IRC logs for #openttd on OFTC at 2024-12-02
⏴ go to previous day
00:22:31 *** tokai has quit IRC (Quit: c('~' )o)
00:35:59 *** talltyler has joined #openttd
00:35:59 <talltyler> Deleted my comment on 13139, failed reading comprehension 🙂
00:55:56 <peter1138> Hmm what did we do for stations.
01:03:32 *** ChanServ sets mode: +v tokai
01:09:22 *** belajalilija has joined #openttd
01:39:21 *** speeder has quit IRC (Ping timeout: 480 seconds)
01:43:26 *** speeder has joined #openttd
02:32:54 *** tokai|noir has joined #openttd
02:32:54 *** ChanServ sets mode: +v tokai|noir
02:39:48 *** tokai has quit IRC (Ping timeout: 480 seconds)
02:56:37 *** Wormnest has quit IRC (Quit: Leaving)
03:10:23 *** gnu_jj_ has joined #openttd
03:13:29 *** gnu_jj has quit IRC (Ping timeout: 480 seconds)
03:56:45 *** godbed_ has joined #openttd
04:00:04 *** D-HUND has quit IRC (Ping timeout: 480 seconds)
04:00:14 *** debdog has quit IRC (Ping timeout: 480 seconds)
06:16:24 *** keikoz has quit IRC (Ping timeout: 480 seconds)
06:23:51 *** Smedles_ has joined #openttd
06:27:00 *** Smedles has quit IRC (Ping timeout: 480 seconds)
07:13:05 *** MarkZwei has quit IRC (Quit: Reconnecting)
07:13:21 *** MarkZwei has joined #openttd
07:18:14 <DorpsGek> - Add: summary for week 48 of 2024 (by OpenTTD Survey)
07:21:51 <Markk> Hai guise, a question. when I downloaded the latest version of OpenTTD two things that I was used with using OpenTTD since 2006. The first was when dragging signals on diagonal tracks it jumps between the outer and inner block betweeen each one, so it looks all weird and my ocd doesn't like that. Can I change some parameter for this?
07:24:58 <Markk> The second one is the moving of the Location button on windows for vehicles, stations and industries. Previously it was on the botton of the window with a nicely sized button, but now it¨s on the top with a very small button, can I change some parameter to get this back? It's quite annoying.
07:25:34 <Markk> two things change since using Openttd* it should say in the first messange.
07:26:52 *** Smedles has joined #openttd
07:34:10 *** Smedles_ has quit IRC (Ping timeout: 480 seconds)
07:57:25 *** godbed is now known as debdog
08:18:01 <kuhnovic> But to answer your questions. 1 - no there is no such setting. 2 - not that I know of. There might be a hotkey for this, the easiest way to find it out is to look at your hotkeys.cfg file that's in your My Documents / OpenTTD folder on windows (not sure about other OS'es). I'm not near a computer atm so I can't have a look myself.
08:26:31 <kuhnovic> You will find that there are quite a few things that have changed since 2006 😉. Sometimes we have to make decisions that alter the user experience slightly to make the UI more logical or prevent clutter. Adding options to maintain both an old and new UI style (or any feature for that matter) is just a giant pain maintenance-wise. Suddenly there are bugs in one mode but not the other, and testing
08:26:31 <kuhnovic> becomes a nightmare because everything has to be checked twice. It's just not worth it. So instead we carefully consider our options, but unfortunately sometimes this means that something has to change.
08:40:48 <LordAro> disregard most of the comments, but you might want to try toggling the 'keep fixed distance between signals'
10:15:34 *** XYZ has quit IRC (Ping timeout: 480 seconds)
11:26:38 *** NGC3982 has quit IRC (Quit: leaving)
11:32:40 *** NGC3982_ has joined #openttd
11:58:52 <xarick> a good AI tool would automatically write commit messages based on what the diff looks like
11:59:12 <xarick> when will we get there?
12:14:07 <LordAro> i'd be surprised if it couldn't in some cases
12:17:22 *** Flygon has quit IRC (Read error: Connection reset by peer)
12:21:56 <xarick> ```bool any_action_occurred = false;
12:21:56 <xarick> for (Direction d : directions) {
12:21:56 <xarick> TileIndex t = AddTileIndexDiffCWrap(tile, TileIndexDiffCByDir(d));
12:21:56 <xarick> any_action_occurred = RiverMakeWider(t, tile);
12:21:57 <xarick> } while (any_action_occurred);```
12:21:57 <peter1138> I recommend understanding what you are doing, and then writing a short summary of that.
