IRC logs for #openttd on OFTC at 2023-12-16
00:03:32 *** esselfe has joined #openttd
00:17:37 <DorpsGek> [OpenTTD/OpenTTD] michicc opened pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
00:20:03 <DorpsGek> [OpenTTD/OpenTTD] Ecmans started discussion #11594: Where Can i Buy Counterfeit Money Online?+33779724106
00:20:51 <DorpsGek> [OpenTTD/OpenTTD] michicc updated pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
00:21:19 <DorpsGek> [OpenTTD/OpenTTD] Ecmans started discussion #11595: Hol vΓ‘sΓ‘rolhatok online hamis pΓ©nzt? +33779724106
00:22:56 <DorpsGek> [OpenTTD/OpenTTD] Ecmans started discussion #11596: Dove posso acquistare denaro contraffatto online?+33779724106
00:27:09 <truebrain> That is new. Blocked user, reported content.
00:27:27 <_glx_> did the same
00:27:41 <_glx_> and deleted discussions
00:28:32 <_glx_> and it's a french mobile number πŸ™‚
00:29:31 <truebrain> Didn't want to delete the discussion till GitHub replied, but too late now πŸ˜›
00:30:28 <_glx_> I deleted only the two last ones
00:30:39 <DorpsGek> [OpenTTD/OpenTTD] michicc updated pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
00:31:14 <_glx_> the first was already gone
00:31:36 <truebrain> Let's hope GitHub Support can see them anyway πŸ™‚
00:32:35 <truebrain> But for next time, let's not delete them till GitHub Support replied πŸ™‚
00:32:54 <truebrain> After blocking the user they become invisible anyway πŸ™‚
00:34:23 <DorpsGek> [OpenTTD/OpenTTD] WenSimEHRP commented on pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
00:37:25 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
00:44:08 <DorpsGek> [OpenTTD/OpenTTD] glx22 commented on pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
00:52:18 <DorpsGek> [OpenTTD/OpenTTD] michicc updated pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
00:54:14 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
01:02:33 <talltyler> This would be nice to see in 14.0 if the schedule allows πŸ™‚
01:03:14 <talltyler> (We have a milestone for it)
01:19:44 <Eddi|zuHause> do we have a schedule?
01:20:22 *** Tirili has quit IRC (Quit: Leaving)
01:30:22 <DorpsGek> [OpenTTD/OpenTTD] michicc updated pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
01:37:02 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
01:40:52 <DorpsGek> [OpenTTD/OpenTTD] michicc updated pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
01:47:22 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 commented on pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
01:54:55 <DorpsGek> [OpenTTD/grfcodec] rubidium42 commented on pull request #30: Codechange: Silence clang warnings about va_start
02:09:56 <talltyler> Eddi: TrueBrain said he’ll have some time to review my NotDaylength PRs next week, and once those are in I’ll put together a changelog. I am hoping for a beta1 release in January.
02:10:30 <talltyler> (Obviously it’ll be past next week when they’re merged, but I’m really hoping we’re nearing the end of this project πŸ™‚ )
02:50:39 <DorpsGek> [OpenTTD/grfcodec] glx22 commented on pull request #30: Codechange: Silence clang warnings about va_start
03:28:40 *** Wormnest has quit IRC (Quit: Leaving)
03:52:13 *** D-HUND has joined #openttd
03:55:39 *** debdog has quit IRC (Ping timeout: 480 seconds)
04:08:25 *** D-HUND is now known as debdog
05:06:36 *** keikoz has joined #openttd
05:11:21 <merni> michiccviaGitHub: With #11593, is it still possible to opt for the pixel font if that's what you want?
07:36:30 *** keikoz has quit IRC ()
08:23:31 *** Flygon_ has joined #openttd
08:30:25 *** Flygon has quit IRC (Ping timeout: 480 seconds)
08:30:29 *** Wolf01 has joined #openttd
08:42:35 *** keikoz has joined #openttd
09:27:10 <truebrain> _glx_: `It does appear that the reported content has been moderated or removed at this time. `. Seems GitHub Support doesn't take action if we already removed the content πŸ™‚
09:35:12 <truebrain> hopefully they still ban the user from the platform ... silly spam bots 😦
09:43:26 *** Flygon_ has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
09:53:02 *** HerzogDeXtEr has joined #openttd
10:03:00 <DorpsGek> [OpenTTD/OpenTTD] zephyris commented on pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
10:08:12 *** nielsm has joined #openttd
11:35:02 *** Wolf01 is now known as Guest10571
11:35:04 *** Wolf01 has joined #openttd
11:36:08 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
11:37:35 <michi_cc[d]> merni: Right now it is not possible. Would need a good UI design idea. But exactly beacuse of things like this, it's so far only a draft PR.
