IRC logs for #openttd on OFTC at 2023-07-30
β΄ go to previous day
00:51:52 *** wallabra has joined #openttd
00:52:31 <wallabra> it involves every pictured bridge, except the ones on the left which are something else
01:17:45 <wallabra> ugh it's so hard to have outgoing trains able to go to a depot without getting in the way of other outgoing trains
01:55:51 *** Wormnest has quit IRC (Quit: Leaving)
02:15:36 *** debdog has quit IRC (Ping timeout: 480 seconds)
02:19:35 *** wallabra has quit IRC (Quit: Page closed)
05:18:41 *** bryjen has quit IRC (Quit: Leaving)
06:22:59 *** D-HUND is now known as debdog
07:35:40 <truebrain> okay, so the workflow problem isn't actualy the fault of the workflow .. it is actually the fault of the upload to CDN part
07:35:53 <truebrain> it created a file called `internal` of type `inode directory`
07:52:29 <truebrain> I leave it to others to do the announcements on the other platforms, like Discord, etc
07:52:55 <truebrain> CDN, GoG and Steam all have the release now; I am trying to make a news post for Steam, but it doesn't allow me to login ..
07:55:27 <truebrain> untested, but I think this is right; could use some extra eyes on it .. I hate `find` π
07:59:53 <truebrain> Steam news done too
08:02:28 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
08:17:14 <truebrain> about 92% of the BaNaNaS traffic goes via http; never knew it was that much of a success π
08:26:32 *** keikoz has quit IRC (Read error: Connection reset by peer)
09:01:46 <truebrain> BaNaNaS in-game content of the last 6 hours
09:34:17 <truebrain> based on 5 minutes, so rubbish, but you get the idea π
10:17:39 <truebrain> TrueBrainviaGitHub: _glx_ in summary, if you could review that PR, would be lovely. And I left it to you to make the Discord announcement, as you kinda deserve the credits π
11:03:51 <_glx_> It's so hard to find the right man page
11:04:14 <truebrain> `If no paths are given, the current directory is used.` yup π
11:04:24 <truebrain> I did test the find and cut part π
11:09:09 <truebrain> so approve yes yes? π
11:18:06 <_glx_> Yeah tested in termux on the phone
11:24:20 *** virtualrandomnumber has joined #openttd
11:25:01 *** virtualrandomnumber has quit IRC ()
11:33:48 <truebrain> funny, even without announcement, you see that 50% of the clients already run 13.4 .. Steam ftw? π
11:34:22 <truebrain> server-list is funny; all 13.4 server are non-dedicated π
11:59:47 <talltyler> _glx_: I gave you the Town Crier role so you should be able to post in Discord channel #announcements now
13:04:05 <truebrain> Stats from last 3 hours
13:19:44 <_glx_> wondering if rerunning cdn upload job is a good idea
13:19:54 <truebrain> why would you want to do that?
13:20:06 *** emperorjake has joined #openttd
13:20:06 <emperorjake> Only 5% JGR players, thought it would be more than that
13:20:11 <truebrain> internal files are already there
13:20:24 <truebrain> emperorjake: only people on this Discord think so π π π
13:21:11 <truebrain> _glx_: I already manually corrected the issue; so no need to rerun anything π
13:22:38 <_glx_> ok I won't touch that then π
13:22:54 <truebrain> a rerun will also fail btw; I made it so files can't be overwritten
13:22:57 <truebrain> just to prevent .. accidents π
13:23:32 <truebrain> Cloudflare is working on decent RBAC for R2 buckets .. but till then, I have added some glue myself that prevents from the most common mistakes π
13:23:40 <truebrain> (overwriting, deleting, wrong folder, etc)
13:50:22 <talltyler> truebrain: Is this public servers?
13:50:48 <talltyler> Old versions donβt have the survey so it wouldnβt be conclusive to everybody playing OpenTTD?
