IRC logs for #openttd on OFTC at 2022-06-01
⏴ go to previous day
01:26:31 *** Wormnest has quit IRC (Quit: Leaving)
02:33:23 *** debdog has quit IRC (Ping timeout: 480 seconds)
02:39:36 *** D-HUND is now known as debdog
05:03:59 *** m1cr0man has quit IRC (Ping timeout: 480 seconds)
05:10:13 *** jpw has quit IRC (Ping timeout: 480 seconds)
05:12:46 *** m1cr0man has joined #openttd
06:27:43 *** sla_ro|master has joined #openttd
06:54:44 *** m1cr0man has quit IRC (Remote host closed the connection)
07:06:55 <TrueBrain> okay, so what is TrueGRF missing for a first beta-release .. hmmm .. changelog, to know what has changed, I guess
07:08:02 <TrueBrain> hard-coded URL in the GRF might need to point to the GitHub instead of TrueGRF itself :P
08:07:33 *** Smedles has joined #openttd
08:22:39 *** m1cr0man has joined #openttd
08:58:27 *** snail_UES_ has joined #openttd
09:33:04 <Samu> initialize in the proper way
09:38:25 <Samu> wow, the first one is also wrong, damn :(
09:43:24 <Samu> std::fill(c->share_owners.begin(), c->share_owners.end(), INVALID_OWNER);
09:53:01 <Samu> Money mon[MAX_COMPANIES] = {}; this initializes with Money being 0 to all companies, what if I wanted to initialize it with another value?
10:09:38 *** m1cr0man has quit IRC (Ping timeout: 480 seconds)
10:18:25 *** m1cr0man has joined #openttd
10:27:59 *** Etua has quit IRC (Quit: Etua)
10:39:47 *** WormnestAndroid has quit IRC (Remote host closed the connection)
10:42:06 *** WormnestAndroid has joined #openttd
11:24:04 <Samu> i have 2 more bugs to report, but now lunch time
12:01:26 *** OsteHovel has quit IRC (Read error: Connection reset by peer)
12:35:51 <TrueBrain> 61 commits, +20k, -40k LoC, to switch to another UI and use GitHub as storage .. but I think I have a good-enough version of TrueGRF to push to production :D
12:37:19 <glx> so annoying to be forced to save .zip from issues, I used to be able to open them on the fly
13:06:14 <glx> ok I see why #9908 happens
13:06:34 <glx> but I don't know how to fix it
13:11:06 <glx> I don't fully understand the new command system :)
13:35:12 <glx> ok don't know how, but I add ": =" in bind addresses
13:46:05 <glx> let's build a 12.2 with asserts to check something
13:48:19 <peter1138> What's this UI then?
13:51:11 <TrueBrain> so .... I have a request to add "introduction date" to TrueGRF ... and I have no clue how to go about it :P
13:52:53 <TrueBrain> I guess callback 22 should be able to deal with that? Hmm
13:57:33 <peter1138> Depends on the feature type? Isn't it normally a basic Action 0 property?
13:57:45 <TrueBrain> sorry, for industries
13:57:51 <TrueBrain> and I cannot find it ..
13:59:20 <peter1138> Ah for those it's a bit special, it'll be a testing a varaction during construction, I believe.
14:00:16 <TrueBrain> meh .... so I need to figure out how to combine CBs with custom defined CBs .. not for today :D
14:00:49 <peter1138> Svelte material UI.
14:01:06 <peter1138> TrueBrain, or just tell them it's a callback-only feature :D
14:01:23 <TrueBrain> Svelte Material UI sucks balls! :P
14:01:30 <TrueBrain> I think I removed all references to that now, didn't I? Hmm
14:01:36 <Samu> 9908 also asserts in 12.2, just tried a debug build
14:01:41 <TrueBrain> peter1138: bit the idea is to make shit easier :D
14:01:43 <peter1138> If it's merged and live, no.
