IRC logs for #openttd on OFTC at 2020-01-15
⏴ go to previous day
00:00:05 <Eddi|zuHause> best docs EU West
00:42:33 <Samu> im running out of time... come on, build fast
00:45:12 <Samu> bah, whitespaces.. well tomorrow i'll fix them
00:53:36 *** Wormnest has joined #openttd
01:11:43 *** hythlodaeus has joined #openttd
01:25:50 *** supermop_Home has joined #openttd
01:26:02 <supermop_Home> peter1138 you have a server going?
02:19:22 *** Wormnest has joined #openttd
02:53:33 *** tokai|noir has joined #openttd
02:53:33 *** ChanServ sets mode: +v tokai|noir
04:58:32 *** OsteHovel has joined #openttd
07:31:33 *** Wormnest has joined #openttd
08:01:35 *** andythenorth has joined #openttd
08:04:35 <Pikka> was it something I said?
08:21:43 *** andythenorth has joined #openttd
08:22:07 <andythenorth> oh it's 1968 again in my game
08:24:16 <andythenorth> I used the built in daylength feature
08:26:52 <Pikka> maybe we should get rid of the "funny" description on that window
08:26:58 <Pikka> it seems to upset some people
08:27:54 <andythenorth> it inhibits sandbox play
08:35:14 <Pikka> oop, gotta go climb a wall!
08:55:33 *** Wormnest has joined #openttd
09:36:34 *** ChanServ sets mode: +v tokai
10:11:11 <peter1138> I... accidentally went to sleep with my client connected to my server ;(
10:19:15 <planetmaker> so... are now many little clients being spawned?
10:54:46 <peter1138> No but it's 2035 or something instead of late 90s :p
11:00:24 *** hythlodaeus has joined #openttd
12:16:29 *** Arveen2 has joined #openttd
12:36:43 *** WormnestAndroid has quit IRC
12:36:56 *** WormnestAndroid has joined #openttd
12:53:23 *** snail_UES_ has joined #openttd
13:19:22 *** andythenorth has joined #openttd
13:52:01 * andythenorth reading about transfer leg profits
14:23:52 <andythenorth> I should turn breakdowns on eh
14:37:48 <peter1138> But yes, breakdowns.
14:38:06 <peter1138> Also I'm using default ships in 2040s... I guess breakdowns would be bad.
14:39:04 <andythenorth> breakdowns mess with my buzz :P
15:12:44 *** supermop_work has joined #openttd
15:25:34 <Samu> oh snap, found NULL's on my code
15:28:57 <andythenorth> oops it's 1973 in my game
15:29:01 * andythenorth turns it back to 1971
15:31:38 <andythenorth> time has to advance I guess :)
15:31:50 <andythenorth> otherwise I'll never get any super OP trains
15:33:45 * andythenorth has a maths puzzle
15:34:04 <andythenorth> 2 trains, both carry 484t cargo
15:34:31 <andythenorth> one weighs 361t empty and 484t loaded
15:35:14 <andythenorth> the other weights 352t empty and 646t loaded
15:35:39 <andythenorth> oof I really needed to sleep
15:36:16 <andythenorth> anyway, something is wrong with the total weight of the 2nd train, but I had no sleep, and don't trust my eyes :P
15:36:27 <andythenorth> self-inflicted, stayed up too late playing openttd
15:49:43 <Samu> well, is it useful? #7937?
15:50:18 <DorpsGek_III_> [OpenTTD/OpenTTD] Eddi-z opened pull request #7938: Feature: Setting for a year that repeats forever https://git.io/JvU08
15:52:40 <Eddi|zuHause> that is the name of the branch :)
15:52:44 <andythenorth> I was wondering what happens if we repeat months or days
15:52:51 <andythenorth> I didn't read industry production code yet though
15:53:02 <Eddi|zuHause> oh, compile failed
15:53:31 <andythenorth> I think I'd want a groundhog threshold
15:53:36 <andythenorth> i.e. repeat n times
15:53:44 <Eddi|zuHause> i forgot a thing adding the setting, apparently
15:53:57 <Eddi|zuHause> andythenorth: just use the date cheat to move on
15:54:01 <andythenorth> so daylength is solved
15:54:13 <andythenorth> and I've solved cdist by turning town growth to minimal
15:54:36 <andythenorth> did we ever stop towns using magic bulldozer to delete things
15:54:55 <andythenorth> cos magic bulldozer solves most of sandbox mode
15:54:59 <Samu> ah, im not the only one that makes non descript PRs
15:55:10 <andythenorth> just need to solve town ratings, game is finished
15:55:19 <Eddi|zuHause> i need to go out shopping, i was like a minute too late yesterday
15:57:48 <DorpsGek_III_> [OpenTTD/OpenTTD] Eddi-z updated pull request #7938: Feature: Setting for a year that repeats forever https://git.io/JvU08
15:58:00 <Eddi|zuHause> (still untested)
16:02:46 <Samu> it still builds, but i get so many errors along the way
16:03:19 *** Wormnest has joined #openttd
16:12:25 <DorpsGek_III_> [OpenTTD/OpenTTD] Eddi-z updated pull request #7938: Feature: Setting for a year that repeats forever https://git.io/JvU08
16:12:36 <Eddi|zuHause> (3rd time is the charm?)
