IRC logs for #openttd on OFTC at 2024-02-07
β΄ go to previous day
00:06:02 <talltyler> (forgot to update a tooltip π )
00:13:27 <talltyler> Oh, #12022 is working as intended. Economy date is not wrong, it just didn't start at 1 because the game was started in calendar mode
00:14:00 <talltyler> Still, it might be possible to resolve more nicel
00:20:53 <talltyler> Goes it throw out implicit orders?
00:22:16 <talltyler> (Joking obviously, but I have a patch to enable/disable them in the GUIβ¦ forget how I got stuck on that π )
00:23:41 <_glx_> well with "new orders are non-stop" being enabled for new openttd.cfg, they should be less frequent
00:24:31 <_glx_> except when ctrl+clicking on a waypoint π
02:02:20 <peter1138[d]> Seems to be spamming for each wagon in the chain, instead of just the consist.
02:03:02 *** herms has quit IRC (Quit: bye)
02:16:39 <peter1138[d]> Well, it's not a new bug. It doesn't crash in 13.4 because... no asserts.
02:52:52 <peter1138[d]> So this will still issue multiple news messages, one for each wagon that fails.
02:53:31 <peter1138[d]> BuildReplacementVehicle does not know there will be more calls, of course.
02:54:35 <_glx_> pff you ask for doxygen, but don't validate π
02:54:59 <talltyler> I saw that it is present π
02:55:05 <peter1138[d]> I can work around it by checking for duplicate news messages in AddNewsItem(), but that's a bit brutal?
02:55:29 <talltyler> Those are some interesting screenshots, I assume from the crash report save. Big GUI and NewCC?
02:55:43 <peter1138[d]> Some horrible ghastly set, yes.
02:55:45 <talltyler> The game looks hideous π
02:55:54 <peter1138[d]> I think they selected every NewGRF possible.
02:56:16 <peter1138[d]> So yes, default engines, some 2cc wagons... delightful π
02:56:19 <talltyler> And that freight car alignment in the depotβ¦
02:57:51 <talltyler> `Fast bridge for the guy on Reddit`
02:58:27 <peter1138[d]> Anyway, the actual message about cargo is nothing to do with recent cargo changes, just they're autoreplacing a refittable wagon carrying goods to a livestock-only wagon.
02:59:18 <peter1138[d]> 3am, I better sleep.
02:59:38 <peter1138[d]> (Something was wrong on the Internet, and it was our code ;))
03:20:05 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:23:40 *** gnu_jj_ has joined #openttd
03:26:50 *** gnu_jj has quit IRC (Ping timeout: 480 seconds)
03:28:37 *** Wormnest has quit IRC (Quit: Leaving)
04:25:12 *** Flygon has quit IRC (Read error: Connection reset by peer)
06:20:39 <jfs> frosch123: Uh, English? If you use pence, shilling and pounds. But yeah that's not relevant to this game.
06:25:30 *** keikoz has quit IRC (Ping timeout: 480 seconds)
07:39:37 <johnfranklin> *ε£Ήθ΄°εθδΌιζζηζΎ*
07:39:37 <johnfranklin> In fact some young generations even don't know how to write Chinese currency in such "formal" way...
08:32:45 <rau117> TallTylerviaGitHub: how about additionally add ctrl+shift+click on any depot to add task "go to nearest depot and unbunch"
08:32:45 <rau117> will be useful on large depot arrays
09:32:28 *** thelounge345 has joined #openttd
09:38:12 <locosage> do h files need to #include all the stuff they use or can they just rely on include order in other place?
09:38:28 *** thelounge345 has joined #openttd
09:38:33 <locosage> like town_land.h requires house.h and table/strings.h but doesn't include them
09:40:24 <locosage> that's a code style question ofc as it obviously works both ways
09:44:00 <peter1138[d]> We don't require it at the moment.
10:55:05 <xarick> is INT32_MAX beaten now?
10:58:02 <LordAro> how long have you been able to found towns in game?
10:59:11 <locosage> looks like tmwftlb tbh
10:59:21 <locosage> you can show/hide industry types by clicking them
11:00:03 <locosage> also can go industry chains -> select cargo -> link to smallmap
11:00:46 <locosage> after that routes seem pretty easy to see even without highlighting them
11:07:49 <locosage> though highlighting routes close to a certain length can be pretty useful...
11:20:41 <locosage> yeah, icons are slowly getting their way in more places but haven't gotten to chains window yet
11:21:15 <LordAro> locosage looks like they're talking to themselves again
11:21:23 <LordAro> am i missing someone?
11:21:50 <locosage> also icons themselves are pretty bad in the current basesets imo
11:33:23 <locosage> peter added them to industry cargo filters recently
11:33:29 <locosage> and maybe somewhere else I don't remember
11:41:15 <emperorjake> at least a decade
11:47:07 <peter1138[d]> Yeah, a fairly new feature.
11:48:08 <reldred> fair short period of time in the ~~cosmic~~openttd scale
11:57:04 <xarick> i'm very surprised AAAHogEx still thrives
11:57:27 <xarick> on an inflation environment with high costs
11:58:55 <xarick> didn't try infrastructure maintentance costs enabled yet
12:02:30 <locosage> LordAro, irc logs look nothing like conversation in discord
12:02:48 <locosage> some messages missing, some in wrong order
12:03:00 <locosage> and it still doesn't understand discord display name :p
12:03:11 <locosage> discord module needs updating afaict
12:14:17 <Eddi|zuHause> i mean, irc was dying long before discord came about
12:15:31 <reldred> video killed the radio star
12:15:33 <merni> xarick: In infinite money mode you will still be able to do it...just that if an AI checks its balance before doing it, it may think it doesn't have enough.
12:16:56 <xarick> what if "someone" use base costs to multiply costs?
12:17:22 <merni> Again, it has no effect on whether they can actually do the action
12:17:37 <merni> Only on the balance value returned by the script api
12:18:53 <peter1138[d]> Just an AI might think it doesn't have enough money to do something.