11:39:31 *** Guest10571 has quit IRC (Ping timeout: 480 seconds)
11:41:58 <locosage> seems kind of pointless if ttf matches bitmap font nearly perfectly anyway
11:44:19 <michi_cc[d]> Some of the usual suspects will complain anyway.
11:55:04 <xarick> river stall 😦
11:55:16 <xarick> it's not as rare as I thought it to be
11:58:19 <xarick> Can I ask for a merge for, it is a mitigation of the issue, but not a complete fix.
11:58:27 <xarick> so don't close the issue
12:01:43 <xarick>
12:01:43 <xarick> There's a question mark at the start of this sentence
12:13:50 <merni> michi_cc[d]: Well, right now font switching is only through the console, so why does it need a UI here? Adding a UI could be a separate thing
12:15:03 <merni> Maybe the console font command with a blank name would mean the pixel font as it does now, with "OpenTTD Sans" or whatever for the new TTF font, but set the latter as default
12:15:30 <merni> also didn't peter share a screenshot of a font GUI here some time ago?
12:22:34 <peter1138[d]> πŸ”₯ pitchforks at the ready
12:23:23 <peter1138[d]> Why do we even need text, just replace the UI with hotkeys πŸ˜‰
12:26:13 <talltyler> xarick: Make the reference to the pathfinder setting a constant and I’ll approve it. Peter doesn’t like the arbitrary 10,000 but whatever number we put in there is going to be arbitrary (unless someone has a better number in mind). It’s just for a sanity check so I’m not particularly worried about it β€” 10,000 is better than infinite. πŸ™‚
12:30:00 <xarick> a const?
12:30:06 <xarick> const finder.max_search_nodes =; ?
12:32:42 <xarick> it does not like that
12:49:23 <andythenorth> michi_cc[d]: I am a big advocate of the pixel font, but that's about how it looks and feels πŸ™‚ It's really really limiting to have it implemented as bitmap sprites, with all the inherent limitations.
12:49:40 *** moll has joined #openttd
12:49:56 <andythenorth> Using an actual font is so much the better implementation, as long as we don't get an Uncanny Valley apperance from 'pixels, but vectors'
12:50:11 <peter1138[d]> I don't mind the 10000, but I'd rather a constant defined amongst other pf constants instead of buried in a function. I think.
12:50:15 <andythenorth> same for the UI icons, vectors would be vastly preferable, if they can avoid uncanny valley
12:50:48 <peter1138[d]> Well. I need to work on SVG sprites again.
12:51:20 <andythenorth> the limits of the pixel font include accessibility, and character support πŸ™‚
12:51:33 <andythenorth> can't even use ' marks in grf strings
12:51:40 <peter1138[d]> The way I did it so far is not ideal. (No colours allowed)
12:52:13 <andythenorth>
12:52:39 <peter1138[d]> Looks fine to me
12:54:38 <peter1138[d]> You could always just define the ' you want on your newgrfs
12:54:52 <peter1138[d]> Don't need to define a whole font.
12:54:58 <_zephyris> peter1138[d]: There's two types of icon - ones that should be binary and recolourable like text and ones that should be full colour.
12:55:12 <peter1138[d]> I know πŸ™‚
12:55:31 <_zephyris> Probably makes sense to use the unicode private range for the former?
12:55:42 <_zephyris> And SVGs for the latter would be fun
12:56:11 <peter1138[d]> Ah, but the only reason those icons can't have colour is because they're defined as glyphs
12:56:22 <DorpsGek> [OpenTTD/OpenTTD] Ufiby commented on pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
12:56:41 <peter1138[d]> Vehicle type icons are an exception, the rest can all be sprites.
12:56:43 <peter1138[d]> But
12:57:03 <peter1138[d]> Currently the only way to have an interface scalable sprite is via glyphs
12:58:36 <peter1138[d]> Some icons were originally glyphs but we switched them over to sprites long ago.
13:00:22 <peter1138[d]> Mind you ttf can have colours these days. Maybe we could make that work.
13:00:56 <peter1138[d]> But then redefining via grf is a pain.
13:02:04 <peter1138[d]> I should get up and think in front of vs code...
13:05:17 <_zephyris> andythenorth: You can define arbitrary unicode characters as addons to the sprite font, right?