16:12:50 <andythenorth> shops Eddi|zuHause :P
16:12:55 <andythenorth> don't want you going hungry
16:13:23 <Eddi|zuHause> i wouldn't be going hungry, but i'm out of sugary drinks
16:14:01 <Samu> music video, no actual video though
16:14:43 <FLHerne> Samu: I like the idea of permanent rivers
16:16:22 <Samu> if only rivers were actually useful
16:16:39 <Eddi|zuHause> the next opportunity to go shopping while on the way to somewhere would be in 2 days
16:16:49 <Samu> my lock friendly rivers patch was rejected
16:16:57 <Samu> would combine well with perma rivers
16:19:39 <Samu> cleaning up stuff that didn't belong to this patch
16:20:03 <Samu> you know me, I always pack everything in a single commit
16:21:49 <Samu> a year that repeats forever is a "good idea"
16:27:01 <Samu> i've been wanting to test AIs in a certain era
16:27:16 <Samu> 1975 era, hoping it would last forever 1975
16:27:25 <Samu> for example, i think that patch would come useful
16:32:04 <Samu> i think i forgot documenting object tiles built on canals build on rivers
16:37:24 <Samu> you know what? locks should be waypoints
16:37:45 <Samu> but that's for another day
16:46:03 <planetmaker> Why should they be waypoints?
16:46:21 <planetmaker> We have buoys as waypoints for water vessels
16:46:59 <planetmaker> when I go from A to B, it doesn't matter whether I use the Eastern or the Western lock in the channel. I use whichever opens first
16:52:49 <Samu> at m6 it stores canal owners
16:53:25 <Samu> should have included shipdepot in the square, but meh...
16:53:58 <Samu> at m8 it stores whether the canal was built on a river
16:56:59 <Samu> i still don't understand what yellow ~ is for
16:57:29 <Samu> ~ - bit is accessed, but does not really have a meaning (e.g. owner of clear land is always OWNER_NONE)
17:09:36 *** sla_ro|master has joined #openttd
17:14:45 *** WormnestAndroid has quit IRC
17:15:08 *** WormnestAndroid has joined #openttd
17:55:11 *** Progman has joined #openttd
17:56:52 *** supermop_work has joined #openttd
18:04:30 <andythenorth> daytime traffic is low in irc Samu
18:04:37 <andythenorth> mostly me and you
18:04:47 <andythenorth> patience grasshopper
18:15:07 <Samu> assert(depth < WATER_DEPTH_MAX); while testing #7924
18:19:43 <Eddi|zuHause> so. i actually was shopping!
18:33:22 <DorpsGek_III_> [OpenTTD/OpenTTD] SJang1 commented on issue #7830: Load font from openttd config file directory, not from working directory. https://git.io/JewMG
18:46:04 <Samu> suddenly "savegame upgrade" pops everywhere
18:46:29 <DorpsGek_III_> [OpenTTD/OpenTTD] nielsmh merged pull request #7791: GS method to control engine availability for a specific company https://git.io/JeREi
19:05:07 *** frosch123 has joined #openttd
19:06:54 *** HerzogDeXtEr has joined #openttd
19:09:01 *** tokai|noir has joined #openttd
19:09:01 *** ChanServ sets mode: +v tokai|noir
19:30:16 *** Wormnest has joined #openttd
19:38:05 *** WormnestAndroid has quit IRC
19:38:42 *** WormnestAndroid has joined #openttd
19:53:10 <DorpsGek_III_> [OpenTTD/OpenTTD] frosch123 commented on pull request #7938: Feature: Setting for a year that repeats forever https://git.io/JvUws
19:55:22 <TrueBrain> I am heavily disappointed only in a comment "groundhog" is mentioned. I vote to rename the variable to GROUNDHOG_YEAR
19:57:16 <frosch123> i never saw that movie
19:57:36 <TrueBrain> I .... am not sure ... how to deal with this
20:00:24 <frosch123> hmm, the shading stuff did not work with palette animation, did it?