14:01:58 <glx> that's exactly what I was testing Samu :)
14:01:59 <TrueBrain> hmm .. where did you find a reference to Materials still?
14:02:40 <peter1138> I saw svelte in the filenames for .css and .js files, along with data-svelte attributes.
14:02:57 <TrueBrain> just not Materials UI :)
14:03:21 <TrueBrain> now using Carbon :)
14:03:47 <glx> at least it's not a bug due to the command rewrite :)
14:04:05 <peter1138> It's cos the first google hit (where google shows a summary) for "svelte ui" says svelte material ui...
14:04:28 <TrueBrain> Svelte is a framework to make reactive websites in a way that does make sense
14:04:33 <TrueBrain> I just cannot understand React, Next, ...
14:09:21 <peter1138> Heh, I'm still doing it the old school way, there the server sends back these things called pages...
14:09:39 <TrueBrain> well, I have to say, this works pretty well
14:10:12 <TrueBrain> was an experiment, but so far .. I am not complaining :)
14:13:47 <TrueBrain> and I like it more than a stand-alone app :) Which was the alternative ...
14:13:49 <TrueBrain> but browsers can do crazy shit these days ..
14:16:07 <TrueBrain> made room for a sprite-editor for industry tiles too .. and I think I know how I want to do that :) But .. enough for today :P
14:34:39 <dihedral> TrueBrain> was an experiment, but so far .. I am not complaining :) <- doubt :-P
14:46:06 <Samu> CMD_CLIENT_ID | CMD_NO_EST
14:46:19 <Samu> should help deal with the no estimate issue
14:46:42 <glx> the issue is command during command
14:46:50 <glx> it's not supposed to happen
14:47:29 <Samu> i checked blame, and somewhere in the past, it used to use DoCommandP
14:47:41 <Samu> then changed to DoCommandPInternal
14:48:06 <Samu> because of pressing shift command
14:48:49 <Samu> maybe I was thinking it could be reverted, and use the appropriate flags, if they exist
14:56:52 *** wallabra_ has joined #openttd
15:00:42 *** wallabra has quit IRC (Ping timeout: 480 seconds)
15:15:16 *** supermop_work has joined #openttd
15:48:58 <glx> ok for #9909 it's a templating issue, it somehow uses the wrong one and discards error_message
15:50:58 *** Wormnest has joined #openttd
16:03:57 *** Tirili has quit IRC (Quit: Leaving)
16:04:51 *** frosch123 has joined #openttd
16:06:19 <glx> CommandCost CmdBuyShareInCompany(DoCommandFlag flags, TileIndex tile, CompanyID target_company);
16:06:43 <TrueBrain> it does read TileIndex :P
16:06:55 <nielsm> TileIndex does not have an implicit constructor from integer types does it?
16:07:40 <glx> vs CommandCost CmdSellShareInCompany(DoCommandFlag flags, CompanyID target_company);
16:07:54 <glx> seems there's an extra useless arg
16:07:56 <TrueBrain> I was hoping there was a property to set :(
16:08:12 <TrueBrain> now I have to think about if and how I want to resolve custom cbs vs helping fields that generate cbs
16:08:14 <frosch123> nope, and i think andy removed all intro-dates from firs
16:08:35 <frosch123> as in: when industries become available, it takes ages for them to appear or vanish, or so
16:09:28 <TrueBrain> not my problem :D :D hihi
16:09:59 <TrueBrain> I keep coming back to templated callbacks, where you can select a predefined callback from a dropdown, which gives you input fields to fine-tune it
16:10:07 <TrueBrain> so I can template the most commonly requested things
16:10:28 *** snail_UES_ has quit IRC (Quit: snail_UES_)
16:10:44 <frosch123> newgrf authors tend to only want the non-common things :p
16:10:48 <nielsm> again: invent a callback on the industrytype to allow it to determine how many of it needs to be on the map
16:11:05 <nielsm> and maybe one for how many are allowed/requested for a town
16:11:33 <TrueBrain> frosch123: that is why they are just templates; there is always an "expert" option which lets you type in your own callback :D
16:11:34 <frosch123> nielsm: albert reworked all the industry spawning 10 years ago
16:11:44 <frosch123> they now track current and desired numbers
16:12:24 <frosch123> but it still makes no sense to shut down old blacksmiths, serviced by the player, and reopen steelmills in other places
16:12:52 <frosch123> all the kids want stable economy, so open/closure of industries or even industry types makes no sense to them
16:13:03 *** HerzogDeXtEr has joined #openttd
16:13:47 <nielsm> that specific case of blacksmith/steel mill I'd say is partially a design thing too, though
16:14:12 <TrueBrain> frosch123: how do you balance, continue extending industries vs adding things like trains?
