IRC logs for #openttd on OFTC at 2019-03-28
⏴ go to previous day
00:00:32 <LordAro> Samu: maybe you should try to understand the code, rather than just adding asserts wherever you feel like
00:02:05 <_dp_> LordAro, well, adding asserts is part of understanding code sometimes :p
00:02:26 <_dp_> probably shouldn't spam about it though
00:05:04 <DorpsGek_II> [OpenTTD/OpenTTD] michicc approved pull request #7426: Fix: Incorrect display of industry production around tiles. https://git.io/fjTk8
00:09:15 <Samu> i have no idea how kdtree works, I just trusted it
00:09:46 <m3henry> LordAro: Easier said than done :v
00:10:04 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN merged pull request #7426: Fix: Incorrect display of industry production around tiles. https://git.io/fjUj9
00:10:33 <LordAro> and it's wrong anyway
00:12:11 <glx> hehe IIRC x - base converts to signed
00:13:02 <LordAro> glx: i was reading the type signatures of base & size, not x
00:14:50 <LordAro> Samu: i rather hope you didn't need a calculator for those
00:15:10 <ST2> he was only making sure :D
00:15:17 <peter1138> assert(40 + 11 == 51)
00:15:53 <DorpsGek_II> [OpenTTD/OpenTTD] michicc updated pull request #7428: Fix #7165: SmallMap::Erase(key) does not work correctly https://git.io/fjTUY
00:16:54 <michi_cc> stormcone: Have a look at my update for #7428?
00:19:01 <Samu> ForAll is giving the wrong town
00:22:20 <Samu> i really need to make sure
00:23:30 <Samu> question, is tile 958 on a 64x64 map inside or outside the map?
00:23:52 <Samu> when i zoom to it, it seems to be one in the border, but could be wrong
00:24:38 <stormcone> michi_cc: i had look at it, but as i said earlier i am not that familiar with c++, so if you say that is better in that way, i probably will agree with you :)
00:25:28 <michi_cc> stormcone: I just noticed that Contains() const was wrong and I couldn't be bothered to create another PR.
00:26:25 <glx> yes not counting the void tiles
00:27:44 <glx> but if you want to be sure I think you can draw tilehighlight on the tile
00:28:43 <Samu> just scrolled to 957 and 959
00:29:10 <stormcone> michi_cc: i don't mind it, should i modify something in the PR's title?
00:29:20 <glx> oh and there's landscape info tool
00:29:40 <glx> should show coords of the clicked tile
00:30:18 <michi_cc> No need, but maybe somebody like m3henry can take a quick look.
00:31:29 <glx> oh old mingw doesn't work
00:31:59 <Samu> the Y lenght of the airport is intercontinental, 11 tiles
00:35:38 <DorpsGek_II> [OpenTTD/OpenTTD] M3Henry commented on pull request #7428: Fix #7165: SmallMap::Erase(key) does not work correctly https://git.io/fjTIl
00:37:22 <Samu> forall is giving the wrong town
00:39:04 <glx> michi_cc: I redirected him to MSYS2, which works fine, MSYS is outdated and doesn't feel maintained
00:39:33 <LordAro> oh hey, it's that wiki page that i wrote
00:40:19 <LordAro> ooh, that still lists --without-freetype
00:40:22 <Samu> how do I report a bug like this?
00:40:40 <glx> I followed it when I installed MSYS2 :)
00:41:31 <glx> yeah wiki pages tends to be created, but not updated later ;)
00:45:21 <DorpsGek_II> [OpenTTD/OpenTTD] michicc commented on pull request #7428: Fix #7165: SmallMap::Erase(key) does not work correctly https://git.io/fjTI1
00:47:58 <michi_cc> LordAro: Review/approve #7428? I should as I have a commit in it (even if GH lets me as it is not my PR).
00:57:19 <LordAro> michi_cc: not sure i understand why your commit is necessary
00:58:12 <michi_cc> The const overload of Find returns a std::vector<Pair>::const_iterator, the const overload of End does not.
00:58:41 <michi_cc> end (small e) OTOH does.
00:59:26 <Samu> this is really funny... they're both wrong!
01:00:30 <Samu> the tile iterator is iterating over the other side of the map
01:01:07 <Samu> tile 959 is the void tile on the bottom left
01:01:15 <Samu> tile 960 is the void tile on the top right
01:04:22 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro approved pull request #7428: Fix #7165: SmallMap::Erase(key) does not work correctly https://git.io/fjTIb
01:04:31 <DorpsGek_II> [OpenTTD/OpenTTD] LordAro merged pull request #7428: Fix #7165: SmallMap::Erase(key) does not work correctly https://git.io/fjTUY
01:06:57 <Samu> tile 958 is out there at the bottom left side of the screen
01:07:21 <Samu> the airport we're trying to build is intercontinental
01:08:05 <Samu> FOR_ALL_TOWNS method says the closest town is Nahill
01:08:21 <Samu> Kdtree method says the closest town is Little Drentfield
01:08:51 <Samu> at naked eye, I can clearly see it's Brinnington
01:09:07 <Samu> yeah... it's the iterator
01:09:18 <Samu> it's passing invalid tiles as valid :(
01:11:56 <_dp_> Samu, Little Drenfield and Brinnington are about the same distance
01:12:18 <_dp_> Samu, naked eye doesn't work well with manhattan distances on isometric projection :p
01:12:57 <Samu> it's close to the bare land at the right side
01:13:09 <Samu> the rigth most bareland, the y width of intercontinental
01:14:33 <peter1138> It's AirportTileTableIterator in this instance, right?
01:15:15 <Samu> tile 960 is being passed as valid, and that already a tile on the top right side
01:15:57 <Samu> not sure, how do I check that?
01:16:21 <peter1138> By following the code :p
01:16:30 <peter1138> It's an unplaced airport, so yes.
01:16:58 <_dp_> ah, it's airport iterator missing boundary check
01:17:11 <_dp_> was wondering how did you manage to get FOR_ALL wrong xD
01:17:39 <glx> missing MP_VOID checks somewhere probably
01:18:19 <peter1138> I'm thinking it probably shouldn't need to bounds check.
01:18:36 <peter1138> At least, not within the iterator itself.
01:18:42 <peter1138> Should be before it.
01:19:01 <peter1138> It makes no sense to iterate for something that is not possible to build.
01:19:06 <Samu> i was so focused on two towns, i missed the 3rd town, which was right in front of my eyes...
01:19:22 <_dp_> peter1138, idk, I see none
01:20:03 <glx> I think there's a check on click, when you actually try to build
01:20:40 <peter1138> Yeah, src/scrfipt/api/script_airport.cpp
01:20:56 <Samu> AIAirport.GetNearestTown(958, AIAirport.AT_INTERCON);
01:21:04 <peter1138> Yeah, but no checks where?
01:21:53 <glx> I guess the AI should first check if it can build an airport before checking for the nearest town
01:22:01 <peter1138> Well, I mean, where were you looking for checks?
01:24:43 <_dp_> peter1138, in GetNearestTown
01:25:05 <_dp_> techically yes, can probably be solved if checked if can build first
01:25:21 <_dp_> still not good that you can make gs do some nonsense
01:26:18 <glx> indeed before trying to get the town the script function should probably check for valid airport location
01:27:27 <_dp_> is there a way to check tho?
01:27:32 <Samu> this function is already heavy, you're gonna make it heavier?
01:27:43 <_dp_> except trying to build one and getting an error
01:28:05 <peter1138> Samu, doing it before any loop means it will be a very lightweight function.
01:28:29 <glx> airport size is maybe accessible (not checked), if it is the check is easy
01:28:38 <stormcone> maybe build rail station also needs some checks, there were a report in JGR's patch pack about AI that were trying to build a station outside of the map
01:28:44 <Samu> but when the map is big...
01:29:26 <peter1138> Samu, what's better, slightly more checks (which are not going to slow anything down) or pointlessly doing lots of checks which are wrong?
01:29:27 *** ChanServ sets mode: +v tokai
01:30:28 <glx> yeah if invalid location return NULL is better than returning wrong town after heavy calculation
01:30:51 <peter1138> So, I'm working on it :-)
01:30:54 <Samu> returning NULL can be bad
01:31:22 <glx> well you can return an error too
01:31:59 <glx> just another precondition
01:32:12 <Samu> I was thinking more like... return the actual town
01:32:31 <peter1138> You are thinking wrong.
01:32:41 <glx> you can't search for the town if the airport can't exist
01:32:42 <peter1138> The airport is not buildable there. It should fail.
01:33:12 <Samu> but what about non ai functions
01:33:45 <glx> when you try to place an airport you don't wrap around the map
01:34:46 <Samu> I suspect that's going to kill many AIs, probably mine too :|
01:35:14 <peter1138> Your AI can't build a town there.
01:35:20 <peter1138> Can't build an airport there.
01:35:45 <glx> so getting the town was a useless call anyway
01:36:21 <_dp_> aren't preconditions part of api? I mean needing compat code
01:36:46 <glx> compat code is only for argument change IIRC
01:37:18 <glx> they prevent "compile" failure
01:37:57 <_dp_> glx, ah, so if shit breaks runtime it's ok?
01:38:59 <glx> I think it won't break anything, in this case an AI is looking for a town and probably tries to build an airport there depending on the result
01:39:16 <glx> but building will fail anyway
01:40:22 <glx> so I guess returning TOWN_INVALID and an error is correct
01:40:56 <_dp_> you're assuming some stuff about AI here, I'm talking compatibility in general
01:41:01 <_dp_> but whatever, not like I care
01:41:44 <_dp_> just curious for the next time I break something in gs xD
01:41:46 <glx> AI "compilation" is done at runtime
01:42:55 <glx> failing at precondition should not kill the script
01:46:43 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 updated pull request #7270: Introduce CMake (and removing all other project-related code) https://git.io/fhbqc
01:59:45 <Samu> no more asserts, but then again... no more towns
02:01:32 <Samu> ah, the table it's retrieveing is [0]
02:01:43 <peter1138> Because you are already testing at that point.
