IRC logs for #openttd on OFTC at 2021-07-03
⏴ go to previous day
00:12:19 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
00:40:22 *** WormnestAndroid has joined #openttd
00:48:24 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
00:49:13 *** WormnestAndroid has joined #openttd
00:49:48 *** tokai has quit IRC (Quit: c('~' )o)
01:15:40 *** yeoldetoast has joined #openttd
01:39:46 *** Wuzzy has quit IRC (Remote host closed the connection)
02:52:26 *** D-HUND has quit IRC (Ping timeout: 480 seconds)
03:41:12 *** roadt__ has joined #openttd
03:47:37 *** roadt_ has quit IRC (Ping timeout: 480 seconds)
04:52:59 *** snail_UES_ has quit IRC (Quit: snail_UES_)
06:13:36 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 opened pull request #9408: Codechange: replace InjectDParam/ShiftParameters https://git.io/JcExw
06:35:32 *** sla_ro|master has joined #openttd
07:37:30 *** andythenorth has joined #openttd
07:37:50 *** Progman has joined #openttd
08:09:21 *** andythenorth has quit IRC (Quit: andythenorth)
08:18:51 *** Progman has quit IRC (Remote host closed the connection)
08:20:17 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #9408: Codechange: replace InjectDParam/ShiftParameters https://git.io/Jcu8R
08:20:43 <TrueBrain> well, "replace"... guess you meant "remove", but what-ever :P
08:58:44 *** andythenorth has joined #openttd
09:01:35 <TrueBrain> morning andythenorth
09:25:37 <Rubidium> replace it with something better ;) Though github does not really like long titles, so I shortened the commit message
09:26:17 <DorpsGek> [OpenTTD/OpenTTD] rubidium42 merged pull request #9408: Codechange: replace InjectDParam/ShiftParameters https://git.io/JcExw
09:30:01 <TrueBrain> yeah, but you don't replace those functions :D
09:30:05 <TrueBrain> you remove their need to exist :D
09:32:09 <Rubidium> sorry I made you feel sick ;D
09:34:45 *** HerzogDeXtEr has joined #openttd
09:36:14 <TrueBrain> can't imagine you weren't by looking at that code :P
09:36:29 <TrueBrain> well, there are good odds you wrote the original version, so in that case, yeah, you should be sorry :P
09:38:48 <Rubidium> the memmove is in r1/r975... which is before I joined
09:40:54 <TrueBrain> well, my odds were still good honestly :P
09:41:14 <TrueBrain> hmm ... with "step 1" of my plan, I do not have all the IPs of a server yet .. so only either the IPv4 or the IPv6 ..
09:41:18 <TrueBrain> I was using STUN to discover the other IPs
09:47:01 <TrueBrain> I knew MacOS could be laggy on the network, but this latency is insane andythenorth :P
09:47:03 <andythenorth> another day, another school closure + covid tests
09:47:08 <andythenorth> stuff and things
09:47:51 <TrueBrain> still that insane over there? Damn ...
09:48:07 *** sla_ro|master has quit IRC ()
09:48:16 <andythenorth> ok it's newgrf coding day
09:49:56 * andythenorth will mostly be complaining about the spec sorry
09:50:43 <TrueBrain> we will prepare /dev/null to accept your complaints
09:50:47 <Rubidium> type checking functionality has been added to StringParameters, though at the moment it is quite useless. There are literally two places where the types get genuinely set, and those are within the string formatting code at places where the least amount of verification is needed. And it's kinda pedantic/unuseful; you can't say it is cargo, you more or less have to say whether it is to be used as
09:50:53 <Rubidium> long/short/tiny string in the end, instead of just "saying" its a cargo (ID)
09:51:15 <Rubidium> so... remove it and replace it in one PR, or split it into two? ;)
09:51:53 <TrueBrain> split into two, as in: remove now, never add later? :D
09:52:27 <Rubidium> potentially yes, though curently its use is severely underrated as well
09:53:22 <TrueBrain> so first PR that you remove it, and if people really object, make them replace it :P
09:54:00 *** WormnestAndroid has quit IRC (Remote host closed the connection)
09:54:13 *** WormnestAndroid has joined #openttd
10:10:22 *** sla_ro|master has joined #openttd
10:19:21 <peter1138> The school is closed because it's Saturday?
10:20:08 *** JacobD88 has joined #openttd
10:28:09 <andythenorth> I was summarising
10:28:38 <andythenorth> the covid test was today's excitement
10:48:14 *** gelignite has joined #openttd
10:51:33 <andythenorth> peter1138 was it lunch yet?
10:51:40 <andythenorth> today is all at sixes and sevens
11:07:17 <Samu> holy crap, already built is such a bad error
11:07:32 <Samu> there is a problem with locks
11:09:09 <Samu> inconsistent error message
11:20:54 <Samu> line 479 water_cmd.cpp is the problem
11:21:23 <Samu> IsTileType(current_tile, MP_WATER), can't use this because ship depots and locks are also water
11:27:56 <Samu> this won't be an easy fix
11:29:54 <nielsm> somewhat a thing I also ran into while working on deepwater, that both clear and built-on water tiles are the same class, yes
11:30:26 <nielsm> there's some odd gymnastics going on to make the various combinations
12:17:39 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
12:19:46 <Samu> DC_FORCE_CLEAR_TILE is in the way...
