IRC logs for #openttd on OFTC at 2020-06-06
⏴ go to previous day
00:31:00 <DorpsGek_III> [OpenTTD/OpenTTD] glx22 opened pull request #8192: Change: Always generate API files at build time https://git.io/Jf1KV
00:36:16 *** virtualrandomnumber has quit IRC
00:45:06 *** Smedles has joined #openttd
01:25:07 <Speeder_> secondary industries have no output limit?
01:25:15 <Speeder_> whatever you give to them, they will convert to output?
01:28:55 <_dp_> Speeder_, on default industry set yes
01:30:53 <Speeder_> so the global production of "STUF" is basically output of primary industries multiplied by the formulas on secondary industries? (for example, farm that takes 8 grain and output 4 food, thus global production is farms production divided by two)
01:32:12 <_dp_> Speeder_, default industries produce 1:1 for any input so idk what you're talking about here
01:32:43 <Speeder_> _dp_, nml 5 allow you to set how much an industry will output, based on input
01:32:56 <Speeder_> if you are using "default" production rules, that is.
01:36:01 <_dp_> Speeder_, well, anything nml is not the default industry set
01:36:39 <_dp_> Speeder_, so idk what default production rules for nml are but my guess is there is no limit on output, if anything input limit is more likely
01:40:23 <DorpsGek_III> [OpenTTD/OpenTTD] glx22 updated pull request #8192: Change: Always generate API files at build time https://git.io/Jf1KV
01:50:41 <Speeder_> any newgrf adds more landmarks?
01:50:53 <Speeder_> without conflicting with other town-related newgrfs that is
02:19:08 <Speeder_> can newgrf customize what information shows up when you query an industry?
02:35:50 <Eddi|zuHause> it can show additional info, but not hide any
02:42:22 <Speeder_> Eddi|zuHause, can you point me to the docs about that? I tried finding but failed
02:43:24 <Eddi|zuHause> my brain wants to call it CB23 but i'm not sure if that is the correct one... something about extra newgrf text callback
02:44:34 <Eddi|zuHause> right, that makes sense
02:45:41 <glx> extra_text_industry and extra_text_fund in nml
02:49:26 <glx> cargo_subtype_display is CB37, extra_text_industry is CB3A and extra_text_fund is CB38
02:52:33 *** m1cr0man has joined #openttd
04:38:09 <DorpsGek_III> [OpenTTD/OpenTTD] glx22 updated pull request #8192: Change: Always generate API files at build time https://git.io/Jf1KV
04:50:07 <DorpsGek_III> [OpenTTD/OpenTTD] glx22 updated pull request #8192: Change: Always generate API files at build time https://git.io/Jf1KV
04:52:23 *** blathijs has joined #openttd
05:50:02 *** Wormnest has joined #openttd
06:29:08 <DorpsGek_III> [OpenTTD/OpenTTD] techgeeknz dismissed a review for pull request #8184: Cleanup: More code comment and doxygen fixes. https://git.io/Jf1Ts
06:29:08 <DorpsGek_III> [OpenTTD/OpenTTD] techgeeknz updated pull request #8184: Cleanup: More code comment and doxygen fixes. https://git.io/JfXst
06:32:53 *** WormnestAndroid has quit IRC
06:34:22 *** WormnestAndroid has joined #openttd
06:35:00 *** WormnestAndroid has quit IRC
06:36:11 <DorpsGek_III> [OpenTTD/OpenTTD] techgeeknz commented on pull request #8184: Cleanup: More code comment and doxygen fixes. https://git.io/Jf1Ss
06:37:22 *** WormnestAndroid has joined #openttd
06:38:03 *** WormnestAndroid has joined #openttd
06:40:46 <DorpsGek_III> [OpenTTD/OpenTTD] techgeeknz commented on pull request #8009: Change #8001: Don't add docking tile cost when ships are still too far from their destination https://git.io/Jf1SC
06:47:08 *** WormnestAndroid has quit IRC
06:47:25 *** WormnestAndroid has joined #openttd
06:50:41 <DorpsGek_III> [OpenTTD/OpenTTD] techgeeknz commented on pull request #8009: Change #8001: Don't add docking tile cost when ships are still too far from their destination https://git.io/Jf1Su
07:34:21 *** Gustavo6046 has joined #openttd
07:47:37 *** andythenorth has joined #openttd
08:00:47 *** Progman has joined #openttd
08:04:38 *** sla_ro|master has joined #openttd
08:47:41 <andythenorth> 28,448 checks of var 61 per vehicle -> 254 checks
08:47:49 <andythenorth> super optimised now, right? :P
09:52:43 <Samu> what are these .log and .tlog files?
09:55:59 <Samu> it's everything in objs folder
09:58:47 <Samu> there is no visual studio file now? i'm slightly confused
10:05:14 <LordAro> correct, the build system has been replaced with cmake
10:05:33 <LordAro> (which will then generate the visual studio projects)
10:05:42 <LordAro> presumably you'll need to install it
10:09:11 <nielsm> if you have visual studio 2019 you don't need to install anything extra
10:09:17 <nielsm> you just use Open Folder (I think it's called)
10:12:53 <DorpsGek_III> [OpenTTD/OpenTTD] nielsmh commented on pull request #8185: Cleanup: StationCargoList::AreMergable doxygen comment references Veh… https://git.io/Jf1Qg
10:15:57 <DorpsGek_III> [OpenTTD/OpenTTD] nielsmh approved pull request #8184: Cleanup: More code comment and doxygen fixes. https://git.io/Jf1QV
10:17:25 <Samu> do i have to configure CMake? Get started with CMake, Open the CMake Settings Editor
10:19:43 <Samu> openttd is built without zlib support? hmm things are different
10:20:54 <nielsm> I remember having some trouble getting cmake and vcpkg to detect things correctly at first
11:36:11 <andythenorth> it's useless as a GS, assuming you want to use an actual GS
11:39:22 <andythenorth> it also doesn't work :(
11:41:27 <andythenorth> well it does, but date cheat is an unfair test of it :)
11:47:43 <andythenorth> maybe I do it in newgrf
11:48:03 <andythenorth> action D read the parameters from a newgrf lib
11:51:10 <DorpsGek_III> [OpenTTD/OpenTTD] techgeeknz commented on pull request #8185: Cleanup: StationCargoList::AreMergable doxygen comment references Veh… https://git.io/Jf15l
12:07:37 <FLHerne> Speeder_: > any newgrf adds more landmarks?
