that wasn't uploaded by "the original author"
23:03:17 *** gelignite has quit IRC (Quit: Stay safe!)
continue to next day ⏵