IRC logs for #openttd on OFTC at 2020-02-16
⏴ go to previous day
00:02:38 <frosch123> andythenorth: 4a is fine, it's also used for railtype
00:03:07 <frosch123> if the names road/tram/rail-type are too specfic, think of them as tracktype
00:03:15 <andythenorth> I understand that :)
00:03:26 <andythenorth> but the names I chose in the nml patch are inappropriate
00:03:29 <andythenorth> or we need a guard
00:04:02 <andythenorth> tracktype would have been better maybe?
00:04:04 <frosch123> anyway, if someone figures out whether it is possible to use the additional action5 tramtrack sprites, one could either add them to opengfx, or remove them from openttd.grf
00:04:54 <frosch123> andythenorth: nml does not check "types" anywhere, you can also compare roadtypes with cargolabels, or at least i think so
00:05:14 <frosch123> nml is on the asm/c level, everything is an integer :p
00:05:43 <andythenorth> if I was the only author I would have moved on from this days ago :)
00:05:50 <andythenorth> but NRT seems popular
00:06:18 <andythenorth> we will just get bug reports that current_roadtype doesn't work for trams and vice versa
00:06:55 * andythenorth renames all the vars 'storm_dennis_type'
00:07:48 * andythenorth considers lol_type
00:08:01 <andythenorth> all the NRT types in Hog are currenly variants of 'LOLZ'
00:09:21 <frosch123> hmm, did pm delete the devzone?
00:10:23 <frosch123> usually when devzone does not work, it's because the container crashed
00:10:31 <frosch123> but this time they are stopped regulary
00:10:56 <andythenorth> webster stopped earlier
00:11:00 <andythenorth> does that run on same box?
00:11:35 <frosch123> well, if they are stopped regulary, i assume that someone is doing something, so i do not interfer
00:16:01 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro merged pull request #8005: Fix: Ignore not printable characters when layouting a string https://git.io/Jvlh7
00:27:00 <andythenorth> so are these NML var names a non-issue?
00:27:03 * andythenorth needs to sleep :P
00:28:13 <glx> names can be anything, as long they are "compiled" to the right var
00:29:35 <andythenorth> ok moving on then :P
00:29:45 *** andythenorth has left #openttd
00:31:22 <DorpsGek_III> [OpenTTD/OpenTTD] frosch123 commented on pull request #8002: Doc: [Script] Add a note about how wagon connectivity works for scripts https://git.io/Jv8JA
00:33:53 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro commented on pull request #8002: Doc: [Script] Add a note about how wagon connectivity works for scripts https://git.io/Jv8Uv
00:38:32 *** D-HUND is now known as debdog
00:42:31 <frosch123> hmm, after you moved a vehicle to a new row using AIVehicle::MoveWagon, how do you get the id of the new chain?
00:43:14 <frosch123> i guess ais never disassemble their trains, so noone ever needed that
00:45:18 <frosch123> lol, admiralai is so old that it checks whether wagons are added to trains automatically
00:51:55 *** Smedles has joined #openttd
01:03:17 <DorpsGek_III> [OpenTTD/OpenTTD] frosch123 commented on pull request #8002: Doc: [Script] Add a note about how wagon connectivity works for scripts https://git.io/Jv8UK
03:27:05 <DorpsGek_III> [OpenTTD/OpenTTD] spnda commented on pull request #7955: WIP Feature: NewGRF Road Stops (Feature 14) https://git.io/Jv8IG
03:28:06 <DorpsGek_III> [OpenTTD/OpenTTD] spnda updated pull request #7955: WIP Feature: NewGRF Road Stops (Feature 14) https://git.io/JvLQL
04:31:10 *** WormnestAndroid has joined #openttd
05:23:46 *** Extrems has joined #openttd
05:28:05 *** tokai|noir has joined #openttd
05:28:05 *** ChanServ sets mode: +v tokai|noir
06:26:56 *** OsteHovel has joined #openttd
07:59:45 *** sla_ro|master has joined #openttd
08:38:24 *** andythenorth has joined #openttd
09:25:24 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro commented on pull request #8002: Doc: [Script] Add a note about how wagon connectivity works for scripts https://git.io/Jv8Oo
09:34:04 <andythenorth> so is it evil to try and import python modules from a parent dir?