13:05:35 <_zephyris> Extending the sprite font by building from TTF was my original plan
13:08:33 <_zephyris> Oh, I also found a bug(?). If _any_ of the fonts is missing a character then _all_ fonts get changed to the fallback system font. Is that intended?
13:09:59 <peter1138> That's known :)
13:11:06 <peter1138> I, er, hope you don't mind that we have taken your font and put it straight into OpenTTD anyway ;)
13:13:30 <_glx_> Yeah the annoying fallback mechanism
13:15:46 <peter1138> Should be simple enough to make it per-FontSize as a "quick fix".
13:16:00 <peter1138> But then a proper fallback system would be nice.
13:23:30 <_zephyris> peter1138: It's great! Well worth the effort πŸ™‚
13:28:53 <andythenorth> _zephyris: Yes…but…real font cuts this step? πŸ˜‰
13:30:46 <peter1138> Kinda. A baseset still needs to define at least the original glyphs that TTD had.
13:31:11 <peter1138> But it means they can be automatically created from the TTF.
13:43:11 <peter1138[d]>
13:43:13 <peter1138> Modern art?
13:45:20 <peter1138> Not entirely sure why the water tiles work :D
13:47:37 <talltyler> What script is that?
13:48:03 <andythenorth> hmm
13:48:26 <andythenorth> can you make it so the blitter can selectively render by tile type always?
13:48:36 <peter1138> No car goal
13:48:46 <talltyler> peter1138[d]: Ah, so following #10623?
13:48:56 <_zephyris> peter1138: Yeah, a sprite font generating tool is on my todo...
13:49:22 <xarick> do I create a NPF_MAX_SEARCH_NODES and a YAPF_MAX_SEARCH_NODES in pathfinder_type.h? or just one MAX_SEARCH_NODES for both?
13:49:49 <peter1138> talltyler, uh...?
13:51:00 <talltyler> I don’t know what’s happening with that PR (approved but not merged, waiting on …somebody? To make new settings), but it creates nice constants instead of using settings.
13:51:20 <talltyler> Oof, too early in the morning to think coherently
13:51:26 <talltyler> Maybe ignore me for now πŸ˜›
13:51:41 <talltyler> It was a long week
13:51:45 <peter1138> That one was "wait for the next release so we get survey data"
13:53:24 <peter1138> talltyler, ah, in Discord there is context, but I am not reading in Discord atm :)
13:55:11 <peter1138> Hmm, right, how to quickly look up the correct size sprite.
13:55:38 <peter1138> Or perhaps split the sprite cache by zoom level.
13:55:44 <peter1138> Did JGRPP do that/
13:57:59 <peter1138> LordAro, definitely a wash-the-bike day :o
13:58:03 <talltyler> Oh you meant my reply context
13:58:59 <peter1138> Yeah, from my pov you were asking about a script :)
13:59:54 <peter1138> Hmm, so those sprites work due to the order they are defined in the baseset.
14:03:25 <xarick> or do I put it in aystar.h?
14:06:52 <_glx_> xarick: common to both, like the depot one I think
14:07:22 <xarick> landscape.h has already an include on aystar.h
14:07:26 <xarick> oops
14:07:29 <xarick> landscape.cpp
14:24:26 <peter1138> Hmm, in theory it's now possible for opengfx2 hi def to not contains normal zoom viewport sprites...
14:24:48 <peter1138> (It was always possible, but nml required it)
14:29:54 <xarick> what kind of magic code is this? `static bool first_init = true;
14:29:54 <xarick> if (first_init) {` and apparently first_init is false
14:31:07 <peter1138> It's static so the first time it's seen it is true, then the conditional sets it to false.
14:31:17 <xarick>
14:31:36 <xarick> ah, the static is doing tricks
14:31:40 <xarick> ok I see
14:31:45 <peter1138> Not exactly tricks.
14:33:07 <xarick> so that's what static does.... it sets things once, then forgets that line ever existed
14:36:01 <peter1138> That's entirely not correct :)
14:45:42 <_glx_> it's just a local variable, but the value is kept between calls
14:46:27 <xarick>
14:46:27 <xarick> benchmark finished... gotta find a way to hide that table being displayed though
14:52:36 <xarick> noise 25 is missing 😦
14:52:41 <xarick> I forgot
14:54:06 *** rau117 has quit IRC (Quit: User went offline on Discord a while ago)
14:57:25 <peter1138> Hmm
15:29:40 <peter1138> Apparently not.