12:19:13 <merni> If you really want it to work perfectly with infinite money mode... is there some kind of exception/status code/return value or something in squirrel when an action fails due to insufficient money? If so you could just do the action and check if it succeeded
12:19:14 <peter1138[d]> But that was already said π
12:19:48 <peter1138[d]> That isn't ideal, sometimes an AI might want to know if multiple actions can be done, and then not do any of them if they can't all be done.
12:19:49 <merni> Yes, I clearly mentioned the limitation in the PR and evidently two core devs thought it still worth accepting, so π€·
12:20:28 <merni> peter1138[d]: Then perhaps the best way is to expose to the API whether this setting is on
12:20:50 <xarick> they can do that already
12:21:05 <xarick> but those old unmaintained AIs.... they're toast
12:21:07 <peter1138[d]> My joke about INT63_MAX was kinda not a joke... `INT64_MAX / 2` (or some other divisor) would still be huge value and less likely overflow.
12:21:28 <merni> xarick: That is to be expected
12:21:31 <xarick> without an update to work around the issue...
12:21:44 <LordAro> definitely nothing that can be done
12:21:45 <xarick> their AIs won't exactly be safe from the problem
12:21:50 <merni> Just don't infinite money mode if you want old AIs to work
12:22:33 <LordAro> AI use with infinite money is dumb anyway
12:22:49 <LordAro> not that i would want to restrict anyone's usage of OTTD of course
12:23:00 <LordAro> AIs are already generally pretty terrible to try to compete against
12:23:12 <LordAro> why would you then also give them infinite money?
12:23:40 <xarick> see what they can do when unrestricted
12:24:05 <merni> peter1138[d]: Well, a comment to that effect on the PR might have been nice before it was merged. I went with glx suggestion and I am not inclined to do another PR just to change one constant :p
12:24:18 <LordAro> sure, but that's just something that's mildly interesting to know, it's not gameplay
12:24:24 <merni> xarick: So just code your ai to take the setting into account
12:24:43 <peter1138[d]> I had other things on π
12:25:02 <peter1138[d]> Anyway it's better currently..
12:25:06 <merni> After all pre-infinite money AIs are not really intended for seeing what they can do when unrestricted
12:26:04 <xarick> I did for mine, but... there's the problem with these radical changes... what used to work can stop working, AIs ... without maintenance, they just keep on getting worse
12:26:15 <peter1138[d]> Anyway, fairly sure Xarick has made a contrived case just to prove a point.
12:26:17 <merni> No, they work exactly as they used to.
12:26:37 <merni> If you enable a new setting that never existed before and will never be default, they work slightly worse
12:26:51 <xarick> years go by, and ppl look at the AI section of openttd and they just see they suck
12:28:02 <merni> If AI authors don't update their AI to take advantage new optional features there is nothing we can do to make those features magically work with AI
12:28:44 <merni> The only other option is to stop development entirely because any new feature, even completely optional, may not work with an old AI
12:29:32 <merni> Savegames from 13.4 will be loadable in 14.0, just as savegames from Transport Tycoon Deluxe will be
12:29:55 <merni> Apart from that not sure what kind of backward compatibility you mean
12:30:16 <merni> For multiplayer players always need to be on the same version as the server
12:31:26 <merni> We haven't broken savegame compatibility even from 1995 TTD, no way we would do it with 13.4 :P
12:31:46 <merni> At least not intentionally
12:31:48 <peter1138[d]> It has broken... but we try to fix it π
12:35:10 <belajalilija> Would TTO save games work?
12:39:33 <belajalilija> I dont have a copy
12:39:46 <Eddi|zuHause> i think there's a best-effort to load TTO savegames
12:40:14 <LordAro> James tells us if they don't
12:40:53 <Eddi|zuHause> there might be some corner cases with renoved vehicles
12:42:33 <Eddi|zuHause> i've no clue how the conversion treats jintys
13:00:17 <johnfranklin> hackish things are geeky and niche
13:07:01 <peter1138[d]> truebrain: IRC/Discord bot is being whacky (delayed/missing messages)
13:19:32 <truebrain> peter1138[d]: Most of the time that is caused by interruption of the Discord webhooks .. let me check
13:22:27 <truebrain> Seems only one person wasn't being relayed properly from IRC to Discord, but others were
13:22:46 <truebrain> That is odd .. on IRC only one bot is reading everything
13:24:04 <truebrain> Will check it out in detail when I am back home, but it looks like the Discord webhook had some issues
13:24:55 <truebrain> But for safety, restarting bridge!