09:34:21 <andythenorth> I currently have a shared library vendored into src/ because otherwise I can't import it
09:34:23 <LordAro> generally indicates you need to restructure your python project
09:34:35 <andythenorth> having shared libraries in src is evil
09:35:15 <andythenorth> importing from parent is evil
09:36:10 <andythenorth> I could install the library to virtualenv
09:36:21 <andythenorth> but it's not on pip, and then nobody else will be able to build my project
09:36:25 <LordAro> that would be what i would expect
09:36:37 <LordAro> pip can install from not-pypi
09:40:09 <andythenorth> what's the current issue count?
09:40:21 <DorpsGek_III> [OpenTTD/OpenTTD] LordAro commented on issue #7337: Audio driver crash on quit - fluidsynth/SDL https://git.io/fhpg0
09:41:43 <andythenorth> I find around 70 is some kind of magic number in project issues, it's a manageable amount
09:42:06 <LordAro> andythenorth: yeah, but doing anything with that involves convincing peter1138 to update it
09:42:16 <andythenorth> well he won't be riding a bike today
10:03:44 *** andythenorth has joined #openttd
10:07:47 <andythenorth> mmv is such a good tool
10:09:01 <andythenorth> is it because of the -1?
10:09:04 <andythenorth> where's the caller? :P
10:10:42 <andythenorth> I'm not very good at using blame :P
10:12:18 <LordAro> so it's definitely been there a while...
10:19:17 *** Progman has joined #openttd
11:21:57 <andythenorth> hmm how to describe Joker trains in Iron Horse?
11:23:59 <andythenorth> they're eye candy or novely trains
11:26:08 <andythenorth> Jokers are trains that are included for novelty value, and may not have much of a transport role. For example snowploughs, driving cab cars and very slow locomotives.
11:27:21 <andythenorth> I guess that works
11:42:40 <Eddi|zuHause> <andythenorth> I currently have a shared library vendored into src/ because otherwise I can't import it <-- sounds like you need a library import path
11:48:28 *** andythenorth has joined #openttd
11:48:40 <andythenorth> Eddi|zuHause: maybe
11:49:02 <andythenorth> SO says that adding to sys path is wrong
11:49:09 <andythenorth> and that shared libraries in src are wrong
11:49:18 <andythenorth> and I'm not packaging this for pypi
11:49:38 <andythenorth> basically what I need to do is wrong, according to SO
11:49:49 <andythenorth> meanwhile Horse 100%
11:50:23 <Eddi|zuHause> sometimes you have only wrong solutions
11:52:28 <andythenorth> the big python web frameworks just manipulate path
11:52:30 <andythenorth> it's pretty standard
12:25:23 <LordAro> andythenorth: yeah, but you're not a big python web framework
12:25:37 <LordAro> what is the shared library?
12:26:16 <andythenorth> it's just a bunch of constants and sprites
12:26:19 <andythenorth> used across all my grfs
12:28:30 <nielsm> make it a git remote and have a configure step that uses python setuptools to install it in the venv
12:28:45 <nielsm> configure or build step
12:30:00 <andythenorth> yes, that's how the big frameworks do this sort of thing
12:30:08 <andythenorth> I don't have any build step at all for newgrfs
12:30:15 <andythenorth> so I'd have to add all that :)
12:31:18 <LordAro> seems like it'd be pretty easy to rework polar-fox into a standard python module
12:31:28 <LordAro> then you can just import it as you would any other python thing
12:32:19 <andythenorth> how would it be vendored in, especially for people who aren't me? :)
12:32:37 <andythenorth> this might all be over-complicated, maybe we should do something more interesting :)
12:33:15 <andythenorth> as a library it has no API versioning strategy
12:33:32 <andythenorth> it relies on the 'correct' version being the one that is currently committed
12:33:50 <andythenorth> it's not really a library in any meaningful way I guess
12:33:55 <andythenorth> it's just a bunch of shared assets
12:34:06 <LordAro> or you could just make sure all GRFs are required to use the latest version
12:34:25 <andythenorth> that would break building older revs
12:34:35 <andythenorth> maybe I have the least-worst solution already
12:34:38 <LordAro> how often do you need to do that?