16:16:19 <frosch123> if you want to be done by next week, don't enter the vehicle rabbit hole :)
16:16:30 <frosch123> objects and houses are very similar to industries
16:16:34 <TrueBrain> well, today I released TrueGRF v2 for industry support :P
16:16:35 <frosch123> lots of things to reuse and generalise
16:17:13 <TrueBrain> yeah, my reasoning for going down the rabbit hole is, that I am afraid I am a bit to narrow in view on the current problems I try to solve
16:17:14 <frosch123> ah, you use browser versioning, v100 by next week :)
16:17:21 <TrueBrain> in result, solve it for industries
16:17:25 <TrueBrain> but doesn't work at all for NNN
16:17:40 <nielsm> IMO the static-ness and predictability of the industries in the Transport Fever makes those games much less interesting, having to respond to unpredictable changes in the world is what makes TT fun
16:17:43 <TrueBrain> well, I had TrueGRF v1, and for sure this was worthy of a major bump :P
16:17:49 <frosch123> stations are also popular, and complicated :)
16:18:12 <TrueBrain> are trains that bad? :P
16:18:58 <TrueBrain> I find it difficult .. I can spend weeks on making industries really polished .. but .. not sure that is very useful :P
16:19:20 <frosch123> that's my attempt of making vehicles more sane. i succeded in making them more consistent, but i failed in making them simpler, so everyone is scared :)
16:19:59 <TrueBrain> I have the feeling I can skip a lot of the bla with TrueGRF, as I could with industries :D
16:20:48 <frosch123> yes, pretty sure 50% of vehicle specs is deprecated :)
16:21:15 <frosch123> sometimes objectively deprecated, sometimes opinonated deprecated
16:21:31 <TrueBrain> we applied both in TrueGRF already for industries, so that is fine ;)
16:21:41 <TrueBrain> I am not building a tool to use everything OpenTTD has to offer
16:21:49 <TrueBrain> I am building one that makes creating NewGRFs easy for everyone :P
16:22:26 <frosch123> fine, go with trains then
16:22:30 <frosch123> but don't cry later :p
16:22:46 <TrueBrain> I just don't know .. neither are good options
16:23:27 <frosch123> you can also go for townnames
16:23:35 <frosch123> everyone wants to make their own schema
16:23:41 <frosch123> and you can shared 0% code with industries
16:23:53 <TrueBrain> how do you even define town names ..
16:24:09 <TrueBrain> it has its own action? Lol
16:24:24 <frosch123> yes, a single huge blob, different from anything else
16:24:35 <TrueBrain> if I do, can we then move all town-names to GRFs and remove them from in-game?