20:01:19 <frosch123> oh, newgrf water. does it work the same as rivers and canals?
20:02:27 <nielsm> no, I figured doing it with something callback-based would be bad for something that exists in as large quantities as flat water
20:02:37 <nielsm> and it may as well just be a static table
20:02:53 <nielsm> but there is a new variable for the river/canals callbacks to check depth of the current tile
20:03:09 <frosch123> he, at some point we wanted to do newlandscape :p
20:03:54 <nielsm> that's RCT-style landscape right?
20:04:13 <frosch123> no, callback based ground tiles
20:04:27 <frosch123> you know, more than just water tiles :)
20:05:01 *** crazystacy has joined #openttd
20:05:08 <frosch123> it was about adding random bits, doodads, fences, hedges, ...
20:05:30 <crazystacy> i always used to full load everything, and specifically put "unload". but now i don't set anything, just goto A, goto B (let's say coal). but i'm not sure what is most efficient
20:05:55 <crazystacy> let's say it never manages to get a full load, then it's just going back and forth. or if the journey is extremely long. if i load 20% and travel for ages, that can't be good
20:06:04 <crazystacy> i was wondering if anyone had some maths on it
20:06:07 <frosch123> you almost never want to explicitly set "unload"
20:06:18 <crazystacy> i used to think so
20:06:25 <crazystacy> it saved my finger a lot of clickin
20:06:47 <frosch123> the time waiting at the station is included in the delivery speed
20:07:12 <frosch123> so you want the wait-for-full-load not be a significant percentage of the travel time
20:07:43 <frosch123> also there is a intermediate between "full load" and "load any". you can timetable the loading time to "5 days" or similar
20:07:57 <frosch123> (but don't timetable travel times, only loading times)
20:08:48 <crazystacy> i can time table loading? that's ice
20:08:53 <crazystacy> i didn't get as far as timetables yet
20:09:05 <crazystacy> what happens if you timetable travel times? it gives up and skips?
20:09:14 <frosch123> well, it's said to be the only useful application of timetables
20:09:41 <frosch123> timetabling travel times causes many issues. noone admits to play with that :)
20:13:02 * _dp_ never knew timetables had a useful application
20:13:35 <_dp_> well, I guess, I don't considered partial load to be a useful application either :p
20:15:15 <frosch123> tgp does not know about max slopes
20:15:27 <nielsm> yeah, I know it needs smoothing
20:17:56 <nielsm> except that deepwater does not require smooth gradients, so it's fine here
20:19:41 <frosch123> does it rerandomise the depth when oilrigs shut down?
20:21:08 *** andythenorth has joined #openttd
20:21:18 <nielsm> does a circular tile search for nearby water tiles and grabs a depth from that
20:22:09 <frosch123> i was thinking of the coal mine disaster
20:23:12 <andythenorth> we flooding coal mines?
20:25:59 <frosch123> some firs industries could leave river/lake tiles behind after they close down
20:26:10 <frosch123> except firs industries do not close down, or something :p
20:26:25 <Samu> wow, how do u get graphics working there?
20:27:26 <Samu> all i get is a ? for every water tile
20:27:53 <nielsm> you need a newgrf with water depth sprites
20:27:54 *** innocenat_ has joined #openttd
20:33:02 <andythenorth> is someone setting up an MP game then? o_O
20:33:12 <andythenorth> so we can find out what's broken in 1.10
20:36:32 <andythenorth> what does clearing water even mean with deep depths? :P
20:36:44 <andythenorth> lots of long discussion there about a possible non thing
20:37:11 <crazystacy> will water depth be added to the game?
20:37:54 <frosch123> maybe that should be a firs industry, producing plastic
20:38:08 <andythenorth> is that the dutch guy?
20:38:27 <nielsm> that's without basecosts mod, without inflation, and at low construction costs :)
20:38:27 <andythenorth> nielsm: and if I canalise it?
