IRC logs for #openttd on OFTC at 2024-03-08
β΄ go to previous day
00:19:20 <peter1138> Okay, what sets other than an ancient UKRS have running sounds...
00:21:51 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
00:25:39 *** emperorjake has joined #openttd
01:45:14 *** wallaby2 has joined #openttd
01:49:44 *** wallabra has quit IRC (Read error: No route to host)
01:56:11 <peter1138> `TTDPatch: Not enough free memory!`
01:56:33 <Eddi|zuHause> is NARS less ancient than UKRS?
01:57:34 <Eddi|zuHause> the germanrv set might have custom tram sounds?
02:01:26 *** D-HUND is now known as debdog
02:06:40 <reldred> Comparatively ancient
02:18:48 <peter1138> Normal sounds are fine, it's running sounds that can have timing issues.
02:28:22 *** Leopold_ has quit IRC (Remote host closed the connection)
02:30:20 *** Leopold has joined #openttd
02:55:28 *** Flygon has quit IRC (Read error: Connection reset by peer)
03:00:05 *** herms has quit IRC (Quit: bye)
03:02:27 *** Wormnest has quit IRC (Quit: Leaving)
03:44:16 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:49:17 *** gnu_jj_ has quit IRC (Ping timeout: 480 seconds)
03:49:24 *** D-HUND is now known as debdog
03:51:56 *** Leopold has quit IRC (Remote host closed the connection)
03:53:10 *** Leopold_ has joined #openttd
04:40:42 <DorpsGek> - Update: Translations from eints (by translators)
06:08:01 *** keikoz has quit IRC (Ping timeout: 480 seconds)
06:38:36 *** Leopold___ has joined #openttd
06:38:41 *** Leopold_ has quit IRC (Remote host closed the connection)
06:49:03 <kuhnovic> xarick: Curious to see what your fix is, do you have a commit somewhere?
08:30:27 *** Leopold___ has quit IRC (Remote host closed the connection)
08:31:57 *** Leopold has joined #openttd
09:02:49 <rau117> Is the absence of icons in the refit window intended?
09:02:49 <rau117> *There is even a small window where the icon just fits...*
09:11:22 <peter1138> Well they don't get drawn automatically, we need to code it.
09:12:45 <peter1138> Hmm, on X, 1387 followers, 84 likes, 24 retweets. On Fosstodon, 415 followers (now, it's gone up a lot since), 237 likes, 155 boosts.
09:13:50 <LordAro> a lot of dead users on twitter, i imagine
09:14:02 <truebrain> you seem alive to me?
09:14:18 <LordAro> that's what you think
09:14:26 <ahyangyi> Alive elsewhere but dead on X
09:14:27 <truebrain> just an AI trained on your conversations, I see
09:14:48 <truebrain> explains the lack of PRs π π π π (sorry, I had to, forgive me π )
09:15:29 <truebrain> you know I love you π
09:15:41 <truebrain> reminds me your last PR introduced a bug; we still need to fix that for 14.0 π
09:15:51 <truebrain> you can no longer put tram tracks on a road bridge π
09:17:12 <truebrain> not bad feature; unintentional consequences π
09:18:50 <peter1138> Do we need a "patch setting" to hide all those extra cargo icons I spammed?
09:19:12 <peter1138> But one (1) player complained!
09:19:25 <truebrain> Just because something is different, doesn't mean it needs toggling
09:20:09 <truebrain> peter1138: yes, and the irony is that the user that agree'd with that player, is now asking to add the icons in more places
09:22:33 <locosage> peter1138: would be nice to have some utility code int the game to add icons easier
09:23:41 <locosage> ideally it would be a string code I think
09:24:21 <locosage> but I so don't want to get into text layouters π΅βπ«
09:38:05 <peter1138> Cargo icons would want (slightly) separate handling from an add-your-own-custom-icon-in-text feature. anyway.
09:41:46 <locosage> yeah, I'm talking cargo icons here
09:41:47 <locosage> well, own-custom-icon can be added as svg-only
09:43:26 <locosage> though I guess it's not impossible to add cargo icons as svg-only too
09:43:41 <locosage> just don't show them for cargos that don't provide svg
10:24:29 <_zephyris> locosage: We need peter1139's svg sprites π
10:25:16 <_zephyris> But it does seem like there's a separate desire for more icons in the private unicode range
10:27:22 <locosage> more icons for the god of icons! xD
10:27:41 <_zephyris> Where can I find which of the unicode private range is currently used? Where's it safe to add more?
10:27:53 <_zephyris> Icons for the icon god, glyphs for the glyph throne.
10:28:15 <_zephyris> The typography warhammer 40k crossover you didn't know you needed.
10:46:03 <peter1138> Yeah, 95 new followers so far from the 20th Birthday fediverse post.
10:48:33 <michi_cc> _zephyris: U+E200 to U+E2FF is special in so far as that range will fallback to sprite glyphs if nothing is present in the loaded TTF. This is part of the normal PUA from U+E000 to U+F8FF.