02:01:51 <peter1138> It's airport table layout 0.
02:01:59 <peter1138> For non-NewGRF airports, that's the only one.
02:02:24 <Samu> intercontinental layout is 0?
02:02:48 <peter1138> Each airport has its own layouts.
02:03:04 <glx> and some can have more layouts
02:04:06 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN opened pull request #7429: Codechange: Check airport table would fit within map bounds before iterating tiles. https://git.io/fjTtQ
02:05:02 <peter1138> Okay, that hot chocolate tasted weird. I suspect the milk was off :/
02:05:31 <Samu> that's a lot of code, eww
02:05:45 <peter1138> It's mostly comments.
02:05:50 <peter1138> But it's also at the right place.
02:08:03 <Samu> ah ya, i remember, there's some newgrf that could rotate
02:08:59 <glx> but AI don't care, they use only the first layout
02:10:03 <glx> there's already enough complexity for AIs with newgrf vehicles
02:10:06 <peter1138> Yup. I still test for rotation, in case the first layout is rotated.
02:10:22 <peter1138> Unlikely but you never know.
02:10:36 <glx> yeah don't trust newgrf authors
02:11:18 <Samu> oh it's not in master yet t.t
02:11:49 <glx> we never push in master without PR
02:12:44 <Samu> i mean 7429, i was doing git fetch rebase stuff, and it wasn't in
02:13:26 <glx> but you can get the PR locally to test it
02:13:56 <peter1138> I'm fairly sure you were testing PRs earlier :p
02:14:14 <Samu> yes, but i can't mix them :(
02:15:22 <Samu> can't make branch on repository that isn't mine
02:15:54 <glx> you can get master PR as branch in your local clone
02:16:27 <peter1138> Now I wonder *how* Samu was testing PRs earlier..
02:17:08 <Samu> navigated to your branch, build
02:17:44 <peter1138> I guess that's a fairly brute-force way of doing it.
02:19:14 <peter1138> "git fetch upstream refs/pull/7429/head:7429" creates a local branch of the PR.
02:20:55 <peter1138> You and your horrible GUIs
02:23:55 <glx> will add 3 useful commands to git
02:25:02 <glx> rb to rebase with master, pr to get a PR, pr-clean to remove all retrieved PRs
02:29:05 <glx> some stuff is not easily doable in GUI
02:29:10 <Samu> i prefer to do rebase with visual studio, the diff tool is greater than that of git CLI crap
02:29:46 <glx> "git mergetool" opens winmerge for me
02:30:30 <Samu> my config has some entries already :(
02:31:03 <glx> my paste is not a replacement, it's an addition
02:32:06 <glx> hehe you can already do "git mergetool" from CLI to fix the merge in VS ;)
02:32:36 <glx> you can paste at the end of the config file
02:34:44 <glx> oh and you should remove the paste, there's your email in it
02:38:47 <Samu> oh, interesting, just that?
02:38:49 <glx> it will create a pr/7429 branch and checkout
02:39:08 <Samu> i must be on my repository, right?
02:39:38 <Samu> in master or anywhere else?
02:39:50 <glx> I usually do it in master
02:41:59 <glx> once you get all pr you want via CLI you can do the rest via VS
02:42:18 <glx> create a test branch and merge pr branches into it
02:42:30 *** Thedarkb-X40 has joined #openttd
02:55:34 <Samu> assert(forall == kdtree && kdtree == kdtper);
02:55:51 <Samu> mixed with 7429, now testing
02:56:04 <glx> of course the result will differ ;)
02:56:18 <Samu> it's being called from the AI
02:59:24 <Samu> on the 13th January, I'll be done
02:59:54 <glx> still testing on giant map ?
03:00:14 <glx> you really should test on smaller maps
03:00:55 <Samu> testing 64x64, then raising
03:01:14 <glx> only one town in 64x64 usually
03:02:09 <Samu> currently at 512x512, it passed so far
03:03:31 <Samu> 1024x1024 is taking its time...
03:06:24 <peter1138> What does "pass" mean?
03:16:21 <Samu_> not sure I wanna test 4096x4096
03:16:26 <Samu_> took 10 minutes for this one
03:17:58 <Samu_> expecting it to finish in 1 hour
03:29:30 <Samu_> I suppose my while tests aren't needed
03:29:49 <Samu_> you say the gui takes care of it previously
03:33:21 *** SimYouLater has joined #openttd
03:34:03 <SimYouLater> Can someone please explain to me what #ifdef and #endif mean? No, they aren't hashtags.
03:34:34 <Eddi|zuHause> they have conditions, and the code between them is ignored if that condition is false
03:34:52 <glx> and it's checked at compile time
03:35:18 <SimYouLater> Why does pybuild.bat not recognize them?
03:36:07 <SimYouLater> Are you telling me I need to use both C+ and Python to make Recycled Infrastructure Set?
03:36:33 <Eddi|zuHause> dunno what that pybuild is
03:37:26 <Eddi|zuHause> usually you would use cpp (or gcc -E) to combine the .pnml files into .nml
03:37:55 <SimYouLater> All I know is that Nutracks has the only GPL v2 narrow gauge, and it's code uses #ifdef and #endif.
03:38:26 <SimYouLater> I have a make.bat and a pybuild.bat, and I no longer remember exactly why but I know they're used to compile the code..
03:38:38 <Eddi|zuHause> well, you can do all the condition checking manually
03:38:45 <SimYouLater> ^ should be one period, end of sentence.
03:38:47 <Eddi|zuHause> and remove the #ifdef appropriately
03:39:36 <SimYouLater> Except I am in over my head. I don't know where the conditions are located.
03:40:04 <SimYouLater> I do have the Nutracks source, so if you could point me in the right direction.
03:40:18 <Eddi|zuHause> i don't think i ever looked at the nutracks sources
03:40:41 <Eddi|zuHause> but lets start with the words following "#ifdef"
03:41:47 <Eddi|zuHause> (which is now your turn, to tell me)
03:49:50 *** psilynt has joined #openttd
03:50:51 <psilynt> Hello. I'm new. That topic is such bait. Is asking to ask to ask a question OK?
03:51:33 *** psilynt has joined #openttd
04:02:14 <psilynt> I have two serious questions:
04:02:36 <psilynt> 1) What are some good settings for games with really long trains (20+ wagons)
04:02:51 <psilynt> 2) Is there some patch or config setting to allow smaller stations to load large trains faster?
04:03:42 <Eddi|zuHause> there is a patch which loads trains in a drive-through fashion. load all wagons that fit into the station, drive forward a bit, load more, etc.
04:06:09 <Eddi|zuHause> don't know right now, but it is probably included in JGR patchpack
04:07:34 <psilynt> I use JGR but haven't found it
04:08:00 <Eddi|zuHause> "Freight train through load (added in v0.24.0) This is an alternative loading mode for freight trains for the case where the train is longer then the platform."
04:08:28 <Eddi|zuHause> it'll be in the order list, there is "stop at beginning/middle/end of platform"
04:08:43 <Eddi|zuHause> if you click on that, it will cycle through the options
04:10:20 <Eddi|zuHause> (i don't think i've ever read the first post of the JGR thread before)
04:13:27 <SimYouLater> Eddi: Sorry, got distracted. Here...
04:13:29 <SimYouLater> #ifdef INTROS introduces_railtype_list: [INTROS]; #endif #ifdef REQS requires_railtype_list: [REQS]; #endif
04:13:52 <SimYouLater> That's literally the whole thing, so hopefull it's easy to solve?
04:14:23 <Eddi|zuHause> yeah, "INTROS" and "REQS" will come from a table somewhere
04:15:03 <Eddi|zuHause> (and that section is probably included repeatedly, with different definitions for "INTROS" and "REQS"
04:15:49 <glx> probably a #define on top
04:16:04 <Eddi|zuHause> you will find multiple definitions of "#define INTROS" in some source files that include that file
04:16:35 <glx> or it's passed as -D flags in some makefile
04:16:59 <SimYouLater> Then maybe I can just edit it to have values that make sense? What does introduces_railtype_list do?
04:17:38 <glx> guess it's some intro date
04:17:54 <Eddi|zuHause> it's a list of additional railtypes that will be made available simultaneously
04:18:09 <Eddi|zuHause> (so you don't have to define vehicles for each one)
04:19:00 <glx> but as eddi said, check the files including this one
04:19:35 <glx> they should have #define INTROS ... #include "file"
04:20:09 <SimYouLater> I found it. tt-table-definitions.pnml contains the definitions, I think.
04:22:10 <glx> BTW I don't see why they do it like that ;)
04:22:56 <SimYouLater> ERROR: "src/graphics_NBAN_NBAE.pnml", line 92: Syntax error, unexpected token "RANDOM_LC_YEAR"
04:23:18 <SimYouLater> RANDOM_LC_YEAR(lc_NBAN_closed_snow_switch,level_crossing_NBAN_closed_snow, level_crossing_NBAN_closed_snow_modern) RANDOM_LC_YEAR(lc_NBAN_open_snow_switch,level_crossing_NBAN_open_snow, level_crossing_NBAN_open_snow_modern) RANDOM_LC_YEAR(lc_NBAN_closed_switch,level_crossing_NBAN_closed, level_crossing_NBAN_closed_modern) RANDOM_LC_YEAR(lc_NBAN_open_switch,level_crossing_NBAN_open, level_crossing_NBAN_open_
04:24:00 <glx> you want to find #define RANDOM_LC_YEAR
04:24:12 <SimYouLater> I know what a macro is, but not what it's doing in a pnml file.
04:24:25 <glx> it's for the preprocessor
04:25:25 <glx> .pnml once preprocessed become .nml I think
04:25:40 <psilynt> OK, the load-through option works amazing. I didn't see that. Thanks.
04:26:44 <SimYouLater> +glx: Don't worry, I at least know what pnml files are for.
