IRC logs for #openttd on OFTC at 2025-03-05
β΄ go to previous day
00:30:31 *** SigHunter_ has joined #openttd
00:30:44 *** SigHunter has quit IRC (Ping timeout: 480 seconds)
02:03:09 *** herms has quit IRC (Quit: bye)
02:06:34 *** Smedles has joined #openttd
03:17:35 *** Wormnest has quit IRC (Quit: Leaving)
03:21:39 *** debdog has quit IRC (Ping timeout: 480 seconds)
04:44:20 <DorpsGek> - Update: Translations from eints (by translators)
06:00:40 *** D-HUND is now known as debdog
06:19:41 *** keikoz has quit IRC (Ping timeout: 480 seconds)
08:06:58 <_zephyris> Quick first go, needs tweaking, but overall looks great.
08:20:48 *** reldred has joined #openttd
08:49:01 *** emperorjake has joined #openttd
08:49:10 <emperorjake> Snowy fields next pls π
10:26:27 <peter1138> Well, the sprites are there.
10:27:22 <peter1138> We probably need to refactor the density data though.
10:57:21 <xarick> please merge 13720 so that I can be happy:)
11:32:20 <_zephyris> emperorjake: Totally agree, but I'd really like to see it done in a more generalised way... Snow covered fields, but also fields as a terrain type, custom fields for different industries, etc.
11:51:30 <_zephyris> Try not to side-quest PeterN though...
12:08:27 *** ahyangyi has joined #openttd
12:08:27 <ahyangyi> _zephyris: I like the rough edge feeling, and also the noisy intermediate sprite
12:08:31 <truebrain> Tnx eints, useful commit you have there π
12:08:33 <ahyangyi> The new sprites seem to not align though
12:08:42 <ahyangyi> (see those white edges around the sprites)
12:09:05 <_zephyris> Yeah, I messed up a few things. There's some almost but not quite transparent blue.
12:09:12 <_zephyris> All fixed now, I was just excited.
12:22:33 <peter1138> Plurals as genders feels cursed.
12:26:34 <frosch123> Many languages conjugate verbs using both genders and plural/singular
13:41:23 *** akimoto has joined #openttd
13:53:09 <_glx_> hmm all savegames made in nightly since f8b1e30 are actually "broken", do we need to fix them ?
13:56:33 <peter1138> Probably not, there's no reliable detection especially for savegames started before but saved during the issue.
13:57:18 <_glx_> yeah they can have a mix of valid flags and broken flags depending on engine introductions
13:58:15 <peter1138> Ironically it's one of the places where the enum bit numbers were not removed, because they need to stable for savegames... d'oh
13:59:03 <_glx_> hmm possible detection could be date check and EngineFlag::Available, if it's not set it's most likely due to the error
14:00:35 <peter1138> I don't like the idea of adding complex conversion for a bug that isn't in a proper release.
14:00:45 <_glx_> but we can also say nighlies may have bugs
14:02:08 <peter1138> Error: String 0xFFFF is invalid. You are probably using an old version of the .lng file.
14:02:12 <peter1138> Okay, that's weird :D
14:03:50 <_glx_> error message is weird, I don't think we ever had 0xFFFF strings π
14:06:26 *** kuka_lie has joined #openttd
14:09:43 <peter1138> Ah, okay, I think we need some conversion, or some fixing up :(
14:13:17 <peter1138> No, we can just add another test.
14:23:48 <peter1138> So anything that was available now gets a preview popup.
14:24:14 <peter1138> It seems that it's possible for engines that aren't normally visible as also flagged as available.
14:24:22 <peter1138> So that's a bit odd.
14:28:37 <peter1138> Hmm, StartupOneEngine sets available even if it isn't. Weirdness.
14:30:45 <peter1138> Okay, I think we can try fixing it quite easily. If EngineFlag::ExclusivePreview is set, but preview_asked is clear, then it's broken data.
15:09:05 *** akimoto has quit IRC (Ping timeout: 480 seconds)
15:19:20 *** michi_cc_ has quit IRC (Ping timeout: 480 seconds)
15:28:37 *** Wormnest has joined #openttd
15:29:15 *** michi_cc_ has joined #openttd
15:40:08 *** Wormnest has quit IRC (Quit: Leaving)
16:16:01 <_glx_> and if broken data shift right the flags
16:23:36 *** Wormnest has joined #openttd
16:58:35 <talltyler> So, title game competition needs to be organized. I guess I can do it again unless someone wants to volunteer.