16:24:40 <frosch123> plus you need some special magic like force-activation
16:24:57 <frosch123> plus townnames are the one thing that really cannot be combined with other stuff in one grf
16:25:35 <frosch123> TrueBrain: iirc the road to remove all the in-game generators was to store the generated names in the savegame instead of the seeds only
16:25:50 <frosch123> after that everything can be trashed
16:25:53 <TrueBrain> yeah, cannot believe we never did that :P
16:27:14 <frosch123> so well, if you want to go for trains, you will have to deal with cargos, railtypes and liveries
16:27:20 <frosch123> all the hot disputes :)
16:28:26 <frosch123> townnames would be popular, but they are really unrelated to anything else, so you could as well create a separate compiler
16:29:09 <TrueBrain> I am surprised how complex townnames are :P
16:29:14 <TrueBrain> reading the specs is hard :D
16:29:44 <glx> townnames grf are loaded in intro menu
16:30:07 <TrueBrain> do these need to be loaded static btw?
16:30:21 <frosch123> no, but you need some magic force-activate actionD
16:30:47 <frosch123> essentially townnames is full of townname-specific magic
16:30:48 *** Smedles has joined #openttd
16:31:11 <frosch123> which is not specced to be townname-specific, but only makes sense for townnames anyway
16:31:39 <frosch123> thus the hard divide: grf defined townnames XOR it defines other stuff
16:31:48 <TrueBrain> "For this reason, your town name style must be able to generate at least 70 different acceptable combinations, since this is the maximum number of towns possible on a map, and TTD may deadlock if it can't find different names for every town on the map."
16:32:36 <frosch123> fun fact: ammler used to build townname-grfs with shell scripts
16:33:31 <TrueBrain> but honestly, town-names sounds like a good idea to do, simply as it paves the road to make TrueGRF generate different types of GRFs :)
16:33:38 <TrueBrain> as I also don't want trains and industries to be in a single GRF :)
16:34:23 <frosch123> that makes townnamed indeed sound like a good idea
16:35:22 <TrueBrain> okay .. so if I understand this correctly, townnames are build up from parts, and each part is selected from a set
16:35:51 <frosch123> i think it's a tree
16:35:56 <glx> many just use a huge list
16:35:56 <frosch123> leafs are single strings
16:36:09 <frosch123> nodes can branch to leafs or further nodes
16:36:27 <frosch123> glx: those are the boring ones
16:37:06 <TrueBrain> do people actually make townnames grfs based on language?
16:37:46 <TrueBrain> well, I am thinking about the example given, a grf for mars townnames
16:37:58 <frosch123> there are two types if town grfs: those who pull in huge lists of real world names, and those who build them from syllables like ottd does
16:38:22 <glx> and also some mix of both types
16:38:52 <frosch123> planetmaker's german townname have a grf parameter to select between "real names from germany", "real names from north/south/...", "fictional names"
16:40:22 <frosch123> 22 townname grfs on devzone
16:40:24 <TrueBrain> how is pointing to another tree useful? (looking for examples)
16:40:57 <TrueBrain> and to loop back to your example, if the GRF gives German townnames, it should ignore any language setting, not?
16:41:15 <TrueBrain> so the language part is only useful if you have a more generic GRF
16:41:49 <frosch123> imagine first level: (prefix, middle part, postfix), and then middle part being (one syllable things, two syllable things, ...)
16:42:06 <frosch123> you create dependencies between parts that way
16:42:31 <TrueBrain> yeah, okay, just a means to make the tree easier to understand
16:42:57 <frosch123> [AB][CDEF] matches different things than (A[CD])|(B[EF])
16:43:58 <frosch123> in the former case first and second syllabel are independent
16:44:09 <frosch123> in the latter case the first syllable A/B restricts what follows
16:45:27 <TrueBrain> yeah, I think I get it .. a machine can just prerender it all, but for humans it is easier to think that way
16:46:08 <TrueBrain> okay, that kinda makes sense
16:46:15 <frosch123> well, sure, you can always expand all 200k cases into a single list :)
16:46:19 <TrueBrain> the language stuff still makes a bit less sense, but okay, also not that important :D
16:46:33 <frosch123> not sure whether thee are lmits on the number of items
16:46:50 <frosch123> what language stuff?
