IRC logs for #openttd on OFTC at 2025-02-12
⏴ go to previous day
00:05:37 <yiffgirl> truebrain: the correct answer, of course, is to nuke windows and put an arch install on it instead. =p
00:21:08 *** akimoto has joined #openttd
00:23:24 *** tokai|noir has joined #openttd
00:23:25 *** ChanServ sets mode: +v tokai|noir
00:30:09 *** tokai has quit IRC (Ping timeout: 480 seconds)
00:31:45 *** k-man has quit IRC (Ping timeout: 480 seconds)
00:31:45 *** k-man_ is now known as k-man
00:44:39 *** tokai|noir has quit IRC (Quit: c('~' )o)
00:54:30 *** k-man has quit IRC (Ping timeout: 480 seconds)
01:05:46 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
01:40:40 *** Wormnest has joined #openttd
02:04:02 *** herms has quit IRC (Quit: bye)
03:07:20 *** Wormnest has quit IRC (Quit: Leaving)
03:53:49 *** debdog has quit IRC (Ping timeout: 480 seconds)
04:43:45 <DorpsGek> - Update: Translations from eints (by translators)
06:07:55 *** akimoto has quit IRC (Remote host closed the connection)
06:20:49 *** keikoz has quit IRC (Ping timeout: 480 seconds)
07:02:33 *** Smedles has joined #openttd
07:16:28 *** D-HUND is now known as debdog
07:49:52 *** HerzogDeXtEr has joined #openttd
08:38:11 *** Smedles has joined #openttd
09:01:59 <_zephyris> Sorry for the churn on the graph gridlines... Thought it was a simple rebase, then introduced a syntax error, then needed to upgrade compiler, then needed to fight git inlcuding whitespace changes on a new computer...
09:02:25 <_zephyris> But updated and looks good to me 🙂
10:24:57 *** kuka_lie has joined #openttd
10:33:16 <xarick> linux gaming looking promising
10:51:16 <xarick> openttd is building faster all of a sudden
11:24:04 <xarick> I forgot why I have `this.force_update`
11:26:09 <xarick> it's confusing the heck out of me
11:45:09 <xarick> they link to vehicles, or to stations at the time it's computed
11:45:34 <xarick> actually, to the tile the vehicles or stations were at the moment
11:46:02 <xarick> with a interval of a month or a year, these links make no sense
11:46:20 <xarick> they'd be scrolling viewport to something in the past
11:48:39 *** concerrox has joined #openttd
11:48:39 <concerrox> Hello, I want to compile OpenTTD myself. It works well for the debug variant, but the compilation fails for release. Can anyone offer some help? (in MSVC) 🤔
11:50:41 <concerrox> It's forked from the latest JGRPP
11:53:06 <LordAro> looks a bit like it's still trying to use debug opus.lib ?
11:53:16 <LordAro> i don't know MSVC though
11:53:49 <_glx_> Clear cache and reconfigure maybe
12:23:27 <xarick> I think I remember what's the purpose of this.force_update
12:26:28 <xarick> it's to update the rating element
12:27:31 <xarick> like for example, company value, everyone has a £1 value at the start, that puts the score to 100% for all
12:28:18 <xarick> but when a company is created, and force update isn't run, it will add an element with a score of 0
12:28:50 <xarick> problem is... the score is dependent on the values of the other companies
12:30:09 <xarick> to get the score, the code is currently set to run the valuations for all companies
12:30:36 <xarick> bad design, limitation 😐
12:31:28 <xarick> so in the case of updating every month, some tables in which new companies are created, the score are just placeholders
12:31:41 <xarick> not the real computed score
12:31:51 <xarick> how would I deal with this?
12:51:57 <concerrox> concerrox: I swapped the vcpkg release lib and debug lib manually and it fixes this, so the release now works trickily. But I still want to know whether there is a proper way is to solve this. (Clearing the cache doesn't work)🤔
12:55:12 <_glx_> CMake is supposed to pick the right libs
13:01:18 <peter1138> xarick, only add the tables when you do the period update.
13:10:31 <kuhnovic> concerrox: By "cleaning cache", do you mean the cmake cache, or the build folders?
13:22:17 *** benjaminv has quit IRC (Quit: Gateway shutdown)
13:24:07 *** benjaminv has joined #openttd
13:42:17 <_glx_> concerrox: hmm are you using "open folder" mode or do you create a solution ?