13:25:03 *** kamnet has quit IRC (Remote host closed the connection)
13:25:03 *** locomotor79 has quit IRC (Remote host closed the connection)
13:25:03 *** therealsimon has quit IRC (Remote host closed the connection)
13:25:03 *** kaomoneus has quit IRC (Remote host closed the connection)
13:25:03 *** gwyd4016 has quit IRC (Remote host closed the connection)
13:25:03 *** gebik4544 has quit IRC (Remote host closed the connection)
13:25:03 *** gilgames has quit IRC (Remote host closed the connection)
13:25:03 *** o0k6 has quit IRC (Remote host closed the connection)
13:25:03 *** cptm1100 has quit IRC (Remote host closed the connection)
13:25:03 *** djdive has quit IRC (Remote host closed the connection)
13:25:03 *** olivercz1324 has quit IRC (Remote host closed the connection)
13:25:03 *** dojando has quit IRC (Remote host closed the connection)
13:25:03 *** lamarr has quit IRC (Remote host closed the connection)
13:25:03 *** xarothbrook has quit IRC (Remote host closed the connection)
13:25:03 *** rau117 has quit IRC (Remote host closed the connection)
13:25:03 *** v453000[d] has quit IRC (Remote host closed the connection)
13:25:03 *** goddess_ishtar has quit IRC (Remote host closed the connection)
13:25:03 *** georgevb has quit IRC (Remote host closed the connection)
13:25:03 *** belajalilija has quit IRC (Remote host closed the connection)
13:25:03 *** emperorjake has quit IRC (Remote host closed the connection)
13:25:03 *** janciothor has quit IRC (Remote host closed the connection)
13:25:03 *** maxshpuntozavr has quit IRC (Remote host closed the connection)
13:25:03 *** efessel has quit IRC (Remote host closed the connection)
13:25:03 *** becky_andrew has quit IRC (Remote host closed the connection)
13:25:03 *** jfs has quit IRC (Remote host closed the connection)
13:25:03 *** locosage has quit IRC (Remote host closed the connection)
13:25:03 *** merni has quit IRC (Remote host closed the connection)
13:25:03 *** eratonysiad has quit IRC (Remote host closed the connection)
13:25:03 *** bigyihsuan has quit IRC (Remote host closed the connection)
13:25:03 *** reldred has quit IRC (Remote host closed the connection)
13:25:03 *** frosch123 has quit IRC (Remote host closed the connection)
13:25:03 *** gwit has quit IRC (Remote host closed the connection)
13:25:03 *** andythenorth has quit IRC (Remote host closed the connection)
13:25:03 *** brickblock19280 has quit IRC (Remote host closed the connection)
13:25:03 *** talltyler has quit IRC (Remote host closed the connection)
13:25:03 *** _zephyris has quit IRC (Remote host closed the connection)
13:25:03 *** _jgr_ has quit IRC (Remote host closed the connection)
13:25:03 *** xarick has quit IRC (Remote host closed the connection)
13:25:03 *** truebrain has quit IRC (Remote host closed the connection)
13:25:03 *** _glx_ has quit IRC (Remote host closed the connection)
13:25:03 *** peter1138[d] has quit IRC (Remote host closed the connection)
13:25:03 *** DorpsGek_vi has quit IRC (Remote host closed the connection)
13:25:03 *** matheusvhs has quit IRC (Remote host closed the connection)
13:25:03 *** lea001 has quit IRC (Remote host closed the connection)
13:25:03 *** kuhnovic has quit IRC (Remote host closed the connection)
13:25:03 *** michi_cc[d] has quit IRC (Remote host closed the connection)
13:25:03 *** wensimehrp has quit IRC (Remote host closed the connection)
13:25:03 *** johnfranklin has quit IRC (Remote host closed the connection)
13:25:05 *** DorpsGek_vi has joined #openttd
13:25:43 *** truebrain has joined #openttd
13:25:43 <truebrain> Discord is getting busy lately .. many puppets active π
13:26:29 *** peter1138[d] has joined #openttd
13:34:34 <peter1138[d]> Queued messages still arriving π
13:47:57 *** andythenorth has joined #openttd
13:47:57 <andythenorth> I've been trialling AWS as an alternative to hosting all infra on my MBP
13:48:02 <andythenorth> the monthly bill arrived
13:48:03 <andythenorth> Total in GBP: 1.04*
13:48:51 <peter1138[d]> Why do you need that level of availability?
13:50:47 <andythenorth> well my MBP has 100% uptime of course
13:51:07 <andythenorth> so it's not really about availability
13:51:13 <andythenorth> it gets a bit warm sometimes though
13:52:51 <xarick> inflation power + base costs!
14:04:53 <peter1138[d]> Only 110 minutes late for that message π
14:06:12 <truebrain> That means it really was the Discord webhook, and not my software π
14:06:59 * LordAro has no idea what is being referred to
14:07:44 <peter1138[d]> Eddi's "not really" comment came through 5 minutes ago.
14:08:56 <xarick> i thought he was replying to me
14:22:42 <LordAro> do you ever think that maybe we should test stuff a bit?
14:23:12 <_glx_> we have users for that, too bad we need to make beta for them to test
14:26:18 *** talltyler has joined #openttd
14:26:19 <talltyler> Why was this not possible before? The hotkey to open the cheat menu didnβt do anything in a network game?
14:29:53 <truebrain> Was about to say, that functionality is unchanged .. internally nothing changed with the rebranding .. so wuth?! π
14:30:16 <xarick> it's in the dropdown options
14:30:18 <peter1138[d]> Presumably it only affects the client, nothing is sent over the network.
14:30:38 <truebrain> I didn't guard the menu entry
14:30:47 <peter1138[d]> So probably it was just the hotkey not working that prevent it π
14:30:52 <truebrain> Lol, that is new ofc π
14:30:57 *** locosage has joined #openttd
14:30:57 <locosage> hotkey still doesn't work in mp afaict xD
14:31:13 <truebrain> Okay, that is an easy fix .. lol, that is actually an oops π
14:31:37 <locosage> but, yeah, cmclient has cheats in mp for ages, they're only useful to desync yourself
14:31:44 <locosage> and maybe confuse some server owners xD
14:32:31 <peter1138[d]> Another argument for moving them to settings, where network is already guarded with appropriate flags.
14:33:16 <peter1138[d]> (But not for 14.0 I think)
14:35:02 <xarick> dissolve into settings
14:37:47 <xarick> can AIs still purchase each other?
14:37:55 <xarick> shares were removed, right?
14:42:26 <xarick> think i found a bug, not sure
14:42:32 <xarick> mail belongs to "freight"?
14:43:13 <peter1138[d]> Freight is "not CC_PASSENGERS"
14:44:19 <xarick> im not sure that was the case, let me check 13.4
14:44:37 <peter1138[d]> Oh you're right, there's an actual attribute for "is_freight"
14:45:26 <peter1138[d]> And MAIL is not freight.
14:51:10 <locosage> mail is not affected by freight multiplier iirc
14:57:02 <xarick> yay, this AI is a serious station spread abuser
14:57:23 <peter1138[d]> Lower station spread then π
14:59:53 <xarick> 34 aircraft going to that airport
15:00:04 <xarick> and still unable to empty the station
15:00:23 <peter1138[d]> Turn off cargodist π
15:00:45 <xarick> and linear cargo too in towns
15:01:25 <peter1138[d]> Oh, using transfers?