12:11:48 *** WormnestAndroid has quit IRC
12:12:00 *** WormnestAndroid has joined #openttd
12:12:25 *** WormnestAndroid has joined #openttd
12:12:29 * andythenorth consider removing progression from grfs
12:16:36 *** Smedles_ has joined #openttd
12:26:28 <FLHerne> andythenorth: You keep saying that, it's still a bad idea :p
12:26:56 <andythenorth> should I switch to JGRPP?
12:27:03 <andythenorth> I'd have to re-engineer FIRS
12:27:11 <andythenorth> but if JGRPP is default now...
12:27:31 <LordAro> andythenorth: see FLHerne's last
12:27:50 <andythenorth> pikka has locked progression in UKRS3
12:28:33 <FLHerne> andythenorth: (a) it isn't, (b) why does that require FIRS changes?
12:28:57 <andythenorth> daylength breaks industry production
12:29:12 <andythenorth> because daylength can never work
12:29:48 <andythenorth> I think this Technology Advancement script does everything I (maybe other people too) want
12:29:56 <andythenorth> but it's a GS so no dice eh
12:30:27 <nielsm> assuming the "production: X units of cargo per month" means that it should produce that in a month of calendar time is misinterpreting the game mechanics
12:30:52 <nielsm> the industry production has always been producing X units per 256 ticks
12:31:05 <nielsm> regardless of the period of calendar time those ticks are presented as
12:31:12 <andythenorth> how many parameters do I need for a newgrf to scale intro dates?
12:31:18 <andythenorth> 1 for scale factor
12:31:35 <andythenorth> do I also need a start date, or can I just assume 1860, and do maths if it's a 1960 game start
12:32:13 <nielsm> a start date would be fine I think, but don't make it "real start" but rather make it a point somewhere in the middle
12:32:22 <nielsm> like "diesel transition year"
12:32:35 <nielsm> the year where diesel overtakes steam
12:32:55 <andythenorth> oddly enough, that's 1950, which is very TTD
12:32:55 <nielsm> so steam engines scale backwards from that and diesel scale forward
12:33:07 <Eddi|zuHause> nielsm: for as long as daylength patches have existed, there has been disagreement whether industry production should be scaled or not
12:33:25 <FLHerne> nielsm: Tell that to the developer of the current daylength patch
12:33:42 <andythenorth> Horse covers 1860-2050
12:33:48 <nielsm> my argument against scaling industry production is that it assumes the game has a scale
12:33:50 <andythenorth> call it 2060, and pin to 1960 as epoch?
12:33:57 <nielsm> because it also implies there is a natural scale to distance in the game
12:34:04 <andythenorth> I am bored of setting the date back
12:34:08 <andythenorth> I keep losing my savegames that way
12:34:24 <andythenorth> I attempt to reload newest, and it's not
12:34:43 <nielsm> the game is balanced around being able to transport X amount of cargo Y distance in tiles over T ticks
12:35:00 <Eddi|zuHause> no, the game has many scales, and if you tune one (length of a day in ticks) you have to tune others as well
12:35:00 <andythenorth> I could just be less bad at computering
12:35:01 <nielsm> if you change the amount of cargo produced per tick you break that balance
12:35:45 <Eddi|zuHause> and calling the game "balanced" is even more of a stretch than assuming a scale
12:37:50 <Eddi|zuHause> ultimately it comes down to the fact that different people want different things out of a daylength patch, and some of these things would benefit from scaling industry production, others would not
12:38:19 <andythenorth> I don't want a daylength patch
12:38:23 <andythenorth> I want to freeze time
12:38:35 <Eddi|zuHause> imho, a daylength patch which misses this switch is incomplete
12:40:26 <andythenorth> my preference is to solve this in newgrf
12:40:33 <andythenorth> ~everything is best solved in newgrf
12:40:42 <andythenorth> I just don't know what the implementation should be yet
12:40:54 <nielsm> but you can't send a signal to a newgrf that you want to start or stop the pace of invention
12:41:02 <nielsm> you can only configure it before game start
12:41:15 <andythenorth> I have reload_newgrfs
12:41:25 <andythenorth> and I can change params on a running game
12:41:54 <andythenorth> I was thinking of making a utility newgrf
12:42:04 <andythenorth> which sets a couple of parameters
12:42:16 <Eddi|zuHause> and the way you should talk about "balance" is: a game is "balanced" if it arrives at an equilibrium, and "unbalanced" if it has exploits that scale towards infinity. it is perfectly valid for a switch like daylength or industry production to come to a different equilibrium
12:42:17 <andythenorth> or how about newgrf control of global state?
12:42:32 <andythenorth> 'advance calendar date' could be a newgrf callback
12:44:17 <andythenorth> someone had the idea of fake month names
12:44:37 <andythenorth> just specify 'how many months per calendar year' newgrf
12:44:43 <andythenorth> and set up fake names, like a town name grf
12:46:08 <Eddi|zuHause> can we just ban andy already? :p
12:49:01 <andythenorth> so how do we get an outcome, not an implementation?
12:49:25 <andythenorth> the obsession with daylength confuses outcome and implementation
12:49:57 <andythenorth> I just want to tinker with a train network for 60 game years before all the trains get replaced, not 30
12:50:13 <andythenorth> seeing as I control the trains, the RVs the ships, the industries and the stations, it seems within reach
12:50:32 <andythenorth> I only don't control planes (AV9) and towns
12:53:08 <LordAro> i feel like a proper "sandbox" mode would go a long way to solve this
12:53:51 <LordAro> remap from 1950 to year 1
12:54:00 <Eddi|zuHause> we need to acknowledge that the game hs many scales, and if possible identify and isolate these scales, so each one can be tweaked independently without offsetting too much of the balance
12:56:13 <andythenorth> Eddi|zuHause seems fair
12:56:17 <andythenorth> but also what LordAro said
12:56:44 <andythenorth> preferably with an option for progression, but via explicit research
12:56:57 <LordAro> that's not full sandbox :p
12:57:06 <andythenorth> what eddi said, many scales
12:57:15 <andythenorth> tech tree is a dimension
12:57:31 <andythenorth> full sandbox is one end of the multi-dimensional matrix :P
12:57:42 <andythenorth> everything else is somewhere in the hyperspace
14:06:43 <andythenorth> Iron Horse tar isn't bananas compatible
14:06:49 <andythenorth> always a new bit of housekeeping :)
14:24:51 <EER> Finally got my old mac to compile OpenTTD, and it works perfectly fine (after changing some minor code to fix a compiler issue).
