IRC logs for #openttd on OFTC at 2025-01-30
⏴ go to previous day
00:03:20 <peter1138> Sometimes the clock at the top of my desktop stops...
00:32:20 *** D-HUND is now known as debdog
00:54:13 *** jinks has quit IRC (Remote host closed the connection)
00:54:13 *** jinks_ is now known as jinks
03:01:13 *** Wormnest has quit IRC (Quit: Leaving)
03:10:45 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:24:16 *** gnu_jj_ has joined #openttd
03:27:36 *** gnu_jj has quit IRC (Ping timeout: 480 seconds)
04:43:11 <DorpsGek> - Update: Translations from eints (by translators)
05:05:46 *** Flygon has quit IRC (Read error: Connection reset by peer)
06:40:33 *** keikoz has quit IRC (Ping timeout: 480 seconds)
08:40:36 *** HerzogDeXtEr has joined #openttd
08:54:41 <peter1138> Hmm, is GetTileTrackStatus' sub_mode parameter ever used with anything other than RoadTramType?
08:57:29 <peter1138> `std::optional<RoadTramType>` could work there, otherwise some variant.
09:00:39 <peter1138> Or expand `TransportType` and sub_mode , but that also affects tunnels/bridges map storage.
09:01:48 <peter1138> *and remove sub_mode.
09:32:39 *** kuka_lie has joined #openttd
09:36:52 *** kuka_lie has joined #openttd
09:47:24 <truebrain> LordAro: You start of with a fun day 😦
09:48:08 <truebrain> wauw, I kinda forgot how horrible the gitlab UI is .. that paginating at the bottom of that commit you reference
09:53:06 <LordAro> truebrain: encouragement to make commits short and to the point!
09:55:53 <peter1138> > 164 files changed, 14461 deletions(-)
09:55:56 <peter1138> This was a good commit.
09:59:23 <andythenorth> I love commits like "Change: deleted legacy app"
10:13:16 <peter1138> Well, such enum class.
10:40:39 <andythenorth> do we have a var for 'variant has child variants' (in purchase scope only)?
10:43:39 <peter1138> > load average: 51.47, 33.93, 19.26
10:43:54 <peter1138> llvm vs code plugin was using 45GB RAM.
10:44:10 <peter1138> I think perhaps installing clang-19 has caused some issue...
10:51:34 *** kuka_lie has quit IRC (Ping timeout: 480 seconds)
11:04:46 *** kuka_lie has joined #openttd
11:13:47 <xarick> well.. more openttd development time
11:14:55 <peter1138> tl;dw, is it deading?
11:16:45 <peter1138> If you have to ask, no.
11:17:00 <peter1138> (Am I going to eat anyway, yes.)
11:24:11 <xarick> I keep doing the same mistakes as ever
11:24:25 <xarick> multiple changes in a single commit
11:25:09 <xarick> splitting the changes is hell
12:21:52 <peter1138> I have a naming problem for CallbackMasks.
12:22:03 <peter1138> Because they are masks, but they are not masks of Callbacks.
12:22:26 <peter1138> CallbackMask / CallbackMasks?
12:23:29 <xarick> Ask CoPilot, he's good at that kind of stuff
12:24:02 <peter1138> I have no access to, and no desire to.
12:29:10 <andythenorth> peter1138: CallbackMaskCabbage
12:29:17 <andythenorth> until you know 😛
12:29:42 <andythenorth> I make a lot more progress when I stop trying to name vars and methods
13:59:27 <xarick> my train of thought derailed
14:01:24 <LordAro> you should've set a waypoint
14:10:14 <xarick> I still have one to go
14:11:28 <LordAro> i'd rather you stopped
14:11:36 <LordAro> you have far, far too many PRs open already
14:24:04 <xarick> oh, something failed here
14:33:15 <xarick> time to test AIs that use buoys
15:13:09 <Rubidium> peter1138: I'd say rename VehicleCallbackMask to VehicleCallback and enum-bitset VehicleCallbacks (or maybe EnabledVehicleCallbacks).