10:50:58 <michi_cc> Of course some software uses parts of the PUA as well, e.g. Apple puts some stuff into U+F400 to U+F8FF.
10:55:19 <michi_cc> Oh, and from the same file, U+E000 to U+E1FF is potentially used for the encoded string control codes.
11:05:25 <peter1138> Yes but everything from E220 to E2FF is mapped to the spritefont's 32-255 range.
11:07:15 <peter1138> I guess most of those are not actually needed though.
11:10:41 <peter1138> We could maybe extend the range which would ensure nothing default uses it. But NewGRFs could still conflict.
11:31:34 <xarick> building openttd takes so long nowadays I forget what I was doing by the time it finishes
11:43:25 <LordAro> surely you're not doing full rebuilds every time?
11:43:55 <LordAro> unless you're modifying strings or some other low-level header, the rebuild time should be fairly low
11:45:45 <_zephyris> peter1138: No worse than newgrfs messing with base set sprites...
11:46:38 <peter1138> No, but when implementing a new solution, if we can do something better that avoids conflicts, we should do that.
11:47:18 <peter1138> 32 bit glyph labels? π
11:48:07 <andythenorth> what is mullering my CPU?
11:48:12 <peter1138> "I reserve the range 0xE300 to 0xE33F for my own personal NewGRFs, nobody else should touch them..."
11:48:27 <michi_cc> peter1138: It is only mapped if the TTF does not provide a matching glyph though.
11:49:37 <xarick> I blame translators, lol
11:49:43 <xarick> they're the ones modifying strings
11:50:14 <peter1138> No, it's NewGRF authors trying to use the string system to draw images π
11:50:54 <peter1138> I did also wonder if we could just provide a default set of glyphs based on what people are currently doing, and build that in.
11:51:25 <peter1138> If they're standard then they could be put into OpenTTD Sans too.
11:51:47 <peter1138> But "I need this other slightly different icon!" would be a thing.
11:52:25 <merni> Well, they also try to use the image system to draw strings... :p
11:52:57 <merni> That makes the image column of the whole list needlessly huge
11:53:31 <merni> and that is not even the worst part of that set...
11:54:26 <merni> Because someone likes their neat alignment with a massive number of spaces, I cannot see whether something is electric or diesel without making the purchase list cover the whole screen :p
11:54:53 <merni> The trains look nice though so that makes up for everything :p
11:54:58 <peter1138> What set is that? I need to see the horror for myself.
11:55:19 <peter1138> I'm not sure, they look very flat and plain. Not really any shading at all.
11:55:46 <merni> Czech Railways Set 1.0 and Czech Railways Set EXTRA 1.1
11:56:56 <merni> I guess what I like is that they look quite close to the real ones :p
11:57:34 <belajalilija> Thereβs another Czech set but its much smaller
11:57:44 <belajalilija> It is a bit better though
11:59:31 <peter1138> "Good job" on that alignment...
12:01:07 <merni> It probably works only on OpenGFX sprite font
12:01:17 <merni> At least, that's what I have and it works
12:02:30 <_zephyris> Support tabs characters?
12:02:56 <merni> I think a better idea is not to do such weird things
12:03:00 <peter1138> Perfect with OpenGFX 2.
12:05:14 <peter1138> It's actually not too bad with OpenTTD Sans.
12:05:24 <peter1138> Arial or Verdana or Tahoma is lolrious though.
12:06:55 <merni> SamuXarickviaGitHub: Is there a useful difference between stuck and lost?
12:07:17 <merni> also, wow, the number of PRs is growing close to the number of issues
12:07:32 <peter1138> So, anyone here know anything about AS2?
12:15:17 <peter1138> And how many CRS326-4C+20G+2Q+RM are appropriate for home use?
12:19:09 <xarick> merni: it would mean ships would go out wandering aimlessly which is "prettier" to look at.
12:20:35 <xarick> ... design philosophy of Kuhnovic
12:21:23 <xarick> ... I'd prefer they'd get closest to the best tile they could get, but that's another story
12:32:23 <peter1138> When there's no path so there's no closest...
12:35:42 <reldred> peter1138: meh, crs518's for home use π
12:40:55 <peter1138> Plugged in to andy's MBP.
12:51:56 <LordAro> peter1138: clearly should automatically collapse repeated space characters
12:52:09 <LordAro> peter1138: also, probably zero
12:53:13 <peter1138> They'd just work around it with different types of space π
12:53:47 <peter1138> If you want your NewGRF to look unsightly, go ahead.
13:05:30 <andythenorth> peter1138: but using powerline-etherner
13:11:48 <andythenorth> I have powerline ethernet for the TV stuff, there are about 5 devices in one cabinet, and there's 2 cast iron radiators directly between them and the nearest AP
13:13:06 <andythenorth> the powerline ethernet is **slower** than the highest real rate on wifi, but probably more stable, and getting those devices off the radio seems to cause other nearby devices to have a happier time
13:20:44 <truebrain> merni: there is a simple fix for that ...... π Well, two actually.