12:34:48 <andythenorth> I build old tags fairly frequently
12:34:51 <andythenorth> dunno about other people
12:34:58 <andythenorth> I see frosch do it to bisect bugs
12:35:12 <LordAro> maybe git submodule would be better?
12:35:23 <LordAro> failing that, you would need some sort of versioning
12:35:34 <nielsm> yeah a git submodule always points at a particular revision of the references repository
12:35:37 <andythenorth> I don't mind that it's committed, but I mind that it's in src
12:35:42 <LordAro> though you can require a particular commit via pip install with a git repo
12:35:52 <nielsm> and the submodule pointer is also part of the git revision of the master repository
12:35:57 <andythenorth> same reason the bin scripts and Makefile aren't in src I guess, it's not src
12:36:12 <andythenorth> I've been told never to modify sys path
12:36:20 <andythenorth> well meaning advice, inappropriately applied?
12:37:03 <LordAro> it's a "never" in the same way you should never use gotos
12:37:11 <LordAro> they generally indicate some structural issue
12:37:41 <LordAro> i'm pretty sure it is some structural issue in this case though :p
12:38:15 <andythenorth> if I was doing this in a work project, it would be a versioned library
12:38:35 <andythenorth> although it might still be committed to the repo, like we commit jquery and bootstrap
12:38:52 <andythenorth> we don't bother relying on building from upstream for such trivial deps
12:39:07 <andythenorth> also the upstreams break and have to be mirrored :P
12:39:53 * andythenorth back to releasing Horse
12:45:23 *** andythenorth has joined #openttd
13:21:40 *** Smedles has joined #openttd
13:55:59 *** Webster has joined #openttd
14:16:49 <FLHerne> andythenorth: That does look like several of the things I was asking for
14:16:56 <andythenorth> not sure if you'll like it
14:17:00 <andythenorth> but the feedback was useful
14:17:12 <andythenorth> as it was considered feedback, not just speak brainz
14:17:46 <FLHerne> Anyone want to play a Horse multiplayer game in a bit, then? :P
14:17:54 * FLHerne needs to walk the dog first
14:18:08 <andythenorth> shall I put FIRS v4 alpha on bananas?
14:18:12 <andythenorth> Steeltown is nuts
14:18:22 <andythenorth> it's the only thing I play now :P
14:18:35 <andythenorth> not sure everyone will like it :D
14:18:50 <FLHerne> I like the gameplay of Steeltown, but not the theme :P
14:19:23 <FLHerne> (it's not very realistic for the world to contain only heavy industry) ;-)
14:20:06 <andythenorth> it's very interconnected, but in a way that builds up
14:20:25 <andythenorth> I'd like to do the same for a town-focussed economy, but I never figured it out yet
14:29:06 *** andythenorth has joined #openttd
15:52:59 <andythenorth> hmm I should do this NRT docs stuff :|
16:48:28 *** Wormnest has joined #openttd
17:10:22 <glx> I'll take a look, some may be easy to do
17:10:56 <glx> the implementation part ;à
17:14:16 <andythenorth> I can draft docs for nml
17:14:25 <andythenorth> I don't like it, but my implementations are probably worse :P
17:14:45 <andythenorth> one day the nml docs go in the repo, as part of the PR :P
17:28:20 <glx> what's best for openttd version ? full, split or both ?