15:01:58 <xarick> only station spread doing its thing
15:08:43 <locosage> isn'n dynamite area clear?
15:09:28 <xarick> that's not an easy fix when you get to locks
15:10:18 <xarick> also, river is gonna disappear now, right?
15:10:37 <_glx_> yes, it's expected when you clear an area
15:11:20 <xarick> it's gonna clear sea too? I somewhat see that being prohibilively expensive
15:11:25 <locosage> is it even possible to remove ship depot without area clear? xD
15:11:36 <rau117> _glx_: *if I want a dock for ships, I want to destroy it without affecting the river*
15:12:08 <_glx_> you can still remove the ship depot without affecting the water tile under it
15:12:31 <_glx_> single tile vs area tile (even if both use the same command)
15:13:54 <locosage> yeah, single tile dynamite has special handling
15:14:10 <locosage> probably fine then xD
15:14:13 <rau117> alt+click/drag to destroy only your infrastructure, without affecting the environment
15:15:01 <locosage> there was some pr with dynamite options but it was rejected
15:47:44 <xarick> AAAHogEx achieved stagnation
15:48:13 <xarick> wondering how it will manage the next ~50 years
15:48:44 <LordAro> doesn't look very stagnant
15:49:05 <xarick> cargo delivery isn't growing, so... inflation will do its job
15:49:25 <xarick> my prediction is a decrease in cargo delivery
15:49:37 <LordAro> doesn't inflation stop at 2050 as well?
15:52:44 <xarick> maybe what I can see happening is a switch to faster aircraft / transportation
15:53:04 <xarick> instead of cargo quantity focus, to keep up with profits? im not sure
15:55:40 <xarick> he had a huge network of railway
15:56:05 <xarick> amazing how he managed to upgrade the majority to monorail
16:05:51 <peter1138[d]> Sounds more like you need a designer, not an HTML coder.
16:30:07 <xarick> he's slowly transitioning from monorail to maglev, but the map is so full now
16:39:34 <Eddi|zuHause> plenty of "space" not used for infrastructure
16:51:55 <xarick> I think AIs can't demolish houses
16:52:15 <xarick> limits their expansion capabilities
16:52:48 <xarick> no room to place rails
16:53:44 <peter1138> They should be able to clear houses if towns let them.
17:03:28 <peter1138[d]> ScriptTile::DemolishTile?
17:09:22 <xarick> strange, they can now?
17:09:49 <peter1138[d]> Heh, such out of order π
17:10:24 <peter1138[d]> Pretty sure that call is not new.
17:11:12 <peter1138[d]> It's in the initial commit for NoAI π
17:16:01 <peter1138[d]> More cargo icons...
17:17:18 <xarick> Am I imagining things?
17:17:30 <xarick> I was almost sure they couldn't demolish houses
17:18:51 <peter1138[d]> CanDeleteHouse() seems to have something about IsValidHumanID
17:19:38 <peter1138[d]> For AIs it checks a callback, or checks a building flag.
17:20:11 <peter1138[d]> But that flag is never set for default houses)
17:22:23 <xarick> there is only one CanDeleteHouse
17:22:36 <xarick> and town_cmd.cpp calls it
17:29:17 <xarick> how drastic would it be, to let AIs demolish houses
17:30:26 *** Extrems has joined #openttd
17:30:36 <xarick> No idea, let me check blame
17:30:59 <peter1138[d]> How about checking actually doing it?
17:32:12 <xarick> I checked the code, town_cmd.cpp has calls to CanDeleteHouse
17:32:55 <peter1138[d]> So? CanDeleteHouse() doesn't just say NO to AIs.
17:34:08 *** HerzogDeXtEr has joined #openttd
17:38:03 <xarick> no more blames prior to that
17:38:20 <peter1138[d]> Why highlight that line?
17:39:09 <xarick> do i look at the callback then?
17:39:30 <peter1138[d]> Are you using NewGRF houses?
17:39:44 <peter1138[d]> Then the callback is irrelevant.
17:40:34 <xarick> okay let me create a quick AI for testing this π
17:50:07 *** Wormnest has joined #openttd
17:53:55 <xarick> gee, there is no IsHouseTile function, that is sad
18:02:40 *** gelignite has joined #openttd
18:04:28 <xarick> wow, they can ! wow that is news to me
18:04:44 <xarick> wow, all this time I thought houses were untouchable
18:06:38 <xarick> or it would only destroy 1 or 2 houses
18:07:04 <xarick> but still, why's the check for IsValidHumanID
18:08:54 <andythenorth> peter1138[d]: I probably need a life coach
18:13:01 <xarick> lets use EternalLove GS to test something
18:14:32 <xarick> with outstanding rating, lenient authority lets it town be fully demolished
18:24:59 *** kuhnovic has joined #openttd
18:24:59 <kuhnovic> Now you can demolish the entire map, place river tiles and create a million ship depots! π
18:30:35 <xarick> i still dont get the IsValidHumanID check, hmm
18:30:46 <xarick> seems humans still have an advantage
18:35:01 <DorpsGek> - Update: Translations from eints (by translators)
18:35:39 <peter1138[d]> Given `CBID_HOUSE_DENY_DESTRUCTION` only applies to AIs, it's probably a misfeature that we implemented it.
18:36:06 <andythenorth> I think I deleted my reddit sock puppet account, but eh
18:36:19 <peter1138[d]> Our implementation predates NoAI.
18:36:41 <peter1138[d]> So it was built in a time when AIs could cheat and what-not.
18:37:15 <peter1138[d]> It's so old, TTDPatch has extra functionality for the callback when OpenTTD doesn't implement.