04:28:02 <SimYouLater> I can't seem to find the rest of the macro. How do I replace RANDOM_LC_YEAR with something I can work with?
04:30:46 *** Supercheese has joined #openttd
04:32:31 <SimYouLater> No one knows how to un-macro-ize RANDOM_LC_YEAR?
04:32:53 <SimYouLater> Should I provide the code?
04:33:12 <SimYouLater> Since you're probably trying to help me blind.
04:34:23 *** SimYouLater has joined #openttd
04:34:34 <SimYouLater> Stupid internet cut out.
04:34:41 <SimYouLater> Did I miss anything?
04:40:45 <SimYouLater> ERROR: "src/graphics_NBAN_NBAE.pnml", line 15: Encountered unknown template identifier: tmpl_tunnel_tracks
04:41:15 <SimYouLater> Not sure where to look to give it what it's looking for
04:41:55 <SimYouLater> What is it, Samu?
04:42:29 <Samu_> took 1h14m to iterate every tile
04:43:08 <SimYouLater> Or something else?
04:43:46 <Samu_> 4096x4096, 13k towns, getting closest town when placing intercontinental airport
04:44:36 <Samu_> now I'm off to bed, cyas
04:45:18 <Samu_> i meant to say 7424 + 7429
05:00:09 <SimYouLater> ERROR: "src/graphics_NBAN_NBAE.pnml", line 92: Unrecognized identifier 'lc_NBAN_closed_snow_switch' encountered
05:00:57 <SimYouLater> I can't remove the section... and I can't use it as-is.
05:01:15 <SimYouLater> RANDOM_LC_YEAR will have to wait, then.
06:18:10 *** Smedles has joined #openttd
07:19:20 *** Hobbyboy|BNC has joined #openttd
07:19:28 *** Hobbyboy|BNC is now known as Hobbyboy
07:19:39 *** Supercheese has joined #openttd
08:16:36 <peter1138> Hmm, crash on removing a station.
08:55:21 *** Laedek_ has joined #openttd
09:37:37 <DorpsGek_II> [OpenTTD/OpenTTD] btzy updated pull request #7028: Feature: Option to group vehicle list by shared orders https://git.io/fhGxd
09:38:28 *** HerzogDeXtEr has joined #openttd
09:42:04 <DorpsGek_II> [OpenTTD/OpenTTD] btzy updated pull request #7028: Feature: Option to group vehicle list by shared orders https://git.io/fhGxd
09:45:52 <DorpsGek_II> [OpenTTD/OpenTTD] btzy updated pull request #7028: Feature: Option to group vehicle list by shared orders https://git.io/fhGxd
09:57:17 <DorpsGek_II> [OpenTTD/OpenTTD] btzy updated pull request #7028: Feature: Option to group vehicle list by shared orders https://git.io/fhGxd
10:03:24 <peter1138> Caused by aa7ca7fe6
10:19:22 <DorpsGek_II> [OpenTTD/OpenTTD] btzy commented on pull request #7028: Feature: Option to group vehicle list by shared orders https://git.io/fjTcJ
10:27:25 *** psilynt_ has joined #openttd
11:54:11 *** Gustavo6046 has joined #openttd
12:30:41 *** andythenorth has joined #openttd
13:25:42 *** gareppa has joined #openttd
14:41:30 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN opened issue #7430: Vehicle waiting at station affects cargo age rating even if it did not pick up cargo https://git.io/fjT4D
14:43:39 <andythenorth> I love that issue
14:43:52 <andythenorth> some things in ottd are just a bit brainfuck eh :)
14:44:37 <peter1138> nielsmh asked them to report it.
14:44:47 <peter1138> But they didn't. They wanted to "understand" it.
14:46:15 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on issue #7430: Vehicle waiting at station affects cargo age rating even if it did not pick up cargo https://git.io/fjTBe
14:46:33 <peter1138> What's a piglet anyway?
14:52:17 <peter1138> Ah the "legitimate" exploitation.
14:52:19 <andythenorth> piglet is an exploit
14:52:30 *** Romazoon has joined #openttd
14:52:47 <peter1138> Update rating on final delivery? o_O
14:53:34 <peter1138> Thing is, with the piglet, that's kinda how the rating works.
14:54:17 <andythenorth> station rating is never a thing I pay attention to
14:55:26 <Romazoon> Hi, i see the "fake" piglet issue made it s way to bugtracker, is that really that important to fix ? i mean why can t we stick to don t use it if you don t want to use it ?
14:56:29 <Romazoon> vehicle with full load orders where always a way to boost ratings, won t it be affected by changing the way cargo pick up affects rating ?
14:57:14 *** supermop_work has joined #openttd
14:57:53 <_dp_> peter1138, you'll soon "fix" all the stuff on our "advanced tricks" page that way xD
14:58:42 *** supermop_work_ has joined #openttd
15:01:52 <Romazoon> also, just want to point out that, if a vehicle that don t has anything to load do not affect station rating i think it might cause some other issues like when cargodist is not in use ? an airport easily get higher ratings due to the max speed of the planes, but why should we penalise the rating of that station in the case only busses managed to pick up some passenger. The ratings are a way to get more cargoes but i find that t
15:04:59 <Romazoon> also a train stations with plenty of trains (deserving good ratings because maybe all those trains allow you to go many different place, i know the rating mecanism don t care of that but with plenty of train you insure an high frequency, now if there is very few cargo to pick up, you don t get any boost on ratings thanks to high frequency
15:05:29 <LordAro> sounds like something we should ask for feedback for in the monthly dev post!
15:12:36 <peter1138> Romazoon, I figured as much.
15:12:58 <peter1138> The "we don't want to report it because it might get fixed" attitude.
15:16:41 *** supermop_work has joined #openttd
15:16:45 <andythenorth> is that a recognised pattern? :P
15:16:50 <andythenorth> does it have a name?
15:19:09 <peter1138> And no, "don't use the cheat if you don't want to cheat" isn't really acceptable.
15:19:16 <peter1138> Not in multiplayer.
15:19:42 <peter1138> I still don't know what you were testing last night.
15:21:10 <Samu> comparing FOR_ALL_TOWNS vs Kdtree vs Kdtree (only perimeter)
15:21:48 <peter1138> Ok, I somehow had it in my head that one of my PRs was being tested as well, hence the confusion :-)
15:22:11 <peter1138> With my map bounds check?
15:22:32 <peter1138> If you do that, then you don't need the 2 loops at the beginning searching for valid tiles
15:22:37 <peter1138> (Those loops are wrong anyway)
15:23:18 <peter1138> with 7429 the bounds have been checked already
15:24:00 <peter1138> You could add an assert but you'd also have to take account of rotation as well.
15:24:02 *** supermop_work_ has quit IRC
15:24:02 *** Wormnest has joined #openttd
15:24:59 <peter1138> I still wonder if just checking the corners would be even simpler.
15:25:21 <peter1138> The distance would be ever so slightly off in the case of a missing corner tile, but otherwise hardly any difference.
15:25:58 <peter1138> Maybe there will be some weird layout in the future where that doesn't work, but in that case just checking perimeter would likely fail.
15:26:11 <peter1138> And then again nobody even knows about newgrf airports :)
15:27:03 <Eddi|zuHause> drawing general conclusions from yourself again? :p
15:27:40 <Eddi|zuHause> so... i had an external HDD... but it isn't in the place where i remember last putting it...
15:30:15 *** supermop_work_ has joined #openttd
15:30:21 <andythenorth> it will be where you look last
15:30:31 <andythenorth> so just stop looking and you'll find it, right?
15:30:36 <andythenorth> language is weird :P
15:32:02 <Eddi|zuHause> i tried that. isn't working.
15:32:03 <Romazoon> Peter, i kinda agree in multiplayer it s not so great, it gives work to server admin to make sure people don t use it. But it s already the case with station spread abuse.
15:33:27 <Samu> question: is if (Town::GetNumItems() == 0) test needed?
15:33:47 <peter1138> Romazoon, station spread has a game setting to reduce it.
15:33:55 <peter1138> Romazoon, two settings, actually.
15:33:59 <Eddi|zuHause> ok, i found it... it was near where i was expecting it to be, just slid between other things
15:34:09 <Romazoon> and like Mczapkie pointed, there is a simple way around that setting
15:34:51 <Romazoon> spread the station while they are attached, then remove the tiles in between the main station tile and the spread abuse tile
15:34:58 <peter1138> Samu, I think it could be an assert().
15:35:44 <Samu> assert(Town::GetNumItems() != 0);
15:37:06 <peter1138> The existing code will return NULL if there's no towns, but the calling code does not check for a NULL return, so will crash anyway.
15:37:56 <peter1138> (they all dereference it )
15:38:34 <peter1138> Romazoon, yes of course. But "Because there are multiple ways to exploit things, we should leave any exploits in" does not hold with me.
15:40:45 <Romazoon> Peter, i agree on that. I just don t want that ottd developers loose time fixing a behavior that is in my opinion totally acceptable. If yu had played on Mczapkie servers you would know that he is looking to make the game "very very hard...and development of companies extremely slow". i think that s the only thing that changing that behavior will achieve. is it worth developers time ? i don t think so (imho)
15:41:34 <Romazoon> rating are quite smart if you ask me : as i pointed on the tt forum, if a station get too much cargo waiting because of the 'fake' piglet, rating will go down.
15:41:49 <Eddi|zuHause> i don't think you should worry about developer time
15:41:55 <Romazoon> piglet cost money : i see that as an investment to upgrade a station
15:41:57 <_dp_> peter1138, thing is, it would've been nice if it was fixed before it became mainstream :p
15:42:10 <_dp_> peter1138, like 5-10 years ago when it was likely first reported :p
15:42:52 <_dp_> peter1138, though whatever you do with loading is probably gonna be ok imo
15:42:56 <peter1138> _dp_, yeah but nobody reported it!
15:43:06 <peter1138> Or it was buried, maybe.