17:36:18 <Eddi|zuHause> i'm concerned that suggesting there's two things when internally it's the same thing might be misleading a few people
17:38:13 <nielsm> agree there, don't pretend there's two distinct properties when it's the same returning different things depending on circumstances
17:38:37 <Eddi|zuHause> Wolf01: i'm afraid the CSUR stuff is a tiny bit too much, memory-wise
17:54:08 <DorpsGek_III> [OpenTTD/OpenTTD] andythenorth commented on issue #7644: Mysteriously poor performance on macOS https://git.io/fjii3
17:55:37 <andythenorth> what are industry vars 88..89?
17:55:41 <andythenorth> they're undocumented
17:57:10 <andythenorth> same for 8A..8D and similar
17:57:17 <andythenorth> nml does not know about them either
17:57:33 <LordAro> do they actually exist?
17:57:44 <LordAro> were they added specifically for something george or mb wanted? :p
17:59:07 <andythenorth> saying they need deprecated
17:59:50 <LordAro> i might be looking in the wrong place
18:00:09 <andythenorth> no they look correct
18:00:28 <andythenorth> I wonder (1) what they're for (2) why they need deprecated (3) why they're not documented
18:01:36 <andythenorth> some of those are more obvious, they relate to 16-cargo industry chagne
18:02:03 <andythenorth> presumably we don't delete them from OpenTTD?
18:02:14 <andythenorth> do we need to mark them as deprecated in src?
18:03:32 <andythenorth> newgrf spec is a bit of a disaster at the moment, I can see why frosch has quit
18:05:22 <andythenorth> we can rebuild him!
18:06:27 <Eddi|zuHause> andythenorth: 8x variables are generally something inherited from TTD(patch)
18:07:12 <DorpsGek_III> [OpenTTD/nml] glx22 opened pull request #82: Feature: global var 0x21/0xA1: openttd_version, openttd_major, openttd_minor, openttd_revision, openttd_build https://git.io/Jv8Rl
18:08:46 *** andythenorth has joined #openttd
18:09:45 <andythenorth> you scared my connection away glx
18:09:50 <andythenorth> it's easily spooked here
18:18:20 <andythenorth> ok so 9E and friends are there
18:18:56 *** D-HUND is now known as debdog
18:21:08 <glx> hmm I guess for house property 0x23 a full replacement of original 0x0D..0x0F and 0x1E is required
18:28:48 <andythenorth> how did nielsm do it for industries?
18:29:10 <glx> industry tiles prop 0x13 is not done yet :)
18:30:05 *** glx is now known as Guest16593
18:30:31 <andythenorth> I think nml magically implements 0x13
18:30:57 <glx> I think it doesn't, I have the code in front of me :)
18:31:16 <andythenorth> something makes it work in FIRS
18:32:14 <andythenorth> special_flags: bitmask(INDTILE_FLAG_ACCEPT_ALL);
18:32:37 <glx> but 0x13 is for the tiles, not the industry itself
18:33:05 <andythenorth> I'm not sure why prop 0x13 even exists
18:33:24 <glx> for specific acceptance per tile
18:33:33 <andythenorth> for industries that's dumb
18:33:57 * andythenorth is a bad person to design the spec, I tend to discount other people's use cases
18:34:07 <glx> original refineries have different acceptance per tile
18:34:23 <andythenorth> meanwhile INDTILE_FLAG_ACCEPT_ALL is not documented, I'll fix that
18:39:42 <andythenorth> why is accepted_cargos still defined in nml?
18:40:14 *** gelignite has joined #openttd
18:45:07 <andythenorth> wow this is fiddly
18:45:38 <andythenorth> INDTILE_FLAG_ACCEPT_ALL has to refer to accept_cargo_types which is nml < 0.5
18:46:03 <andythenorth> and also to cargo_types which is >= 0.5
18:46:43 <andythenorth> and it can't actually consider accept_cargo_types because INDTILE_FLAG_ACCEPT_ALL only exists in nml >= 0.5
18:47:07 <andythenorth> anybody care to verify this? Strong chance I am writing broken docs here
18:48:56 <andythenorth> should tile prop 'accepted_cargos' be deprecated in nml?