16:47:14 <frosch123> TrueBrain: nope, way less. 256 per actionF, and at most 256 actionF in total
16:47:25 <frosch123> iirc actionF ids are not reusble
16:48:39 <TrueBrain> hmm .. why 256 per actionF btw?
16:48:45 <TrueBrain> the wiki reads as it can be infinite?
16:49:00 <TrueBrain> well, no, 256 * parts, per actionF
16:49:13 <TrueBrain> owh, and 256 parts max, so 65536 per actionF?
16:49:52 <frosch123> still unsure what you mean with "language stuff"
16:50:04 <frosch123> do you mean the name of the townname-set?
16:50:42 <TrueBrain> hmm .. it might be :P
16:51:03 <TrueBrain> ah, okay, so you can translate the style names here
16:51:21 <frosch123> so 8 million parts in total
16:51:39 <TrueBrain> lol, will that be enough? :P
16:51:49 <frosch123> yeah, the name of the style is translated, the townnames themself are not
16:51:51 <TrueBrain> okay, [<langs> <name>] ...
16:51:54 <TrueBrain> did not understand it like that :P
16:51:58 <TrueBrain> that makes more sense
16:52:15 <TrueBrain> so how do I figure out how to load such GRF exactly?
16:53:25 <frosch123> you mean you want an example grf?
16:53:38 *** tokai|noir has joined #openttd
16:53:38 *** ChanServ sets mode: +v tokai|noir
16:53:42 <TrueBrain> that would work; but mostly I am trying to understand what it is saying about action9 :P
16:54:39 <frosch123> that's only important if you want to support grf parameters
16:54:41 <TrueBrain> doesn't seem to be something I actually care about :P
16:55:16 <TrueBrain> and what do you mean with "force-activation"?
16:56:19 <frosch123> magic action E: "0E 01 <own grfid>" has to appear in townname grfs
16:56:34 <frosch123> normally grfs are loaded in-game only
16:56:46 <Samu> okay I dunno how to fix the order back up deal...
16:56:46 <frosch123> with that action they are loaded in intro-screen
16:57:01 <frosch123> which is important for ttdp to be able to show them in the settings
16:57:13 <frosch123> and i believe ottd choosed to emulate that
16:57:26 <TrueBrain> okay .. but ActionE disables it?
16:57:42 <frosch123> yes, except with <own grfid> it means something else :p
16:58:13 <TrueBrain> "Forced activation is not supported by OpenTTD. "
16:59:27 <frosch123> ok, ignore all glx and I said :)
16:59:38 <frosch123> nml does not generate that actionE either
16:59:49 <frosch123> so, i guess it does not matter for ottd
16:59:59 <glx> openttd always activate townnames
17:00:26 <glx> else you can't show them in the dropdown
17:00:28 *** tokai has quit IRC (Ping timeout: 480 seconds)
17:00:38 <TrueBrain> okay, so I only have to do ActionF
17:00:41 <TrueBrain> that makes it easier
17:00:46 <frosch123> yeah, now that i think about it... people used to complain about ECS landscape icons not showing up in ottd
17:00:47 <TrueBrain> but I guess I really do have to add parameters for this
17:00:52 <TrueBrain> as otherwise you get a lot of small GRFs :)
17:01:17 <frosch123> do you expect people to create more than one?
17:01:27 <TrueBrain> a random google tells me people already started that
17:01:50 <frosch123> oh, actually, i was wrong about grf parameters as well...
17:02:00 <frosch123> you can just define multiple townname schemas per grf
17:02:06 <frosch123> no need to use parameters to choose them
17:02:25 <TrueBrain> you can have 13 or something the specs say?
17:02:36 <frosch123> sound slike a ttdp limit
17:02:43 <frosch123> i would say you can have 128
17:02:50 <TrueBrain> "The town name style dropdown menu doesn't support more than 19 entries, and since six entries are already reserved for the old styles, only 13 remain for new styles."