20:39:00 <nielsm> that still costs over a million per tile :)
20:39:15 <frosch123> andythenorth: it's some marketing sketch from some 2017 startup. no idea whether that startup still exists
20:39:17 <crazystacy> now they won't grief my boats anymore :D
20:39:31 <nielsm> canals have a max depth of 3 (right now) and converting a deeper tile to canal will require demolishing the water and rebuilding it as canal
20:39:55 <crazystacy> well it's a much nicer startup than Juicero that's for sure
20:40:26 <crazystacy> so there will be some sort of canal tanker? so that i can ferry my oil tanker oil up the river
20:41:15 <nielsm> right now there are no restrictions on how deep or shallow water any boat can sail on
20:41:32 <supermop_work> BUT MY GALLEY WILL SINK
20:41:45 <crazystacy> is it "sail" if it's an oil tanker? :P
20:41:45 <nielsm> all ships can sail on canals and rivers, as long as there are locks built for every elevation change
20:42:02 <nielsm> ships sail even when they have combustion engines :P
20:42:17 <crazystacy> how much sail would a sailboat sail if a sailboat didn't have sails?
20:42:52 <nielsm> also wow debug build performance is terrible with 32bpp sprites
20:43:05 <nielsm> but only when the game is unpaused
20:43:32 <crazystacy> i'm trying to manually call NetworkClientConnectGame from the code, but it crashes
20:43:39 <crazystacy> and sometimes it does work O_o
20:44:32 <crazystacy> by manually i mean directly as opposed to through the existing methods like console or gui
20:45:12 <crazystacy> - and the debug output is not meaningful :/
20:45:51 <andythenorth> nielsm: full animation is over
20:45:58 <andythenorth> we should possibly remove it?
20:46:10 <nielsm> choices: build this bridge https://0x0.st/zhuQ.jpg - or build around the bay - or spend 2 or 4 million on making a series of shorter bridges spanning
20:46:24 <andythenorth> canal cheat of course
20:46:26 <nielsm> (making the islands costs that much)
20:46:30 <andythenorth> build canals, demolish the inner part
20:46:57 <nielsm> building canals over the deep water still costs the water clearing cost
20:47:09 <nielsm> and demolishing deep canals is still expensive
20:47:32 <crazystacy> clear water = purify the ocean at that exact spot
20:48:42 <crazystacy> that looks like a meme waiting to happen
20:48:52 <crazystacy> it just needs a half-transparent face of a crying man in the corner
20:49:17 <andythenorth> reddit will love it
20:50:08 <frosch123> andythenorth: it's only osx that fails with animation, isn't it?
20:50:44 <andythenorth> there are other reports of slow with no final cause
20:51:03 <nielsm> 32bpp sprites with full animation is also stupid slow on win32 in debug builds, I just discovered, but that's not an issue in itself
20:51:59 <andythenorth> I tried to find the docs for 'not a debug build'
20:52:02 <andythenorth> but I couldn't figure it out
20:52:12 <andythenorth> I even read the build farm stuff on GH, but no light shone
20:52:22 <nielsm> -O2 or -O3 in the CXXFLAGS
20:52:25 <andythenorth> my local compile is ~unusable
20:53:30 <andythenorth> so are these configure flags or make flags?
20:53:34 * andythenorth is clueless sorry
20:54:16 <milek7_> --enable-debug[=LVL] enable debug-mode (LVL=[0123], 0 is release)
20:56:35 <nielsm> is "water_clearing_cost_power" a good setting name?
20:56:56 <nielsm> depth raised to the n'th power as multiplier for clearing cost
20:57:33 <nielsm> n=0, same clearing cost for any depth, n=1 clearing cost is linear dependency on depth, n=2 clearing cost is squared dependency on depth
20:58:09 <Samu> oh, even toyland water is dark too
20:58:17 <frosch123> water_clearing_cost_exponent
20:58:25 <crazystacy> what difference does --enable-debug=[level] do in configure?
20:58:39 <crazystacy> i set it to 1 and 2 and didn't notice any difference. 3 made my laptop lag to death during building
20:58:47 <crazystacy> i mean it's not the same as openttd -d 1/2/3 right?
20:58:49 <milek7_> water_clearing_cost_exponent?
20:59:02 <crazystacy> -d [lvl] is for debug logging?