18:49:01 <andythenorth> in favour of prop 0x13?
18:49:10 <andythenorth> we deprecated the old industry cargo acceptance properties
18:51:02 <TrueBrain> Ugh, having shitty internet sucks .. how do people live like this?
18:51:44 <TrueBrain> I can do 200 kbit/s ... joy
18:51:57 <TrueBrain> I wanna go home now
18:52:11 <glx> no need to deprecate andythenorth, just improve
18:53:53 <andythenorth> well we deprecated for the industry props
18:54:36 <glx> but accepted_cargos can just accept 16 pairs instead of 3
18:55:01 <glx> of course it's valid only for openttd >1.9
18:56:41 *** gelignite_ has joined #openttd
18:57:16 <glx> oh I need to update regression output I think
18:58:51 <LordAro> are we deprecating andythenorth now?
18:59:05 <andythenorth> can we deprecate my role in nml?
19:00:46 <andythenorth> can we deprecate the nml docs :)
19:00:50 <andythenorth> I really hate them
19:01:50 <glx> hmm ok no regression checks for industries it seems
19:02:57 <LordAro> sounds like you should write some ;)
19:09:24 <andythenorth> removed one red item :P
19:11:55 *** Smedles has joined #openttd
19:16:04 * andythenorth tries to figure out industry vars 69..71
19:17:57 <andythenorth> somewhat documented here
19:29:42 <andythenorth> possibly should be a table not a list, I looked for examples, but usually it's one heading per var
19:37:55 <andythenorth> nml docs list vars like produced_this_month_1
19:38:05 <andythenorth> but they don't exist in nml src
20:03:05 <andythenorth> I haven't added the new vars yet
20:12:04 *** Wormnest has joined #openttd
20:17:05 *** frosch123 has joined #openttd
20:22:44 <andythenorth> nml master doesn't know industry vars 0x8B or 0x8D
20:23:21 <DorpsGek_III> [OpenTTD/OpenTTD] frosch123 commented on pull request #8002: Doc: [Script] Add a note about how wagon connectivity works for scripts https://git.io/Jv8z0
20:23:28 <DorpsGek_III> [OpenTTD/OpenTTD] frosch123 approved pull request #8002: Doc: [Script] Add a note about how wagon connectivity works for scripts https://git.io/Jv8zE
20:23:55 <andythenorth> can anyone figure out what 0x8B and 0x8D are?
20:24:42 <andythenorth> frosch123: are 8B and 8D just an oversight?
20:25:21 <Wolf01> Shit, I didn't watch star trek friday
20:25:31 <andythenorth> same for 0x95 and 0x97
20:25:37 <andythenorth> I can't find those vars
20:26:44 <andythenorth> and 0xA3 and 0xA5
20:27:34 <DorpsGek_III> [OpenTTD/nml] frosch123 commented on pull request #82: Feature: global var 0x21/0xA1: openttd_version, openttd_major, openttd_minor, openttd_revision, openttd_build https://git.io/Jv8zo
20:29:34 <frosch123> 8B is the second byte for 8A, 8D is the second byte to 8C
20:29:48 <frosch123> you wouldn't document that
20:29:56 <frosch123> that is openttd staying compatible to ttdp weirdness
20:30:33 <andythenorth> so in terms of deprecating it in nml and documenting it?
20:30:36 <andythenorth> they're a no-op?