15:46:49 <peter1138> That's what I did initialy, but VehicleCallback is wrong. The bit number of the mask has no relevance to the actual callback that is masked.
15:48:04 <peter1138> And for whatever reason, the callback numbers are not separated by feature.
15:49:14 <peter1138> (I've already done all the changes, it's just the naming :))
15:50:10 *** Wormnest has joined #openttd
15:51:13 <Rubidium> EnableableVehicleCallback
16:04:17 <xarick> Search "GetStationID" (3462 hits in 305 files of 1101 searched) [Normal]
16:12:41 <xarick> i just discovered i can't change AI settings and reload
16:12:52 <xarick> it restarts with the previous settings
16:19:30 <xarick> I blame myself for not picking this up earlier
16:35:56 <_glx_> Seems correct, when you change settings for an active slot it doesn't affect the original values
16:37:06 <_glx_> You need to stop the AI first if you want to change slot settings
17:18:00 *** virtualrandomnumber has joined #openttd
17:19:56 *** virtualrandomnumber has quit IRC ()
17:27:08 <xarick> someone forgot to set WindowClass
17:29:10 <xarick> WC_TRAINS_LIST, WC_ROADVEH_LIST, WC_SHIPS_LIST, WC_AIRCRAFT_LIST
17:37:43 <xarick> Im looking for the commit
17:38:16 <peter1138> So the fault is in two places.
17:45:01 <xarick> my poor english, let me ask copilot
17:47:02 <peter1138> "Fix: Incorrect WindowClasses for deduplicated window descs."
17:49:55 <LordAro> xarick: pro-tip, instead of typing "let me ask copilot" into irc, just ask copilot
17:50:45 <merni> don't think he's typing anything into irc :p
17:52:03 <LordAro> merni: an interesting theory
17:52:48 <xarick> bad habit, I feel kinda lonely otherwise
17:58:40 <peter1138> Ah of course, I didn't do it wrong in one place.
17:59:51 <peter1138> std::numeric_limits not working for enums is an interesting choice. Clearly deliberate, but...
18:02:58 *** toktik is now known as Guest7501
18:04:08 <peter1138> That it returns zero instead of failing to compile is also an interesting choice.
18:06:25 *** Soni has quit IRC (autokilled: This host violated network policy. - Contact support@oftc.net for help. (2025-01-30 18:06:25))
18:10:11 *** Extrems has quit IRC (Read error: Connection reset by peer)
18:11:52 <Rubidium> could also only do the enum specialisation for the bit what's happening in pool_type.hpp, as that's the only place that needs it now and wouldn't need to in the future
18:20:05 *** Extrems has joined #openttd
18:48:34 *** ChanServ sets mode: +v tokai
18:55:28 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
19:45:57 <xarick> can you define your own numeric_limit?
19:46:21 <xarick> like if I want 100 to be the limit
20:04:04 <peter1138> Oh god these macros to define default tables are killing me :S
20:14:25 <xarick> if I commit to a detached HEAD, where will my commit go?
20:15:03 <peter1138> In your detached HEAD.
20:15:46 <xarick> is it gonna poof if i switch back to a branch
20:16:25 <peter1138> Pretty much. You can just create a new branch where you are though.
20:59:48 <Rubidium> I really see no way to not go through enums to make the pool ID types stronger / available in std::variant ;(
21:00:47 <peter1138> Because of unions? One way is to de-unionise, but that might be more and less clear work.
21:06:22 <peter1138> I have never seen an error like that before.
21:08:09 <peter1138> Seems to be missing a strongtype.hpp include though.
21:08:27 <Rubidium> it's not the compilation errors, that's just a compiler wanting includes that others don't need. It's about all the nasty casts that are required
21:13:51 <peter1138> Ah yes, switches and array indices are painful for StrongType.
21:14:48 <peter1138> I did one wonder if we could have a Mixin to optionally allow implicit converison to and from the base type, so that it works a bit more like before.
21:14:54 <peter1138> But I think that's not possible.
21:15:14 <peter1138> So probably a whole new WeakerStrongType would need to be created for that.