14:25:16 <EER> 1.10.2 however still crashes, but I just tried the latest nightly, and it appears that works fine :)
14:25:38 <EER> 20200604-master-gcdd2892c49-macosx
14:29:22 *** otetede has joined #openttd
14:29:24 *** virtualrandomnumber has joined #openttd
14:32:12 <EER> now trying to build from branch release/1.10, if that does crash, I'm satisfied that either a code change or the buildchain change has solved it
14:41:32 *** WormnestAndroid has quit IRC
14:42:15 *** WormnestAndroid has joined #openttd
14:43:01 <Timberwolf> Does NML/NewGRF have a value for "don't draw anything for this vehicle", or do I need to have a spriteset that is a 1x1 transparent sprite if I want to have a "vehicle" that doesn't have any graphics?
14:43:02 *** WormnestAndroid has quit IRC
14:44:28 * Timberwolf is pondering fixing some glitches with trailers being >8 units in length by adding an extra invisible articulated part on the end.
14:45:41 *** WormnestAndroid has joined #openttd
14:52:13 *** WormnestAndroid has quit IRC
14:52:30 *** WormnestAndroid has joined #openttd
15:19:07 <DorpsGek_III> [OpenTTD/OpenTTD] glx22 updated pull request #8192: Change: Always generate API files at build time https://git.io/Jf1KV
15:27:44 <glx> wow silly, I finally found why compile fails on linux CI
15:29:57 <EER> is that different on CI vs local?
15:31:05 <glx> locally I'm not on linux, and I don't have SDL2 headers
15:32:14 <EER> ah, I'm using WSL but didn't have any issues compiling the current master branch
15:32:43 <FLHerne> Timberwolf: You could try just not defining the graphics callback
15:32:54 <FLHerne> But you might get some default sprite, idk
15:33:14 <Timberwolf> Indeed, I got a nice SH 8P for my troubles when I tried that :)
15:33:49 <EER> Also, on mac, I have tried compiling 1.10.2 and 1.10.0, no problems starting those builds. The release build 1.10.2 did not work, but latest nightly build does, so for now I can only assume that the mac os x (<10.13) crash is resolved in the latest nightly.
15:33:56 <Timberwolf> 1x1 transparant sprite seems to work well enough, it feels a bit hacky but then I guess the whole "your vehicle is longer than 8 units" idea is a bit hacky.
15:35:07 <EER> 20200604-master-gcdd2892c49-macosx <- works on mac os x 10.11.6
15:36:45 <EER> I haven't tested other nightlies so far, combing through the archives was a bit much for me at this point :)
15:37:17 <andythenorth> Timberwolf transparent sprite is accepted solution
15:38:06 <andythenorth> I think if you just do 'graphics: return;' you might get the right outcome, but I don't know if it's wise
15:38:32 <andythenorth> explicit transparent sprite safest
15:39:49 <EER> glx: I'll try those and get back to you :)
15:40:37 <glx> if they both fail it will be a very useful info btw :)
15:41:10 *** frosch123 has joined #openttd
15:41:11 <EER> I was hoping I could contribute a fix, but new information is a good second ;)
15:43:24 <andythenorth> called once, passing up to 16 vehicle IDs
15:43:43 <andythenorth> stores into 3 registers, which can then be used like vars elsewhere
15:44:11 <andythenorth> procedures FTW :D
15:44:20 <andythenorth> so many things now possible without destroying compile time
15:48:05 <EER> glx: both start up. I re-downloaded 1.10.2 again just to verify that I am not entirely crazy, or that something has changed on my machine allowing it to start, and that (luckily) still fails.
15:50:44 <EER> Are both nightly and release versions built on the same server? Perhaps it's an optimization setting?
15:51:07 <glx> it's the exact same build system
15:52:21 <EER> I'm going to try some nightlies with one week intervals, starting from april 1st as that's the release of 1.10
15:56:56 <glx> I can see 2 commits in src/os/macosx from april 10th not included in release
16:00:40 <glx> and some from april 12th and may 21 in src/video/
16:07:23 <andythenorth> is it time for next nml release?
16:07:31 <EER> I just tried a bunch of versions, interestingly: 1.10.0 works, 1.10.1 crashes. So I tried some nightlies in between, all worked: 31/03, 07/04, 08/04, 10/04, 12/04 and 26/04
16:08:26 <EER> right now I'm compiling the version that was released as 1.10.1, I doubt that will crash, but if it's something in the code rather than build environment hopefully it's reproducible
16:08:43 <glx> but 1.10.0 is supposed to fail
16:08:55 <EER> then perhaps I am having a different crash issue than the one reported
16:10:16 <EER> [00] openttd 0x00000001098499c8 (SelectCompanyLiveryWindow::GroupN ameSorter(Group const* const&, Group const* const&) + 68168)
16:10:37 <glx> yeah looks like the 1.10.0 crash
16:10:54 <glx> (and it's useless corrupted stack)
16:13:12 *** sla_ro|master has joined #openttd
16:13:42 <EER> I could attach lldb to the crashing process? I tried before but because it didn't have symbols I didn't really look into that any further.
16:16:23 <glx> I fail to see how 1.10.0 works for you while 1.10.1 fails, the changelog between 1.10.0 and 1.10.1 is so short and contains nothing related
16:19:41 <EER> Just retried, and 1.10 just works ... My only guess would be "something changed in the build environment", but even that doesn't make sense since nightlies continue to work.