13:22:54 <michi_cc> Oh no, it's growing even more π
13:23:37 <truebrain> there, made room for you
13:24:03 <michi_cc> Now there's even more room π
13:28:53 <LordAro> so much room for activities!
13:29:14 <truebrain> that PR isn't actually helping .. does things on both sides ...
13:31:48 <truebrain> some PRs are marked as "backport requested", but I am not always sure that is a good idea π
13:34:38 <peter1138> Unflag it then π
13:36:10 <truebrain> I am always so scared of introducing stupid bugs so close to release, because the fix "looked fine" π
13:40:27 <peter1138> I think that particular one was not flagged by me π
13:42:06 <truebrain> dunno .. do we want to? No clue π
13:42:11 <LordAro> any titlegame submissions yet? :p
13:42:59 *** TrueBrain_ii has joined #openttd
13:43:05 <_glx_> #12205 needs reviewing π
13:45:08 <truebrain> _glx_: I assume that review thing is resolved?
13:46:34 <merni> truebrain: well, if all these backport-requested PRs are going to be backported (33 atm), that might improve the playing experience quite a bit
13:46:53 <truebrain> yup; but we have to be careful to not add crashing bugs π
13:46:54 <merni> Especially ones such as 12158
13:47:16 <merni> truebrain: Isn't it better that happens in RC2 than in the final release? :P
13:47:34 <truebrain> merni: bugs we introduced in 14.0 should be fixed, yes π That is always easy π
13:47:40 <_glx_> it's better to not backport more than necessary π
13:47:41 <truebrain> it is more those fixes that are ... 10+ years old bugs, etc π
13:48:59 <merni> Yeah... even fixing 10 year old bugs is a good thing and why deprive 14.0 users of that and make them wait a year?
13:49:05 *** TrueBrain_ii has quit IRC (Quit: Page closed)
13:49:15 <truebrain> that is the always ending argument
13:49:18 <truebrain> "why not add this feature"
13:49:28 <truebrain> but in the end, stability is more important than perfection π
13:49:36 <_glx_> depends on bug severity too
13:50:19 <truebrain> and yes, that makes fixes around March each year always a bit meh, as it takes a year before the end-user sees them
13:50:22 <truebrain> solution: release more often
13:50:26 <truebrain> owh boy, how often we said that π
13:50:47 <_glx_> yeah "nobody" plays nightlies
13:52:06 <LordAro> last time we tried that we had a beta for about 6 months then just released in April anyway
13:52:17 <truebrain> we had an early release once!
13:52:30 <truebrain> because I really did not want to wait 6 months to test out the new multiplayer on a big group of people
13:56:34 <truebrain> but the only thing that caused was that the next release took 18 months, not 12 π
14:03:44 <talltyler> I will work on my blog post today, so you can tease it in tomorrowβs post.
14:37:17 <michi_cc> talltyler: There's a teaser for the survey blog in it right now. Can be changed of course, but please "fight" with TB over the spot π
14:40:16 <xarick> wow, yapf changes got merged, that is so scary
14:42:00 <LordAro> it's going to leap out and attack you
14:42:49 <xarick> I should notice an immediate difference in performance
14:42:53 <xarick> on these large savegames
14:44:14 <truebrain> michi_cc: no no, you can pick who is next π You are the author of the blog π We all picked the next person π
14:44:45 <michi_cc> Well, then it stays as it is. I want to keep my pun π
14:48:18 <kuhnovic> xarick: Why is that?
14:48:54 <truebrain> kuhnovic: more importantly, check your DMs π
14:58:30 <belajalilija> peter1138: oh god where is this
14:59:20 <xarick> I thought he hated OpenTTD existence
14:59:39 <michi_cc> You expect qualified comments on the Steam forums?
15:00:29 <truebrain> they should send CS a letter, asking if he is proud
15:00:33 <truebrain> and publish the answer
15:02:19 <LordAro> iirc the phrasing used was less harsh than that
15:03:46 <LordAro> more along the lines of "i don't like it as TTD was 'finished' and any further changes does not fit my vision"
15:04:02 <LordAro> but maybe that was overinterpretation and wishful thinking too
15:04:35 <michi_cc> kuhnovic: Finally paying back your debt from being at the Brussel meeting I see? π
15:05:19 <kuhnovic> All I had to do is show up to a meetup! Oh yeah and build a new ship PF π
15:05:32 <truebrain> Gratz kuhnovic ; it is well deserved π
15:05:42 <michi_cc> Now find some button to press.
15:05:56 <kuhnovic> Thank you guys, I feel honored
15:07:03 <xarick> maybe better performance, as detect destination is called less often now, am I right?
15:08:13 <peter1138> CREDITS.md is wrong.