15:43:07 <Romazoon> Eddi, developpers time is not mine, and it s seems to be a limited ressource, sorry to care about that ;)
15:43:19 <Samu> about the rotation thing...
15:44:48 <Romazoon> for me that rating problem is not a problem. but i m all open to others opinion, after all i d be totally ok with the way to go via a 'setting' to turn on or off that behavior
15:45:20 <peter1138> Samu, i don't know if it needs fixing.
15:45:29 <Eddi|zuHause> my favourite approach so far was a setting in FIRS: "make rating 100% always"
15:46:18 <peter1138> Romazoon, "i don't mind that rating is broken" is not really valid.
15:46:31 <Romazoon> eddi, i like it too ! especially for the oil platform with neutral station....now it s barely possible to get high ratings (unless you terraform and use some maglev station next to oil platform)
15:46:43 <Romazoon> it s even a serious problem for ECS
15:46:46 <peter1138> Samu, it'll be wrong for airports with missing corners, I think.
15:47:00 <Eddi|zuHause> Romazoon: we "fixed" the maglev station thing, as far as i followed :p
15:47:22 <Samu> there are airports with missing corners already? :(
15:47:28 <peter1138> And also for a very large airport, possible wrong for towns half-way down a side.
15:47:36 <Romazoon> rating are not broken ! if a station have plenty of vehicles stopping by it, it should get higher ratings, it doesn t matter if there is somthing to load or not
15:48:04 <Eddi|zuHause> we didn't say "broken", we said "exploit"...
15:48:11 <peter1138> I said broken, sorry.
15:48:15 <Samu> newgrf airports? or vanilla?
15:48:24 <Eddi|zuHause> ok, maybe i was too quick on that :p
15:48:35 <peter1138> Anyway station rating and station cargo rating are two different things.
15:49:05 <peter1138> Samu, tell me about it. I'm still mad that we ever added the extra layouts rather than deferring to newgrf.
15:49:18 <Romazoon> eddi, i mean to get high level of production with ECS i think it s needed to reach 80% of station ratings (or 80% of the oil taken from oil platform), what else than a maglev oil train can reach that rating ?
15:49:29 <_dp_> Romazoon, and who's changing that?
15:49:33 <Eddi|zuHause> Romazoon: there's two ways to "fix" an exploit: a) "nerf" it, meaning you can't do it anymore, or b) making it "official" and redundant
15:49:48 <_dp_> as I understand issue is with vehicles that can't load anything still bumping rating
15:50:16 <Romazoon> dp : the future behavior apparently want to make station rating no going up if a vehicle arrives and find nothing to load
15:50:59 <Eddi|zuHause> Romazoon: but how is that a problem? an industry will frequently produce something, that can be loaded
15:51:23 <Romazoon> so in a small town, if i set plenty of trains only the one actually picking up something will up the ratings, maybe it will be the slowest train and so rating boost will be low
15:51:33 <_dp_> Eddi|zuHause, that actually could've been a problem. e.g. with seasonal industries like fruit plantations
15:51:48 <Samu> if ((TileX(cur_tile) == it_x || TileX(cur_tile) == max_it_x) && (TileY(cur_tile) == it_y || TileY(cur_tile) == max_it_y)) {
15:51:55 <Samu> i hope these are the corners
15:51:56 <peter1138> Romazoon, nope. I was only considering touching the time_since_pickup variable.
15:52:04 <peter1138> Not the vehicle speed/age/whatever stuff.
15:52:14 <Eddi|zuHause> so maybe: station rating lowers, when cargo is moved to station, but nothing there to pick it up?
15:52:25 <supermop_work_> rating my linkgraph would be cool
15:52:38 <supermop_work_> as would town growth by link graph
15:52:52 <Eddi|zuHause> no cargo at station means no rating change
15:52:53 <supermop_work_> does that already exist?
15:53:01 <peter1138> This is just about vehicles sitting on a station and waiting and waiting and waiting and not loading because they're full, but not leaving because they are told not to leave yet.
15:53:17 <supermop_work_> eg can get to 20 towns from here = good rating
15:53:25 <peter1138> supermop_work_, Oh, I was looking for a "rate my linkgraph" website :p
15:53:48 <peter1138> NewStationCargoRating!
15:53:52 <supermop_work_> takes 10 intermediate stops to get to a nearby big destination = less good
15:53:57 <peter1138> I've not done that for a long time.
15:54:00 <_dp_> peter1138, another option is when already full vehicle arrives at the station
15:54:20 <peter1138> In the bug report I list a couple of options.
15:54:23 <Romazoon> peter, ok for no change on th vehicle speed/age/whatever, i m less worrying already
15:54:43 <peter1138> Like only reseting it once per vehicle.
15:54:46 <andythenorth> newgrf station rating? o_O
15:54:54 <peter1138> andythenorth, no, not NewGRF.
15:55:03 <peter1138> Although is that a thing? Probably is it.
15:55:06 <andythenorth> yes it's a thing
15:55:14 <andythenorth> content-ish station ratings
15:55:18 <andythenorth> make everything content
15:55:27 <andythenorth> and make the non-content vanilla defaults Not Shit ®
15:55:29 <_dp_> peter1138, no, not once, you need to constantly update it if there is no cargo on station
15:55:32 <supermop_work_> i tried a (Polish?) GS once that tried to make growth dependent on passenger connections
15:55:40 <Eddi|zuHause> iirc the industry newgrf can change ratings, but not the station newgrf
15:55:48 <Romazoon> once per vehicle sound a good way to no t break the 'seasonal industries'...when player don t use full load
15:55:57 <andythenorth> Eddi|zuHause: let me introduce you to a new concept
15:56:04 <peter1138> _dp_, so that rating goes up, otherwise it never will, right?
15:56:34 <_dp_> peter1138, yeah, would be a pity to loose acceptance while train is waiting for cargo ;)
15:56:40 <supermop_work_> andythenorth: i do feel that stations that are major nodes, even if few people ever board or alight there should get good ratings
15:56:45 <peter1138> _dp_, at least, that is what i gathered about the original bugfix.
15:56:49 <andythenorth> make a newgrf for that supermop_work_
15:57:05 <andythenorth> I mean., obviously everything is broken with cdist
15:57:06 <supermop_work_> though if every train that pulls in is already full then yes, i get why it should have a bad rating
15:57:09 <andythenorth> but that's because cdist is broken
15:57:16 <Eddi|zuHause> nothing good ever came out of a sentence starting with "To understand how to use it, you must first understand"
15:57:18 <andythenorth> I love cdist, but it is broken
15:57:31 <supermop_work_> i won't ever play without it
15:58:25 <Romazoon> i also can t play without CD (for passengers and mail), but i just never use it for freight (too broken)
15:58:26 <_dp_> also it's always possible to build station and send trains before building an industry
15:58:29 <Eddi|zuHause> supermop_work_: the problem with "you must deliver X destinations for good ratings" was industry newgrfs that have overly strict reactions to low ratings (ECS)
15:59:00 <Eddi|zuHause> also, finding out the number of valid destinations is nontrivial
15:59:38 <andythenorth> is cdist broken for freight? :P
15:59:55 <FLHerne> Mostly it works fairly well
15:59:55 * andythenorth just uses a distorted play style to account for it
16:00:11 <andythenorth> - never more then 2 destinations from one pickup station
16:00:19 <FLHerne> It's broken for supplies, because it doesn't understand the industries' demand levels
16:00:22 <andythenorth> - always set 'do not load' orders at destination
16:00:33 <_dp_> Eddi|zuHause, it's also goal-related. e.g. it makes no sense to require x destinations in cb where everything goes to one city
16:01:03 *** Wormnest has joined #openttd
16:01:03 <FLHerne> andythenorth: Where's the fun in that? :P
16:01:12 <Eddi|zuHause> _dp_: well, that could be tuned
16:01:50 <_dp_> Eddi|zuHause, also easily exploitable with multiple stations per destination xD
16:01:56 <FLHerne> Overcomplicated systems based on autorefit between several cargoes in both directions are !FUN!
16:02:09 <andythenorth> autorefit doesn't really work
16:02:09 <Eddi|zuHause> andythenorth: freight cdist is not broken, but the global "distance bias" is interfering, because for freight you want it low, but for passengers high
16:02:31 <andythenorth> the inability to serve a 3rd destination from a station is pretty broken
16:02:38 <Eddi|zuHause> autorefit works, but you need to jump-start it
16:02:50 <FLHerne> Actually, my current game isn't too absurd
16:03:02 <Romazoon> so the simple solution would beto make vehicles that are already full or just became full to have no effect on the last cargo picked rating
16:03:04 <Eddi|zuHause> generally, starting new lines can be a bit tricky with cdist
16:03:18 <FLHerne> I've got a few lines where sand/stone/coal all share the same trains for a leg
16:03:21 <andythenorth> it's inevitable, with the current approach
16:03:31 <andythenorth> not sure how it could be fixed
16:03:37 <nielsm> the big issue with cdist is that it does not model demand
16:03:42 <Eddi|zuHause> Romazoon: that doesn't cover the variant involving cdist, where the vehicle never loads anything because it doesn't serve a link
16:03:45 <nielsm> but invents demand based on supply
16:03:52 <andythenorth> nielsm: that's not an issue, that's by design :)
16:04:07 <FLHerne> And one where trains drop about half their wood to a sawmill, then pick up the corresponding timber, then drop the other wood at a paper mill and the timber at a yard
16:04:10 <supermop_work_> Romazoon: i actually don't mind cdist with firs
16:04:25 <supermop_work_> it's never really bothered me that much
16:04:31 <andythenorth> the big misnomer about cdist is that it provides destinations for cargos
16:04:35 <Romazoon> Eddi, but why that would be an exploit, people doing that (me) are paying for the running cost of said vehicle
16:04:41 <andythenorth> clue is in the name though, it's not destinations
16:04:44 <FLHerne> One tiny shuttle between docks and an industrial plant that carries pretty much everything in both directions
16:05:14 <Eddi|zuHause> Romazoon: because those running costs are negligible
16:05:44 <FLHerne> Oh, and lots of trains that carry fruit/goods/whatever in one direction and supplies in the other
16:05:49 <Eddi|zuHause> (and nobody cares about the company rating for minimum vehicle profit)
16:06:14 <Romazoon> eddi, true, cargodist broke that company rating XD
16:06:31 <Eddi|zuHause> Romazoon: no, that was "broken" long before...