20:59:07 <frosch123> crazystacy: >0 enables debug symbols, higher values disable optimisations
20:59:39 <andythenorth> so yeah, the official binary runs my savegame stably at 1.02x
20:59:59 <andythenorth> my local build fluctuates same savegame, same date, same map position 0.9-0.95x
21:00:15 <andythenorth> the local build falls on it's knees further and faster as well
21:00:21 <andythenorth> when I go to a busy area of map
21:00:46 <andythenorth> I gave it ./configure --enable-debug=0
21:01:09 <andythenorth> so something in my libs, or compiler or something is fucked
21:01:11 <DorpsGek_III_> [OpenTTD/OpenTTD] Phx01 opened issue #7939: Feature Request: Moving/Relocating/Offsetting (Certain) Industries https://git.io/JvUr7
21:01:16 <andythenorth> this will be fun to diagnose :(
21:01:51 <frosch123> andythenorth: i think stable builds add --disable-assert
21:02:20 <nielsm> what setting value names? "original/expensive/very much expensiver"?
21:02:36 <nielsm> or just boring "original/linear/square"?
21:03:08 <crazystacy> "troll/hardmode/avoid ocean"
21:03:15 <nielsm> or maybe "original"/"cost*depth"/"cost*depth*depth"
21:03:19 <frosch123> Influence of water depths on clearing cost: None/linear/quadratic
21:05:04 <frosch123> every game should contain some math. otherwise people are afraid because their moms forbid them to use ctrl+alt+c
21:05:45 <andythenorth> pikka suggested rewording the cheat menu
21:05:55 <andythenorth> I suggest 'sandbox mode options'
21:05:59 <andythenorth> or just 'have fun!'
21:06:24 <andythenorth> I am very uninterested in any of the original game's restrictions, after this long playing it :P
21:06:44 <frosch123> crazystacy: there are some people who want a "feature" that is already available via the cheat menu. but they refuse to use the cheat menu and instead request that it should be regular game behaviour
21:07:04 <crazystacy> what feature is that?
21:07:10 <frosch123> somehow they are afraid to use "cheats"
21:07:22 <crazystacy> call it "handicap"
21:07:41 <frosch123> recently it was about fast aircraft having extra crash chance on short runways
21:07:54 <crazystacy> oh i saw that. my whole fleet crashed
21:08:14 <crazystacy> i had the smallest airport with that 1500 km/h plane
21:08:15 <_dp_> just move all cheats to settings :p
21:08:23 <andythenorth> I wanted full sandbox mode, but it turns out to be mostly magic bulldozer
21:08:36 <andythenorth> I wanted daylength, but it turns out to be 'reset the year every year'
21:08:48 <nielsm> magic bulldozer and neutered town councils
21:08:52 <andythenorth> _dp_: any idea if magic bulldozer got fixed?
21:09:08 <andythenorth> it used to be super dangerous to leave on as towns would just delete stuff
21:09:10 <_dp_> andythenorth, you mean for towns? I didn't even know it was broken xD
21:09:19 <frosch123> it was fixed in 0.6
21:09:29 <andythenorth> so towns no longer just destroy arbitrary things? o_O
21:09:39 <andythenorth> I have been turning it off in fear
21:09:44 <andythenorth> and then on again
21:09:48 <crazystacy> tbh cities should be decimated
21:09:55 <frosch123> the fix was part of one of my first patches, a side effect of terraform under infrastructure
21:10:00 <crazystacy> make way for the coal
21:10:17 <andythenorth> if we can just nerf local authority, sandbox mode is done
21:10:27 <frosch123> andythenorth: so the issue has been fixed for longer than it ever was a thing :)
21:10:28 <crazystacy> i want a grf which changes passengers to meatbags and disables town councils
21:10:57 <frosch123> issue for 4 years, fixed for 12 years :)
21:11:04 <andythenorth> this is how long rumours persist :P
21:11:07 <andythenorth> bad half-life :P
21:11:29 <andythenorth> well ./configure --enable-debug=0 --disable-assert is running at 0.88x
21:11:47 <milek7_> "Magic bulldozer has some side effects though. As not only you as player gain higher bulldozer rights, also the towns get the ability to remove industries. "
21:12:12 <frosch123> just as outdated :)
21:12:38 <frosch123> stuff got removed accidentially by terraforming nearby
21:12:50 <frosch123> but terraforming no longer does that
21:13:56 <nielsm> uh do we have a simple "int power" function?
21:15:01 <andythenorth> wiki is fixed then?
21:16:22 <TrueBrain> any suggestions / additions / changes
21:16:31 <TrueBrain> (this would be publishes just after a new master release is created)
21:16:36 <andythenorth> TrueBrain: I love it!
21:16:48 <frosch123> TrueBrain: earlier it only said "hello" :)
21:16:57 <TrueBrain> I know :P That is called developing :)
21:17:25 <TrueBrain> it will do "poef" if I roll this out :P
21:17:25 <andythenorth> or to ask a better question...