15:08:26 <kuhnovic> xarick: It actually has very little to do with performance, I wonder if it would even be measurable. It's more about pathfinder correctness. Starting nodes can be destination nodes too, and that case was handled poorly
15:08:27 <truebrain> I ordered him to fix that first, yes π
15:10:02 <peter1138> Also it might not hurt to credit some 'minor' contributors too.
15:11:15 <xarick> first impressions seems to be improved performance
15:12:27 <kuhnovic> That's unforeseen then π
15:12:49 <xarick> not actually measuring it
15:13:10 <xarick> maybe I should use TicToc
15:16:07 <truebrain> no errors on the backend .. hmmm
15:17:05 <truebrain> it does work if you log in
15:17:35 <peter1138> Also Zephyris made some improvements to OpenTTD Sans for Cyrillic, would be nice to get a new release merged π
15:18:02 <truebrain> better font support!
15:19:07 <peter1138> I think there were some Hebrew characters as well, not sure if they are done yet.
15:19:21 <talltyler> michi_cc: No, thatβs perfect. Gives me another week to procrastinate! π
15:20:16 <_glx_> truebrain: 14.0 table looks wrong when logged in
15:20:33 <truebrain> michi_cc: for some reason the caching cached a white page π
15:20:57 <truebrain> now I need to remember how I could remove it from the cache π
15:21:33 <truebrain> or delete the cached page
15:21:45 <truebrain> _glx_: that is a "not-me" problem π
15:22:35 <peter1138> `CBID_VEHICLE_BUILD_PROBABILITY` is making it too wide.
15:22:51 <peter1138> Probably just remove that text π
15:22:52 <truebrain> right, so I should fix my caching script to not store 0 byte caches π
15:23:39 <truebrain> seems it is a rare bug, as I can't find any other 0 byte cache entry π
15:23:41 <xarick> hmm, strange to see AAAHogEx not performing. I'm using Time features...
15:23:53 <_zephyris> peter1138: Yup, but I was waiting to see if any new translations introduced any more missing characters.
15:24:20 <_zephyris> peter1138: Some sketches, nothing confirmed as readable by a Hebrew speaker!
15:24:35 <truebrain> almost 27 languages are fully translated; damn
15:25:16 <truebrain> (14 are actually fully translated, but 13 are very close; can't blame them for not staying up-to-date constantly π )
15:25:41 <truebrain> still leaves 23 languages that are not complete, some not even close
15:25:42 <peter1138> _zephyris: Is that because it's unreadable, or no response?
15:25:57 <xarick> What does Time do to scripts? They're not performing
15:26:35 <_glx_> it doesn't change anything on script side
15:26:46 <peter1138> IIRC it's not as complex and picky about shapes as something like Arabic.
15:27:28 <xarick> AAHog at the bottom of profits is not normal
15:27:29 <peter1138> Maybe we can get some in game comparison screenshots?
15:27:48 <_glx_> except some part of the game go slower
15:28:56 <_zephyris> peter1138: Yup, seems much more flexible. I've broken design convention by having heavier vertical rather than horizontal strokes - that's the TTD style, but I can't find any Hebrew fonts that do that
15:29:52 *** gelignite has joined #openttd
15:30:37 <xarick> I need to check how Time was implemented for scripts, seems that it breaks their flow
15:31:38 <_glx_> nothing special was needed for scripts
15:32:05 <truebrain> They are already special β€οΈ
15:35:04 <peter1138> Hmm, anyone able to confirm audio on Windows isn't totally broken by #12227 being merged?
15:35:18 <peter1138> I forgot that, uh, it's not actually something I tested, or can test.
15:35:57 <merni> is it in the current nightly?
15:37:23 <merni> I can hear music in latest nightly, anyway
15:37:25 <xarick> oh, I have to use AITimeMode for everything that is time related now? π¦
15:37:50 <merni> and also cash registers
15:38:26 <xarick> let me check what my own AI is getting from AIDate
15:38:28 <_glx_> no, default is economy mode, which is fine
15:40:55 <kuhnovic> For CREDITS.md: `- Koen Bussemaker (Kuhnovic) - General / Ship pathfinder (since 14)` ?
15:41:25 <michi_cc> Make a PR with whatever you like π
15:42:35 <xarick> AIDate.GetCurrentDate() returns the date from the bottom left screen or the date from that period thingy?
15:42:46 <talltyler> You could be the Buoy Butcher π
15:46:34 <xarick> let's experiment setting economy mode for the entire script
15:47:11 <_glx_> no need to set anything for economy mode
15:49:07 <xarick> something's not working for my script, and I suspect other scripts as well
15:49:24 <xarick> I use AIDate.GetCurrentDate() massively everywhere
15:53:12 <xarick> I'm the bringer of bad news...
15:53:28 <merni> You know where the report goes :p
15:53:35 <peter1138> Come on you know how to properly report π
15:54:29 <_glx_> if it's a custom build we can't do anything though
15:55:23 <xarick> it's a Release build without debugs... gotta redo with debug
15:55:55 <_glx_> won't help, we can't open dmp without the corresponding exe
15:56:15 <merni> can't it be reproduced on a nightly/RC?