18:58:47 <xarick> no signs of slowing down
19:00:59 *** olionkey has joined #openttd
19:00:59 <olionkey> They were an active account too
19:01:53 <talltyler> Huh, maybe lost control of their account
19:02:05 <olionkey> So i kicked them instead of banning
19:02:16 <olionkey> though they didn't write anything since 2018 so :Shrug:
19:09:00 *** Webster has quit IRC (Ping timeout: 480 seconds)
19:21:19 *** johnfranklin has joined #openttd
19:35:28 *** wallabra has quit IRC (Quit: Bowserinator is wrong don't blame iczero)
19:36:08 *** wallabra has joined #openttd
19:51:01 *** michi_cc[d] has joined #openttd
19:51:01 <michi_cc[d]> Hmm, reddit for beta 3 is mostly (well, 2/3): debunching doesn't work.
19:51:29 <michi_cc[d]> Could very likely be user error too, like not having shared orders or stuff.
19:52:36 <jfs> someone in a steam discussion thread:
19:53:19 <michi_cc[d]> If the order window wasn't already a complete UI mess, it might be good to have some kind of warning shown when the order list is only a single vehicle.
19:53:40 <michi_cc[d]> jfs: Well, not anymore since beta3 π
20:01:19 <andythenorth> orders window is good
20:01:26 <andythenorth> you just have to learn not to see most of it
20:16:05 *** gilgames has joined #openttd
20:16:05 <gilgames> In the beta, you can't drag signals over a split/junction and that's a huge problem for me at least. Also it took away the simple drag option, if I want to have a path signal before the station and blocks on the line, now I have to put down a new signal and change that. Instead of drag the path a bit and click on the new one and Ctrl click that to be a block
20:16:20 <gilgames> Another thing is from past versions too, if you place a depo where straight and diagonal meets, it doesn't connect one bit of tracks
20:19:34 <_glx_> removing the randomness ?
20:19:44 <kuhnovic> Read the description π
20:22:05 <kuhnovic> Sorry xarick , I might be making your PR redundant. But you were right on this one: returning INVALID_TRACKDIR was the wrong thing to do in this case. I was too set on "if there is a high level path then there must be a low level path", but there's always an edge case!
20:23:37 <truebrain> lol, that was the wrong ticket number
20:24:41 <kuhnovic> _glx_: The TL;DR is that ships can get stuck when no low-level path is found. And it turns out it's not a good idea to return INVALID_TRACKDIR, that can lead the ship to reversing back-and-forth indefinitely.
20:25:16 <kuhnovic> I'm actually making it a tiny bit more random! π
20:26:21 <truebrain> one bug per report btw π
20:26:50 <gilgames> truebrain: Idk, if it's a bug or not. Also won't do that
20:29:13 <_glx_> drag signal behaviour is intended IIRC
20:29:43 <_glx_> old ctrl-drag became drag
20:38:09 <truebrain> _glx_: it is weird as fuck btw; but reading the PR, I am also a bit clueless why we merged it in the first place π
20:38:47 <LordAro> gilgames: this isn't a good place to file issues, can you do it on github please?
20:39:13 <truebrain> think LordAro was scrolled up π
20:39:30 <gilgames> No, GitHub is Chinese to me
20:39:54 <LordAro> Github is many things, Chinese it is not
20:40:41 <gilgames> I can play OpenTTD, I'm noob to any forums and similar things
20:40:48 <gilgames> Barely can use discord
21:02:28 *** gelignite has quit IRC (Quit: Stay safe!)
21:18:38 <talltyler> Signal autofill has always stopped at junctions and stations. It says so in the tooltip.
21:21:29 <_glx_> seems some users are used to the old non-autofill behaviour
21:32:15 *** craz_oatmeal has joined #openttd
21:32:15 <craz_oatmeal> gilgames: This is literally the entire form dude
21:32:15 <craz_oatmeal> Tbf though it *would* be nice to have an option that doesn't require making a Github account - maybe a form right on openttd.org that has a bot file the issue on the submitter's behalf. But then the submitter wouldn't be able to comment with additional info if needed, so I can see it just leading to a lot of wasted time
21:32:22 <gilgames> _glx_: This version came out in a week, why would anyone be used to it and not the old one?
21:33:23 <_glx_> I think many users knew about ctrl drag
21:36:04 <truebrain> talltyler: The problem is that the autofill is on by default with no visual indicator that the functionality changed in any way, shape or form to this new default. So you drag like always, and no signals appear. That is just bad UX π
21:36:13 <_glx_> but yeah, autofill might be surprising
21:36:37 <truebrain> A simple solution would be to not drag the white stuff this far
21:36:48 <_glx_> like you drag 3 tiles and you get signals everywhere
21:36:48 <truebrain> Just stop that after 3 tiles or so
21:37:17 <truebrain> But there is zero visual indicator the CTRL behaviour is now default
21:37:38 <_glx_> except the tooltip, and nobody reads them π
21:37:50 <truebrain> Text is not a visual indicator π
21:38:41 <truebrain> This is just a poorly thought out change of default forced on everyone
21:39:06 <truebrain> Different from changing a default setting, which can be changed back
21:39:32 <gilgames> _glx_: Wdym???? Most players never heard of it.. eh, whatever
21:39:57 <gilgames> _glx_: That's the problem
21:40:11 <truebrain> Ideally we make the visual indicator show the whole rail that will be signalled
21:40:18 <truebrain> That way it is clear what is going on
21:41:08 <gilgames> Yeah that's one part of it
21:41:13 <truebrain> But honestly, I am more in favour of just reverting this PR and first fix the visual issues
21:41:14 <_glx_> original intent of the PR was to invert behaviour (without setting), and it ended with total removal of old drag
21:41:51 <truebrain> Even inverting is just hostile to current players, as there is no visual indication which mode you are in
21:42:25 <truebrain> Like changing D with R in your car without telling you overnight, or something
21:42:48 <_glx_> dunno I don't have D in mine π
21:42:56 <truebrain> You get what I mean π
21:43:15 <gilgames> truebrain: The idea was fine, 80% of the player base had no idea about Ctrl drag.