13:43:42 <_glx_> because looking at CI output I can see a potential issue in solution mode
13:44:33 <_glx_> screenshot from nightly build
13:45:29 <_glx_> and from a normal CI run
14:17:18 <xarick> how do I rename a member name in a table
14:18:47 <xarick> example I got this in the save game `local table = { badname = 0 }`
14:19:04 <xarick> but on the new version i decided to rename the badname
14:19:38 <xarick> local table = { goodname = 0 }, but when it loads, it's still badname
14:22:06 *** kuka_lie has quit IRC (Quit: leaving)
14:22:21 *** kuka_lie has joined #openttd
14:23:31 <_glx_> you need to manually handle the conversion
14:24:56 <_glx_> `table<-goodname = table.badname;` I think
14:29:43 <_glx_> and for a really safe version it probably should be ```if ("badname" in table) {
14:29:43 <_glx_> table<-goodname = table.badname;
14:36:21 <xarick> a bit too much work for the Load function
14:38:16 <_glx_> you can also just create new tables and discard the loaded one
14:39:26 <_glx_> but it would be simpler to say new version don't support old version data
14:40:28 <_glx_> or you can do the conversion outside of Load()
14:40:44 <_glx_> before really using the data
15:18:27 <xarick> GS crashing in the load part doesn't print anything
15:18:59 <xarick> report "the error" which is not shown
15:43:48 <xarick> ScriptInstance::CallLoad() is returning false
15:47:11 <xarick> and i'm not sure what the error is
15:47:25 <xarick> i just know it errors in load
15:48:07 <xarick> I placed an assert there which is failing
15:49:26 <_glx_> try replacing second SQFalse with SQTrue
15:57:24 *** gelignite has joined #openttd
16:07:03 <truebrain> LordAro: I deny everything!
16:13:18 <LordAro> truebrain: incidentally, thanks for your work on Azure Trusted Signing. I'm about to steal it all for work purposes :)
16:13:40 <LordAro> and by 'steal' i mean use it as a starting point
16:13:42 <truebrain> Redirect that tnx to orudge please 😄
16:14:16 <LordAro> $1000 for a 3 year EV codesigning cert, i mean good grief
16:14:49 <truebrain> Yeah, those prices boomed a bit
16:15:53 <LordAro> of course, it also turns out that we have approximately 70000 signings a month
16:16:16 <LordAro> which, tbf we really don't need to be signing all our debug builds
16:16:39 <LordAro> or all the (112) executables within the installer for non-releases
16:16:44 <LordAro> so that should cut things down nicely
16:20:47 <_glx_> so when I opened VS it told me there's a new 17.2.5 version, so I installed it, and now I needed to open another instance and it tells me there's a new 17.3.0 version
16:23:30 <merni> SetDParam(5, STR_EMPTY);
16:23:30 <merni> SetDParam(8, STR_EMPTY);
16:23:30 <merni> SetDParam(9, STR_EMPTY);
16:23:31 <merni> Why is this necessary here? Does the string parameter array not get cleared after a string is drawn?
16:24:30 <_glx_> and that's one of the reason peter is rewriting the system
16:24:41 <merni> can't wait for it to get merged :p
16:25:05 <merni> but in that case, how come such manual clearing is not more frequent and not done for all the parameters?
16:25:25 <_glx_> it's an issue in very rare occasions
16:26:20 <_glx_> here the order string is composed of multiple optional elements
16:27:15 <peter1138> Yes, in most cases it does not matter than parameters are set to some other value.
16:27:54 <peter1138> It used to matter a bit if those parameters were `char *` pointing to something that no long exists.
16:28:16 <_glx_> and drawn via a single STR_ORDER_TEXT, so properly init stuff is the right thing to do
16:28:25 <peter1138> But we've gone from "do the minimum to reduce work" to "do what's necessary to be safe"
16:29:19 <merni> so... it's a good idea to initialise every parameter to empty? :p
16:29:42 <_glx_> because it will access them
16:30:34 <_glx_> and the different paths in the function only set the params they care bout
16:30:43 <xarick> rats... the placeholder effect
16:31:11 <peter1138> It won't be unmerged for much longer.
16:31:52 <xarick> this table didn't exist in version 2, it version 3 it was pushed, it got all stats filled, but then didn't update the score because that's not the job for pushing tables
16:32:27 <merni> (all this because I am considering somewhat refactoring the whole order string generation business as suggested by Tyler's comment in 12267)
16:33:43 <peter1138> Sure, it's being split into smaller PRs, which have been merged already.
16:34:43 <merni> maybe I should base my branch on yours then to avoid a tedious rebase soon
16:35:52 <merni> why do I always write "paramenters"
16:41:20 <_glx_> _glx_: my intuition was correct, it happens in "solution" mode and I have a working fix for it
16:41:21 <peter1138> Pronounced correctly, or that weird way Americans say it?