21:17:36 <frosch123> we should add a proper front page though
21:17:54 <LordAro> TrueBrain: looks good
21:18:03 <Samu> this water depth stuff is so pretty :)
21:18:09 <LordAro> though it would be nice if the doxygen for the game itself was there
21:18:12 <LordAro> as useless as it may be
21:18:17 <TrueBrain> it really really is useless
21:18:23 <TrueBrain> so if you promise me to make it less useless
21:18:25 <TrueBrain> I am willing to add it :P
21:18:47 <TrueBrain> frosch123: yeah, I was thinking the same .. just not sure how yet
21:19:09 <frosch123> by adding a @mainpage in the source?
21:19:17 <TrueBrain> yeah ... I wish it was that easy
21:19:29 <TrueBrain> but what I was thinking: add a dummy page to the website, and curl it into here
21:19:39 <Samu> currently toying with pathfinder + water depth. I'm trying higher pf costs on deeper depths, but the opposite as well
21:19:46 <frosch123> are we talking about different things?
21:19:51 <frosch123> why does @mainpage not work?
21:19:54 <nielsm> frosch123: it looks like the MS stdlib implementation just calls the C pow() function with doubles
21:19:57 <Samu> lower costs on deeper depths vs higher costs on deeper depths
21:20:19 <TrueBrain> frosch123: we are talking about different things I guess; no clue what you mean with @mainpage in that case :)
21:20:59 <andythenorth> I just didn't fancy the inevitable PR review cycle, and 10 people with 10 ideas who didn't do anything themselves :)
21:21:29 <TrueBrain> ah, doxygen frontpage :)
21:21:54 <andythenorth> it's not unusable or anything
21:22:06 <TrueBrain> okay, you guys fix the source doxygen, I will add it to docs :)
21:22:18 <Samu> I can see water depth help "guide" the pathfinder to the correct places
21:23:16 <Samu> I envision it can visit more nodes which are closer to land
21:23:24 <Samu> than searching deep into the ocean
21:24:27 <milek7_> nitpick: it's xhtml served with text/html content-type
21:24:39 <frosch123> do you intentionally not use "nogo" and "noai" in the url, in favour of "xyz-scripting"?
21:24:46 <nielsm> Samu: water depth does not necessarily have anything to do with distance to land
21:24:54 <frosch123> because noone says "-scripting" on gs/ai context :)
21:25:20 <Samu> oh, right :( it's not instant :(
21:25:25 <Samu> it's an over time effect
21:25:44 <nielsm> yes and the current erosion might still be too fast
21:25:53 <TrueBrain> damn, source docs is HUGE :P
21:26:09 <TrueBrain> I used the lingo used there
21:26:33 <TrueBrain> and gs is "Game Scripting", so .... :P
21:26:50 <frosch123> well, "game script"
21:27:18 <TrueBrain> Game API and NoAI API ?
21:27:23 <TrueBrain> (as it is named in Doxygen titles)
21:27:43 <frosch123> the "todo list" is the best part of every doxygen output
21:28:03 <TrueBrain> or how do you refer to noai/nogo these days most often?
21:28:47 <milek7_> something looks off with truetype fonts in ottd
21:30:43 <frosch123> TrueBrain: "AI API" and "Game API" sounds fine
21:30:56 <nielsm> I prefer just "AI" and "GS"
21:30:56 <TrueBrain> (I have to fix the "index.html" part btw, but this needs an Lambda@Edge .. need to read up on that :P)
21:31:16 <frosch123> "ai" and "gs" would match the class prefix
21:31:22 <nielsm> that the features were called "noai" and "nogo" during development is a weird artifact I don't think is important for those urls
21:31:38 <TrueBrain> so ai-api/ and gs-api/ ?
21:31:41 <frosch123> and you do not have to worry about space/dash/underscore in front of api :)
21:32:53 <frosch123> nielsm: "OpenTTD NoAI API" is the title of the doxygen page :)
21:33:07 <TrueBrain> so fix the Doxygen :)
21:33:33 <TrueBrain> it also reads "Game API"
21:33:48 <frosch123> nielsm: it looks like gcc inlines pow when the exponent is known, otherwise it calls the c function
21:34:50 <frosch123> TrueBrain: urls are fine. i am not sure whether it needs the "docs" front page
21:35:03 <frosch123> the subpages could be linked directly from the website
21:35:22 <TrueBrain> yeah, but people tend to try the root domain too
21:35:29 <TrueBrain> so if it doesn't hurt, it is fine, I guess :)
21:35:36 <nielsm> frosch123: but I want a function that does not invoke floating point
21:35:36 <frosch123> well, the this basic layout is fine
21:35:55 <TrueBrain> but always to the subfolders :)
21:36:07 <TrueBrain> I will also redirect noai... to the right folder
21:36:28 <TrueBrain> if it helps, I can also attach "ai.docs.openttd.org" to it?