21:43:25 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:43:30 <truebrain> That 80% is made up
21:43:49 <truebrain> Don't strongforce your point by making shit up
21:44:06 <gilgames> 95% of players I met didn't knew it
21:44:31 <truebrain> Again, don't do that; it is not helpful to the conversation π
21:44:50 <gilgames> Then let's end it. Bye
21:46:02 *** nanapipirara has joined #openttd
21:46:02 <nanapipirara> I only learned about ctrl drag recently by watching someone play on youtube and constantly think βhow the hell did they do that!?β
21:46:08 <nanapipirara> I want to do that too!
21:46:13 <truebrain> What a way to have a conversation, lol. Anyway, I think a revert is in order and a bit more thought should be put in this. I am not against changing the default, but this is just a bit weird way of going about it π
21:46:27 <nanapipirara> And had to search a bunch for the shortcut. It wasnt super obvious
21:46:40 <truebrain> nanapipirara: Haha, sounds like how most of us learn anything about OpenTTD π
21:46:41 <nanapipirara> And have been playing on and off for years π₯Έ
21:46:55 <nanapipirara> Thatβs just me though π€
21:47:03 <_glx_> when you know something exist but don't know how, just try with ctrl π
21:47:24 <truebrain> nanapipirara: Also why there is a good argument to change this default behavior π
21:47:42 <truebrain> We just need to do that with a visual indicator what is actually going on π
21:47:51 <nanapipirara> Whatever the new direction is going to be. I feel it needs to be undo-able in some way. Or at least be suuuper clear in a preview before it fills the rails with items you cant simply delete again
21:48:36 <_glx_> you can delete the same way
21:48:50 <truebrain> But you can't delete a small portion anymore
21:49:05 <truebrain> (Again, without visual feedback)
21:49:32 <truebrain> It is really weird you select 2 signals for deletion and they are all removed
21:49:55 <truebrain> (Without telling me that is happening)
21:50:44 <nanapipirara> Sounds like clear visual feedback is whatβs needed for this whole thing π€
21:50:58 <nanapipirara> Good luck devs ππββοΈ
21:51:43 <_glx_> ideally the selection should show where it will apply
21:52:10 <truebrain> if we want this to be a success, that is one of the ways to fix it
21:52:14 <truebrain> I still don't like it, personally
21:52:18 <truebrain> but at least it is visually clear what happens
21:52:47 <_glx_> and it should be fixed to allow drag removal without auto fill
21:53:11 <_glx_> it's clearly a needed thing
21:53:52 <andythenorth> if in doubt revert? π
21:56:32 <_glx_> if we have a nice visual feedback it might be possible to do autofill with minimal drag, but switch to non autofill when draging more
21:59:19 <xarick> I'm out of touch with this feature
21:59:34 <xarick> didn't know it existed
21:59:53 <truebrain> if anyone can produce a better solution before 14.0, I am not against that at all
22:00:02 <truebrain> that PR I just made is a: OWH SHIT WE RUN OUT OF TIME, PR
22:00:04 <truebrain> more than anything else
22:00:20 <truebrain> locosage: this is nothing against your work; I see in the PR you fought for a better solution
22:00:24 <locosage> if only all other features were reverted after a single user complaint π
22:00:30 <truebrain> but ... seems the devs were a bit too focused on the change, more than the current players π
22:01:41 <truebrain> (which, to be clear, I could have been one of π )
22:01:58 <_glx_> yeah autofill is clearly too hidden
22:01:59 <rau117> truebrain: Well... personally I like the "**just** invert ctrl action on signals" solution, preferably with a setting that will be enabled for all new openttd.cfg
22:01:59 <rau117> This will help beginners not to build signals manually, and old, accustomed players will not complain βoh, thatβs inconvenientβ
22:02:29 <truebrain> rau117: does not fix the problem at all; there still won't be any visual indicator that the behaviour would be changed
22:02:53 <truebrain> that only works, as the PR originally was, when it adds a setting which is different from current games vs new games, or something odd
22:03:01 <truebrain> but rightfully, we don't want more settings, and we don't need to have them
22:03:07 <rau117> truebrain: But it also won't create new problems, which may give us more time to think of a better solution.
22:03:10 <truebrain> we just need to make it a bit more visual what autofill is actually doing, imo π
22:03:35 <truebrain> rau117: so having one problem, that is alien to current players, is fine? That is a weird argument π
22:04:15 <locosage> indicator is not an issue here, ctrl-drag behaviour wasn't changed and those using non-ctrl-drag are probably experienced enough to figure it out
22:05:11 <truebrain> from the PR, ironically: "A different usability concern related to that: many users won't read the changelog, so will continue pressing Ctrl. This will now result in "covered area" dragging, which looks similar (some signals built) but isn't what they expect. There will definitely be bug reports."
22:05:32 <_glx_> but really autofill was unknown to many experienced players π
22:05:47 <truebrain> you had to know to know π
22:06:06 <truebrain> one can easily argue that the UX of the autofill was always broken π
22:06:57 <rau117> truebrain: Wait⦠complete impossibility of building signals to a given length is what was intended?..
22:07:02 <_glx_> yeah never been a good UX, but it was hidden behind ctrl so it was "fine"
22:07:30 <_glx_> no, the intention of the PR was just to invert ctrl
22:08:09 *** _zephyris has joined #openttd
22:08:09 <_zephyris> FWIW, I was confused by this behaviour too. I thought it was just a muscle memory problem. Proper UI feedback (eg. autorail highlight of the area to gain signals) was my immediate thought for improvement too.