13:52:10 <_glx_> that's only for bananas accesses IIRC
13:52:17 <truebrain> If you scrolled slightly up, you would have remembered I was talking about BaNaNaS server stats π π
13:52:29 <talltyler> Ah, I did not scroll slightly up π
13:52:40 <truebrain> That was obvious π π
13:53:04 <truebrain> And this is a low sample size; better stats in a week π
13:53:13 <talltyler> Is JGRPP participating in the survey? That would be the best way to know how many people play it va vanilla
13:54:14 <truebrain> Anyway, not processing any survey results till we release 14.0, so that will be a whole before I say anything about that
13:54:41 <truebrain> Whole = while, autocorrect, grrr
13:56:07 <truebrain> So for now BaNaNaS and multiplayer will have to be an indication π
13:56:27 <talltyler> Heh, I am apparently not awake enough to participate in an intelligent discussion yet π
14:00:50 <truebrain> Assuming anything I said is intelligent is a bit of a stretch honestly
14:23:59 <belajalilija> I imagine you guys get this a lot
14:24:09 <belajalilija> But is shunting technically possible
14:24:43 <Eddi|zuHause> depends on your definition of "shunting", "technically", and "possible"
14:25:08 <Eddi|zuHause> there was once a shunting patch, which is probably very out of date
14:25:13 <belajalilija> I mean like having marshalling yards or adding trains to the game
14:25:28 <emperorjake> There was an experimental patch for it
14:27:03 <belajalilija> Is it something that is having any work on it or is it just one of those βthatβd be coolβ things like DVTs?
14:35:51 <talltyler> Having played around with the shunting patch, my opinion is that code-wise itβs feasible but trying to figure out a user-friendly interface to control it is the biggest problem. I had lots of problems with trains driving backwards, crashing into themselves, etc., because my orders were not bulletproof enough for cases when all station tracks were full, another train was using the wye (triangle),
14:42:01 <peter1138> Yes, UX is the main issue... for something as simple as "shunting, woo!" it is a complex task with ultimately no gameplay purpose.
15:01:16 <andythenorth> peter1138: Fake it with vehicles-in-vehicles π
15:01:26 <talltyler> βGameplayβ as defined by Chris Sawyer anyway
15:01:39 <andythenorth> There was a patch for v-in-v once
15:01:40 <talltyler> I think for many people the roleplay is the gameplay
15:01:49 <peter1138> Yes. it's model-railway gameplay. Fun, but doesn't add anything economy wise π
15:02:05 <talltyler> Yes, but we all gripe about how the economy is broken π
15:02:35 <talltyler> Itβs just pixel trains, as Andy helpfully reminds us sometimes π
15:02:42 <peter1138> And I don't mean it's useless or pointless, but I can't see how you would begin to design a UI for shunting.
15:02:57 <peter1138> (Can I shunt trams?)
15:03:09 <talltyler> Agreed. The existing patch uses orders, but it is really frustrating
15:03:53 <peter1138> To me shunting seems to be a "do what's needed with whatever is available" option.
15:04:27 <Eddi|zuHause> drive-backwards-trams when?
15:04:30 <peter1138> It needs extra tracks that you don't need without shunting.
15:04:48 <peter1138> And the scale is kinda wrong too.
15:05:13 <Eddi|zuHause> tracks are too wide
15:05:21 <talltyler> Yes. Iβve done a lot of shunting in real life and it takes a while to develop the mental skills to plan the entire operation in your head, in advance. When itβs people doing it, thatβs fine because you can adapt. But with orders, the train will keep following your instructions even if theyβre wrong.
15:05:23 <Eddi|zuHause> space between tracks is wasteful
15:06:06 <Eddi|zuHause> theoretically, 4 vehicles fit side-by-side on one tile
15:11:07 <peter1138> Hmm, actually there is one gameplay reason I can think of for shunting...
15:11:52 <peter1138> Instead of having multiple trains for a slow-loading industry, have loads of wagons, shunt them around so that some are always loading, and have just one engine doing the actual transporting.