17:03:16 <frosch123> only the "bit 7 must be set" is important for ottd
17:03:24 <frosch123> so i would think you can have 128 styles per grf
17:03:50 <TrueBrain> okay, that makes it stupidly easy :P
17:03:56 <TrueBrain> now to think about a good UI
17:04:23 <TrueBrain> as I guess you want two different UIs for each type of usage .. the "drop a list of cities in here" and the "let's be fancy and generate names based on parts"
17:06:22 <frosch123> looks like the comment above that is on a conservation order
17:06:35 <glx> actionF is loaded very early compared to other actions
17:08:28 <TrueBrain> it always amazes me how much more clear things become when I read FeatureTownName for example over the specs :P
17:08:28 <TrueBrain> not reading about all the TTDp shit really helps :D
17:09:38 <TrueBrain> all the (now outdated) TTDp references .. they don't mean anything to me :)
17:10:32 <frosch123> oh, did you see that townname-video on reddit?
17:10:37 <frosch123> you could link it as tutorial :p
17:11:20 <TrueBrain> dude has a nice voice, so that helps
17:11:20 <frosch123> that guy creates ottd-codereviews regulary now
17:11:33 <frosch123> i think it's the same guy who was on the podcast with nielsm
17:12:04 <TrueBrain> he actually goes into the code?! Haha :D
17:14:02 <TrueBrain> I like how the specs make you select bits to base the decision on, but also adds the probabilities together and handle that for you :)
17:14:11 <TrueBrain> I was afraid it would have to add up till the bits selected
17:14:19 <TrueBrain> feels like half-way someone was like: no, we could do better :D
17:15:30 <TrueBrain> okay, I hope this dude can explain more parts of OpenTTD for us .. as that town name generator is coolly done :)
17:18:25 <frosch123> there is at least one another video about the tileloop and the way ottd picks the tiles
17:18:44 <frosch123> maybe we could ask him to explain tgp :p
17:20:04 *** wallabra has joined #openttd
17:44:53 <Samu> why does order backup need to sync with clientid
17:45:42 <glx> the issue might be in a totally different location
17:45:50 <Samu> i'm testing this Command<CMD_CLEAR_ORDER_BACKUP>::Do(DC_EXEC, ob->tile, static_cast<ClientID>(user));
17:46:35 <Samu> but i assume this is gonna desync
17:46:40 <glx> maybe #9908 is just a symptom triggered by a special case that should have been handled way before the company deletion
17:50:49 <Samu> it's closing the depot window, the depot still exists
17:50:57 <Samu> what if the order is swapped
17:51:40 <Samu> deleting the depot would also cause the window to be closed, hmm probably still gonna assert
17:59:12 <Samu> I moved CloseCompanyWindows(c->index); after ChangeOwnershipOfCompanyItems(c->index, INVALID_OWNER); in CmdCompanyCtrl
17:59:46 <Samu> might break something else now
18:01:55 *** gelignite has joined #openttd
18:11:40 <TrueBrain> owh boy ... spam incoming
18:12:10 <frosch123> new month, new spam
18:12:43 <TrueBrain> owh, seems it is once again failing to make PRs
18:13:00 <TrueBrain> but dependabot still doesn't have grouping, and it is rather noisy otherwise :P
18:15:30 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
18:29:36 <TrueBrain> hmm .. I might just switch to dependabot, and take the spam that comes with it .. still think best to filter it out from this channel, but we can see how it goes :)
18:29:53 <TrueBrain> it is roughly 5 to 10 PRs per repo per month
18:30:48 <TrueBrain> we have around 10 repos :P
18:31:36 <TrueBrain> well, that maybe tomorrow .. now time for some telly :D
18:34:42 *** gelignite has quit IRC (Read error: Connection reset by peer)
18:34:53 *** gelignite has joined #openttd
18:47:07 <supermop_work> got a 2nd 6x6 back - it's in nicer shape than the one that came with the camera
19:13:45 <peter1138> andythenorth[m], what the heck is Lego's new road layout system...