22:08:25 *** keikoz has quit IRC (Ping timeout: 480 seconds)
22:08:42 <truebrain> yeah .. and now we did the default before the visual part
22:08:45 <truebrain> which is just awkward π
22:09:11 <truebrain> it just isn't easy to implement a visualiser like that, I think
22:09:15 <truebrain> at least, I don't see an easy way π
22:09:53 <truebrain> anyway, easy for me to talk, looking back on something done 1.5 years ago π But it is what it is π
22:09:56 <locosage> even in cmclient I haven't gotted to signal preview yet
22:10:19 <locosage> needs so many things to be implemented first
22:10:44 <truebrain> honestly, what mostly tripped me up, is that it was also done for removing signals
22:10:52 <truebrain> I just can't remove 2 signals anymore .. have to click them one by one
22:11:31 <locosage> how often do you need to remove 2 signals? ;p
22:11:43 <truebrain> when building junctions, as it turns out, often enough to be confused
22:11:52 <truebrain> (again, nothing visually telling me the behaviour changed, and mussle-memory etc π )
22:12:04 <locosage> there is signal autoremoval option for junctions
22:12:19 <rau117> locosage: well... if I made a mistake, for example. Or in some βadvanced coop-like gamestyleβ.
22:15:36 <locosage> rau117: for advanced coop my plan was to restore area-drag in cmclient
22:16:37 <_glx_> clearly plain removal of simple drag was a mistake (and you were kind forced to do it π )
22:17:16 <rau117> locosage: hmm⦠area-drag?
22:18:04 <locosage> simple drag, fixed-length drag whatever you call it
22:19:38 <truebrain> now .. how do I make a movie .... hmmm
22:19:55 <_glx_> snipping tools can do it
22:19:57 *** xarothbrook has joined #openttd
22:19:57 <xarothbrook> TrueBrain gone wild?
22:20:17 <truebrain> xarothbrook: be gone!
22:21:29 <truebrain> haha, looking close, it is a tiny bit worse: you do see: "length: NN" for the amount you dragged
22:21:32 <locosage> _glx_: tbh I don't see why it was "clearly" a mistake, so far there has been the only complaint and it wasn't due to the lack of indication
22:21:49 <locosage> if that's a clear mistake then hiding block signals was a failure of the century :p
22:22:32 <truebrain> yes, as forcing a change upon people is totally comparable to having settings to make it however you like
22:22:35 <truebrain> makes sense, good argument
22:23:21 <truebrain> ha, I managed to make a movie \o/
22:23:30 <truebrain> I hope it didn't record sound π
22:23:32 <locosage> well, changing tick rate was also forced and I've been complaining ever since, revert? ;)
22:23:51 <truebrain> no, I disabled it; but that actually works
22:24:37 <truebrain> Mind the "length: 11" π
22:24:41 <rau117> locosage: The question is, why remove such an opportunity in the first place?
22:24:41 <rau117> Even with an indication, this will remain a problem.
22:24:41 <rau117> βAre you liked this function? Well, **we** decided that **you** don't need it."
22:25:50 <xarothbrook> You do realize that 90% of game design is exactly that?
22:25:52 <locosage> well, the answer was to preserve ctrl-drag behavior, not that I particularly like it
22:26:06 <xarothbrook> trying to figure out what players need, and want, and how to make the game most enjoyable for the largest group.
22:26:19 <truebrain> and please don't ask me why I placed down a semaphore π
22:28:28 <truebrain> yes, that is a movie, very good *puzzled look*
22:29:40 <truebrain> That is the other example of what confuses even me π
22:30:36 <peter1138[d]> It was wet & muddy
22:32:01 <truebrain> I drive 40km a day; I think that is a lot π
22:33:30 <truebrain> haha, what could be a cheesy solution instead of #12034, is to just change the label to read "Till next junction" π
22:34:05 <peter1138[d]> I think revert is a good solution :p
22:34:23 <truebrain> I tried to change the highlighter to show the whole path .. that .. is ... euh, yes. It is.
22:34:27 <_glx_> hmm we have AdvanceSignalAutoFill(), dunno if the GUI could use it
22:34:43 <peter1138[d]> After reading scrollback, a good visual indicator would be a new button in the signal UI.
22:35:14 <truebrain> didn't even think about a solution like that, but yeah, that too would work
22:36:16 <truebrain> _glx_: finding the path isn't the biggest issue, from what I can tell. Making the selection update correctly seems to be the tricky part π
22:36:40 <truebrain> the current implementation only allows going from point A to point B, where B is your mouse
22:36:53 <truebrain> with some options on what to do .. X, Y, X and Y, ..
22:37:00 <truebrain> `VpSelectTilesWithMethod` for context
22:37:19 <truebrain> or more exact, `CalcRaildirsDrawstyle`
22:37:49 <truebrain> owh, the tooltip is a setting; did not know π
22:38:32 <truebrain> anyway, going to leave the PR open till just before RC1 (or next beta), in case someone manages to cook up something nice before that time π
22:40:57 <truebrain> "complicated direction", always a good comment π
22:41:12 <truebrain> "What direction are you going?" "Complicated!" π
22:44:55 <rau117> It seems that I once saw the implementation of previews for signal-dragging but I donβt remember whether it was animation or finished code.
22:44:55 <truebrain> wow, found an artifact? Nice
22:46:20 <truebrain> ah, the linguistic issue with or and and π
22:46:51 <truebrain> seems very reasonable to me _glx_ ; PR it! π
22:48:10 <_glx_> oh and it's recent enough to not affect translators too much
22:48:48 <truebrain> owh, wow, many translators have been busy π Nice π
22:49:29 <_glx_> peter1138[d]: what do you think ? (as it's your addition)
22:53:00 <peter1138[d]> It was in the context of "these options are selected" but if that's confusing or awkward in other languages... not fussed
22:53:59 <peter1138[d]> Oh, forgot to mention, 12035 only affects some NewGRF houses, default houses don't set the things that prohibit deleting.