16:21:40 <glx> but 1.10.0 fails on 10.2.8 and all nigthlies since #8023
16:22:05 *** Wormnest has joined #openttd
16:22:07 <EER> I am running on Mac OS 10.11.6, so that might be another difference with the reporters
16:25:26 <glx> build env changed in #8028 but nigthlies built between this change and #8023 works on 10.12.8
16:25:43 <glx> and they start to fail with #8023
16:28:42 <Speeder_> FLHerne, I meant stuff like parks, capsule hotel "office" building, etc...
16:29:32 <EER> Perhaps I have something dev-related installed that allows these versions to run?
16:31:19 <EER> my home-compiled 1.10.1 also works btw, there is one change I did, which was necessary to workaround an issue with clang:
16:32:09 <EER> I really doubt that this would make the difference between a working and crashing setup, especially since the nightlies work on my mac.
16:33:52 <glx> yeah it seems specific to 10.12
17:13:05 <FLHerne> Speeder_: You can use HOUSE_FLAG_PROTECTED to prevent a building being replaced
17:14:15 <FLHerne> Speeder_: If you want to ensure there's only one, set HOUSE_FLAG_CHURCH or HOUSE_FLAG_STADIUM (only one of each can exist per town)
17:14:54 <FLHerne> Speeder_: Or set building_class and check for it in construction_check
17:15:12 <FLHerne> (or just check by id)
17:36:35 *** Gustavo6046 has joined #openttd
17:55:56 <Speeder_> FLHerne, thus, I am asking if any newgrf add more landmarks like that ;)
17:56:12 <Speeder_> the "improved town layouts" one use existing ones, don't add any
18:03:05 <FLHerne> I think it's UK Town Set that has some very nice churches
18:10:33 <FLHerne> Actually, I think they might be Swedish
18:10:46 <Speeder_> those look quite good
18:17:07 <FLHerne> Agh, or are they TaI?
18:17:15 <FLHerne> Too many town grfs :p
18:18:38 <andythenorth> oh that's interesting
18:18:48 <andythenorth> nmlc with #155 fails on FIRS
18:19:53 <andythenorth> I'm getting a repro
18:20:31 <glx> nielsm: when you tried clang in VS did it warn a lot ?
18:22:05 <andythenorth> doesn't happen with nml master, just 155
18:22:32 <glx> andythenorth: you know this is almost useless without -s
18:22:59 <andythenorth> hangon I'll re-paste
18:23:50 <andythenorth> hmm it's hung when it should be printing stacktrace
18:24:19 <glx> but looking at the line I have an idea of what the error is
18:24:53 <andythenorth> wow I think nml has crashed my mac
18:25:03 <andythenorth> got a runaway kernel process using 1000% of CPU now
18:26:39 <glx> nielsm: I guess I need to build all libs with clang too
18:31:20 <nielsm> glx yeah I remember having some big issues making vcpkg use clang for build
18:31:21 <EER> glx: found the cause of the crash on my mac (but not a solution). It may or may not be the same as others have reported. When stepping through the release version with the debugger I saw that the crash is accompanied by the message "EXC_BAD_INSTRUCTION", the instruction that fails is `PINSRQ` which appears to be an SSE4 instruction and my processor does not support SSE4 yet.
18:32:36 <Eddi|zuHause> so messed up SSE4 detection?
18:32:56 <glx> detection is done on build
18:33:04 <nielsm> more like the compiler might just be emitting SSE4 opcodes for float stuff
18:33:18 <EER> yeah, that explains why my own builds of the same code always work
18:34:02 <EER> and I suppose the nightly builds are not optimized so perhaps don't contain these instructions? If that's the case, then my crash is definitely not the same as the mac 10.12 users are having.
18:34:02 <glx> and detection is used only for blitters, we don't set any SSE flags for the rest of the source
18:34:03 <Eddi|zuHause> then why would that be solved by upgrading to 10.13?
18:34:49 <EER> Eddi, I don't think it's the same issue. I was hoping, but it looks like my mac is just too old to have the same problem. May be something 10.12 specific
18:37:25 <glx> andythenorth: FIRS master is enough to test ?
18:39:58 <glx> of course deps are still broken, so manual make clean required
18:44:19 * andythenorth finds all the bugs :)
18:44:57 <EER> Just verified and my mac does not support an upgrade to 10.12, if it was my own mac I would happily try to upgrade anyway, but since I "borrowed" it from the office, I'd like to not accidentally brick it. So that concludes my mac saga for this week :)
18:46:02 <glx> andythenorth: crash confirmed :)
18:47:06 <Eddi|zuHause> glx: if deps are wrong, make -B :)
18:47:20 *** virtualrandomnumber has quit IRC
18:53:08 <nielsm> glx: so should we fix things to make clang-cl compile cleanly?