15:56:38 <xarick> it was master, but it was a Release, not even RelWiThDebInfo, no symbols
15:56:51 <_glx_> you can open the dmp yourself
15:57:51 <_glx_> (if you still have the corresponding exe)
15:59:21 <_jgr_> xarick: If you're doing local testing, it's probably a good idea to stick to build types which have symbols in them
16:00:04 <_glx_> symbols are outside of exe with MSVC anyway
16:02:58 *** Wormnest has joined #openttd
16:05:25 <xarick> waiting for the crash to happen again
16:05:45 <xarick> I'm in RelWithDebInfo now
16:20:10 <kuhnovic> Do you have a savegame to reproduce it?
16:22:17 <kuhnovic> That could very well be it *hides under desk*
16:24:27 <_glx_> dmp is useless in this report, but json has the decoded trace π
16:25:21 <_glx_> anyway it's an assert so we know where it happens
16:25:54 <_glx_> and most likely the "why" won't be in the dmp
16:26:53 <_jgr_> Looks like it is #12181
16:34:56 <michi_cc> Hmm, interesting. In #12234 I moved some code between two commits without changing the combined diff, and GitHub didn't dismiss the approval.
16:35:30 <_glx_> if global diff is the same it's not a chang π
16:37:24 <kuhnovic> I will look at that pathfinder issue tonight. Gotta put the little human to bed first.
16:52:47 *** Smedles has quit IRC (Read error: Connection reset by peer)
16:53:28 *** Smedles has joined #openttd
16:58:58 <xarick> Owner 3, that's actually LuDiAIAfterFix
17:05:17 <kuhnovic> Do you know which specific ship triggered the assert?
17:06:32 <xarick> ship wants to go to where it is
17:07:49 <xarick> order is not LEAVINGSTATION this time
17:08:17 <xarick> it's really looking for a docking tile, which is where the ship is atm
17:10:30 <xarick> let's try... some detached heads
17:12:54 *** Leopold has quit IRC (Remote host closed the connection)
17:15:36 <xarick> commit prior to this doesn't crash
17:18:58 *** Leopold has joined #openttd
17:19:51 <xarick> meanwhile I forgot what I was doing
17:19:56 <xarick> something about time...
17:24:37 <kuhnovic> Clearly I messed something up hehe
17:27:17 <michi_cc> kuhnovic: Please do your work yourself π (Hint: there's a PR ready)
17:30:43 *** Leopold has quit IRC (Remote host closed the connection)
17:33:09 <michi_cc> Hmm, also, we have a new dev, wasn't there something about RC2... ? π
17:35:09 *** HerzogDeXtEr has joined #openttd
17:35:28 <kuhnovic> Backport requested label?
17:35:51 <michi_cc> No, doing the release π
17:36:55 <kuhnovic> I hope you're not expecting me to do that π
17:38:44 <michi_cc> It's actually easier than you think. We made 2TT do it, too.
17:39:41 <michi_cc> If you look at the channel pins, there's a message from me that starts some post. Most if should be in the link 2TT posted I think.
17:40:06 <talltyler> I wrote the guide based on your messages here, with additional info I asked about when writing the guide π
17:40:40 <talltyler> Thanks again for that π
17:40:40 <michi_cc> kuhnovic: So yes, maybe not today or tomorrow, but you *will* release RC2 π
17:41:25 <kuhnovic> Haha alright, baptism of fire π₯
17:42:02 <michi_cc> Backport PR can be updated incrementally, so you may start at any time π
17:42:24 <michi_cc> talltyler: That's dangerous in case someone presses buttons too soon again π
17:42:59 <talltyler> I can un-approve if you want π
17:46:53 <xarick> AIDate.GetCurrentDate() = 361
17:47:04 <xarick> it starts at 361? very odd number
17:47:35 <xarick> I expected 0, but hmm...
17:48:12 <talltyler> Economy year 1, day 1
17:49:19 <talltyler> This is the first day of the economy time system, unlike calendar date which is zero-indexed but generally starts at some date that trains exist π
17:49:42 <talltyler> Economy years have 360 days, because economy months all have 30 days
17:52:38 <xarick> anyway, I find my AI very shy on adding ships
17:53:11 <xarick> still figuring out what's wrong
17:59:47 *** nielsm has quit IRC (Ping timeout: 480 seconds)
18:01:32 <xarick> I suspect something with vehicle age...
18:06:49 <xarick> I only amass vehicles after vehicle age is > 730
18:07:10 <xarick> AIVehicle.GetAge returns what now?
18:09:30 <xarick> I also use AIVehicle.GetAgeLeft() <= 365
18:09:52 <xarick> to detect old veihcles
18:10:50 <xarick> Is there a function to convert this time to the other time?
18:24:45 <kuhnovic> XAARRIIIIIIIICK! π
18:26:03 <xarick> AIVehicle.GetProfitLastYear what does this return in relation to AIVehicle.GetAge?