21:15:24 <peter1138> In which case, maybe enum is fine?
21:16:26 <Rubidium> all the GroupID{static_cast<uint16_t>(vli.index)} ones are horribly as well
21:17:19 <peter1138> No GroupID(vli.index)?
21:17:47 <peter1138> But yes, if we have to keep casting then it means the idea is a bit broken.
21:17:49 <Rubidium> you'd be narrowing there
21:18:03 <Rubidium> and some compilers really dislike that
21:20:17 <peter1138> Or, in this case, stuffing a GroupID in vli.index is the problem, and it should be something else. But I dunno how that would work.
21:22:05 <Rubidium> make it a variant and it'll become std::get<GroupID>(vli) which is much nicer
21:23:05 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:25:30 <xarick> there's something strange in Regression::Order()
21:25:54 <xarick> I see too many failures
21:26:13 <xarick> doesn't look to be intended
21:28:30 <andythenorth> what's the correct term for a vehicle if it's a part of an articulated vehicle?
21:28:50 <andythenorth> (grf spec or OpenTTD internals, either/both)
21:29:28 <_glx_> internally it's a vehicle (nicknamed articulated part IIRC)
21:29:36 <andythenorth> unless it's an engine
21:30:01 <peter1138> Is it an instance or a definition?
21:30:21 <peter1138> GRF Spec goes with vehicle.
21:30:29 <andythenorth> Horse haas consists, which are articulated vehicles
21:30:44 <andythenorth> made up of entities referred to as units / vehicles / trains depending where in the code
21:31:22 <andythenorth> the consists are used to generate buyablevariants
21:31:30 <andythenorth> which are then put into variantgroups
21:32:41 <andythenorth> I couldn't draw the hierarchy, I tried 😛
21:32:45 <andythenorth> "it works though"
21:39:21 <andythenorth> all the buyablevariants are just instances of liveries
21:39:57 <andythenorth> I considered variants that change stats, but I can't see any reason that would be good
21:40:35 <xarick> oh... the vehicle really has 0 orders initially
21:41:02 <xarick> the intentions sometimes are weird
22:01:53 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:11:57 <andythenorth> Is there any single English word that covers all rail vehicles? (Can’t use vehicle)
22:13:09 <andythenorth> Engine, wagon, train are all inappropriate
22:13:26 <andythenorth> Vehicle doesn’t work
22:14:31 <andythenorth> Train terms are stupid
22:14:52 <andythenorth> A ‘unit’ is made up of ‘units’
22:15:01 <andythenorth> A train is made up of trains
22:15:14 <belajalilija> rail apparatuses
22:15:33 <andythenorth> Maybe I should convert Horse internals to German
22:15:47 <andythenorth> Probably has an accurate word
22:16:36 <belajalilija> schutzengartenbrauerheit
22:19:38 <belajalilija> i almost named a real thing when i was just making up a german sounding word
22:26:45 <peter1138> "Your script made an error: There is no passenger cargo."
22:27:12 <peter1138> But PASS and TOUR exist, both in passengers class.
22:27:24 <peter1138> cargo_list.KeepValue(AICargo.TE_PASSENGERS);
22:27:25 <peter1138> if(cargo_list.Count() == 0) throw("There is no passenger cargo.");
22:27:46 <peter1138> I guess town effects removed.
22:41:05 *** keikoz has quit IRC (Ping timeout: 480 seconds)
22:45:00 <xarick> Testing road waypoint orders, what needs fixing
22:48:40 <xarick> conditional orders are a pain to set up
22:49:00 <xarick> well, goodnight, tomorrow I'll look at it
22:54:45 <peter1138> Hmm, HouseZones is kinda two enums in one.
22:55:24 <peter1138> Town zone, climate and subartic information all combined.
23:10:36 <michi_cc> andythenorth: Well, German does have a single word, but only because you can string nouns together like no man's business 🙂 Would be Schienenfahrzeug or Rollmaterial, which is rail vehicle or rolling stock.
23:11:50 *** kuka_lie has quit IRC (Remote host closed the connection)
23:14:54 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
continue to next day ⏵