22:54:14 <_glx_> yeah it means both options are selected, but it can be seen as what the filter will be filtering
22:54:27 <peter1138[d]> Silly GitHub logged out on phone again π¦
22:54:27 <truebrain> forgot to press approve π
22:55:04 <truebrain> now `IsValidHumanID` is only used for visual things \o/
22:55:17 <truebrain> (and a bit of network filtering)
22:55:22 <peter1138[d]> I didn't check that, heh
22:55:55 <truebrain> makes it even more clear that your PR is the right solution π
22:56:55 <peter1138[d]> I was surprised it predates NoAI π
22:57:05 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:57:24 <truebrain> silly me for not finding that π
22:57:53 <truebrain> well, I guess we were mostly focused on removing "the AI can do these extra things" π
22:58:03 <_glx_> yup IsValidHumanID and IsValidAID are just IsValidID subsets
22:58:10 <truebrain> terraform for AIs was free ... still can't believe that was needed to keep AIs afloat ... π
22:59:52 <truebrain> I guess that quote predates NoAI π
23:01:28 *** frosch123 has joined #openttd
23:01:28 <frosch123> i am not sure whether ttdp had magic bulldozer
23:01:31 <_glx_> now AI are paying to demolish stuff, so if it's expensive they won't do it
23:01:45 <peter1138[d]> NewGRF spec specifically says players can demolish, but hey
23:01:54 <frosch123> i think the intention is: this building show not be removed in normal gameplay, except if the role-playing player knows better
23:04:06 <frosch123> i can't find any "remove tile" sign-cheat in ttdp
23:04:22 <frosch123> so i think ttdp did not have any "sandbox" mode
23:04:28 <peter1138[d]> Oh gosh, sign cheats π
23:05:28 <peter1138[d]> I wonder if anything uses those flags/callbacks.
23:05:52 <truebrain> sign cheats were the best!
23:05:59 <peter1138[d]> And if that flag is set does that mean it's there permanently for all eternity...
23:06:05 <truebrain> can we revert their removal?
23:06:22 <truebrain> I read we only need one user to complain
23:06:58 <frosch123> there is also a `minimum_life` property, which prevents towns from removing the house for up to 255 years
23:07:25 <frosch123> peter1138[d]: there is also a callback
23:07:49 <peter1138[d]> Yes, but the callback result can vary, being a callback.
23:08:06 <peter1138[d]> That flag, not so much.
23:08:40 <frosch123> are there any other cases for #12035? as it is now, I think it only blocks for towns and nothing else, so it is a `if (_current_company != OWNER_TOWN) return true;`
23:09:13 <frosch123> which IMO is the purpose of `minimum_life` and not `BULDING_IS_PROTECTED`
23:10:14 <frosch123> the callback can actually not check the company, i kind of expected that :p
23:10:36 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
23:11:34 <peter1138[d]> I figured just allow AIs to do what players can do. But if we think players shouldn't be allowed either, that works too.
23:12:31 <frosch123> i claim it was only allowed for players, because ttdp did not have magic-bulldozer / sandbox mode
23:12:35 <peter1138[d]> Or: find out what uses it and see if it breaks π
23:15:50 <xarick> towns demolishing houses of a nearby town?
23:15:58 <xarick> to make a house for their town?
23:18:10 <xarick> towns competing for space
23:24:59 <frosch123> that's with the human-exception removed
23:27:02 <frosch123> ah silly me. i paused for the screenshot π
23:27:32 <frosch123> when unpaused the error is different
23:27:46 <_glx_> or allow build on pause π
23:28:25 <frosch123> well "different" means "only first line", no detailed reason
23:28:32 <frosch123> i guess that could be improved, too
23:29:39 <xarick> AI can't delete some buildings
23:29:48 <xarick> in current master, neither me as Human
23:34:58 <_glx_> haha and now the question is why it is ONE_SANDBOX while all other start with OME
23:35:30 <_glx_> we are so bad at reviewing
23:36:12 <truebrain> Owh my, it reads OME?! Haha
23:36:20 <truebrain> Really misread that π
23:36:48 <frosch123> oh, i assumed you promoted magic bulldozer to a regular multiplayer-safe setting
23:37:09 <truebrain> It was a rebrand only π
23:37:44 <truebrain> But most of that window needs to move to settings imo π
23:37:54 <frosch123> well, but then my "protect houses from players" argument does not work in multiplayer :p
23:38:34 <truebrain> Move the setting first! Or already assume it happened in the future π
23:38:47 <frosch123> i think some of the things were added to the cheats, because they do not work well, e.g. the max-map-height thingie
23:39:22 <frosch123> but magic-bulldozer, crossing-tunnels, and strong-aircraft-brakes should work
23:39:26 <truebrain> Max map thing we can just remove tbh, and make it 255 .. I don't understand why it is a variable value
23:39:34 <frosch123> well, date change is an action, no setting :p
23:39:59 <xarick> nevermind, I don't have enough rating to demolish a cathedral
23:40:18 <xarick> EternalLove GS failing
23:40:29 <_glx_> there's a setting for that, no need to use a GS
23:42:20 <xarick> that may be a problem, depending on where you're heading with it
23:42:36 <xarick> allow for all, disallow for all
23:43:37 <xarick> is deity able to demolish that cathedral? let's test
23:43:55 *** ChanServ sets mode: +v tokai
23:44:38 <xarick> this is what the AI can't demolish
23:44:55 <xarick> with Eternal Love constantly feeding Outstanding rating
23:46:34 *** doppel_r_dwr has joined #openttd
23:46:34 <doppel_r_dwr> xarick: well atleast you have 252 people living in the cathedral
23:48:55 <xarick> oh snap... can use Eternal Love with my test GS together
23:50:52 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
23:53:26 <xarick> GS can't demolish those
23:55:21 <truebrain> One thing I learnt from the 14.0 changelog, is that we tend to use "fix" for codechanges, and they are annoying to filter out π styleguide implies a fix is visible to the user, but isn't really clear in the wording. Maybe we should make that a bit stronger worded π
continue to next day β΅