18:54:07 <glx> but for now it doesn't even link for me ;)
18:55:11 <glx> I tried mingw64 targets too, but a fix in our cmake files is needed
18:56:09 <glx> mingw64 works fine from command line, but not from VS because some assumptions are made
18:56:26 <FLHerne> EER: It's not as if your crash *isn't* worth fixing
18:57:50 <EER> FLHerne: I agree, but it's quite certainly not the crash that other people are having :(
19:00:18 <DorpsGek_III> [OpenTTD/OpenTTD] James103 commented on issue #8076: Odd Cursor Glitch When Moving Through Windows and Game https://git.io/JvhZm
19:06:58 <DorpsGek_III> [OpenTTD/OpenTTD] erenes opened issue #8194: Releases from version 1.10.1 no longer work on mid-2007 iMac https://git.io/JfMJz
19:09:19 <nielsm> microsoft: "fopen() is deprecated because it's unsafe, use fopen_s() instead (which is not part of any defined standard)"
19:10:42 <nielsm> also microsoft: "do not use stricmp(), it is not a function defined in any standard, instead use _stricmp() which is an allowed name for it"
19:10:46 <glx> hmm I think we silence those in stdafx
19:11:00 <nielsm> yeah for some reason when building with clang those are ignored
19:11:41 <nielsm> nope, clang-cl does defined _MSC_VER
19:12:20 <nielsm> oh, managed to shut it up
19:13:36 <nielsm> also the abort build button does not work for some reason
19:15:34 <nielsm> yeah that does nothing
19:20:02 <glx> if I could merge #8193 before nightly it would be nice
19:27:47 <DorpsGek_III> [OpenTTD/OpenTTD] nielsmh opened pull request #8195: Fix: Silence some warnings when building with clang-cl on VS 2019 https://git.io/JfMUv
19:39:11 <nielsm> I need to update my local commit checker scripts... they complain about the exact opposite thing with preprocessor indentation
19:40:40 *** arikover has joined #openttd
19:43:01 <DorpsGek_III> [OpenTTD/OpenTTD] nielsmh updated pull request #8195: Fix: Silence some warnings when building with clang-cl on VS 2019 https://git.io/JfMUv
19:43:30 <glx> andythenorth: should be ok now
19:50:52 <andythenorth> glx confirmed working, thanks :D
19:51:13 <glx> was again a stupid error I made :)
19:54:11 <andythenorth> happy to help test :D
19:54:26 <andythenorth> if we built grfs on github actions
19:54:39 <andythenorth> and we built nmlc on github actions (oh wait...we do)
19:54:46 <andythenorth> we could build the grfs with a build of nmlc :P
19:55:29 <andythenorth> probably need a gate so it doesn't try to build grfs with nml head that are known to fail already in nml release (due to broken grf)
20:00:17 <andythenorth> or I can just build them locally :)
20:05:07 <Eddi|zuHause> <andythenorth> if we built grfs on github actions <-- if i moved CETS to github, would you set that up? :p
20:05:21 <andythenorth> I'd try to avoid it :)
20:05:33 <andythenorth> I am quite happy just making newgrfs at the moment
20:05:47 <andythenorth> I did a lot of tooling stuff this last year or so
20:05:57 <andythenorth> also, probably better at making grfs :P
20:06:11 <andythenorth> I did set up FIRS on Azure Pipelines
20:06:19 <Eddi|zuHause> build on push hasn't worked for years, in part because it failed to get eddi-nml
20:53:44 *** arikover has joined #openttd
20:57:53 <milek7> SDL autoconfig is broken
20:58:22 <nielsm> is it on the issue tracker?
20:58:29 <LordAro> i thought it was fixed
20:59:23 <milek7> -- SDL2 found -- -DWITH_SDL2 -- --
20:59:27 <milek7> /home/milek7/OpenTTD/src/os/unix/unix.cpp:28:10: fatal error: SDL.h: Nie ma takiego pliku ani katalogu
21:00:26 <LordAro> sounds like something weird with your system
21:01:25 *** trendynick has joined #openttd
21:04:29 <milek7> $ pkg-config --cflags sdl2
21:04:29 <milek7> -I/usr/include/SDL2 -D_REENTRANT
21:04:47 <glx> that's not what use cmake
21:05:07 <glx> cmake use a cmake script provided by sdl2
21:14:33 <milek7> target_link_libraries(openttd SDL2::SDL2)
21:15:08 <glx> target is created by the sdl2 script
21:17:41 <glx> it's probably a bug in the distrib
21:23:06 <milek7> there's no sdl2-config.cmake, but there is SDL2Targets.cmake
21:26:32 <Speeder_> al lcargos use same unit, and other units are just display effects?
21:26:51 <Speeder_> for example, 1oil in the nml code is same as 1 goods or whatever?
21:27:01 <Speeder_> but in game might show as 1000 litres, and 1 goods show as 1 ton?
21:27:29 <glx> milek7: in src/CMakeFiles.list, line 160 add "TARGET SDL2::SDL2" to link_package()
21:32:11 *** gelignite has joined #openttd
21:33:31 <Speeder_> another question then: there are any standard for amount of cargo in a generic train wagon?
21:34:19 <nielsm> no, but there are multipliers for how refittable cargos fit in space
21:35:01 <nielsm> e.g. a wagon that can refit between wood, goods, and mail, will hold 2x as many units of goods as units of wood, because goods has a 2x multiplier
21:35:12 <nielsm> and mail iirc. has a 4x multiplier
21:36:41 <Speeder_> and popular newgrfs are how c lose t o that?
21:38:31 <nielsm> most stick to having between 20 and 40 units of cargo per car
21:39:20 <nielsm> some having shorter cars (especially in earlier eras) with less capacity and bigger cars (later on) with more capacity
21:42:27 <frosch123> hmm, creating source packages takes longer than compiling :p
21:43:38 <andythenorth> what nielsm said
21:43:46 <andythenorth> 8/8 wagon is ideally 40
21:44:42 <andythenorth> length of vehicle
21:46:36 <DorpsGek_III> [OpenTTD/OpenTTD] frosch123 commented on pull request #8195: Fix: Silence some warnings when building with clang-cl on VS 2019 https://git.io/JfMI6
21:47:36 <FLHerne> Speeder_: Vehicle lengths are in 16ths of a tile, where the default vehicles are half a tile long
21:48:54 <Speeder_> so 8/8 is half a tile?
21:49:03 <Speeder_> thus you can fit theoretically, 80 cargo per tile?
21:49:22 <andythenorth> it's roughly what default game does
21:49:32 <Speeder_> any idea what is the most popular train size?
21:49:54 <Speeder_> I wanna figure out how many cargo you can haul, total, with the default 500 train limit, using typical sized trains.
21:50:18 <andythenorth> pure guessing, but I'd think TL5 or TL6
21:50:35 <andythenorth> spectate on the reddit servers if you want a better sample
21:50:38 <frosch123> ok, "bundle_source" is completely broken. it bundles the "build" directory
21:50:47 <frosch123> s/bundle_source/package_source/
21:51:16 *** virtualrandomnumber has joined #openttd
21:52:08 <_dp_> TL7 ofc, that's the default
21:54:05 <glx> ok an installer is generated for windows nightly, probably not a good idea
21:54:26 <FLHerne> Speeder_: If you're trying to balance stuff, I'd say that FIRS has maybe 30% more production than would be ideal for a 'realistic' game
21:54:53 <FLHerne> And the default industry production can grow to ludicrous levels
21:55:33 <Speeder_> FLHerne, thing is... how much would be ideal for a 'realistic' game?