18:41:43 <xertov> someone explain wtf means prospect this industry
18:43:42 <brickblock19280> try to build specifically
18:44:03 <brickblock19280> funding is guaranteed prospecting is not
18:48:39 <michi_cc> Literally speaking prospecting is the act of searching for natural resources (like .e.g oil), which you might or might not find in a specific place.
19:00:01 <xarick> Age is in days, not years
19:00:41 <talltyler> Be warned that engine age is in calendar time, not economy time, so if you freeze the calendar they wonβt get older.
19:02:08 <_glx_> AITimeMode only affects AIDate, all other functions use economy or calendar depending on the function
19:04:03 <talltyler> If needed that could probably be exposed to scripts
19:04:31 <xarick> 10980, this engine has 25 years of life... 10980/25 = 439 days, this looks weir
19:30:35 <truebrain> owh, and kuhnovic, you are no longer allowed to play the game. So please stop this nonsense π π π
19:31:11 <truebrain> context for IRC π
19:31:44 <kuhnovic> I'm debugging and comparing against 13.1. Can hardly call that "playing" π
19:32:37 * kuhnovic Turns off steam hehe
19:33:28 <truebrain> what was I going to do tonight .. hmm .. I had a plan earlier. I kinda forgot.
19:34:08 <truebrain> how on earth did you manage to do that π
19:34:56 <truebrain> ah, your title was self-referencing π On squashing, GitHub does that for you; no need to do those plebs tasks yourself π
19:37:56 <kuhnovic> I was wondering how people did that haha. I looked at the latest issue / PR and did that number + 1. I was overthinking it π
19:40:40 <peter1138> Hmm, I've got (an AI played) savegame where which keeps cycling between slow and fast...
19:41:53 <truebrain> slow and fast what?
19:42:11 <peter1138> Hmm, seems to be road vehicle ticks that surge.
19:42:55 <peter1138> Okay that's annoying, when you pause to look at a chart the chart disappears too because it's paused...
19:43:21 <peter1138> Doesn't go far back enough to show a pattern.
19:43:34 <truebrain> make a "zoom out" button π
19:44:48 <kuhnovic> Wow. Conditional orders for ships is starting to compete for "bane of my existence" when it comes to working on the ship pathfinder.
19:45:34 <peter1138> Hmm, surprised they have any bearing on the pathfinder.
19:45:48 <truebrain> ugh, uploading crashdumps of custom binaries .. ugh
19:47:26 <peter1138> Hmm, where is the framerate graph period determined... (the one that likes to jump around a bit.)
19:47:35 <kuhnovic> They seem to trigger annoying edge cases. For example, a ship is at a dock, but the next conditional order says "go back to that dock if 0% loaded".
19:48:00 <xarick> it was master, not custom
19:48:09 <truebrain> kuhnovic: why does a PF care? π
19:48:31 <peter1138> This might be the "special case" I removed...
19:48:54 <truebrain> well, that wouldn't be a terrible thing, at least means we can document it correctly π
19:52:26 <kuhnovic> peter1138: I thought so too, so I put that code back, and it doesn't seem to be the case.
19:54:31 <peter1138> Maybe this condition should be detected before calling the pathfinder.
19:54:37 <xarick> oh, because it wasn't built by your servers, it's a custom build...
19:55:43 <peter1138> Speculating, no idea.
19:56:04 <kuhnovic> xarick: I have created an isolated savegame in 13.1 that reproduces the problem
19:56:07 <peter1138> Well, not quite the plan...
19:57:32 <kuhnovic> And I'm quite sure that my recent changes deep in yapf_base have something to do with it. Even though I still think those were good changes. But yeah, side effects...
19:58:03 <xarick> it was crashing in 13.1?
20:00:46 <kuhnovic> No it was not, but I used 13.1 to compare what happens in the same situation.
20:01:56 <truebrain> testing runner changes by just making the PR ... let's see how well this goes π
20:02:01 <kuhnovic> 13.1 has the old ship pathfinder, and doesn't have my recent yapf destination check fix. This means the ship never consideres the current tile to be a destination tile, and will therefore always look for a differnt tile.
20:02:25 <truebrain> one obvious failure .. I used the Windows commands .. lol
20:03:25 <michi_cc> I just wanted to comment, but you were faster π
20:03:40 <truebrain> I don't need nobody to tell me I am an idiot π
20:06:00 <peter1138> That was too much effort just to make this visible :p
20:06:24 <truebrain> can you make it the default now? π
20:06:32 <truebrain> that seems so much more useful π
20:07:16 <xarick> okay, I seem to understand this time stuff
20:07:46 <xarick> but a day takes so long with 10080
20:08:37 <truebrain> lol, we are building x64 for the CI on Mac .. I was wondering why it wasn't all that fast π
20:10:34 <xarick> talltyler: I think I know what's the problem. There is no conversion from calendar to economy for at least these two functions GetAge, GetAgeLeft
20:11:00 <truebrain> let's see which is faster .. 4m30s for x64
20:11:18 <xertov> Select the action to take at this depot = select the action you want to take right ?