16:41:48 <merni> geez peter has a lot of branches
16:42:24 <merni> well not that many on github thankfully
16:44:27 <peter1138> merni, I could push everything... hehe.
16:44:53 <peter1138> Many of those are irrelevant though of course.
16:45:29 <peter1138> I could send them via email, like Linux kernel maintenance.
16:46:19 <merni> everyone flood TrueBrain's inbox with patches
16:46:53 <peter1138> Is truebrain angry and tetchy enough to be our torvalds?
16:47:27 <peter1138> I know the answer to that but I can't say.
16:50:20 <truebrain> A PR that only adds code and doesn't change any existing line .. does that actually do something?
16:54:14 <Rubidium> truebrain: not yet, but #13510 uses it
16:55:29 <truebrain> Rubidium: Ah; the PR description wasn't clear to me it was meant like that 😄
16:55:51 <truebrain> I thought it was some template magic 😛
16:57:23 *** kuka_lie has quit IRC (Read error: Connection reset by peer)
16:57:36 <peter1138> I guess technically it is :D
16:57:52 <peter1138> uint32_t p1, p2. Much easier...
16:58:15 <peter1138> Remember when we had to worry about having enough bits in the network protocol...?
16:59:56 <merni> Hmm, I am thinking... is it really the best approach to have orders as simple strings in the order/timetable GUIs?
17:01:51 *** kuka_lie has joined #openttd
17:12:16 <merni> For instance, which is more readable?
17:12:16 <merni> 1: Go non-stop to Sanford (No loading)
17:12:16 <merni> 2: Go to Sanford Southblablahblah (Transfer and leave empty)
17:12:17 <merni> Destination Unloading Loading
17:12:17 <merni> 1: Go to Sanford Unload if accepted No loading
17:12:19 <merni> 2: (Stop at all intermediate stations)
17:12:19 <merni> 3: Go to Sanford Southblahblah Transfer all No loading
17:14:30 <merni> Not relevant right now, anyway...
17:15:54 <_glx_> don't forget the translations
17:16:25 <merni> the current strings make coherent phrases in English and in translations
17:16:45 <merni> the idea I proposed would essentially make it into a table with each line not necessarily reading as a single phrase/sentence
17:17:01 <merni> so translations would translate the individual parts rather than the entire thing
17:17:14 <_glx_> might actually be better indeed
17:17:44 <peter1138[d]> Fix the separation between order and timetable.
17:17:53 <merni> would also make it more explicit since some "default" options are currently hidden (like "unload if accepted", and of course non-non-stop orders)
17:18:47 <peter1138[d]> The two views of the same data.
17:19:06 <merni> they should be one screen?
17:19:34 <peter1138[d]> Always feels a bit off to me.
17:19:40 <merni> I feel like that would be a bit big, especially if we are using strings rather than icons for things like "unload if accepted"
17:20:21 <merni> Though I guess vanilla timetabling just has wait and travel times... so the TT data doesn't actually need a great deal of space
17:23:47 <merni> Could do something like
17:23:47 <merni> Destination Unloading Loading Timetable
17:23:47 <merni> 1: Go to Sanford Unload if accepted No loading Wait 75 sec
17:23:47 <merni> | (Stop at all intermediate stations) Travel 342 sec
17:23:47 <merni> 2: Go to Sanford Southblahblah Transfer all No loading Wait 10 sec
17:23:47 <merni> 3: Go to Sanford Northfoo Transfer all No loading Wait 5 sec
17:25:30 <merni> I think I saw a similar mock-up somewhere
17:25:56 <merni> wrt the way travel between orders is represented
17:27:16 <merni> The other issue would be button overload if both the order and TT buttons are in the same screen... but I guess many things could be changed to not need a button (for instance, clicking on "Unload if accepted" could open a dropdown right there with all the unloading options)
17:29:41 <xarick> is it british english?
17:29:53 <merni> no, it's misspelled english
17:33:04 <xarick> it's not a british spelling fixer
17:34:46 <merni> _glx_: One thing that might not work is when some languages (like Tamil and Hindi) would naturally write the equivalent of "Sanford to go" rather than "Go to Sanford"... in that case, if we put "Go to" at the start of the line forcibly they would have to do some awkward phrasing. But then, the Tamil translation already has very awkward phrasing that effectively says "Go here: Sanford" even though
17:34:46 <merni> a more natural phrasing could probably be achieved somehow
17:35:24 <merni> I don't think that issue will be a big problem, languages would be able to come up with something that can fit in that slot
17:36:05 <peter1138> Generally strings should not be composed from different strings.