16:30:22 *** Wormnest has joined #openttd
16:56:50 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
17:29:28 <DorpsGek> [OpenTTD/OpenTTD] michicc updated pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
17:31:24 <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
17:33:03 <michi_cc[d]> Now with a very-ppor-man's font selection toggle in the game options window.
17:34:35 <michi_cc[d]> merni: The PR got a sprite font toggle.
17:43:40 <peter1138> Sounds reasonabl.
17:43:46 <peter1138> And reasonable.
17:44:19 <peter1138> Although I built a font selector UI, it feels out of place.
17:44:32 <peter1138> And with our own custom font it seems entirely unnecessary.
17:48:16 *** Wormnest has joined #openttd
17:59:31 <michi_cc[d]> Yeah, finally managed to pass all checks πŸ™‚
18:01:22 <xarick> english question: initiated, or initialised?
18:01:51 <xarick> > Fix: Don't let AyStar max_search_nodes unattended when initialising / initiated
18:03:37 <xarick> with s or z? what is more british?
18:05:07 <_glx_>
18:05:53 <talltyler> Ship pathfinder test game is live! Server name is `Oops, all boats`
18:06:02 <_zephyris> peter1138: Still has a place IMO. Currently the TTF covers the same glyph set as the sprite fonts, so still the same problem for unsupported languages.
18:08:04 <peter1138> Test build for ship pathfinder test game:
18:09:32 <xarick> why did I read this article.
18:10:07 <peter1138> You like wasting your time reading AI-spewed nonsense?
18:10:26 <xarick> yes, kinda, sometimes
18:11:24 <_zephyris> Might be weird, but could have font section window as an option only if characters are missing?
18:12:05 <_glx_>
18:12:05 <_glx_> still some hard to spot slopes πŸ™‚
18:12:21 <talltyler> Not on OpenGFX2 πŸ˜„
18:13:26 <_glx_>
18:13:26 <_glx_> here too πŸ™‚
18:13:45 <_glx_> only temperate and arctic are improved IIRC
18:14:08 <peter1138> Huh
18:14:21 <peter1138> I didn't know the water was different...
18:14:21 <_zephyris> Huh^2
18:14:40 <peter1138> No, it's working for me.
18:14:44 <peter1138> Are you using 8bpp?
18:14:45 <_zephyris> It's not AFAIK, there's only one set of sprites and one nml for river slopes
18:15:44 <LordAro> peter1138: yup
18:16:14 <_glx_> shouldn't, blitter line is empty in config, and it's on opengl driver
18:18:21 <peter1138> Out-of-date files?
18:19:38 <_glx_> I'm using the files from the zip
18:19:38 <michi_cc[d]> _glx_: Do you have `support8bpp = yes` in your config file?
18:20:01 <_glx_> support8bpp = hardware
18:20:06 <_glx_> yeah that could be it
18:20:12 <michi_cc[d]> If yes, it seems that we prefer `8bpp-optimized` even with the ogl driver.
18:20:44 <michi_cc[d]> Oh, yes, it's not a bool toggle. Anything that is not none will prefer the 8bpp blitter.
18:21:49 <_glx_> it's better with none
18:23:53 <peter1138> Yeah the default extra graphics having 32bpp won't automatically trigger 32bpp mode, I think.
18:24:03 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #11544: Fix #10452: Limit river generation max_search_nodes
18:33:34 <talltyler> What does the "Processing map took too long" error indicate?
18:36:57 <DorpsGek> [OpenTTD/OpenTTD] eints-sync[bot] pushed 1 commits to master
18:36:58 <DorpsGek> - Update: Translations from eints (by translators)
18:37:32 <DorpsGek> [OpenTTD/OpenTTD] zephyris commented on pull request #11593: Feature: Include scalable OpenTTD TrueType fonts by default.
18:38:57 <peter1138> It means the client took too long downloading the map and then catching up to the current state on the server.
18:39:16 <talltyler> So probably not a new pathfinder issue?
18:41:03 <peter1138> Well it could be if updating the state takes too long, but probably not.
18:41:46 <Eddi|zuHause> i thought pause on join was suppose to mitigate that
18:43:38 <peter1138> It should.
18:52:36 <xarick> I had a "great" idea! Make Valuate function multi-threaded!