12:27:50 <Samu> the problem im having is that I'm overbuilding a canal on a lock, but the code treats that lock as if it were a canal
12:53:18 <Samu> but haven't tested the other cases
12:59:19 *** gelignite has quit IRC (Quit: Stay safe!)
13:17:53 *** Elouin has quit IRC (Quit: So long and thanks for all the fish!)
13:31:40 <Samu> just tested some other cases, and ... meh
13:32:25 <Samu> building a canal on a canal gives me an error that i must demolish the canal
13:32:41 <Samu> because there is an object on the canal
13:33:06 <Samu> it should tell me to remove the object, not the canal
13:50:26 *** sla_ro|master has quit IRC ()
14:32:47 <Samu> i don't understand, the fix is basically going against the purpose of DC_FORCE_CLEAR_TILE
14:53:12 *** Progman has joined #openttd
15:43:15 *** Eddi|zuHause has quit IRC ()
15:47:34 *** jottyfan has joined #openttd
16:02:07 *** gelignite has joined #openttd
16:07:39 *** snail_UES_ has joined #openttd
16:11:06 *** Eddi|zuHause has joined #openttd
16:13:07 <TrueBrain> owh, right, I Was going to fix that .. forgot :P
16:13:27 <glx> I think it's important before nightly
16:14:01 <glx> only new saves should be affected if I understand correctly
16:17:54 <TrueBrain> nothing is broken or will break
16:17:58 <TrueBrain> it is just not converted properly :P
16:18:49 <TrueBrain> GOAL chunk is one of those few that didn't have a SLE_CONDNULL, so the order of the description is identical to the order of the header currently ;)
16:19:16 <glx> yeah this one is safe, indeed
16:19:36 <TrueBrain> also the reason I didn't notice it ;)
16:19:43 <TrueBrain> the unsafe ones would have come up during testing :)
16:21:10 <glx> come on msys2, build faster
16:23:12 <DorpsGek> [OpenTTD/BaNaNaS] TrueBrain opened pull request #101: Change: [GitHub] switch default branch to "main" https://git.io/JcgUy
16:23:23 <TrueBrain> look at that, found two more wrong instances :D
16:23:51 <TrueBrain> useful in some times, annoying in others
16:25:21 <DorpsGek> [OpenTTD/BaNaNaS] glx22 approved pull request #101: Change: [GitHub] switch default branch to "main" https://git.io/JcgTq
16:26:51 <DorpsGek> [OpenTTD/BaNaNaS] TrueBrain merged pull request #101: Change: [GitHub] switch default branch to "main" https://git.io/JcgUy
16:26:51 <glx> (yes I squashed to rename ;) )
16:26:58 <TrueBrain> nobody cares about staging :P
16:27:17 <TrueBrain> glx: I wasn't doubting that for a sec honestly :)
16:27:28 <TrueBrain> I love that we can approve PRs and still change the commit message .. so fucking useful :)
16:27:58 <glx> haha and of course I get a conflict, because I already fixed it in "saveload"
16:28:32 <DorpsGek> [OpenTTD/OpenTTD] SamuXarick opened pull request #9410: Fix: Inconsistent error message when overbuilding canal on a lock/ship depot tile https://git.io/JcgTi
16:29:15 <glx> (just because everything is indented differently)
16:30:06 <glx> ignoring whitespace changes when looking at diffs in github app is nice for these kind of PR
16:31:05 <glx> easy to spot rebase errors
16:33:55 <TrueBrain> I wonder why people think sending offers to abuse@ is a good idea
16:34:03 <TrueBrain> maybe it translates poorly in their language
16:35:27 <glx> maybe I should just create static variables and a array of references to them
16:36:33 <TrueBrain> interesting problem in your case .. std::variant is not going to fly
16:36:40 <TrueBrain> and because of the vtable, you kinda do need to do a "new"
16:36:59 <TrueBrain> or indeed what you suggest
16:38:44 <glx> if (_sl.expect_table_header && ch.load_check) { <-- and I need a some kind of hack to remove load_check and fix_pointers bools
16:40:52 <TrueBrain> yeah, that is also a bit weird anyway
16:41:05 <TrueBrain> the fact that SlSkipArray skips the header
16:41:10 <TrueBrain> but otherwise it doesn't
16:42:10 <TrueBrain> I guess now you abstracted LoadCheck like this
16:42:19 <TrueBrain> you could call SlTableHeader for CH_TABLE before SlSkipArray
16:42:27 <TrueBrain> and remove that "ch.load_check" part from that if
16:43:04 <TrueBrain> that should do the right thing too, I think
16:44:16 <TrueBrain> and fix_pointers .. isn't that just making it a default empty function and always calling it?