20:30:47 <frosch123> on that page you can see that there is a gap after every 80+x variable with size "W"
20:31:12 <frosch123> i added "see also variable 6xy" to the nfo wiki
20:31:31 <andythenorth> I just expanded those to explain what each var is
20:31:33 <frosch123> but i only added it to those, which were already documented before
20:31:48 <andythenorth> we don't really have a format for 'list of multiple vars under a heading'
20:31:52 <andythenorth> I just used a list :P
20:32:00 <frosch123> i didn't feel like adding docs for variables getting deprecated, when they were not documented before :p
20:32:08 <frosch123> kind of post-mortem-docs
20:32:27 <andythenorth> I propose we never change newgrf ever again :P
20:32:33 <andythenorth> it's way too painful to document
20:36:38 <andythenorth> 88..89 have me baffled
20:37:30 <andythenorth> Types of cargo produced?
20:38:43 <andythenorth> can't find any historical nml support for 0x88 or 0x89
20:39:11 <andythenorth> nor 90...92 which I assume are Types of cargo accepted
20:39:58 <frosch123> that's good then :)
20:40:21 <frosch123> those vars do not exist as 16 cargo-versions
20:40:40 <frosch123> they were pretty pointless anyway
20:41:11 <frosch123> if nml does not know them, there is nothing to update
20:44:01 <andythenorth> I'll add a note for nfo spec column
20:45:34 <andythenorth> frosch123: can those cyan notes just be N/A then?
20:47:48 <andythenorth> I will have to write docs for the new nml industry vars
20:48:39 <frosch123> some docs already seem be present
20:50:14 <andythenorth> incoming_cargo_waiting and friends need added
20:52:13 *** Smedles has joined #openttd
20:54:42 *** ChanServ sets mode: +v tokai
21:01:59 <andythenorth> under "Variables that require some parameters."
21:05:26 <andythenorth> I have not added tests or example use
21:05:35 <andythenorth> I have NFI what it's for, or why anyone would ever ever want it
21:14:41 <andythenorth> or is it % versus absolute?
21:15:53 <DorpsGek_III> [OpenTTD/nml] glx22 commented on pull request #82: Feature: global var 0x21/0xA1: openttd_version, openttd_major, openttd_minor, openttd_revision, openttd_build https://git.io/Jv8gP
21:15:58 <andythenorth> ok, I fix the docs for 6D :P
21:17:43 <andythenorth> want to edit that then? :)
21:18:12 <DorpsGek_III> [OpenTTD/nml] glx22 closed pull request #82: Feature: global var 0x21/0xA1: openttd_version, openttd_major, openttd_minor, openttd_revision, openttd_build https://git.io/Jv8Rl
21:18:47 <andythenorth> sorry if you wasted time
21:20:50 <glx> basically there's nothing to change with git, maybe at one point build will be increased but the meaning of the variable won't change
21:22:03 <glx> and yes it was indeed mentioned in #80, I just overlooked the info :)
21:23:25 <andythenorth> so just need to update docs, wiki spec table and close #80 :P
21:23:48 <andythenorth> I have so many open tabs for nml stuff now :P
21:26:18 <andythenorth> glx fancy doing 83? o_O
21:35:30 *** WormnestAndroid has quit IRC
21:36:00 *** WormnestAndroid has joined #openttd
21:50:38 <DorpsGek_III> [OpenTTD/nml] andythenorth commented on issue #83: Support for industry vars 0x70 and 0x71 needed. https://git.io/Jv8gN
22:00:45 *** andythenorth has joined #openttd
22:23:21 *** y2kboy23 has joined #openttd
22:50:29 <DorpsGek_III> [OpenTTD/nml] andythenorth commented on issue #80: openttd_version: OpenTTD var 21/1A result has changed since OpenTTD moved to git https://git.io/JvchK
22:55:00 *** WormnestAndroid has quit IRC
23:00:42 *** WormnestAndroid has joined #openttd
23:12:50 *** WormnestAndroid has quit IRC
23:16:05 *** tokai|noir has joined #openttd
23:16:05 *** ChanServ sets mode: +v tokai|noir
23:37:18 *** WormnestAndroid has joined #openttd
23:47:31 *** mindlesstux has joined #openttd
continue to next day ⏵