12:25:19 <xarick> it should be any_action_occurred |= RiverMakeWider
12:27:57 <xarick> ```bool any_action_occurred;
12:27:57 <xarick> any_action_occurred = false;
12:27:57 <xarick> for (Direction d : directions) {
12:27:57 <xarick> TileIndex t = AddTileIndexDiffCWrap(tile, TileIndexDiffCByDir(d));
12:27:59 <xarick> any_action_occurred |= RiverMakeWider(t, tile);
12:28:01 <xarick> } while (any_action_occurred);```
12:31:07 <xarick> well.. infinite loop detected 😦
12:58:02 <_glx_> A for loop enclosed in a useless do-while with a silly condition
12:59:21 <_glx_> (the for loop is always doing the same thing because the entry values don't change)
12:59:57 <peter1138> RiverMakeWider might do something different if it's already done it before.
13:01:19 <_glx_> True, but it's still a silly construct
13:02:44 <_glx_> First step is to determine the actions in normal language, then you can try to convert to real code
13:04:09 <truebrain> LordAro: currently I can't look at the work you did; just so you know I am not ignoring you, but only can look at it next week 🙂
13:04:38 <truebrain> I do appreciate your effort btw 🙂
13:09:23 <peter1138> "Why is this code so much slower?"
13:09:52 <peter1138> "Oh, I was still compiling master"
13:27:34 *** benjaminv has quit IRC (Ping timeout: 480 seconds)
13:28:49 <xarick> tile at DIR_S and DIR_W seem to be at odds
13:36:39 *** keikoz has quit IRC (Ping timeout: 480 seconds)
13:48:18 *** kuka_lie has joined #openttd
14:16:54 <xarick> now it's leaving river peaces behind 😦
14:23:59 <xarick> "at least it's connected "
14:25:23 <xarick> gonna give up on this circular reinforcement
14:32:18 *** kuka_lie has quit IRC (Quit: Reconnecting)
14:32:32 *** kuka_lie has joined #openttd
14:38:34 *** kuka_lie has quit IRC (Quit: leaving)
14:38:48 *** kuka_lie has joined #openttd
15:02:21 <xarick> to lambdify or not to lambdify
15:16:55 <peter1138> Well, I do think #13116 could have value but it does now have a higher bar for complexity vs improvement since #13094.
15:17:54 <peter1138> And there's no longer any need to do comparisons against master before #13094 🙂
15:23:06 <talltyler> xarick: You can still do it, but get it working before you open the PR 🙂
15:42:13 <xarick> can I invent a GetTileSlopeMaxZ function?
15:43:20 <talltyler> How can a slope have a max Z? It is relative.
15:43:37 <talltyler> I could be misunderstanding what you’re asking 🙂
15:43:48 <talltyler> If you need a function that doesn’t exist, just write one 😉
15:43:51 <xarick> GetTileSlopeZ already exists
15:44:06 <xarick> helps reducing double computations when i also want to get height
15:44:07 <talltyler> Then sure, go for it
15:46:02 <xarick> there's a GetTileMaxPixelZ interesting
15:48:17 <xarick> maybe.... GetTileSlopeGivenHeight could return 3 values for simplicity
15:48:31 <xarick> the slope, the min height and the max height, since it is already computing both
16:06:35 *** kuka_lie has quit IRC (Quit: leaving)
16:06:50 *** kuka_lie has joined #openttd
16:09:50 *** kuka_lie has joined #openttd
16:27:34 <xarick> 13 files touched just to add a zmax 😦
16:31:55 <xarick> how much slower did I make openttd
16:32:29 <xarick> looks like I'm doing this wrong
16:38:59 <xarick> I'm pretty sure I'm ruining performance somewhere
17:23:34 <kurzov> Minor QoL improvement suggestion ... Being able to click on a cargo in the "Accepted cargo" section of a station box, and it would bring up the Industry Chain window for that cargo. Maybe it exists in one of the external patches, I wouldn't know since I've only played on vanilla.
17:23:34 <kurzov> I wouldn't know how to implement such a feature myself, and I wouldn't exactly have the time to do so within the next few months (new job).