18:52:48 <peter1138> lol
18:53:10 <xarick> the game state is paused at that point, I guess
18:53:13 <_glx_> Won't happen
18:53:38 <xarick> okay :/
19:00:56 *** Tirili has joined #openttd
19:10:27 <andythenorth> generally try not to valuate
19:10:47 <andythenorth> I did try caching to avoid valuating, but it bloated memory sooooo fast πŸ™‚
19:10:58 <andythenorth> I think I managed to use about 500 MB or so on a large JGRPP map
19:11:26 <peter1138> Doesn't valuate do funny things to instruction counting?
19:11:33 <peter1138> Or is it just that it can't be paused?
19:13:11 <Eddi|zuHause> i would think that each valuate step counts towards instructions, as that is counted deeper in the interpreter
19:13:47 <Eddi|zuHause> what didn't work was things inside the interpreter, like qsort or something that get run directly
19:13:51 <andythenorth> such Squirrel, very GS
19:14:11 <andythenorth> /me wrote a GS once
19:16:56 <andythenorth> actual proof
19:18:22 <andythenorth> it does pretty much everything I want, at least since async mode was added
19:18:25 <andythenorth> it's probably a bit buggy
19:18:36 <peter1138> That's just proof it exists, not that you wrote it.
19:18:43 <andythenorth> could have been the AI
19:19:02 <andythenorth>
19:19:02 <andythenorth> the only thing I didn't solve yet ^
19:19:13 <andythenorth> can't link story book pages
19:19:43 <andythenorth> the GS is prone to crashing on stuff like "I magic-bulldozed an industry whilst it was in a monthly industry loop"
19:20:02 <andythenorth> error-handling stuff in loops that might be above my pay grade as a devloloper
19:20:37 <andythenorth> I think things like that probably just need a guard every time e.g. an industry is accessed for get/set
19:21:17 <Eddi|zuHause> we should add a browser and reimplement the wohle ui in html5 :p
19:21:52 <andythenorth> 'that escalated fast'
19:21:56 <peter1138> Why?
19:21:58 <andythenorth> there's already a command to open a story page
19:22:07 <andythenorth> just no text string way to access it
19:22:22 <Eddi|zuHause> i heard games do that nowadays.
19:22:30 <andythenorth> oh and GS ability to respond to clicks in user-friendly time is ..... unreliable πŸ˜„
19:27:14 *** Flygon has joined #openttd
19:28:03 <_zephyris>
19:28:03 <_zephyris> Not too bad...
19:30:05 <peter1138> *100
19:31:04 <andythenorth> shall I play OpenTTD?
19:31:09 <peter1138> Join the ship game.
19:31:20 <andythenorth> oh, I'll need a build?
19:31:45 <andythenorth> ok downloading
19:31:52 <andythenorth> are you all on voice, being witty?
19:32:08 <peter1138> No
19:32:13 <talltyler> No, just text chat in-game
20:06:01 <_glx_> oh I was about to start my second line
20:06:24 *** kuhnovic has joined #openttd
20:06:24 <kuhnovic> People seem to enjoy using ships now hehe
20:06:44 <_glx_> this kind of map is only viable with ships
20:06:56 <peter1138> Especially when other vehicle types are disabled :)
20:07:04 <kuhnovic> Fair enough πŸ˜›
20:08:11 <peter1138> I added penalties to make ships prefer odd / even tiles depending on direction.
20:08:16 <peter1138> It's... not ideal :)
20:09:01 <peter1138> Needs the distance-from-coast penalty but that's going to be expensive without map support.
20:13:27 <peter1138> Ah, I guess the odd / even thing needs to be different for diagonal directions.
20:14:32 <andythenorth> deep water anyone? πŸ˜›
20:14:45 <talltyler> Not ideal how? Performance or effect?
20:15:48 <peter1138[d]>
20:16:10 <peter1138[d]> Because of diagonals (well, horizontal/vertical)
20:17:43 <DorpsGek> [OpenTTD/OpenTTD] glx22 opened pull request #11597: Add: Horizontal scroll for script debug log
20:22:33 <kuhnovic> peter1138[d]: Yeah I guess you'd have to flip the prefered directions for each odd tile
20:25:28 <kuhnovic> Btw my finder still uses the vanilla pathfinder to some extent, It first uses a high-level search to find an intermediate destination instead of the final destination. It then users the regular YAPF ship pathfinder to find a path to that intermediate destination. So any tweaks or changes to the YAPF ship would still have an effect.