17:36:19 <peter1138> So the order text completely breaks that rule.
17:36:24 <merni> well, orders presently do a helluva lot of that :p
17:37:09 <merni> so much so that I've written 1.5 pages of notes in Word just to figure out which parameters end up where and produce what
17:37:55 <merni> and am transforming that into a list of all possible final order text combinations to use for testing
17:42:57 <merni> With the current model of order texts I don't think it is possible to eliminate string composition, but I intend to at least limit it to one level
17:43:44 <merni> because atm there are strings made out of strings made out of strings
17:43:44 <peter1138> Didn't you want to change it to a table-based layout anyway, which doesn't need composition?
17:43:58 <merni> But that is likely to be much harder :p
17:44:18 <merni> Especially considering that I really have no idea how to do GUIs in openttd 😆
17:45:04 <merni> I want to refactor the existing code so at least I can understand it better before doing anything more drastic
17:45:41 <peter1138> I wouldn't bother refactoring that section. It's a candidate for being completely replaced when doing drastic (or even not drastic) changes.
17:47:41 <merni> Maybe I should do a Discussion to see if a table-based layout actually makes sense to people, before investing a bunch of time into it
17:48:07 <merni> First things first, I want to understand all the possible combinations of orders that GUI can show 😆
17:51:02 *** Wormnest has joined #openttd
17:52:37 <merni> "Service non-stop at"... that is pretty unclear
17:53:50 <_jgr_> Ideally, non-stop behaviour should be the default and shorter text
17:56:07 <merni> More than that, the non-non-stop behaviour, though set on the next order, is actually a behaviour of the path between two orders
17:56:38 <merni> And imo would make more sense to indicate "between orders" rather than on an order
17:59:01 <xarick> are you rewriting orders window
17:59:37 <peter1138> Timetables already have a "between orders" line.
17:59:56 <merni> but that doesn't indicate non-nonstop-ness
18:00:07 <merni> jeez we need a better word for non-non-stop
18:00:21 <peter1138> Yeah, that wasn't the point. The point is having two lines per order has precedent and works fine.
18:01:01 <merni> Yeah as far as that goes I agree
18:01:27 <xarick> add icons instead of text
18:01:39 <xarick> each icon reflects what's happening in that order
18:01:40 <peter1138> (Probably they end up touching the line before for one, and the line after for the other, which is a pain)
18:01:53 <merni> xarick: they would have to be very clear
18:02:40 <peter1138> 17:56 < merni> More than that, the non-non-stop behaviour, though set on the next order, is actually a behaviour of the path between two orders
18:03:16 <merni> Yeah, so that would touch the line before. What'd touch the line after?
18:03:30 <merni> Are TT travel times stored in the previous order?
18:04:23 <xarick> oh, and the icons could be a menu
18:04:42 <peter1138> I mean the travel times are 'after' the order, and the "non-stop" part is 'before' the order.
18:05:08 <merni> I wasn't thinking of travel times as being part of an order at all
18:05:11 <peter1138> I mean it's not a massive issue but consistency...
18:05:50 <peter1138> Well, they are. They are not separate things at all.
18:06:16 <peter1138> That was why I mentioned the weird split earlier :)
18:07:08 <merni> I haven't looked at the structure of orders at all so far so there will probably be some other ugly surprised if I try to do this :p
18:07:09 <frosch123> Implicit orders will complicate your life :p
18:07:09 <frosch123> It was suggested to add a toggle to hide them
18:07:35 <peter1138> Implicit orders are just orders that exist but weren't explicitly created.
18:07:50 <peter1138> I don't think they'd complicate things that much.
18:08:26 <peter1138> merni, oh. Yeah, if you're trying to restructure the order window then you really need to understand the underlying structure at least .
18:08:29 <peter1138> It's not that hard.
18:08:40 <frosch123> They have no timetable. And essentially occupy the between-order space you suggested for non-stop
18:08:53 <merni> I wasn't trying to do that at all unitl very recently :p
18:09:20 <frosch123> Hmm, maybe they have timetables... can't remember
18:10:26 <merni> implicit orders can disappear whenever the train happens to take a different route between explicit orders, right?
18:11:00 <merni> in that case, does it really make sense to give them timetables without transforming them into explicit orders?
18:11:16 <merni> (ie. is there any use-case for that?)
18:11:39 <merni> as someone who never uses implicit orders in gameplay I only have a vague idea of what they do
18:12:01 <peter1138> Are you redesigning orders, or just the UI, here/ :p
18:12:44 <merni> but i need to understand what the orders do to give the a logical place in the UI, no?