21:56:14 <FLHerne> The nice thing about FIRS' supply thing is that you can decide how much you want
21:56:24 <FLHerne> But I find the base production a bit high
21:56:52 <FLHerne> I think ~100t/month total for a primary industry would be ideal for what I want to do
21:57:08 <FLHerne> Of course, everyone plays in different ways, so ymmv
22:02:09 <DorpsGek_III> [OpenTTD/OpenTTD] frosch123 opened pull request #8196: Fix: Adjust .gitignore to CMake build system. https://git.io/JfMLt
22:24:21 *** arikover has joined #openttd
22:24:21 *** sla_ro|master has joined #openttd
22:24:21 *** Progman has joined #openttd
22:24:21 *** m1cr0man has joined #openttd
22:24:21 *** Hobbyboy has joined #openttd
22:24:21 *** johanna[m] has joined #openttd
22:24:21 *** Alkel__U3 has joined #openttd
22:24:21 *** coulomb.oftc.net sets mode: +v glx
22:24:34 *** cyberjunkie[m] has quit IRC
22:24:34 *** ad5twoknebor[m] has quit IRC
22:24:35 *** WormnestAndroid has quit IRC
22:24:35 *** mindlesstux6662 has quit IRC
22:24:35 *** supermop_Home_ has quit IRC
22:24:35 *** virtualrandomnumber has quit IRC
22:24:37 *** Alkel__U3 has joined #openttd
22:24:37 *** johanna[m] has joined #openttd
22:24:37 *** Hobbyboy has joined #openttd
22:24:37 *** m1cr0man has joined #openttd
22:24:37 *** Progman has joined #openttd
22:24:37 *** sla_ro|master has joined #openttd
22:24:37 *** arikover has joined #openttd
22:24:37 *** daspork has joined #openttd
22:24:37 *** DorpsGek_III has joined #openttd
22:24:37 *** mikegrb has joined #openttd
22:24:37 *** TrueBrain_ii has joined #openttd
22:24:37 *** TrueBrain has joined #openttd
22:24:37 *** ircer[m] has joined #openttd
22:24:37 *** jeeg[m] has joined #openttd
22:24:37 *** menelaos[m] has joined #openttd
22:24:37 *** rudolfs[m] has joined #openttd
22:24:37 *** dwfreed has joined #openttd
22:24:37 *** Eddi|zuHause has joined #openttd
22:24:37 *** Maarten has joined #openttd
22:24:37 *** Extrems has joined #openttd
22:24:37 *** patrick[m]1 has joined #openttd
22:24:37 *** albert[m] has joined #openttd
22:24:37 *** natalie[m]1 has joined #openttd
22:24:37 *** joey[m] has joined #openttd
22:24:37 *** BekJut[m] has joined #openttd
22:24:37 *** tonyfinn[m] has joined #openttd
22:24:37 *** michi_cc has joined #openttd
22:24:37 *** reticulum.oftc.net sets mode: +vovv glx orudge orudge michi_cc
22:24:37 *** LordAro has joined #openttd
22:24:37 *** Timberwolf has joined #openttd
22:24:37 *** greeter has joined #openttd
22:24:37 *** grossing has joined #openttd
22:24:37 *** DecapitatedO has joined #openttd
22:24:37 *** lastmikoi has joined #openttd
22:24:37 *** grag[m] has joined #openttd
22:24:37 *** ciet[m] has joined #openttd
22:24:37 *** paulus[m] has joined #openttd
22:24:37 *** Heiki[m] has joined #openttd
22:24:37 *** dekeract[m] has joined #openttd
22:24:37 *** magdalena[m] has joined #openttd
22:24:37 *** Webster has joined #openttd
22:24:37 *** XeryusTC has joined #openttd
22:24:37 *** planetmaker has joined #openttd
22:24:37 *** Hazzard has joined #openttd
22:24:37 *** V453000 has joined #openttd
22:24:37 *** Hirundo has joined #openttd
22:24:37 *** ^Spike^ has joined #openttd
22:24:37 *** leward[m] has joined #openttd
22:24:37 *** cyberjunkie[m] has joined #openttd
22:24:37 *** ad5twoknebor[m] has joined #openttd
22:24:37 *** ZirconiumX has joined #openttd
22:24:37 *** Terkhen has joined #openttd
22:24:37 *** FLHerne has joined #openttd
22:24:37 *** ericnoan has joined #openttd
22:24:37 *** julie[m] has joined #openttd
22:24:37 *** robert[m]1 has joined #openttd
22:24:37 *** gooodger has joined #openttd
22:24:37 *** fonsinchen has joined #openttd
22:24:37 *** TinoDidriksen has joined #openttd
22:24:37 *** KenjiE20 has joined #openttd
22:24:37 *** y2kboy23 has joined #openttd
22:24:37 *** luaduck has joined #openttd
22:24:37 *** reticulum.oftc.net sets mode: +oov planetmaker Terkhen Terkhen
22:24:37 *** supermop_Home_ has joined #openttd
22:24:37 *** Lejving has joined #openttd
22:24:38 *** Speeder_ has joined #openttd
22:24:38 *** mindlesstux6662 has joined #openttd
22:24:38 *** innocenat_ has joined #openttd
22:24:38 *** blathijs has joined #openttd
22:24:38 *** andythenorth has joined #openttd
22:24:38 *** Smedles_ has joined #openttd
22:24:38 *** WormnestAndroid has joined #openttd
22:24:38 *** frosch123 has joined #openttd
22:24:38 *** Wormnest has joined #openttd
22:24:38 *** Gustavo6046 has joined #openttd
22:24:38 *** gelignite has joined #openttd
22:24:38 *** virtualrandomnumber has joined #openttd
22:25:00 *** cyberjunkie[m] has quit IRC
22:25:00 *** ad5twoknebor[m] has quit IRC
22:25:02 *** gelignite has joined #openttd
22:25:02 *** blathijs has joined #openttd
22:25:02 *** Lejving has joined #openttd
22:25:02 *** luaduck has joined #openttd
22:25:02 *** KenjiE20 has joined #openttd
22:25:02 *** gooodger has joined #openttd
22:25:02 *** robert[m]1 has joined #openttd
22:25:02 *** ericnoan has joined #openttd
22:25:02 *** FLHerne has joined #openttd
22:25:02 *** Terkhen has joined #openttd
22:25:02 *** ZirconiumX has joined #openttd
22:25:02 *** ad5twoknebor[m] has joined #openttd
22:25:02 *** cyberjunkie[m] has joined #openttd
22:25:02 *** ^Spike^ has joined #openttd
22:25:02 *** Hirundo has joined #openttd
22:25:02 *** V453000 has joined #openttd