20:25:53 <peter1138> Yes, I remember that being the case :)
20:28:05 <andythenorth> very ships
20:28:18 <_glx_> hmm I didn't expect a crash during regression
20:28:20 <Eddi|zuHause> you can probably "fix" diagonals by adding the track(dir) to the tile id in creative ways
20:28:41 <andythenorth> maybe I pull vanilla and play a cdist game
20:28:50 <Eddi|zuHause> so the upper track on an even tile makes it odd, or something
20:29:03 <peter1138> I'm trying for diagonals to add the penalty based on direction. For NE/NW/SE/SW, it's based on tile.
20:29:40 <peter1138> because for diagonals the "path" is encoded into the trackdir.
20:29:45 <Eddi|zuHause> well, the nature of diagonals is that you alternate between even and odd tiles
20:29:50 <peter1138> UPPER_E vs LOWER_E
20:30:06 <Eddi|zuHause> so the penalty averages out
20:31:27 <Eddi|zuHause> so upper and lower need different parity, and add that to the parity of the tile, then both upper/lower and tile parity alternate the same way
20:32:19 <xarick> `TimerGameTick::counter % Ticks::DAY_TICKS == 0`
20:32:19 <xarick> I'm looking for an alternative to this. It's for delaying the pathfinder to run a search for nearest depot. I wouldn't like it to create spikes
20:32:51 <Eddi|zuHause> read "add" as "xor" in the context of parity
20:33:14 <xarick> don't want all vehicles trigger the search at the same time.
20:33:36 <peter1138> Ah yes, of course, it alternates between upper & lower for each tile. Hmm.
20:34:54 <peter1138> Just use one of x or y I guess, as only 1 changes each tile.
20:36:07 <peter1138> Eh, as both change each tile, sorry.
20:36:58 <Eddi|zuHause> maybe you can use the parity of (x+y)/2
20:36:59 <_glx_> xarick: replace 0 with something like `v->unit_number % Ticks::DAY_TICKS` or `v->index % Ticks::DAY_TICKS`
20:37:13 <Eddi|zuHause> haven't thought through how that looks, though
20:38:26 <xarick> oh, good idea!
20:44:36 *** Xaroth92 has quit IRC ()
20:45:01 *** Xaroth92 has joined #openttd
20:45:49 *** rau117 has joined #openttd
20:45:49 <rau117> *For some reason, the ship-test version has terrible performance. FPS drops to a few units when I move the map on max unzoom. In 13.4 and the current master version everything works fine. I hope this is a ship-test specific issue*
20:47:02 <peter1138> I think that's a you think. No slow down here.
20:47:35 <locosage> may be different settings like max unzoom level
20:48:40 <andythenorth> how does JGRPP do ship track avoidance?
20:51:04 <xarick> now I just need 74 road vehicles searching for the nearest depot that doesn't exist!
20:51:25 <_jgr_> It's fairly crude direction changes when another shop is in front
21:00:41 <rau117> rau117: Idk, if this is significant enough to create a real bug report. After all, there is no such thing in the current master.
21:01:16 <DorpsGek> [OpenTTD/OpenTTD] glx22 updated pull request #11597: Add: Horizontal scroll for script debug log
21:03:30 <_glx_> #11440 helped a lot to fix the regression crash
21:06:12 <peter1138> Hmm?
21:06:39 *** tokai has joined #openttd
21:06:39 *** ChanServ sets mode: +v tokai
21:06:53 <_glx_> a change from my PR cause crash during ctest run because it needed fontcache
21:07:27 <_glx_> and you had the same issue in #11440
21:07:37 <_glx_> with a fix included
21:09:25 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #11548: Fix #10511: Delay 'go to nearest depot' orders
21:09:38 <peter1138> Ah.
21:09:44 <peter1138> I will say, that's the wrong place to get the string width.
21:11:08 <_glx_> yeah and almost redo it in InvalidateWindowData(), but it's to not recalculate string bounding boxes everytime a new one is added
21:11:16 <peter1138> Width should be calculated in Invalidat... yeah, there.
21:11:25 <peter1138> You can *store* it in the string.
21:11:28 <peter1138> er, in the log.
21:11:32 <_glx_> log can have at most 400 strings
21:11:34 <peter1138> But I wouldn't calculate when logging.
21:11:51 <peter1138> Basically, on invalidate, check for width = 0 and calculate it.
21:12:09 <peter1138> And also, if font is changed, invalidate all widths.
21:12:22 <peter1138> (font changed or interface scale changed)
21:12:31 <peter1138> might be simpler just to do that on a reinit.
21:12:46 <peter1138> On ReInit update all widths, on invalidate update just the width of the new lines.