21:36:34 <TrueBrain> not sure that helps :D
21:36:59 <frosch123> nielsm: "pow(num,2)" is optimised to "mulsd", "pow(num 3)" results in a call
21:37:14 <TrueBrain> I hate it doesn't mention anywhere on what version it is based .. that is just bad
21:37:37 <frosch123> oh wait, it uses the xmm registers
21:39:06 <TrueBrain> I wonder if we can change Doxygen to add the version in the footer ..
21:39:13 <frosch123> TrueBrain: doxygen has a timestamp of the gernation
21:39:46 <frosch123> oh, it doesn't in our layout....
21:40:06 <frosch123> well, i am pretty sure i have seen that, so it should be possible
21:41:21 <frosch123> yeah, there is $datetime
21:41:55 <TrueBrain> means we need to make a footer html file, I guess
21:43:09 <TrueBrain> and if we do that, we might as well add the exact OpenTTD version it was build for in there
21:43:38 <frosch123> then you need an additional builtstep
21:43:59 <frosch123> datetime can be added with static source and doxygen
21:44:04 <frosch123> but doxygen does not know git hashes
21:44:35 <TrueBrain> but the docs-builder does
21:44:42 <TrueBrain> and doxygen can use env-variables :)
21:45:01 <TrueBrain> so I think we can simply do both
21:45:24 <TrueBrain> if you disagree that would be best, let me know :)
21:46:45 <Samu> testing extreme penalties that basically just tells ships to walk near coasts
21:48:07 <DorpsGek_III_> [OpenTTD/OpenTTD] nielsmh commented on issue #7939: Feature Request: Moving/Relocating/Offsetting (Certain) Industries https://git.io/JvUr7
21:48:36 <frosch123> TrueBrain: apparently doxygen has a PROJECT_NUMBER variable for that
21:49:10 <frosch123> though that is set in Doxyfile
21:49:31 <TrueBrain> which can accept env-variables :)
21:49:34 <TrueBrain> so that is fine for me
21:50:00 <frosch123> oh, did not know that
21:50:12 <frosch123> so, then it is probably the intended method to do that
21:51:49 <TrueBrain> lets test that out :)
21:52:26 <Samu> have a gamescript "move" the industry for u
21:52:37 <frosch123> why do we generate a tag file ...
21:53:25 <TrueBrain> bit ugly in the title, the project number .. but .. I guess it is exactly what we need
21:56:48 <Samu> moving an industry to the left 1 tile
21:56:55 <Samu> how doable is that in code? :o
21:58:29 <Samu> even if it's a water tower
21:58:29 <nielsm> you need to remove all the industry tiles, then place them again at the new location while doing all checks for being valid at the new location, and if those fail you need to make sure the industry is put back
21:58:40 <DorpsGek_III_> [OpenTTD/OpenTTD] TrueBrain opened pull request #7940: Add: [Doxygen] Add the current version of documentation in header https://git.io/JvUoh
21:58:42 <nielsm> yeah town industries are probably worse
21:58:53 <nielsm> since they need to replace houses to build
21:59:25 <Samu> i'd just build a new industry, remove the older one, and call it "moved"
22:00:33 <DorpsGek_III_> [OpenTTD/CompileFarm] TrueBrain opened pull request #41: Add: [release-docs] tell Doxygen about the version we are creating https://git.io/JvUKe
22:00:59 <Samu> i wish there was a way to pathfind from every tile to every tile and tell the success rate
22:02:47 <Samu> is it enough for 4kx4k maps?