15:12:14 <peter1138> No idea how to express that in an order list π
15:15:32 <Eddi|zuHause> shunting can be used to distribute FIRS supplies
15:17:13 <_jgr_> I expect that that would be excessively fiddly
15:18:50 <_jgr_> I always assumed that the main use case would be moving the locomotive to the other end for locomotive-hauled services, instead of the magic flip that you get at the moment
15:19:22 <_jgr_> Some players really don't like magic reverse behaviour
15:23:24 <merni> I would personally use it for through-coach services on passenger trains because I am a "model-railway gameplay"er :p
15:40:37 <Eddi|zuHause> one of the use cases i always want is switching electric/diesel services for long-running trains, when i don't want to electrify the whole network
15:44:28 <merni> long diesel under wire is pain
15:51:34 <emperorjake> Or attaching a banking engine on steep sections
15:52:12 <andythenorth> Consist change rules
15:52:44 <andythenorth> βGo to A and use consist xyzβ
15:53:14 <andythenorth> βGo to B and change to consist opqβ
15:57:44 <andythenorth> Shunting means we can throw out 50% of grf? π
15:59:30 <Eddi|zuHause> and then add 250% more?
16:00:09 <Eddi|zuHause> shunting means you can't reliably query position in consist
16:02:37 <Eddi|zuHause> you'd probably have to divide the train into immutable units, and the "legacy" consist queries (including PARENT) would have to stay within that unit
16:09:05 <andythenorth> Also shunting means most of cb 36 needs banned
16:09:11 <belajalilija> Iβd honestly really love up simulate the Danish rubber nose trains splitting
16:09:29 <belajalilija> talltyler: Exactly
16:09:43 <andythenorth> Start stop cb and wagon attach cb need banned
16:10:13 <andythenorth> And weβd want shunting for RVs also
16:11:10 <andythenorth> Do vehicles load cargo in a station if thereβs no engine attached?
16:14:05 <Eddi|zuHause> why shouldn't they?
16:14:39 <alfagamma_0007> shunting orders seem very compllicated
16:14:51 <alfagamma_0007> Now I get it why it's impractical here
16:15:21 <alfagamma_0007> since you can spam as many locos you want anyway in openttd , it actually has no use case
16:15:49 <alfagamma_0007> I do find shorter freight trains frustrating at times
16:23:53 <andythenorth> Trailer trucks via shunting
16:24:26 <andythenorth> Also branch lines with one engine, but always wagons waiting at pickup station to load
16:24:34 <andythenorth> Also boat shunting
16:29:25 <Eddi|zuHause> loading wagons on a boat, and ferrying them across..
16:30:52 <andythenorth> I mean tugs and unpowered barges
16:31:29 <andythenorth> Hmm canβt shunting just be done in grf? With invisible vehicles?
16:32:01 <merni> how would the player instruct them to couple and decouple?
16:32:44 <andythenorth> i.e. shunting is just a livery
16:32:49 <merni> wouldn't that just delete cargo rather than transferring it to a different train
16:33:44 <andythenorth> I think GS could do shunting
16:34:42 <andythenorth> Yeah GS can do it in depots. Not at stations.