22:25:02 *** Hazzard has joined #openttd
22:25:02 *** planetmaker has joined #openttd
22:25:02 *** XeryusTC has joined #openttd
22:25:02 *** Webster has joined #openttd
22:25:02 *** magdalena[m] has joined #openttd
22:25:02 *** dekeract[m] has joined #openttd
22:25:02 *** Heiki[m] has joined #openttd
22:25:02 *** paulus[m] has joined #openttd
22:25:02 *** ciet[m] has joined #openttd
22:25:02 *** grag[m] has joined #openttd
22:25:02 *** lastmikoi has joined #openttd
22:25:02 *** DecapitatedO has joined #openttd
22:25:02 *** grossing has joined #openttd
22:25:02 *** liquid.oftc.net sets mode: +ovo Terkhen Terkhen planetmaker
22:25:21 *** virtualrandomnumber has quit IRC
22:25:37 *** virtualrandomnumber has joined #openttd
22:25:37 *** frosch123 has joined #openttd
22:25:37 *** andythenorth has joined #openttd
22:25:37 *** TinoDidriksen has joined #openttd
22:25:37 *** fonsinchen has joined #openttd
22:25:37 *** Timberwolf has joined #openttd
22:25:37 *** LordAro has joined #openttd
22:25:37 *** michi_cc has joined #openttd
22:25:37 *** tonyfinn[m] has joined #openttd
22:25:37 *** BekJut[m] has joined #openttd
22:25:37 *** joey[m] has joined #openttd
22:25:37 *** natalie[m]1 has joined #openttd
22:25:37 *** albert[m] has joined #openttd
22:25:37 *** patrick[m]1 has joined #openttd
22:25:37 *** Eddi|zuHause has joined #openttd
22:25:37 *** rudolfs[m] has joined #openttd
22:25:37 *** menelaos[m] has joined #openttd
22:25:37 *** jeeg[m] has joined #openttd
22:25:37 *** DorpsGek_III has joined #openttd
22:25:37 *** dacia.oftc.net sets mode: +vov michi_cc orudge orudge
22:31:06 *** ChanServ sets mode: +v planetmaker
22:34:50 <glx> frosch123: I think I'll add more stuff to your PR if you don't mind
22:36:04 <frosch123> i am currently making "package_source" work, FYI
22:41:44 <andythenorth> 'Improved Town Layouts'
22:45:07 <supermop_Home_> the executive livery freight is nice
22:48:35 <supermop_Home_> now this is a train
22:49:23 <supermop_Home_> not familiar with the livery on the 2nd locomotive
22:49:49 <glx> hmm should we include a default CMakeSettings.json ?
22:50:11 <andythenorth> supermop_Home_ 'dutch' - nominally dedicated to civil engineering trains
22:50:19 <andythenorth> supposedly on reduced maintenance
22:50:41 <andythenorth> looks quite cool on most engines, not all
22:51:34 <supermop_Home_> i'd like to get something N scale large logo
22:51:43 <supermop_Home_> never got super into 37s
22:51:54 <supermop_Home_> but i do find them very british looking
22:52:27 <supermop_Home_> to my mind i always think of large logo on 47s
22:52:39 <andythenorth> I am trying to limit purchases
22:52:43 <andythenorth> due to space in my house :P
22:53:00 <supermop_Home_> i had that problem with sneakers a couple months ago
22:53:11 <supermop_Home_> danger of wife getting a job at nike
22:53:55 <andythenorth> oof, I have a pre-montreal waffle obsession
22:53:58 <andythenorth> so many colourways
22:54:54 <supermop_Home_> the incongruity of the 'old fashioned' nose of a 37 etc with the supergraphic boldness of Large Logo is a nice wabi-sabi
22:55:18 <andythenorth> the black roof is best
22:55:25 <andythenorth> and the split box front
22:55:40 <supermop_Home_> well now that everyone is getting into 90s air maxes suddenly its like a new colorway ever fortnight that feels needed
22:56:39 <supermop_Home_> i could go for some large logo Cortezs
22:56:44 <andythenorth> I have stopped buying trainers as I have stopped wearing shoes :P
22:56:58 <supermop_Home_> i was a bit slow on that andy
22:57:12 <supermop_Home_> the first month of lockdown i still bought a couple
22:57:41 <supermop_Home_> then had to come to terms with the fact that i am not wearing any
22:58:09 <supermop_Home_> and not wearing any out as i no longer spend the weekends walking 8-12 miles around the island
22:58:25 <andythenorth> I am not making scenic model railways (it's an oval on bare wood)
22:58:46 <supermop_Home_> maybe railfreight Air Max 90s
22:59:13 <supermop_Home_> i don't have room for an oval
22:59:37 <supermop_Home_> i just have two tracks and some switches, about 4 feet long at my parents
22:59:45 <andythenorth> railfreight trainers
22:59:49 <andythenorth> the crossover needs to happen
23:00:12 <supermop_Home_> with a 225, 158, and 08 on them
23:00:41 <supermop_Home_> here i could manage maybe a platform or fiddle yard on one vitsoe shelf at most
23:02:02 <supermop_Home_> maybe a pair of GNER AM97s
23:06:43 <FLHerne> andythenorth: It's nice compared to the base game, but all a bit bland
23:06:53 <FLHerne> Pikka's TaI is much nicer :-)
23:07:12 <andythenorth> I like the monotony of the suburbs :)
23:07:40 <FLHerne> andythenorth: Also, TaI proves that you *can* control town growth rate/style in NewGRF :-)
23:07:42 <andythenorth> I do normally use TaI
23:07:55 <FLHerne> Even if it's in a rather roundabout way
23:08:02 <supermop_Home_> i do like improved town layouts
23:08:04 <andythenorth> well that's the question eh :P
23:08:18 <andythenorth> do I change FIRS to include a house set
23:08:31 <FLHerne> No, FIRS is way too big already
23:08:53 <FLHerne> TBH, I'd seriously suggest distributing FIRS as separate grfs
23:09:01 <andythenorth> nah, that's just confusing