19:15:48 <supermop_work> ooh should make a NRT road set with the 90's blacktron baseplate roads
19:16:24 <supermop_work> not really a good fluorescent green in the palette though
20:11:40 *** gelignite has quit IRC (Read error: No route to host)
20:11:52 <frosch123> peter1138: the reason why people switch to non-lego bricks
20:11:56 *** gelignite has joined #openttd
20:22:05 <supermop_work> my brother sent me a cute little technic pallet tele-handler a couple weeks ago when i had covid
20:28:29 <Samu> i just triggered the crash in another manner
20:29:11 <Samu> i was in the company, with depot window open, i waited for it to bankrupt
20:29:30 <Samu> and it asserts the same way
20:33:33 <frosch123> sometimes you need forceps, but i like that
20:36:36 <glx> peter1138: seems more modular but looks ugly
20:37:06 <frosch123> glx: they are too small to put two vehicles next to each other
20:37:51 <frosch123> maybe they copied that from ottd :)
20:37:58 <glx> oh width 6 should be fine it seems
20:39:28 <glx> but newer vehicles seem to be width 8
20:52:24 <Samu> client and server can have different OrderBackup pools, it's not synced
20:52:39 <Samu> i just tested 2 openttd, client + server
20:53:02 <Samu> client was in the company, created depot, did those steps, and crashed
20:53:29 <Samu> server was in another company, but still opened the depot window of the client
20:53:48 <Samu> client crashed, server didn't
20:54:08 <Samu> server's backup was empty
20:56:41 *** gelignite has quit IRC (Quit: Stay safe!)
20:58:08 <Samu> I still don't understand why it needs to be synced
21:03:20 *** sla_ro|master has quit IRC ()
21:04:08 <Samu> nevermind, apparently is synced
21:07:41 <glx> OrderBackup::ResetUser has @note Must not be used from a command. I guess OrderBackup::Reset should follow the same rule
21:08:54 <Rubidium> Samu: IIRC if you sell a vehicle in the depot and then build another, the orders of just sold vehicle are placed inthe new vehicle. For this every clients needs to know the orders that need to be set. Having said that, maybe with the new command stuff you can just pass the order in the command if that needs to be done, and thus not having to bother about retaining the backup except locally
21:08:56 <glx> btw in both cases the crash/assert happens when the company is removed
21:10:07 <glx> so maybe it should be possible to auto remove the backup everywhere without issuing a command
21:10:59 <Samu> i created an orderbackup in a client, and visual studio does also createed it
21:22:11 <Samu> waiting for bankrupt again
21:23:03 <glx> probably faster to use the remove company method
21:25:40 <Samu> doesn't let me delete company with someone in it
21:28:41 <glx> I think I understand what happens
21:33:08 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:34:05 <glx> if the depot is manually deleted then OrderBackup::Reset() is called with from_gui = false, then the window is close and OrderBackup::Reset() is called again with from_gui = true but does nothing (and doesn't crash) as the backup is already removed
21:35:01 <glx> but in the case of company deletion (bankrupt or other) the windows are closed before removing assets, so with from_gui = true and a backup exists
21:35:28 <glx> and that means command call in command, and it's a big no
21:43:41 <Samu> gotta go sleep, cyas goodnight
21:43:45 *** Samu has quit IRC (Quit: Leaving)
21:56:19 *** frosch123 has quit IRC (Quit: be yourself, except: if you have the opportunity to be a unicorn, then be a unicorn)
21:58:25 *** Etua has quit IRC (Quit: Etua)
22:02:47 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
22:14:20 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:16:55 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
22:17:32 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
22:18:27 *** snail_UES_ has joined #openttd
22:36:43 *** Wormnest has joined #openttd
22:51:51 *** snail_UES_ has quit IRC (Quit: snail_UES_)
continue to next day ⏵