16:06:50 <Romazoon> you are right, transfert were already a thing before
16:06:52 <supermop_work_> for what it's worth, when i go to a subway station and have to wait for 2 trains to go by in rush hour because they are full
16:07:04 <supermop_work_> i have a bad rating of that station in my mind
16:07:16 <Romazoon> i agree on that supermop
16:07:27 <supermop_work_> and i may next time chose a different station where i am more likely able to board
16:08:00 <supermop_work_> similarly, on the lexington avenue line (most crowded line here), in the evening rush hour
16:08:02 <Eddi|zuHause> Romazoon: it's also "broken" without transfers. like almost never you can have road vehicles meeting that barrier, whereas trains is trivial. and ships taking more than a year for a roundtrip
16:08:57 <supermop_work_> downtown local trains often skip 28th, 23rd, and astor place stations, because they are too full and dont wan't to mess up the tight headway
16:09:05 <Romazoon> i see my 'fake piglet' as machinery that equip the station. And so i find that it make sense to boost rating in those case. I mean a company would probably prefer a station that has a forklift to load it s train than a station that let the company to deal with the laoding of freight (i mention forklift cause that often what i use)
16:09:35 <supermop_work_> so if you are waiting at 28th street, you are pissed (bad rating). and might walk instead to 33rd(better rating)
16:11:01 <supermop_work_> Eddi|zuHause: in the past i have occasionally had a lonely overland truck duplicate a train or ship route just to keep a link active
16:11:54 *** Alberth has joined #openttd
16:11:54 *** ChanServ sets mode: +o Alberth
16:12:25 <peter1138> Samu, yeah I figured it would not give the same results.
16:12:49 <peter1138> Just perimeter will work mostly unless it's a strange layout which probably doesn't exist (yet)
16:13:03 <Romazoon> Eddi, ok so minimum profit per vehicle is broken, is it on the bug list ? or is there any plan to fix that ? (an personally i never had that problem, back in the days of TTD i was taking that as a challenge and would modify my bus route if it would not achieve the profit to get max company rating
16:13:07 <peter1138> Samu, on the other hand, maybe it's not wrong, just different? ;)
16:13:35 <peter1138> Romazoon, the bug list is searchable.
16:14:04 <Eddi|zuHause> Romazoon: back in my TT days, i didn't even know how that rating was calculated. the detailed rating window did not exist
16:14:16 <Romazoon> now, i just don t care of company rating alltogether, i grew up and play for having nice and working large network
16:14:41 <peter1138> nielsm, yeah... i think simutrans did it differently.
16:15:07 <peter1138> nielsm, industries had explicit destinations already, regardless of routes.
16:15:10 <Romazoon> back in the TTD days i guess i played long enough to figure it out somehow
16:15:18 <peter1138> nielsm, so if you just randomly hooked things up you'd not get anywhere.
16:15:37 <peter1138> But that is the difference between cargo*dest* and cargo*dist*
16:16:39 <Romazoon> YACD was very interesting for OTTD, it just could not cope with a map like Cindini.
16:17:09 <Eddi|zuHause> YACD was a performance bottleneck waiting to happen
16:17:18 <Romazoon> and cargodist while it has default, once you have a huge network, it kinda do exactly what YACD was doing
16:17:42 <peter1138> Eddi|zuHause, yeah but *my* computer got faster since then ;)
16:17:53 <peter1138> Should I have a chocolate bar or a kiwifruit?
16:18:01 <Eddi|zuHause> peter1138: now try YACD on a 4096^2 map :p
16:18:20 <peter1138> I don't know where to find YACD these days :(
16:18:34 <peter1138> Romazoon, probaby sensible, the chocolate will still be there tomorrow...
16:19:09 <peter1138> My snack draw is terrible. The only thing that saves it is that all that stuff is still in it, rather than having been eaten.
16:20:13 <Eddi|zuHause> i still have a yacd checkout
16:21:07 <peter1138> I probably don't have that scaler patch any more.
16:21:22 <peter1138> It's certainly an... interesting effect.
16:22:00 <Eddi|zuHause> it reminds me of the days of MiniIN with the double zoom filters
16:24:17 <peter1138> Chocolate coated kiwifruit?
16:26:03 <andythenorth> YACD was so good
16:26:08 <andythenorth> really fricking awesome
16:26:23 <andythenorth> I didn't have the 'it eats CPU' problem
16:26:34 <andythenorth> probably because I don't play idiot-sized maps
16:26:51 <andythenorth> why is openttd distorted by demands from people who have no taste? :(
16:27:00 <andythenorth> it stops us making it nice
16:27:09 <andythenorth> peter1138: drawer not draw :P
16:28:37 <peter1138> Yes, I... lazied it.
16:29:11 <Romazoon> i m an idiot with no taste. Yet i m having lot of fun playing ottd, thank you :P
16:30:28 <peter1138> andythenorth, okay, how do we bring back YACD, bearing in mind that CargoDist exists.
16:30:46 <peter1138> I was told it was possible to build it on top of cargodist, at the time.
16:31:00 <Romazoon> and to be honest, i only play idiot sized map on multiplayer, so when there is enough idiots to play with me xD
16:31:45 <andythenorth> peter1138: NFI, fork it, remove cdist?
16:31:52 <andythenorth> fragment the community?
16:32:12 <andythenorth> to be completely fair to fonso, I don't hate cdist, it's really added something to the game
16:32:16 <andythenorth> it's just not YACD
16:32:22 <andythenorth> and YACD wasn't viable at the time
16:32:50 <peter1138> yet simutrans has it ;)
16:33:00 <Romazoon> we should have cardodist for passenger and mail (because once you have a big network it kinda work the same as YACD), and YACD for freight/indutries
16:33:00 <peter1138> Maybe it just needed k-d trees.
16:33:08 <andythenorth> k-d trees for all
16:33:18 <andythenorth> michi_cc would know more what the problem was
16:33:33 <andythenorth> iirc, it was because packets are tracked individually, not statistically, but I might be wrong
16:34:16 <Eddi|zuHause> we need to rewrite the game loop so things like vehicles and cargopackets can be concurrent
16:34:38 <peter1138> They are intertwined...
16:34:47 <Eddi|zuHause> yeah, we need to cut those
16:35:29 <Eddi|zuHause> vehicles must be able to move independently from each other
16:35:45 <Eddi|zuHause> it works in the real world :p
16:35:59 <peter1138> A thread for each vehicle :D
16:36:15 <andythenorth> can we offload it to cloud lambdas?
16:36:15 <Eddi|zuHause> i don't think that is feasible :p
16:36:37 <peter1138> Okay so you need to break out cargo loading from the vehicle loops.
16:36:50 <Eddi|zuHause> just because you design it to be concurrent doesn't mean it must run in parallel
16:36:51 <andythenorth> now it gets interesting
16:37:08 <peter1138> And cargo unloading, I guess.
16:37:22 <peter1138> But once you do that, what's left of the vehicle loops? Just a bit of movement really.
16:37:41 <peter1138> Oh yeah, pathfinding :)
16:42:57 <Samu> there's still something I wanna test
16:43:05 <Samu> different airport types!
16:43:13 <Samu> I only tested intercontinental
16:49:05 <peter1138> I suppose you could run vehicle types separately, except for level crossings.
16:49:18 <peter1138> But probably most games only have tons of trains and not much else :p
16:49:34 <peter1138> Split per company? :p
16:53:30 <_dp_> ha, looks like I'm not the only one with too many idle cores :p
16:53:38 <Eddi|zuHause> split per company doesn't make a lot of sense, you still have interdependencies, especially RV, and you just kill infra sharing for no benefit
16:54:17 <Samu> if I waited 1h25m yesterday, how much am i gonna wait now?
16:55:41 <peter1138> Do it on a small map.
17:04:00 *** synchris has joined #openttd
17:05:30 *** Progman has joined #openttd
17:08:07 *** sla_ro|master has joined #openttd
17:20:43 <Samu> I added some % so that I dont get bored quickly
17:21:08 <Samu> 1024x1024 and below have passed the test
17:21:18 <Samu> waiting for these two, 2048x2048
17:41:58 <Samu> while this is running, im gonna play patch of exile
17:45:10 <peter1138> Samu, oh, I was wrong, none of the default airports have missing tiles.
17:45:43 <peter1138> I thought the intercontinental did, but it's just a bare airport tile visible as grass.
17:47:06 <peter1138> It is a missing tile in the OpenGFX+ airports version of it, though.
17:48:02 <Samu> im not testing opengfx :(
17:48:18 <Samu> patch of exile is patching... typical
17:51:56 <peter1138> Test the opengfx+ intercontinental :)
17:52:01 <peter1138> It has one corner missing.
17:52:13 <Samu> im gonna assume intercontinental is still 7
17:53:48 <peter1138> I would not assume.
17:54:02 <peter1138> They may be APIs to query but I dunno
17:54:53 <Samu> there's AIAirport.INTER_CON things
17:57:00 <peter1138> NewGRFs can add more airports, so...
17:59:31 <Samu> for this test I know I'm gonna have to wait 1h25m or less if it asserts
17:59:49 <Samu> cpu usage 51%, I can still play poe
18:03:00 <Samu> if i were on the i5-2500, this would mean 100% cpu usage, could no longer play poe
18:08:13 <Eddi|zuHause> so my yacd checkout, how would i move that over to the new git?
18:09:28 <Eddi|zuHause> apparently i also have a daylength patch applied on top of it :)
18:10:05 <Eddi|zuHause> can i export stuff as like a patch queue?