16:35:21 <merni> I mean you can do shunting in depots manually too :p
16:50:39 <Eddi|zuHause> abbreviation for "callback", a very common NewGRF feature
17:43:29 *** Wormnest has joined #openttd
18:38:19 <DorpsGek> - Update: Translations from eints (by translators)
18:53:55 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
19:15:51 *** Wormnest has joined #openttd
19:36:05 *** virtualrandomnumber has joined #openttd
19:36:18 *** virtualrandomnumber has quit IRC ()
19:58:01 <andythenorth> So is there actually a meta to 2048? Iβve just won it 3 times in a row
19:59:32 *** shrekshellraiser has joined #openttd
19:59:32 <shrekshellraiser> Yes, don't know if it guarantees success but the strategy is to keep you big pieces in a corner
20:08:18 <Eddi|zuHause> andythenorth: you just ban one of the 4 directions
20:09:17 <Eddi|zuHause> so like "never go down"
20:09:34 <Eddi|zuHause> then naturally the highest tile will be top-left or top-right
20:10:08 <andythenorth> Thatβs fine as a basic approach, but I think thereβs more to the meta than that
20:11:40 <andythenorth> I canβt describe accurately the approach I found, but it involves using both ends of the second row (lower value tiles) and making sure thereβs always a way to clear any 2 or 4 that drops
20:12:27 <andythenorth> Thereβs probably a video if I bothered to search π
20:14:06 <truebrain> optimizing a game meant to bore your brain
20:26:07 <Eddi|zuHause> that both 2 and 4 can spawn is what makes the game hard
20:29:36 <_glx_> that was an easy to fix bug π
20:31:49 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:04:48 *** Smedles_ has quit IRC (Read error: Connection reset by peer)
21:06:47 *** Smedles has joined #openttd
21:13:08 <LordAro> wait, wasn't that done years ago?
21:14:19 <talltyler> It came up in a help thread here on Discord and I was surprised, then decided to fix it π
21:14:52 <LordAro> i predict a forum thread by morning
21:15:13 <_jgr_> I'm curious why the post_cb is `InvalidateShipPathCache`
21:15:18 <_jgr_> That doens;t sound relevant at all
21:17:10 <_glx_> IIRC the setting applies to all pathfinders
21:18:16 <truebrain> talltyler: we could wait till 14.0 is released to see what people use as settings π
21:20:46 <_jgr_> _glx_: It doesn't appear to be referenced in any non-train contexts that I can see
21:21:18 <truebrain> I was about to comment the same π
21:22:44 <truebrain> very odd invalidation
21:23:28 <emperorjake> I thought the setting said to forbid trains and ships from doing 90 degree turns
21:24:06 <truebrain> guess it got removed from ships at some point π
21:24:33 <truebrain> `fb6e31ca43b10b40fcaf1db0596dfaa48a1ecf20` removes the setting from ships
21:24:59 <truebrain> and not uncommon for that patch author, it is a very local patch π
21:25:19 <truebrain> `3f327116d` is the actual commit removing it
21:26:19 <truebrain> so yeah, that Invalidate can be removed too π
21:26:28 <_glx_> my memory just forgot this change π
21:26:38 <_glx_> while I even commented in the PR
21:27:41 <truebrain> didn't have to be done in this PR π
21:27:56 <talltyler> Would you like me to open a separate one?
21:28:05 <truebrain> I counted .. zero fucks were given either way π
21:28:42 <truebrain> just remember to rebase when merge, not squash π
21:29:26 <truebrain> counting to zero was really difficult π¦
21:29:51 <truebrain> I kept falling asleep!
21:30:14 <_glx_> be sure to not overdlow
21:30:21 <truebrain> underflowing was the issue π¦
21:31:15 <_glx_> steam really helps updating on new release
21:32:31 <truebrain> just a random person or two in the last 30 minutes
21:32:39 <truebrain> downside of such low amount of time
21:33:00 <truebrain> can't wait till next week, and see some actual representable values π
21:33:48 <truebrain> but still, the transition to new versions is insanely fast these days
21:34:25 <_glx_> too bad the bananas protocol doesn't care about the OS π
21:34:33 <truebrain> yeah ... not much data to go on π
21:34:36 <truebrain> but for that we have the surveys
21:35:01 <_glx_> most likely the oldest version are from linux app managers
21:35:26 <truebrain> what is funny about these stats, this means someone is actively using that client version
21:35:35 <truebrain> I mean, someone had to press the "Online Content" button
22:05:59 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
22:13:58 *** Kitrana has joined #openttd
22:19:51 *** Kitrana1 has quit IRC (Ping timeout: 480 seconds)
22:32:42 *** keikoz has quit IRC (Ping timeout: 480 seconds)
22:34:48 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
23:50:22 *** ChanServ sets mode: +v tokai
23:57:27 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
continue to next day β΅