22:05:20 <nielsm> (single-source shortest paths algorithm, from a single tile find the shortest path to any other reachable tile)
22:05:32 <nielsm> (and you'd have to repeat it for every source tile you're interested in)
22:05:53 <TrueBrain> okay, I have most of the code ready to put these "docs" in production .. but that is something for another day :)
22:07:44 <nielsm> well, A* is single-pair shortest path, it requires a fixed destination too
22:07:59 <Samu> i wonder wether pathfinding could be multi-threaded, because apparently it already waits for the result
22:08:01 <nielsm> it's been a while since I looked at this
22:08:01 <andythenorth> so who's writing the dev blog post then? o_O
22:11:52 *** ToBeFree has joined #openttd
22:12:22 *** ToBeFree is now known as Guest13852
22:18:42 <nielsm> multithreaded pathfinding is problematic
22:19:30 <nielsm> you'll end up doing duplicate work if you try putting multiple threads to work on the same vehicle, and splitting up relatively short jobs like that is usually not efficient either
22:20:04 <nielsm> you can't pathfind multiple land vehicles in parallel, since the path of one can affect the path another one must choose, and then you end up with non-determinism
22:20:57 <nielsm> if you accept locking yourself out of ever having ships depend on each others' paths (like ships avoiding other ships) then you could pathfind multiple ships in parallel
22:21:04 <andythenorth> so daylength 1.5 seems about right
22:21:07 <andythenorth> so 18 months / year
22:21:38 <Samu> just increase the number of months
22:22:14 <Samu> january will now have 90 days
22:22:25 <nielsm> you could probably also run pathfinding of all land vehicles in parallel with pathfinding of all ships, i.e. two parallel threads, sine they should never interfere
22:22:32 <andythenorth> switch from RL dates to 'elapsed years'
22:23:21 <Samu> i still don't know how to make threads
22:23:51 <nielsm> but doing that would still mean having to split vehicle ticks up in pathfinding and "doing other things", since vehicle ticks also involve some of the station loading/unloading logic iirc, and you still want the order vehicles process loading/unloading in to be deterministic
22:24:08 <DorpsGek_III_> [OpenTTD/OpenTTD] frosch123 approved pull request #7940: Add: [Doxygen] Add the current version of documentation in header https://git.io/JvUKc
22:25:05 <frosch123> i never know whether the "ENV=foo binary" syntax is proper shell
22:25:17 <frosch123> i always use "env ENV=foo binary"
22:26:23 <andythenorth> oof is it bedtime?
22:26:26 <andythenorth> today seems long :P
22:26:34 <Samu> with multiple docking points now
22:27:05 <nielsm> sleep sounds like a good plan
22:27:17 <Samu> rip multithreading opentttd
22:28:36 <Samu> anyway back to my depth/pathfind tests
22:29:00 <Samu> gonna start taking notes
22:31:00 *** andythenorth has left #openttd
22:39:18 <Samu> oh :( it's actually worse than i thought
22:40:03 <Samu> searches more nodes with the different costs values per depth
22:40:16 <Samu> i expected it to search less
22:41:20 <Samu> well, it walks more tiles, so maybe it's expected? must think
22:46:44 <Samu> 57435 <-> 31768 unchanged
22:46:44 <Samu> 83024 <-> 56269 cheaper costs on deeper water
22:46:44 <Samu> 63820 <-> 38350 cheaper costs on shallower water
22:47:11 <Samu> going from A to B, then B to A, the number of search nodes
22:48:20 <Samu> I am disappointed with myself, I honestly expected it to search with less nodes
22:49:15 <Samu> especially disappointed with shallower water results
22:59:38 <DorpsGek_III_> [OpenTTD/OpenTTD] Eddi-z commented on pull request #7938: Feature: Setting for a year that repeats forever https://git.io/JvU6l
23:07:20 <TrueBrain> frosch123: they both work; with the letter, an extra process is started
23:07:27 <TrueBrain> not sure if there is much more difference besides that
23:07:41 <DorpsGek_III_> [OpenTTD/OpenTTD] TrueBrain merged pull request #7940: Add: [Doxygen] Add the current version of documentation in header https://git.io/JvUoh
23:08:20 <DorpsGek_III_> [OpenTTD/CompileFarm] frosch123 approved pull request #41: Add: [release-docs] tell Doxygen about the version we are creating https://git.io/JvU60
23:08:37 <DorpsGek_III_> [OpenTTD/CompileFarm] TrueBrain merged pull request #41: Add: [release-docs] tell Doxygen about the version we are creating https://git.io/JvUKe
23:08:38 <DorpsGek_III_> - Add: [release-docs] tell Doxygen about the version we are creating (#41) (by TrueBrain)
23:32:48 <DorpsGek_III_> [OpenTTD/OpenTTD] esigra opened issue #7941: train pathfinder acts as if station exits had built-in signals https://git.io/JvU6p
continue to next day ⏵
![https://i.imgur.com/vv4TTg0.png[/img]](https://i.imgur.com/vv4TTg0.png%5B/img%5D){kind=link}