18:10:42 <Eddi|zuHause> so i export all stuff from the last common trunk revision, check out that revision in the new git, and reapply it?
18:10:43 <nielsm> if the old code is also in git, I guess you'd add the new repos as a remote to the old, rebase the patch onto the matching base revision from the new repos, and then carefully bring the code up to present HEAD
18:11:04 <Eddi|zuHause> let's ignore the rebase to head for now
18:11:16 <Samu> that test was testing all vanilla airports on all tiles
18:11:30 <Eddi|zuHause> there's a problem that the old git mirror and the new git have unrelated revisions that are essentially the same
18:11:54 <Eddi|zuHause> the old code is in a git repo
18:12:02 <Eddi|zuHause> but one based on the old mirror
18:12:50 <Eddi|zuHause> i would need to teach it that those are the same
18:12:50 <nielsm> you want to use "git rebase --onto"
18:13:14 <Eddi|zuHause> so, in the old repo, i fetch from the new git?
18:13:39 <nielsm> which likely requires three revision hashes, starting point of old branch, end point of series you want to move, and destination revision
18:13:55 <nielsm> yeah add the new repo as a new remote to the old
18:14:22 <nielsm> or do it the other way around
18:14:30 <nielsm> add the old repo as a remote in the new
18:14:48 <nielsm> the important part is getting both histories available in a single repo
18:16:17 *** supermop_work_ has quit IRC
18:18:32 <Eddi|zuHause> so, last common revision is r22439
18:18:47 <Samu> server address doesn't fit
18:19:06 <Samu> server name on the right in orange, doesn't fit
18:21:05 <Eddi|zuHause> which in our current timeline is b7c6424104e6261599a00f81e7e2aa1754454a18
18:21:14 <Eddi|zuHause> so i now git rebase --onto b7c6424104e6261599a00f81e7e2aa1754454a18?
18:21:33 <nielsm> and you need one or two revision names more
18:22:03 <nielsm> start and end revision you want to rebase from
18:22:34 <Eddi|zuHause> i have no clue what you're saying
18:22:53 <nielsm> I barely remember this either
18:23:00 <nielsm> did it twice or so last year
18:23:41 <nielsm> but rebase --onto with just the destination revision is going to do the wrong thing
18:28:15 <nielsm> you want to name three revisions on the rebase --onto commandline
18:28:18 <Eddi|zuHause> but where would i put those things into a command line?
18:28:41 <Eddi|zuHause> not finding the right info
18:29:16 <Eddi|zuHause> "git help rebase" seems to suggest just "rebase" should do it
18:29:37 <Eddi|zuHause> but it gives only a simple example there
18:33:41 <Samu> that's how my brain interprets that
18:34:25 <Eddi|zuHause> ok, i think just plain rebase worked
18:36:18 <Eddi|zuHause> hm, git has somewhere stored wrong login/password pair
18:36:26 <Eddi|zuHause> and i don't know where that is stored to delete it
18:37:14 *** HerzogDeXtEr has joined #openttd
18:39:28 <nielsm> actually that looks like the wrong thing
18:39:45 *** andythenorth has joined #openttd
18:39:47 <nielsm> how you configure a system to manage credentials
18:40:10 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN opened pull request #7431: Fix aa7ca7fe6: Linkgraph node index order must be maintained due to other references. https://git.io/fjTgY
18:42:46 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN commented on pull request #7431: Fix aa7ca7fe6: Linkgraph node index order must be maintained due to other references. https://git.io/fjTgn
18:44:37 <Eddi|zuHause> bah... i can't find it
18:46:16 <peter1138> Hmm, I guess it's not impossible that ca2f33c6d needs more eyes.
18:50:37 <Eddi|zuHause> what's ca2f33c6d?
18:52:43 <Eddi|zuHause> hm, it might be broken
18:53:55 <peter1138> old erase did a swap & pop from the end
18:54:08 <peter1138> std::vector erase shifts all the elements down
18:54:30 <peter1138> so #7431 fixes one occurrence that definitely mattered.
18:54:54 <peter1138> I think this other commit is okay.
18:55:33 <peter1138> Although I haven't looked at the SmallMap callers.
18:56:13 <peter1138> All checks passed, woo :D
18:58:52 <Eddi|zuHause> "This branch is 66 commits ahead, 5899 commits behind OpenTTD:master. "
18:59:32 <Eddi|zuHause> where can i view history on github?
18:59:57 *** supermop_work has joined #openttd
19:00:35 <Eddi|zuHause> it seems to look right
19:00:54 <peter1138> I'm getting used to 'tig' but gitg is okay.
19:02:07 <peter1138> Oh, does NML need to be updated for those new variables?
19:02:38 <peter1138> Hmm, and I guess newgrf-specs wiki
19:03:29 *** supermop_work_ has joined #openttd
19:03:51 <Eddi|zuHause> this is based on yacd 2.2, dunno if there was a newer version
19:05:48 <peter1138> Well it's going to go horribly wrong at cargodist, certainly.
19:06:35 <Eddi|zuHause> then i don't have that
19:08:37 <Eddi|zuHause> can you "bisect" a rebase? like "try rebasing over X revisions, if no errors, continue"?
19:08:58 <nielsm> I don't think there's anything built in to do that
19:19:30 <andythenorth> time for the test game
19:19:43 <andythenorth> which inevitably ends after 7 years when I discover some horrible FIRS bug :P
19:19:55 <andythenorth> if only there was a way to short cut that part :P
19:20:09 <peter1138> AIs + fast-forward.
19:20:15 <andythenorth> 15 AIs or 16? o_O
19:20:47 *** supermop_work__ has joined #openttd
19:21:11 <andythenorth> pikka hasn't done much industry stuff in Civil AI yet
19:22:36 *** supermop_work___ has joined #openttd
19:24:53 <andythenorth> anyone played Villages is Villages?
19:25:38 *** supermop_work____ has joined #openttd
19:26:07 *** supermop_work_ has quit IRC
19:28:11 *** supermop_work_ has joined #openttd
19:28:51 *** supermop_work__ has quit IRC
19:31:42 *** supermop_work___ has quit IRC
19:32:40 <Wolf01> fillLeft("supermop_work", "_", random()) fix your connection :P
19:34:27 <Eddi|zuHause> andythenorth: you mean, you want to just have the horrible bugs, and not play a game?
19:35:13 *** supermop_work____ has quit IRC
19:35:14 *** andythenorth is now known as Guest4443
19:35:15 *** andythenorth has joined #openttd
19:35:33 *** Supercheese has joined #openttd
19:37:01 <supermop_work_> andythenorth: ive played a lot of viv
19:37:27 <supermop_work_> if that's waht you want
19:37:37 <supermop_work_> if certainly does what it says on the tin
19:37:44 <andythenorth> I want GS with some kind of narrative
19:38:06 <peter1138> I've loaded it, but wasn't sure what I was meant to do
19:38:26 <andythenorth> I need to try more GS
19:38:27 <supermop_work_> it is still limited by the map generator in the sense that if you want big city surrounded by smaller towns that stay small you get that
19:38:31 <andythenorth> instead of just declaring GS broken
19:38:39 <peter1138> What I want to do is play a multiplayer game with NRT
19:38:45 <peter1138> It's not been merged yet.
19:39:03 <andythenorth> I need to test unreleased FIRS :P
19:39:05 <supermop_work_> even if a small town is the most natural center of your network, and the only big city is on top of some remote mountain in the corner of the map
19:39:09 <andythenorth> but we have no way to do that
19:39:26 <andythenorth> how will Villages* interact with pikka's TAI?
19:39:31 <andythenorth> both try to control towns
19:39:33 <peter1138> "At least 1 approving review is required by reviewers with write access."
19:39:42 <supermop_work_> viv just limits growth
19:39:44 <peter1138> ARE PEOPLE EXPECTING ME TO REVIEW THIS?
19:39:55 <peter1138> When all the commits have my NAME on them.
19:40:03 <andythenorth> TB also reviewed it
19:40:18 <supermop_work_> i imagine it would work with tai, just dampen some of its effects in smaller towns
19:40:25 <andythenorth> if only it had already been available in a binary for like 2 years, peter1138
19:40:35 <andythenorth> it might have had play testing and stuff
19:41:45 <Eddi|zuHause> "43c8cd1d5e7d0fa2b7e369eac0239a41ae6c0bb1 is the first bad commit"
19:42:27 <Eddi|zuHause> that's the first rev that causes a conflict
19:42:31 <peter1138> That's automatic -> implicit orders
19:43:02 <peter1138> Fairly invasive search & replace, tbh.
19:43:27 <Eddi|zuHause> but that shouldn't be a code problem, just a conflict resolving headache
19:44:41 <peter1138> Hm, maybe I should fake the commits to be from andythenorth, and then approve it myself, and then lol
19:45:12 <peter1138> And we'd already had updates before it was PR'd.
19:46:38 *** supermop_work has joined #openttd
19:47:26 <andythenorth> peter1138: create a sock puppet account, gain our trust, then get approval rights, then commit it?
19:47:43 <peter1138> Maybe I already did!
19:48:30 <FLHerne> Eddi|zuHause: Wasn't there an attempt somewhere at implementing YACD by manipulating the cdist graph?
19:48:59 <Eddi|zuHause> "configure: error: gcc older than 3.3 can't compile OpenTTD because of its poor template support" well, that makes testing the changes difficult
19:49:14 <peter1138> FLHerne, that was the alleged way of doing it
19:49:49 <peter1138> FLHerne, not sure if anyone actually did it.
19:50:18 <Eddi|zuHause> glx: it's misdetecting the version
19:51:27 <glx> at least it's explicit :)
19:54:04 <peter1138> Maybe we can merge it before nrt
19:54:13 *** supermop_work_ has quit IRC
19:55:07 <Eddi|zuHause> well, let's ignore that :)
19:59:18 *** supermop_work_ has joined #openttd
20:01:25 *** supermop_work__ has joined #openttd
20:03:12 <Eddi|zuHause> next one: "5bb79b1f46f2428f6416ad6d85b0e903ca9629b4 is the first bad commit"
20:03:17 <Eddi|zuHause> another rename commit
20:04:31 <Eddi|zuHause> that conflict doesn't seem to have anything to do with that?