16:58:37 *** HerzogDeXtEr has joined #openttd
17:00:24 <talltyler> How late do we want to push the release? Iβd like at least two weeks for entries, although Iβm fine with a short voting window β most votes happen in the first few days.
17:00:24 <talltyler> So it could potentially wrap up before April 1st but itβs a bit tight considering we always struggle to get enough good entries for a real competition.
17:06:17 *** locosage has joined #openttd
17:06:17 <locosage> omg, I just realized it's already April soon π
17:07:08 <peter1138> Please make sure the rules mention the map size (256x256 is the maximum) and no NewGRFs (including if NewGRFs were present but were removed.)
17:08:32 <talltyler> Yes, and thanks for the clarification map size β itβs always been βuse your best judgementβ but not everybody has that π
17:23:41 *** digitalfox has joined #openttd
17:23:41 <digitalfox> talltyler: Will the title game competition be using OpenGFX2 for 15?
17:23:41 <digitalfox> I know OpenGFX2 is not yet used by master, but I hoped it would be present on final release.
17:24:04 *** tokai|noir has joined #openttd
17:24:04 *** ChanServ sets mode: +v tokai|noir
17:24:09 <peter1138> OpenGFX is a user setting, not something the savegame cares about.
17:24:43 <talltyler> We had discussed shipping OpenGFX2 with OpenTTD 15 but I can't recall the result. Something about a hacky way the default baseset is loaded in Bananas, maybe?
17:25:14 <peter1138> For bootstrap, there's a hacky way, yes.
17:25:32 <_glx_> yeah hardcoded somewhere in bananas
17:25:52 <peter1138> However, bootstrap is mostly irrelevant. For releases on storefonts, OpenGFX is specifically included as part of the bundle.
17:29:10 <_glx_> ha it's a command line parameter of bananas server
17:31:01 *** tokai has quit IRC (Ping timeout: 480 seconds)
17:31:03 <digitalfox> So there is hope for GFX2 on 15 by default?
17:38:36 <LordAro> peter1138: could follow_vehicle be made private? (not logged into GH)
17:39:02 *** gelignite has joined #openttd
17:57:20 <xarick> openttd at 8k resolution is "interesting"
18:02:25 <frosch123> digitalfox: We can switch to ogfx2 any time we want/when it's ready. This is not tied to any ottd release
18:07:38 <digitalfox> IMHO I just think you would be cool to be part of the release of 15 π
18:11:23 <xarick> fix the hovercraft amount of coloured colour
18:11:36 <xarick> give it more areas it can be company coloured
18:15:09 <xarick> 256x256 is enough space for 1 company, 512x512 is enough space for 2, 1024x1024 for 4, 2048x2048 for 8, 4096x4096 for 15
18:21:36 <_zephyris> Bugs on github pls, but I've gone through the ogfx1 bugs and I think I've addressed them all in ogfx2... including hovercraft
19:32:45 *** draegdrwg has joined #openttd
19:38:21 *** dwfreed has joined #openttd
19:39:31 <_zephyris> FWIW, I think OGFX2 is essentially ready.
19:39:53 <_zephyris> Things can be improved, there's always room for improvement, but I'm feeling good about it.
19:43:35 <andythenorth> it's quite transformational π
19:43:48 <andythenorth> can we unlink zbase from bananas soon? π
19:46:20 <_zephyris> My penance is almost complete
19:46:35 <gwyd4016> xarick: I would argue it goes 1/4/16/64/256
19:46:39 <gwyd4016> Depends on preference
19:58:24 <xarick> oh, frozen time service interval... does it still needs fixing?
20:00:23 <xarick> it was fixed apparently, i missed it
20:03:11 <xarick> yeah, the defaults don't switch
20:04:31 <xarick> I wonder, what happens when i join a wallclock server
20:04:50 <xarick> when my current local setting is calendar
20:08:38 <_glx_> I think nobody touched that
20:09:00 <peter1138> The 'fix' worked at long as the window is not massive.
20:09:12 <_glx_> Ha Andy's debugger hack
20:09:44 <peter1138> Yeah, it's faster than it was, but drawing 5 very long lines every tick is apparently slow enough to be a problem still.
20:09:59 <peter1138> s/drawing/layouting.
20:11:13 <peter1138> Let's try it with my other patch.
20:11:21 <andythenorth> it's kind of fun
20:11:40 <andythenorth> if I remember correctly, it's also a manual refresh on story book page
20:12:25 <peter1138> I implemented progressive update, so it doesn't try to layout the whole thing at once.