20:11:20 <xarick> version 13 versus version 14
20:12:35 <xarick> that is actually not version 13 expects
20:13:52 <_glx_> and this PR doesn't touch API but I think it should
20:14:01 *** Leopold has joined #openttd
20:14:41 <xarick> I wonder what GetProfitLastYear return, let's see
20:15:38 <peter1138> Hmm, might just be all the RVs queueing.
20:15:50 <truebrain> they all queue at the same time?
20:16:12 <truebrain> okay, x64 and arm64 are both equally fast on the CI. Both faster than all other targets, lol
20:16:36 <truebrain> but also no change with macos-latest
20:16:39 <truebrain> so that is good-ish
20:17:31 <truebrain> that is one hell of a traffic jam π
20:17:38 <andythenorth> peter1138: I am ignoring the RVs and looking at the flat docks
20:17:52 <_glx_> check the string name xertov (be it's billion)
20:18:06 <xertov> STR_CURRENCY_SHORT_GIGA
20:18:26 <xertov> thanks , i will check string names from now on
20:18:31 <truebrain> mostly our strings are more telling than the english text π
20:18:51 <xertov> however i think i ll keep this bn
20:19:16 <xertov> because there is not short for of διΟΡκαΟΞΏΞΌΞΌΟΟΞΉΞΏ that i know of
20:19:41 <peter1138> Nobody wants British nudes.
20:19:56 <peter1138> Especially not British OpenTTD developer nudes.
20:19:57 <xertov> nevermind we got διΟ
20:20:08 <xarick> for (local go = false; go != true; go = !go) { very smart coding
20:20:49 <truebrain> peter1138: maybe LordAro is making a lot of money with his OnlyFans .. you don't know! DONT JUDGE
20:21:52 <truebrain> lol, Rb wrote in a comment " for me it would be mostly generating failed builds until something sticks. I doubt that's the way to go" .. and I ended up doing exactly that with #12237 .. lalalala π The irony sometimes π
20:22:08 <truebrain> like any of us actually understands GitHub workflows π
20:22:09 <xertov> british is the first nationality that came to mind with a B . in handsight i should had said Brazilian
20:22:30 <truebrain> xertov: where can I sign up for this .... site? book? video?
20:25:11 <xertov> idk do ur research i guess XD
20:27:17 <frosch123> are peoples with B more likely to have no government?
20:27:25 <frosch123> brits, belgians, ...
20:28:22 <frosch123> ah, right, you joined the belgians
20:28:52 <frosch123> at least you do no longer have to worry about the german wheat tourists
20:28:52 <peter1138> We have a government. They're too busy lining their own pockets though.
20:35:01 <andythenorth> should I play OpenTTD?
20:35:16 <truebrain> Are you a developer?
20:35:37 <andythenorth> never wanted the responsibility
20:37:33 <frosch123> andy even left when we force-joined him
20:41:57 *** Leopold has quit IRC (Remote host closed the connection)
20:44:23 *** sherlockholmes1904 has joined #openttd
20:44:23 <sherlockholmes1904> andythenorth: Then who is developer?
20:44:50 <andythenorth> frosch123: oh yes, I used to have nml approval rights
20:45:15 <frosch123> sherlockholmes1904: ask watson
20:47:54 <xarick> for (local go = 0; go <= 1; go++) {
20:47:54 <xarick> local calendar_mode = AITimeMode(go.tobool());
20:50:14 <andythenorth> if I want the game to progress at half the normal speed, I set 24 minutes?
20:50:32 <peter1138> These RVs need to get Waze or something.
20:52:23 <peter1138> Is there no road occupancy penalty for RVs...
20:53:34 <xarick> calendar mode doesn't do anything, or I don't understand
20:55:13 <xertov> ee i still dont get this
20:55:26 <xertov> maybe it means all cargo types with no ratign ?
21:02:05 <xarick> should test v->tile == src_tile, which probably is the case
21:02:52 <xarick> or just revert peter's recent edge case removal?
21:03:54 <kuhnovic> xarick: I did that locally, but that's actually not the cause. It's our good friend OT_LEAVESTATION again.
21:04:54 <kuhnovic> And on one hand I'm inclined to accept your fix, as it does make issue go away. However, I also feel that we are fixing symptoms and not causes.
21:06:21 <peter1138> I've got a different pattern now, nice.
21:06:24 <xarick> eh I don't think the fix i have there solves it....
21:07:16 <xarick> or does it? hmm must test, though I was a bit busy with the Time feature breaking my AI
21:08:16 <peter1138> Hmm, quantum vehicle queueing.