20:05:02 *** supermop_work___ has joined #openttd
20:05:43 *** andythenorth is now known as Guest4446
20:05:43 *** andythenorth has joined #openttd
20:06:10 <andythenorth> some way to built train stations in cities :P
20:06:16 <andythenorth> built / build /s
20:06:54 *** supermop_work has joined #openttd
20:08:23 *** supermop_work_ has quit IRC
20:08:39 <Eddi|zuHause> ok, i think that was right... but i'm not sure
20:12:23 *** supermop_work__ has quit IRC
20:13:06 *** supermop_work___ has quit IRC
20:16:42 <nielsm> have anyone tried to solve the problem of income distribution for transferred cargo by using a chain of only partially resolved cargo packets that builds up until the cargo reaches the final destination?
20:17:16 <nielsm> or other way of keeping track of every leg of the route travelled
20:19:35 *** supermop_work_ has joined #openttd
20:22:06 <andythenorth> diagonal canals? o_O
20:22:22 *** supermop_work__ has joined #openttd
20:22:30 <Eddi|zuHause> nielsm: problem will be that loading/unloading may split and join cargopackets
20:23:32 <Eddi|zuHause> (and memory consumption will be multiplied by average transfer chain length)
20:23:57 <Samu> i think the rail station is uglier
20:24:19 <Eddi|zuHause> andythenorth: not sure what your problem is
20:24:32 <Eddi|zuHause> other than the raised landscape
20:24:38 <Eddi|zuHause> and the stub river
20:25:12 *** gareppa has joined #openttd
20:25:15 <andythenorth> ugly square canal
20:26:03 <peter1138> Gu for desert tonight
20:27:48 <peter1138> Hmm, okay, why are trams available in the scenario editor?
20:29:38 *** supermop_work_ has quit IRC
20:30:50 <Samu> peter1138, 7424 + 7429, passed the assert, intercontinental opengfx+ newgrf
20:30:50 <Eddi|zuHause> the question should be: why not rails?
20:31:01 <Eddi|zuHause> and a company switch?
20:31:27 <Eddi|zuHause> similar to gamescripts, scenario editor user should be able to build as "world" or as "company <X>"
20:31:43 <LordAro> Eddi|zuHause: traditionally scenarios don't contain any companies
20:31:50 <peter1138> Yeah? Well, that isn't the task I am tackling.
20:31:56 <Samu> on a 4096x4096 map, about 12k towns
20:32:04 <LordAro> i imagine it could, but it's not exactly been designed with that in mind
20:32:06 <Eddi|zuHause> LordAro: the orginal featured scenarios with prebuilt infrastructure
20:32:21 <peter1138> Samu, try with low towns on a 4k map :-)
20:32:52 <peter1138> That Gu was very nice but pretty damn small for the amount of carolies
20:32:56 *** supermop_work_ has joined #openttd
20:36:09 <JGR> @nielsm, to answer your earlier question on income distribution, yes, I've implemented this
20:37:35 *** supermop_work___ has joined #openttd
20:38:41 *** supermop_work__ has quit IRC
20:38:58 <Samu> i wonder, how would an AI access those other airports
20:39:09 <Samu> time to run visual studio
20:45:12 *** supermop_work_ has quit IRC
20:45:43 <andythenorth> it was briefly JGR
20:47:15 *** supermop_work has joined #openttd
20:50:00 *** supermop_work_ has joined #openttd
20:54:11 *** supermop_work__ has joined #openttd
20:54:47 *** supermop_work___ has quit IRC
20:57:40 <Samu> so this is how an AI sees the airports
20:57:50 <Samu> the opengfx+ airports thing
20:58:44 <peter1138> I mean, what's the API call?
20:59:15 <Samu> type 9 apparently isn't a valid airport type
20:59:29 <peter1138> Or are you blindly querying airport information?
20:59:54 <peter1138> IsAirportInformationAvailable?
21:00:12 <peter1138> We're on 1.9 now of course.
21:01:02 *** supermop_work_ has quit IRC
21:01:07 <Samu> gonna add that to the list
21:02:10 *** supermop_work_ has joined #openttd
21:02:22 <TrueBrain> glx: that cmake commit, will you backport that to master?
21:03:44 <glx> the warning happens only with cmake, because settingsgen is now 64bit for win64
21:04:45 <Samu> gonna test without opengfx
21:05:00 <TrueBrain> glx: okay, let me rephrase: please backport this to master
21:05:19 <TrueBrain> the smaller the diff with the cmake branch, the better :)
21:05:35 <TrueBrain> having these random things in there only make the review more difficult :)
21:05:45 <TrueBrain> current code is wrong either way :P :D
21:06:11 <Samu> then from 10 to 127 it says it's available
21:06:15 <peter1138> You mean after 8 you get the same?
21:06:24 <peter1138> Yeah, I think that's a bug.
21:07:22 *** supermop_work___ has joined #openttd
21:08:07 <Samu> I wonder what happens if I actually try to place airport type 127 lol
21:08:47 <peter1138> If it is actually allowed, that's another bug.
21:08:53 *** supermop_work__ has quit IRC
21:13:22 <peter1138> Watching trees grow and die in fast-forward is mesmerising.
21:13:35 <TrueBrain> make into a gif? :D
21:13:42 <Eddi|zuHause> if only tree growth had a vison...
21:13:50 <peter1138> Also: too many trees.
21:13:58 <DorpsGek_II> [OpenTTD/OpenTTD] glx22 opened pull request #7432: Fix: [Win64] settingsgen MSVC compile warnings https://git.io/fjTao
21:14:16 *** supermop_work_ has quit IRC
21:14:26 *** supermop_work has joined #openttd
21:15:01 *** andythenorth is now known as Guest4455
21:15:01 *** andythenorth has joined #openttd
21:15:20 <TrueBrain> that would be awesome :D
21:16:00 <TrueBrain> means we can customize andythenorth with actions :D
21:16:10 <andythenorth> content > settings
21:16:26 <andythenorth> TrueBrain: approve NRT pls, thx bai
21:16:39 <peter1138> andythenorth, nah, I found a bug :p
21:16:39 <TrueBrain> sure. You want more pink, or more purple?
21:16:50 <TrueBrain> owh, approve? I was reading improve
21:17:01 <DorpsGek_II> [OpenTTD/OpenTTD] PeterN approved pull request #7432: Fix: [Win64] settingsgen MSVC compile warnings https://git.io/fjTa6
21:18:17 <TrueBrain> without checking if the CI generates warnings :P He has faith in you :D
21:20:27 *** supermop_work_ has joined #openttd
21:20:56 *** supermop_work___ has quit IRC
21:25:24 <DorpsGek_II> [OpenTTD/OpenTTD] TrueBrain merged pull request #7432: Fix: [Win64] settingsgen MSVC compile warnings https://git.io/fjTao
21:28:54 <glx> and what's the bad news ?
21:29:35 <peter1138> Undefined airports are... buildable. They fallback to the first type.
21:30:02 <peter1138> So the airport type restrictions are GUI-only at the moment, I guess.
21:30:28 <peter1138> Hmm, does it let you build that even if the small airport would otherwise be unavailable?
21:31:11 <Samu> let me test in year 2000
21:31:45 <andythenorth> I need to make industries bigger
21:31:48 <andythenorth> just to fit stations around htem
21:33:24 <supermop_work_> andythenorth: i want newgrf trees
21:33:53 <supermop_work_> particularly i want a British Trees set
21:34:10 <supermop_work_> that will never be released despite generating 10 years of bad blood
21:34:39 <andythenorth> or can I have stations with bigger catchment?
21:34:57 *** frosch123 has joined #openttd
21:36:03 <peter1138> Everything spread out further...
21:36:35 <peter1138> Right, I'm guessing trams shouldn't be in the scenario editor, but roads are fair game.
21:38:13 *** supermop_work has joined #openttd
21:38:46 <andythenorth> 5 years into game; all my industries start to close
21:39:00 <andythenorth> I should restore that feature in FIRS :(
21:40:27 *** supermop_work__ has joined #openttd
21:41:07 <supermop_work__> why is the office internet so bad this week
21:42:24 *** supermop_work___ has joined #openttd
21:42:28 <Eddi|zuHause> i thought that was office standard
21:43:50 <andythenorth> these closures :(
21:44:02 <andythenorth> with FIRS Steeltown, the industry chains are about 6 levels deep
21:44:11 <andythenorth> so I planned a network based on industry locations
21:44:16 <andythenorth> and now they're closing :(
21:44:33 <andythenorth> is this just how people normally play? :P
21:45:37 *** supermop_work_ has quit IRC
21:46:36 *** supermop_work_ has joined #openttd
21:48:04 <_dp_> andythenorth, so do you see now why I want no closing setting? :p
21:48:37 <andythenorth> write a special GS
21:48:43 <andythenorth> to use with all the other special GS
21:48:46 <andythenorth> that nobody writes
21:49:10 <andythenorth> can GS make a call to an AWS lambda? o_O
21:49:11 *** supermop_work__ has quit IRC
21:49:20 <_dp_> andythenorth, GS can't quite help with closing :p
21:50:04 <andythenorth> oof I might have to fix this FIRS
21:50:41 <peter1138> Closing after how many years?
21:51:00 <andythenorth> it's behaving as expected
21:51:04 <peter1138> Should've hooked more up by then.
21:51:13 <peter1138> New New Road Vehicle. Hm.
21:51:18 <andythenorth> I solved this years ago in FIRS, but 16-cargos breaks it
21:51:25 <_dp_> btw, technically speaking newgrf can't modify default industries either
21:51:32 <peter1138> Why does it break it?