20:12:53 <peter1138> But gets rid of the lag when switching page or resizing the window.
20:13:13 <peter1138> My other patch caches the layouter result.
20:13:28 <peter1138> So now I'm compiling a combination of both.
20:13:52 <_glx_> but it needs to layout to know if a multiline is visible or not
20:15:16 <peter1138> That's somewhat better.
20:15:49 <andythenorth> I thought we filed it under author error π
20:16:39 <_glx_> from 60ms average to 7ms
20:17:01 <peter1138> Oh, no, the average is wrong.
20:17:15 <peter1138> The rendering is so slow the average is still catching up.
20:17:24 <_glx_> ah then if we only look current it's more that 10 π
20:17:28 <peter1138> 400ms is about right.
20:18:24 <peter1138> Now it only needs to relayout if the content is changed, or if the window size changes.
20:18:28 <_glx_> and it's funny when you compare to viewport
20:18:43 <xarick> opengfx2 looks kinda sharp and bright
20:19:10 <_glx_> opengfx2 is closest to original
20:20:46 <peter1138> xarick, yes, OpenGFX 1 is drab, and brown.
20:21:01 <peter1138> Not as base as zBase, but.
20:21:18 <truebrain> We do not talk about zBase
20:21:30 <_glx_> zbase if full of colours, but flat
20:21:51 <truebrain> peter1138[d]: That is a nice difference π
20:22:49 <xarick> i only have access to the bananas version
20:24:26 <peter1138> Hmm, should probably drop the cache for non-visible pages.
20:26:02 <peter1138> truebrain, my main objection is that this is not a general purpose solution :(
20:26:11 <peter1138> It only works on this window.
20:26:22 <truebrain> and to fix one window because andy can't behave, is a bit silly π
20:30:59 <_zephyris> Still want to push the OGFX2 contrast a little more, but it's soo much nicer than OGFX2
20:31:06 <peter1138> Yeah. The fix I already did, to only draw visible content, was a sensible and worthwhile fix.
20:31:31 <peter1138> But adding caches and progressive layouting stuff just to one window is meh.
20:31:50 <truebrain> just restrict the pages to a sane number of characters
20:31:53 <truebrain> also solves the issue π
20:32:25 <_glx_> IIRC industry list can be super expensive to draw
20:32:32 <peter1138> Disable line wrapping :)
20:32:40 <truebrain> I like how you think π
20:32:53 <peter1138> _glx_, I think I fixed that one.
20:34:49 <peter1138> Nice, all these snowy rocks were invisible before.
20:36:20 <peter1138> 6b9dc8cfcf8fce79c93f8f14b2244f1d2d5305b2 was the industry list fix.
20:37:31 <_glx_> I'm old, my memory has issues π
20:37:40 <peter1138> Nah, it was a thing :)
20:37:50 <peter1138> Usually spotted in one of Xarick's savegames.
20:38:53 <andythenorth> hmm could we just 'update' zBase to OGFX2?
20:39:11 <peter1138> The author of zBase could do whatever they wanted with it.
20:41:42 <_zephyris> Maximum requirement < 15.0?
20:41:45 <xarick> it considers mail as freight still
20:42:02 <xarick> unless, that's not a filter
20:42:13 <_zephyris> Doesn't kill aBase of course...
20:43:45 <xarick> I better start creating issues
20:43:55 <peter1138> It just tests for not passengers.
20:44:31 <andythenorth> I would be less scathing about zBase but eh π
20:44:39 <andythenorth> if the creator is averse to it ....
20:44:45 <peter1138> Mail is freight, but often it's used a mixed mode consist.
20:44:54 <peter1138> That doesn't make it not freight.
20:47:37 <LordAro> _zephyris: as "unfinished" as it is, i think it'd be a shame to lose it entirely
20:48:37 <_zephyris> It did its job, but did't gather any real interest to make it into a nicer product.
20:48:45 <_zephyris> Not sure I've even got the source...
20:48:48 <xarick> getting the version of openttd is a pita
20:52:21 <_glx_> you don't need the version of openttd, requested API version matches openttd version
20:55:16 *** nielsm has quit IRC (Remote host closed the connection)
21:03:45 <peter1138> What the heck is "freighting"?
21:10:06 <xarick> at a minimum it's inconsistent
21:10:24 <_glx_> you load mail in truck station, not in bus station
21:12:10 <andythenorth> mail is ambiguous
21:17:53 <Rubidium_> so change the string to "Not passengers"?