23:09:10 <andythenorth> I've considered it, but it wins nothing
23:09:13 <FLHerne> My suspicion is that a *lot* of the player base don't know about parameters
23:09:33 <nielsm> you will make the compiled GRF simpler and the user interface simpler
23:09:41 <andythenorth> it would be trivial to compile 5 grfs
23:09:44 <nielsm> and possibly win performance by being able to simplify the callbacks
23:09:54 <FLHerne> Also, you could have 'FIRS Core' and then sector-specific addons like ECS, and I really like that model
23:10:05 <andythenorth> I'd just write the same grf 5 times, with a hard-coded economy parameter
23:10:30 <FLHerne> In theory, nmlc should be able to optimize that out
23:10:31 <andythenorth> there's no other benefit to splitting them, they're not recombinable
23:10:57 <andythenorth> when originally developed 'not being confusing separate grfs' was a much requested FIRS attribute
23:11:13 <andythenorth> I'm potato / potato about it currently
23:11:37 <FLHerne> Hm, and I /still/ need to bake a cake before tomorrow :-/
23:11:41 <andythenorth> more likely is splitting up Horse when I do a US version
23:11:48 <andythenorth> Horse is still so sloooooow to compile
23:11:50 * FLHerne should stop procrastinating, and find a whisk :p
23:12:37 <andythenorth> maybe I should redesign FIRS
23:12:42 <DorpsGek_III> [OpenTTD/OpenTTD] glx22 updated pull request #8196: Fix: Adjust .gitignore to CMake build system. https://git.io/JfMLt
23:12:52 <andythenorth> the 'deliver cargos to towns' thing really doesn't work at all
23:13:19 <FLHerne> I don't think it really needs to, it's just a nice sink
23:13:19 <Timberwolf> A full rebuild of Road Vehicles and HD when I've changed something in the renderer and thus 0 sprites are cached again...
23:13:22 <andythenorth> and the only effect it could have is growing popn.
23:13:28 <andythenorth> and growing popn. is undesirable
23:13:30 <Timberwolf> That ends up being a "walk away, come back later" job.
23:13:38 <supermop_Home_> has anyone tried to make enamel pins of railfreight sector logos?
23:13:45 <andythenorth> supermop_Home_ probably exist
23:13:51 <andythenorth> Timberwolf how long does it take to compile?
23:14:11 <supermop_Home_> i need a boonie or bucket hat, maybe with a nice logo on it
23:14:12 <Timberwolf> On the desktop it's a few minutes, on my old laptop it gets closer to 20.
23:14:15 <FLHerne> Timberwolf: Hm, does your version of nmlc have the C lz77 module functioning?
23:14:24 <Timberwolf> Most of the time is rebuilding the sprite cache.
23:14:25 <FLHerne> The sprite encoder shouldn't be *that* slow
23:14:30 <andythenorth> is it also rendering + compositing the sprites from CGI?
23:14:43 <Timberwolf> Not sure, I have the all-in-one Windows download.
23:14:49 <Timberwolf> That's just the NML times.
23:15:01 <andythenorth> it's 32bpp extra zoom?
23:15:10 <Timberwolf> GoRender is pretty fast, although I added a bunch of higher-quality options to slow it down again.
23:15:20 <FLHerne> Hm, I committed a patch that would tell you, but it's not in a released version yet...
23:15:24 <Timberwolf> Yeah, the regular one isn't too bad (2x only), HD has the 4x sprites.
23:15:47 <andythenorth> I found that using multi-processing really does make a difference with PIL graphics processing
23:15:57 <andythenorth> generally, processing sprites is very parellelisable
23:16:01 <andythenorth> but nmlc can't do that
23:16:13 <Timberwolf> I don't think I'll do anything else at 4x, the difference using base game graphics is just too much at maximum zoom.
23:16:20 <supermop_Home_> incessant helicopter noise
23:17:45 <Timberwolf> Also when I first did 4x I hadn't figured out a lot of things about a voxel-based workflow, so I was losing a lot of detail at 2x there was no technical reason I couldn't have.
23:18:03 <glx> Timberwolf: what says "nmlc --version" ?
23:18:53 <Timberwolf> 0.5.2 \n Library versions encountered: \n PIL: 7.1.2 \n PLY: 3.11
23:20:17 <glx> ah right the native or not status is not shown in 0.5.2
23:20:31 <FLHerne> Yeah, we need to do 0.5.3 soonish :p
23:21:01 <glx> it should at least include #156
23:21:19 <glx> #155 can wait more testing/breaking by andy
23:22:03 <andythenorth> I'm all out of cases
23:22:08 <andythenorth> but we should ship what we have
23:22:28 <Timberwolf> I think a lot is the fault of the laptop, it desperately needs either an SSD or replacing.
23:22:35 <FLHerne> I feel like perhaps we should make a stable 0.5.x branch before merging the various feature/mass-code-shuffle patches
23:22:38 <andythenorth> we won't find some bugs unless we push releases
23:22:51 <FLHerne> btw I'm sorry I've not looked at 155/6 yet, busy :-/
23:22:59 <andythenorth> does pypi have any convention of rc releases?
23:23:14 <glx> #156 is a fix, #155 is a feature
23:24:30 <andythenorth> I'm probably the only newgrf author who's checking out PRs to test things
23:24:31 <glx> I think a branch is needed when compatibility is broken
23:24:45 <andythenorth> we need more exposure :)
23:25:38 <glx> anyway there's already a major change ready to ship, the 0.4 compatibility :)
23:25:52 <FLHerne> Eh, that's basically a fix :p
continue to next day ⏵
![https://i.imgur.com/0JhhRB6.png[/img]](https://i.imgur.com/0JhhRB6.png%5B/img%5D){kind=link}