21:51:37 <_dp_> only define a new set that looks like default and works like default
21:51:55 <andythenorth> it made assumptions about 3 cargos
21:52:01 <andythenorth> needs adjusting for 16
21:52:06 <andythenorth> so I just commented it out
21:52:34 <andythenorth> I figured it would be interesting to play a test game with original closure behaviour also
21:52:45 <andythenorth> it's just annoying mostly :P
21:53:36 *** supermop_work___ has quit IRC
21:53:46 <peter1138> Hmm, AIAI places very odd road stops everywhere.
21:54:51 <andythenorth> 'too far from previous destination'
21:55:19 *** supermop_work has joined #openttd
21:55:34 <peter1138> Ah, AIAI is abusing stationspread :p
21:56:48 <andythenorth> have you nerfed it? o_O
21:57:09 <peter1138> No, it just explains the very odd road stop everywhere.
21:58:49 <Eddi|zuHause> but, did you break it with your new acceptance radius?
21:59:11 <peter1138> No, it'll still work.
21:59:55 <_dp_> will need much more spreading if station size is big enough
22:00:18 <glx> just needs more road stops :)
22:01:10 <peter1138> Yeah basically the "only place two opposite corners" trick no longer works
22:01:21 <peter1138> But you can still stuff a city with road stops to get full catchment
22:02:47 *** supermop_work_ has quit IRC
22:04:34 <Eddi|zuHause> yeah, my question was, does the existing implementation of AIAI use a "sparse" approach that breaks or a "stuffed" approach that is unaffected?
22:05:20 <glx> it probably already handle different catchment settings
22:06:15 <Eddi|zuHause> yeah, but it won't know about this change
22:06:27 <Eddi|zuHause> (unless someone updated the code)
22:06:57 *** supermop_work_ has joined #openttd
22:18:47 <michi_cc> And for the record, the main problem of YACD is/was performance, especially generating cargo packets where source and destination have a station but no valid path between.
22:18:58 *** supermop_work has joined #openttd
22:20:41 <michi_cc> Caching might work, but negative caching is always tricky.
22:22:14 <peter1138> I imagine fitting it in around cargodist is going to be tricky as well
22:24:02 <peter1138> Just got the random hang on exit bug...
22:26:16 *** supermop_work_ has quit IRC
22:27:21 *** supermop_work_ has joined #openttd
22:27:30 <Samu> no, I'm not doing anything atm
22:28:24 <Samu> also, the documentation is kinda misleading :(
22:29:11 <LordAro> Samu: stop complaining and fix it, or at least document that it's an issue
22:29:15 <andythenorth> so does HEQS get merged to Road Hog, or what? :P
22:29:36 <Eddi|zuHause> andythenorth: i'd vote no
22:29:49 <Eddi|zuHause> keep them separate
22:30:19 <Eddi|zuHause> HEQS should be a set of special purpose vehicles that can be used with other vehicle sets
22:30:44 <Eddi|zuHause> no need to vendor-lock HEQS users
22:30:54 <Samu> I think Zuu talked to me once about api stuff code he was preparing for trunk, about these issues
22:31:18 <andythenorth> it's just a bit weird that HEQS can't be used with Road Hog
22:31:41 <peter1138> trunk is now master. It's not over, it's better.
22:32:11 <glx> we can't break things without approval now :)
22:34:01 <Samu> i really wanted a website to see the new functions all in one place
22:34:31 <andythenorth> why would it be updated for 1.9?
22:34:35 <andythenorth> 1.9 isn't released
22:34:37 <Samu> think this trunk, is still the old trunk
22:34:54 <Samu> it doesn't have the new functions
22:35:21 <LordAro> (also available for nightlies)
22:35:55 <LordAro> but yes, someone (read: TB) should work out how to get the noai/nogs/docs sites updated
22:36:45 <peter1138> Poor TB, having stuff dumped on him.
22:36:51 <glx> not many functions are missing I think
22:37:36 <glx> but we don't change API so often
22:38:26 <Samu> enum AIAirport::AirportType is described as: "The types of airports available in the game. "
22:38:34 <Samu> lists the 9 vanilla airports
22:38:48 <Samu> and nothing about possible newgrf airports :o
22:39:25 <glx> but you can still build newgrf airports
22:40:32 <glx> I mean you can't put newgrfs airports in the enum as the index are not constant
22:41:19 <glx> and you can have at most 128 airport types
22:41:39 <glx> if you load more than one newgrf airport
22:41:54 <Samu> then from 24 to 127, are the buggy airports that shouldn't be listed
22:42:08 <glx> depending on the newgrf order index will change
22:43:25 <peter1138> glx, samu is crap at explaining anything
22:43:44 <peter1138> glx, basically, the "out of range" indexes are marked as available, and can even be built.
22:44:02 <peter1138> So if an AI wants to query airport types, it gets a list of 127 of them.
22:44:08 <peter1138> Instead of, say, 10.
22:44:29 <andythenorth> FIRS design is probably too much dictated by neato algorithm for graphviz :P
22:44:36 <peter1138> You could, but what it really needs is just an issue on github.
22:45:01 <peter1138> Samu, thems the rules.
22:45:30 <peter1138> Samu, if you are not going to fix it yourself and make a PR (which is fine, you don't need to) then you really should file an issue instead. Don't just keep moaning about it on here.
22:46:00 <Samu> ok I will create an issue, I don't think i know how to fix this, it's newgrf stuff, which I rarely use/test
22:46:48 <glx> it's not newgrf stuff, it's how AirportSpec::Get(byte type) to always return a valid spec
22:48:51 <andythenorth> yair, FIRS definitely too much designed by 'what makes graphviz look less crap'
22:49:31 *** supermop_work has joined #openttd
22:49:57 *** supermop_work__ has joined #openttd
22:56:27 <DorpsGek_II> [OpenTTD/OpenTTD] SamuXarick opened issue #7433: AI Airports: the "out of range" indexes are marked as available, and can even be built. https://git.io/fjTwi
22:56:52 *** supermop_work_ has quit IRC
22:57:48 <Samu> copy pasted the explanation
22:59:08 <peter1138> I have no idea how to "create an ai with this" :p
22:59:19 <peter1138> But thanks, that's good.
22:59:53 <glx> I'll check that, but not now :)
23:04:20 *** supermop_work has joined #openttd
23:06:43 *** supermop_work_ has joined #openttd
23:08:51 *** supermop_work___ has joined #openttd
23:10:56 *** supermop_work__ has quit IRC
23:11:30 *** supermop_work__ has joined #openttd
23:13:56 *** rocky11384497 has joined #openttd
23:15:03 <glx> you can print the value, but it's 8
23:15:47 *** supermop_work_ has quit IRC
23:15:58 <glx> they are not ordered by value
23:16:21 <Samu> the one before was commuter
23:16:37 <glx> but it's not really important to know the value as there's an enum
23:17:05 <Samu> dang, this 4096x4096 test still got 7 and 8 to go
23:17:36 <glx> 7 intercon, 8 helistation
23:18:21 <glx> but if something works on smaller maps it will usually work on huge ones
23:18:57 *** supermop_work___ has quit IRC
23:19:38 *** supermop_work has joined #openttd
23:20:20 <peter1138> Are you still testing the airportgetnearesttown function for giant maps for all airport types?
23:20:44 <Samu> because I'm making sure perimeter check is enough test
23:21:13 <Eddi|zuHause> are we talking about a function that gets called once on a user click?
23:21:36 <peter1138> Eddi|zuHause, a bit more when AIs test it.
23:21:56 <Samu> talking about 7424 + 7429 combined
23:22:17 <Samu> without 7429, 7424 fails
23:22:27 <Eddi|zuHause> if we're implementing an optimized-for-area kdtree lookup, then it shouldn't matter much whether we're checking the perimeter or the full area
23:23:33 *** supermop_work_ has joined #openttd
23:23:42 <peter1138> It's not, it still loops.
23:25:38 *** supermop_work__ has quit IRC
23:26:21 *** rocky11384497 has joined #openttd
23:32:38 <Samu> just started testing intercontinental... again :(
23:33:04 <Samu> if everything goes according to plan, in 3 hours I'm finished
23:33:54 <Samu> % indicators were a good addition, at least I'm not looking at a blank screen
23:39:43 *** supermop_work_ has quit IRC
23:44:10 *** rocky11384497 has joined #openttd
23:48:32 *** SimYouLater has joined #openttd
23:50:52 <SimYouLater> So, can someone help me figure out how to deal with "RANDOM_LC_YEAR" for the Recycled Infrastructure Set?
23:51:26 <SimYouLater> I can't delete it because it's needed for some of the code I took from Nutracks, but it spits up an error.
23:51:50 <SimYouLater> ERROR: "src/graphics_NBAN_NBAE.pnml", line 92: Syntax error, unexpected token "RANDOM_LC_YEAR" Included from: "ris.pnml", line 2
23:51:52 <Eddi|zuHause> have you tried compiling NuTracks the official way?
23:52:44 <SimYouLater> I just said I'm not trying to compile Nutracks. I'm trying to borrow it's code. RIS is basically "use this instead of 10 other NewGRFs".
23:52:57 <Eddi|zuHause> yes. but have you tried?
23:55:18 <SimYouLater> I have no idea how to complie Nutracks. It has a couple "Makefile" files, but nothing for compiling. Do I just do it with the command line, or what?
23:55:46 <LordAro> you might find something more interesting inside those files
23:55:47 <Eddi|zuHause> there is a command line tool called "make"
23:55:57 <Eddi|zuHause> that will process the "Makefile" files
23:57:12 <SimYouLater> There's a "make" tool in my RIS folder. How do I copy and use it for Nutracks?
23:58:25 <LordAro> that would be... unexpected
23:59:04 <LordAro> what's the current structure of your RIS folder?
23:59:40 <SimYouLater> Just tried to copy it over and edit with notepad. Even after adding the subfolder the pnmls are in, it won't compile.
continue to next day ⏵