21:12:55 <peter1138> But don't calculate the width when it's being logged
21:12:56 <_glx_> the font change part should be done
21:13:01 <peter1138> (It might never get displayed)
21:13:35 <_glx_> font change cause ReInit, and I invalidate with data=-1 there
21:19:41 <DorpsGek> [OpenTTD/OpenTTD] glx22 updated pull request #11597: Add: Horizontal scroll for script debug log
21:20:07 <_glx_> yeah it's cleaner when done only in OnInvalidate
21:21:42 <peter1138> Yes :)
21:22:11 <peter1138> GetStringWidth is basically GUI code so should stay in GUI land :)
21:22:57 *** Smedles has quit IRC (Quit: - Chat comfortably. Anywhere.)
21:23:24 <_glx_> BTW quick word wrap test was actually incorrect due to how vscroll is used (it supposes only 1 line at a time)
21:23:45 <_glx_> so the bottom lines where invisible
21:24:01 <_glx_> *were
21:24:32 <peter1138> Yes, word wrap is a pain. horizontal scroll is easier
21:25:08 <_glx_> I think the scrollbar under the break stuff is the cleanest location
21:26:28 <_glx_> it would look ugly above the break stuff with vscroll
21:27:23 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick commented on pull request #11548: Fix #10511: Delay 'go to nearest depot' orders
21:31:16 <xarick> what if there were 5000 road vehicles
21:31:50 <xarick> 74 ticks becomes too small
21:33:01 <_glx_> if you have 5000 lost vehicles you have more problems πŸ™‚
21:34:24 <xarick> someone may do it to make servers unplayable
21:34:51 <_glx_> then it will be kicked/ban and the company removed
21:34:59 <_glx_> non issue πŸ™‚
21:35:56 <xarick> i wish pathfinders didn't need to give an answer right away
21:36:47 <_glx_> it's not really the pathfinder, but the fact "go to closest" needs to find a destination
21:37:21 <xarick> an interruptible pathfinder
21:38:05 <xarick> search that depot for 10 loops or so, if you don't find the answer in 10 loops, resume game then come back later and try 10 more loops
21:38:40 <_glx_> but starting point changed so you need to restart from begin
21:38:55 <xarick> oh yeah 😦
21:39:07 <peter1138[d]>
21:39:17 <peter1138[d]> That seems wrong 😦
21:39:37 <peter1138[d]> Oh it's a gap of 2, not 1.
21:40:21 <peter1138[d]> Separation based on tile index. Problem is it increases all path costs, so any cost limits are hit sooner.
21:41:52 <andythenorth> looks nice though eh
21:43:10 <andythenorth>
21:43:10 <andythenorth> I miss this
21:43:48 <andythenorth> Flyspray 6145 apparently
21:43:50 <peter1138> Ah the old random.
21:44:17 <andythenorth>
21:44:17 <andythenorth> oh this was the old ctrl-click implementation of Variants
21:44:28 <andythenorth> used the first / last vehicles as bits
21:44:56 <xarick> so much quality in the videos
21:45:45 <_glx_> andythenorth: typical bad feature from newgrf authors πŸ˜‰
21:45:53 <andythenorth>
21:45:53 <andythenorth> oh this was the lolz of windowshading the frame rate monitor
21:45:54 <andythenorth> 2019
21:46:21 <peter1138> ?
21:47:49 <_glx_> yeah it's faster when it doesn't need to redraw the window
21:47:59 <_glx_> drawing strings is slow
21:48:15 <peter1138[d]>
21:48:59 <peter1138[d]> Not perfect
21:50:19 <xarick> Am I the only one that can't create quality openttd videos?
21:50:26 <peter1138[d]> Yes.
21:50:27 <xarick> image quality
21:52:45 <andythenorth>
21:52:45 <andythenorth> did that m4v work?
21:53:05 <andythenorth> I guess it's ok if we only want relative FPS πŸ™‚
21:53:09 <peter1138> m4v, no. mov, yes.
21:53:26 <andythenorth> as long as we all measure fps with the fps window open, we all get the same hit?
21:53:42 <andythenorth> can I use this to inflate my reported CPU performance? πŸ˜›
21:54:00 <_glx_> no, many things influence openttd fps
21:54:29 <peter1138> I'm using Gnome's build in screen recorder. There's no config, and it just works ;)
21:54:51 <peter1138> Gamescript paused the game, and... I can't unpause it :D
21:54:58 <peter1138> GAME OVER FOLKS
21:54:58 <andythenorth> xarick: I just use Quicktime on macOS
21:55:01 <andythenorth> GAME OVER
21:55:06 <andythenorth> "that's all folks"
22:00:15 <_glx_> yeah GS pause has higher priority
22:01:50 <andythenorth> hmm FIRS station names πŸ™‚
22:02:03 <andythenorth> need some for recently added industries
22:02:04 <andythenorth> "Wire Rope Factory"
22:02:18 <andythenorth> GPT? πŸ˜›
22:11:40 <xarick>
22:11:40 <xarick> terribad quality again 😦
22:12:05 <_glx_> too much compression
22:17:52 <peter1138> Oh god what have we unleashed with AsyncMode...