21:20:38 <peter1138> Freight is not passengers :)
21:22:37 <peter1138> Hmm, how many strings do we cache...
21:23:18 <peter1138> /* TODO LRU cache would be fancy, but not exactly necessary */
21:23:25 <peter1138> Well... maybe we could?
21:24:15 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:26:43 <Rubidium_> peter1138: so should NewGRF cargo property 15 (freight status) be deprecated in favour of not having cargo class 'passengers'?
21:28:01 <peter1138> You could update the Freight cargo filter to use that flag instead of cargo class.
21:30:03 <frosch123> The freight property is about the weight multiplier in the settings, not about passengers
21:30:37 *** Wormnest has quit IRC (Quit: Leaving)
21:32:47 * peter1138 compares with master.
21:34:46 <Rubidium_> frosch123: tell that to xarick :D Why call it freight in one place, and not apply the 'freight multiplier' in the other place...
21:35:56 <frosch123> We could rename it :) "is_heavy_freight" or something
21:36:44 <peter1138> NewGRF specs are not things you can rationalise 20+ years later :p
21:36:45 <Rubidium_> that'd apply to script API as well, and then... is it really "is_heavy_freight" as I can set it for passengers as well
21:39:17 <Rubidium_> that's why changing the 'Freight' filter string is the easiest way. We wouldn't have a second definition of 'freight', so whatever NewGRF defines is freight is freight
21:41:18 <frosch123> Never saw that filter, is it new?
21:42:38 <Rubidium_> looks like November 2022, but then I haven't really noticed/used that filter
21:45:39 <peter1138> Hmm, so this is not bad, and IS a generic solution.
21:47:05 <peter1138> (Not quite as performant as my story-book only method.)
21:48:44 <frosch123> No idea why you would want to filter by passenger class. I can only think of busses and liveries. Both are weird
21:49:37 *** Wormnest has joined #openttd
21:50:09 <frosch123> We could remove that drop down entry, and instead add a bus/truck filter for the rv list only, and maybe a tram filter?
21:51:02 <frosch123> Or better: rename the freight filter to "bus" and only offer for rv
21:52:24 *** rightnut has quit IRC (Server closed connection)
21:52:25 *** rightnut has joined #openttd
21:53:41 <frosch123> Either way, I am pretty sure no-one ever wanted to filter for "weight multiplier applies"
21:57:32 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
22:04:04 <peter1138> Apparently the PR author did.
22:06:44 *** kuka_lie has quit IRC (Quit: Lost terminal)
22:16:53 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:23:05 <_glx_> the global string array was so permissive
22:23:26 <peter1138> Yes. I don't really like this way either, but at least it works now.
22:23:52 <peter1138> Problem is that wrapping the errors of sub-commands would make "doing it properly" a bit trickier.
22:26:43 <xarick> there's some issues with implicit orders indeed
22:33:38 <xarick> how would you resolve order position for the jump to order position
22:33:47 <xarick> which isn't a current order
22:35:04 <xarick> ResolveOrderPosition is incomplete
22:39:28 <xarick> not even sure what's expected
22:39:44 *** keikoz has quit IRC (Ping timeout: 480 seconds)
22:45:53 <xarick> InsertConditionalOrder also needs that
22:54:37 <_glx_> now it needs proper testing, but it should work
22:54:50 <_glx_> IIRC scripts don't see implicit orders
22:55:24 <xarick> i suspect GetOrderJumpTo is gonna not give the exact value that was given by SetOrderJumpTo
22:56:25 <_glx_> GetOrderJumpTo uses ResolveOrder so it should return the position passed the SetOrderJumpTo
23:00:31 <_glx_> for safety I did a quick search for "jump_to" in the file
23:02:10 <xarick> GetOrderJumpTo should have returned 2
23:07:48 <peter1138> Bah. Stupid reviewer.
23:08:27 <_glx_> well my fault too, the code is weird
23:09:07 *** gelignite has quit IRC (Quit: Stay safe!)
23:11:04 *** izhirahider has quit IRC (Server closed connection)
23:11:15 *** izhirahider has joined #openttd
23:18:09 <_glx_> ok there's no function to convert real index to manual index (yet)
23:21:32 *** Flygon has quit IRC (Read error: Connection reset by peer)
23:39:00 <_glx_> xarick: since you already have a script for testing, please test
23:49:54 <_glx_> oh and maybe fix alignment in tables, would be easier to read π
continue to next day β΅