21:09:05 <peter1138> That lorry is like YOLO
21:09:52 <xarick> it's to prevents deadlocks if they wait too much, unelegant though
21:11:29 <truebrain> "I don't believe in physics" π
21:15:05 <xarick> confirmed, 12156 doesn't fix this new crash
21:17:06 <xarick> now testing with 12192 reverted
21:18:42 <xarick> 12156 + 12192 reverted
21:19:59 <xarick> oh, right... this is so confusing
21:20:34 <xarick> tile is the tile the vehicle enters, v->tile is the tile the ship is at, of course reverting 12192 wouldn't solve anything
21:21:26 <xarick> v->dest_tile is matching v->tile, not tile
21:21:38 <peter1138> Who knows where they are going π
21:22:30 <belajalilija> you're redoing rv ai now?
21:23:01 <belajalilija> it would be cool if trains would turn around to find a depot
21:24:00 <belajalilija> for ex in a case like this
21:24:27 <belajalilija> a train could go to a theoretical station, turn around, enter the depot
21:25:58 <talltyler> Assuming youβre playing nightly. RC1 is quite broken. π
21:29:56 <_glx_> xarick: I told you, AITimeMode affects only AIDate class
21:30:35 <_glx_> AIEngine and AIVehicle uses the appropriate one
21:30:37 <andythenorth> I built recent master
21:30:55 <_glx_> though in some case they might use the wrong one
21:31:15 *** Leopold__ has joined #openttd
21:39:16 *** Leopold__ has quit IRC (Remote host closed the connection)
21:40:29 *** Leopold has joined #openttd
21:42:34 <xarick> I really wish brazillians would stop translating everything on youtube!
21:42:55 <xarick> even voice overs Linus Tech Tips, why do you do that!!
21:48:05 <xarick> We understand english very well, unlike Brazil
21:48:23 <xarick> we despise voice overs when the source is english
21:48:54 <xarick> but google thinks Portugal and Brazil are the same, so we're fed with that content by default...
21:52:20 <Rubidium> it's not like American, British and Australian English are the same, yet Google only "knows" one of them...
21:57:59 *** Leopold has quit IRC (Remote host closed the connection)
21:59:50 *** Leopold_ has joined #openttd
22:07:12 *** SigHunter has joined #openttd
22:11:59 *** SigHunter has joined #openttd
22:12:22 *** nielsm has quit IRC (Ping timeout: 480 seconds)
22:13:08 *** Leopold_ has quit IRC (Remote host closed the connection)
22:13:31 *** keikoz has quit IRC (Ping timeout: 480 seconds)
22:14:18 *** Leopold_ has joined #openttd
22:16:52 <xarick> Im receiving emails about 12136, interesting
22:20:54 <talltyler> I should stop trying to do critical thinking tasks on Friday evening π
22:21:47 <xarick> version 13 scripts, and before need vehicle age related stuff to be what it used to be before
22:22:19 <xarick> version 14 however... not sure
22:23:25 <_zephyris> peter1138: No, no, that's totally intentional!
22:25:57 *** SigHunter has joined #openttd
22:26:01 <_zephyris> Hmm, I might have already fixed it.
22:27:08 <xarick> I use GetAgeLeft to know whether my vehicles need replacement, due to their reliability decrease being faster in the final year of their lifetime. What timer is reliability counters based on?
22:30:26 <peter1138> Also the background gradient of the company president image is not company coloured.
22:30:39 <xarick> and I use GetAge on vehicles to check if the route they're servicing is profitable. Maybe I have to change this and store the date of construction of the route instead of looking for the oldest vehicle in the route
22:31:00 <_zephyris> peter1138: Oh yeah...
22:31:42 <truebrain> talltyler: I just stopped with thinking in general. Easier that way.
22:36:38 <peter1138> Thinking? What's that?
22:51:21 *** Leopold_ has quit IRC (Remote host closed the connection)
22:52:37 *** Leopold has joined #openttd
22:58:06 *** SigHunter has joined #openttd
23:03:54 <_glx_> xarick: age stuff should use economy not calendar
23:05:17 <_glx_> but you can't do anything script related without #12142
23:09:22 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
23:18:30 <xarick> I'm afraid I won't get my AI ready for 14 release
23:22:20 <xarick> this code is messy enough
23:23:55 <peter1139> That's okay, you don't have to wait until 15 just because you missed it :)
23:26:30 <xarick> is there an easy way to convert a date to another?
23:32:39 <_glx_> no, they are not meant to be converted
23:35:40 <xarick> going to need a m_constructionDate for all 4 route types
23:39:52 <xarick> and then... look over 96 + 28 results for m_lastVehicleRemoved + m_lastVehicleAdded to update to using m_constructionDate where appropriate, 17 for GetAge + GetAgeLeft + 72 results for GetCurrentDate + 18 for GetProfitLastYear + GetProfitThisYear... it's basically the entire AI affected
23:42:01 <xarick> oh well, cyas goodnight
23:49:47 <xertov> xarick: thats hilarious
23:56:12 <matusguy> xarick: as a Portuguese person i hate that. I purposely put my youtube in english for such reason
continue to next day β΅