18:13:30 <andythenorth> implicit orders are 'visited stations graph'
18:20:36 <frosch123> They exist for cargodist. To visualize/support the routing of cargos.
18:20:36 <frosch123> For the same reason they have a readonly timetable, so cargodist can estimate the link capacity.
18:21:23 <merni> readonly timetable : how is it set? automatically somehow?
18:21:49 <merni> if that is the case, is there any reason to show them (and particularly their timetables) in the GUI?
18:23:00 <andythenorth> we can draw cdist overlays on the map, maybe we should draw the stations visited graph on the map 😛
18:23:01 <peter1138> Yes, because either you want them there, so it's useful to know, or you misconfigured things and don't want them there, so it's useful to know.
18:23:16 <peter1138> andythenorth, jgrpp probably does.
18:23:42 <andythenorth> we could use badges? 🙂
18:23:49 <merni> peter1138: Isn't it more useful to know "whether you have non-non-stop orders that can result in implicit orders" than the specific implicit orders that exist?
18:24:28 <merni> Perhaps the implicit order destinations could be listed in the "between orders" text but I really don't see a need for displaying their timetables
18:25:14 <merni> though trying to aggregate the timetables across implicit orders is probably harder than just leaving them in 😆
18:30:53 <xarick> in my league tables script
18:32:02 <xarick> most difficult rename is going to be the GSText as Text tables
18:32:12 <xarick> still unsure how to proceed
18:32:32 <xarick> text.str -> text.string
18:35:37 <peter1138> What's the script's name?
18:41:27 <xarick> oh, link is not a text, it's a link type + link target
18:41:50 <xarick> what will I call this combination as a single word?
18:44:06 <xarick> link -> link_key / link_data / link_pair / link_params / link_array
18:48:38 <xarick> gonna ask mr Copilot for ideas lol
18:57:24 <xarick> must have two words, or it's a bit ambiguous since I am putting it inside an array of other stuff
18:57:53 <xarick> three words is already too much
18:59:32 <xarick> i think link info is fitting maybe
19:00:27 <xarick> ` return [ rating, score_text, element_text, link_info ];`
19:50:16 <xarick> why do I have "rating" twice
19:50:35 <xarick> why did I do things this way
19:55:46 <Rubidium> cause mr. Co told you to?
20:02:09 <xarick> so much gymnastics with the variables here
20:03:34 <xarick> best_value is equal to rating
20:05:04 <xarick> one rating is the value that I send to be printed, the other rating is the value that is used as the value used as score for sorting the table by value
20:10:46 <xarick> i need to make sure vehicle_list.Count() is never 0
20:34:09 <xarick> wow, openttd has no settings!
20:55:59 *** k-man has quit IRC (Ping timeout: 480 seconds)
21:23:12 *** gelignite has quit IRC (Quit: Stay safe!)
21:35:16 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:36:30 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
21:39:15 *** k-man has quit IRC (Read error: Connection reset by peer)
21:39:16 *** k-man_ is now known as k-man
21:40:41 <johnfranklin> oops, english native speakers also need to fix spelling errors
21:42:34 <johnfranklin> meanwhile one of our member always misspells "default" by "defeat"...
21:56:24 <xarick> how do I traverse a table to print its content?
21:57:13 <xarick> its content can be integer, array, string, table, keep expanding and printing
22:01:50 *** k-man has quit IRC (Ping timeout: 480 seconds)
22:05:03 <andythenorth> call a method recursively
22:05:36 <andythenorth> try something like
22:06:41 <andythenorth> FIRS GS uses it to print debug to story book
22:08:10 *** Flygon has quit IRC (Remote host closed the connection)
22:15:24 *** kuka_lie has quit IRC (Quit: Lost terminal)
22:16:42 <andythenorth> pff not the result I want after a refactor that is supposed to be zero diff 🙂
22:16:58 <andythenorth> also macOS counts bytes 'interestingly'
22:17:07 <andythenorth> probably some disk sector thing
22:19:29 <andythenorth> pff really naptime
22:22:49 *** keikoz has quit IRC (Ping timeout: 480 seconds)
22:33:48 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
23:14:06 <xarick> recursive lambas are possible?
23:17:59 <_jgr_> (Some light reading for you)
23:19:52 <_jgr_> I expect that if you declare the local and assign the functor in separate statements, the problem may go away
23:24:29 <xarick> ```local rename_text = function(text, version, rename_text) {
23:24:29 <xarick> rename_text(stat[k], version, rename_text);```
23:26:01 <xarick> the vm must be fuming at this
23:26:31 <xarick> this was the hardest conversion i had
23:37:45 <peter1138> I imagine your user based will be very happy.
continue to next day ⏵