17:32:53 <talltyler> Oh, that’s an interesting idea
17:33:36 <talltyler> Do we have clickable text anywhere or just buttons…? Can GS make clickable text or am I hallucinating things?
17:35:17 <kuhnovic> Isn't a click event handled at window level, where we then check which widget was clicked? If that's the case then it should be relatively easy to add
17:42:38 <peter1138> The list of accepted cargoes is a formatted string. There's currently no reasonable way to get the position of each cargo out of a formatted string, so it's not possible to know which cargo name is being clicked on.
17:43:09 <talltyler> I’m not thinking about how, but whether it would fit our design norms (which are by no means “standards”) 🙂
17:43:10 <peter1138> The industry chain window has 'clickable text' but it only does that because the position of each string is known.
17:44:07 <peter1138> Text in the order list is clickable (if you press ctrl, you'll more the viewport to a station, etc)
17:44:58 <peter1138> And the stations within the waiting cargo list as clickable too.
17:45:03 <peter1138> So yeah, we have precedent.
17:45:19 <peter1138> They just by row though, not text position.
17:50:22 <peter1138> GS can't make clickable text, but it can do some kind of 'link' in the story book windows.
17:52:46 <andythenorth> GS Goal window auto-adds clickable to some things (town names?)
17:52:56 <andythenorth> does chat window do that also?
18:42:24 *** gelignite has joined #openttd
20:14:06 <andythenorth> hmm ` nmlc ERROR: "generated/firs.nml", line 436: Redefinition of constant 'CC_OPEN_BULK'.`
20:14:33 <andythenorth> should I commit to FIRS _with_ the nml constants file, without which it won't compile
20:14:49 <andythenorth> or with the constants file, which conflicts with the forked nmlc I am using?
21:28:35 *** kuka_lie has quit IRC (Quit: leaving)
21:28:53 <andythenorth> there are a couple of approved but not merged PRs also
21:44:17 <peter1138> Hmm, hunting the annotation.
21:49:30 <xarick> guys, do you oppose on this:
21:51:04 <xarick> should I credit _dp_ ?
21:52:32 <xarick> it was mostly his idea
21:53:27 <xarick> 4 years ago you would say no
21:58:39 <xarick> copilot helped me convert it to fully standalone
21:58:53 <xarick> no longer reliant on the table header
22:01:34 <peter1138> Changing one algorithm to another algorithm is not algorithmic, right?
22:08:19 <peter1138> Nope, I've no idea what that is.
22:09:09 <xarick> the table header is a set of tilediffs. If starting from a centre tile and going over all the tilediffs, it takes the form of a circle
22:09:44 <xarick> i (or _dp_) replaced the circle with Distance based
22:10:00 <xarick> DistanceSquare + DistanceManhattan
22:12:11 *** gelignite has quit IRC (Quit: Stay safe!)
22:13:12 <xarick> > loop variable ‘normal_tile’ creates a copy from type ‘const TileIndex’ {aka ‘const StrongType::Typedef<unsigned int, TileIndexTag, StrongType::Compare, StrongType::Integer, StrongType::Compatible<int>, StrongType::Compatible<long int> >’} [-Wrange-loop-construct]
22:14:04 *** nielsm has quit IRC (Ping timeout: 480 seconds)
22:15:18 <xarick> it does not like normal_tile being const?
22:18:54 <xarick> nobody else complained, only linux
22:30:37 <xarick> that progress updating is a bit of a mess
22:33:54 *** keikoz has quit IRC (Ping timeout: 480 seconds)
22:43:46 <peter1138> Yeah, everything is failing with check annotations. Hmm.
22:53:42 <xarick> normal_tiles.size() has an arbitrary size 😦
22:56:33 <LordAro> peter1138: i can't see any warnings in any of the build logs
23:14:40 <LordAro> that was only added in 3.20
23:14:52 <LordAro> which somewhat contradicts your other PR :p
23:15:15 <peter1138> If this fixes it then I guess I bump that one to 3.20.
23:15:30 <LordAro> or just disable the warning...
23:18:20 <peter1138> 3.20 is only 3 years old.
23:28:57 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
23:34:42 <peter1138> Well, that was useful.
23:36:21 <LordAro> annotation check output being unclear has been an issue for a while...
23:37:03 <peter1138> So like 59 is the first install() that adds a path to the destination dir.
23:39:17 <peter1138> CMake is turning into autotools with all its quirks :p
continue to next day ⏵