22:18:10 <_glx_> visually it's a nice effect πŸ™‚
22:19:44 <andythenorth> I should revisit nuking cities, but with async
22:20:14 <andythenorth>
22:20:17 <andythenorth> also the buttons are quite unresponsive
22:20:49 <_glx_> buttons are fire and forget πŸ™‚
22:21:01 <andythenorth> think that was running on ffwd
22:21:09 <_glx_> or more likely fire and hope the script will catch it
22:21:15 <andythenorth> ^ that
22:27:50 <peter1138[d]> CircularTileSearch in the pathfinder, what could go wrong? πŸ˜„
22:32:19 <peter1138[d]> Hehe, stuttery path finder πŸ˜„
22:32:41 <peter1138[d]>
22:32:43 <peter1138[d]> Looks cool though.
22:33:18 <_glx_> using road driving side to decide ?
22:33:25 <peter1138[d]> Nope
22:37:37 <peter1138[d]>
22:37:55 *** nielsm has quit IRC (Ping timeout: 480 seconds)
22:37:58 <peter1138[d]> Hmm, that one glitches a lot.
22:38:12 <talltyler> Road driving side would require knowing which parallel track is free, I think
22:38:18 <talltyler> At least in two-tile-wide canals
22:38:53 <peter1138[d]>
22:40:28 <andythenorth> such nice
22:40:45 <andythenorth> reminds me of funfair dodgemts
22:41:12 <peter1138[d]> Okay, just testing the 4 extremes is enough, no need for a circular tile search.
22:41:24 <peter1138[d]> (We don't have teleporting ships)
22:41:55 <peter1138[d]> Be interesting to try that MP test save, but... won't load 😦
22:42:28 <_glx_> bump in the PR ?
22:42:37 <_glx_> oh of course bump
22:43:22 <peter1138[d]> Oh they avoid buoys too, oops.
22:45:22 <peter1138[d]> Hmm, station vehicle list ... would be nice if you could filter for vehicles currently at the station.
22:53:06 <DorpsGek> [OpenTTD/OpenTTD] 2TallTyler commented on pull request #11544: Fix #10452: Limit river generation max_search_nodes
22:53:58 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick commented on pull request #11544: Fix #10452: Limit river generation max_search_nodes
22:54:23 <xarick> my head is forgetting things...
22:55:43 <xarick> I was testing if it was actually applying the real 10000, so I changed that one temporarily, then forgot
22:55:49 <peter1138[d]> Ah, fails in specific case 😦
22:56:10 <Eddi|zuHause> <_glx_> using road driving side to decide ? <-- ships always go on the same side worldwide, and does not depend on road side
22:58:21 <andythenorth> what's the rule? turn right?
22:58:49 <andythenorth>,should%20make%20the%20appropriate%20move.
22:58:54 <andythenorth> ship overtaking πŸ˜›
22:59:02 <andythenorth> can articulated ships overtake?
23:03:33 <xarick> starboard
23:06:30 <peter1138[d]> Bah, annoying edge cases 😦
23:06:34 <peter1138[d]> Now it's slow again.
23:07:10 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #11544: Fix #10452: Limit river generation max_search_nodes
23:08:46 <DorpsGek> [OpenTTD/OpenTTD] PeterN merged pull request #11591: Fix: Make compact picker windows expand to fill if necessary.
23:09:02 <xarick> I'm bad at comments
23:11:00 <xarick> and I forgot to edit the comment too... It's alzheimer
23:13:19 <Eddi|zuHause> the best code is self-commenting.
23:15:36 <xarick> I was attempting to imply the default is a fail-safe mechanism, but it's still perfectly fine to be changed later. I need english lessons.
23:15:59 <_glx_> default implies that usually
23:20:56 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick updated pull request #11544: Fix #10452: Limit river generation max_search_nodes
23:22:07 <Eddi|zuHause> comments tend to explain things that don't need explaining, and miss out on the things that do need explaining
23:48:16 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)