16:44:44 <glx> yeah it's there for the debug message
16:45:02 <TrueBrain> just remove the debug message
16:45:06 <TrueBrain> it is pointless with that change :P
16:45:22 <TrueBrain> I doubt anyone in the last 10 years had any interest in that debug message ;)
16:49:32 *** WormnestAndroid has joined #openttd
16:58:12 <glx> pff I was getting a weird error when lauching debug, VS decided to select arm64 target
16:58:25 <glx> of course I can't debug that
17:01:59 *** yeoldetoast has quit IRC (Quit: My Mac Mini has gone to sleep. ZZZzzz…)
17:12:10 *** y2kboy23_ has joined #openttd
17:16:57 *** y2kboy23 has quit IRC (Ping timeout: 480 seconds)
17:27:22 <glx> strange I have a save with CH_TABLE, didn't expected that
17:30:46 *** jottyfan has joined #openttd
17:33:08 <andythenorth> seems FIRS does things
17:47:48 *** justdroppedin[m] has joined #openttd
18:02:18 *** andythenorth has quit IRC (Quit: andythenorth)
18:17:24 *** andythenorth has joined #openttd
18:24:37 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
18:33:46 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
18:43:19 *** WormnestAndroid has joined #openttd
18:49:58 <DorpsGek> - Update: Translations from eints (by translators)
19:04:56 <TrueBrain> lets see if I did everything correct ..
19:06:21 <TrueBrain> "fun" fact .. you can add an API token on PyPi for the whole repo without further login step-up
19:06:28 <TrueBrain> but removing an API token, that requires a step-up
19:06:35 <TrueBrain> not sure in what world that makes sense, but okay ..
19:36:59 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #9411: Replace master-server (UDP-based) with Game Coordinator (TCP-based) https://git.io/JcgaE
19:37:07 <TrueBrain> especially what is tagged is surprisingly interesting :P
19:37:27 <TrueBrain> right, part 1 of #9017 .. tomorrow I will deploy the Game Coordinator on AWS
19:37:45 <TrueBrain> I just need to learn the master-server to also record his changes in redis .. and that should be it
19:39:24 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9411: Replace master-server (UDP-based) with Game Coordinator (TCP-based) https://git.io/JcgaE
19:39:33 <TrueBrain> hopefully this is easier to review .. it is still pretty big, but not as big at least :D
19:45:08 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #9411: Replace master-server (UDP-based) with Game Coordinator (TCP-based) https://git.io/JcgaE
19:45:31 <Samu> remember ScriptMarine::AreWaterTilesConnected ?
19:46:07 <Samu> checking if a 2-tile wide aqueduct is connected returns false
19:47:12 <Samu> one tile is the start, the other is the end
19:50:10 <Samu> 'tile_from' and 'tile_to' are directly neighbouring tiles.
19:57:35 *** nielsm has quit IRC (Ping timeout: 480 seconds)
20:08:11 <andythenorth> && is boolean AND in nml?
20:08:25 <andythenorth> 1 AND 0 gives me what also?
20:19:39 <TrueBrain> I would hope those two mean the same in any language ;)
20:21:40 <TrueBrain> And these and are exclusive ands .. so it only returns true if both sides are true. Otherwise false.
20:21:59 *** jottyfan has quit IRC (Quit: jottyfan)
20:22:00 <andythenorth> I even looked up one of those 1 and 0 tables like in school :P
20:22:12 <TrueBrain> In contrast to the 'and' in the English language
20:22:17 <andythenorth> my problem was elsewhere, not the && :P
20:23:50 <andythenorth> this is good, using the nml procedures
20:24:11 <andythenorth> I have been able to convert more code and delete a whole load of stupid book-keeping code from my compile
20:24:24 <andythenorth> calculating names for 'next switch'
20:36:02 * andythenorth doing lots of industry location crap
20:37:46 <glx> so loading of FIRS in OTTD is faster ?
20:38:38 <glx> because the time between pressing "generate" and begin of generation is quite high with FIRS enabled
20:39:03 <glx> it's really the loading of the grf :)
20:39:31 <andythenorth> I am concerned about map gen time with more checks
20:39:36 <glx> debug build of course doesn't help
20:39:39 <andythenorth> ECS manages to be really really slow somehow
20:39:53 <glx> ECS as weird landscape requirements
20:40:09 <glx> especially for tourist centers
20:41:45 <andythenorth> hmm I have a branch testing FIRS locations
20:42:00 <andythenorth> instead of clustering industries by distance, it defines 'town type'
20:42:02 <andythenorth> so coal town etc
20:42:10 <andythenorth> it might be worse
20:43:35 <andythenorth> I thought it could be nice to give towns a flavour
20:43:41 <andythenorth> like 'this is a mining town'
20:43:52 <andythenorth> then the house grf might adjust buildings also to match
20:44:34 <glx> but openttd expects at least one type per town
20:44:56 <glx> and depending on settings it's at most one per town
20:45:15 <andythenorth> lol that setting :)
20:45:35 <andythenorth> anyway it doesn't work as well as the existing FIRS clustering
20:45:51 <glx> hmm maybe I'm wrong about the one per town, I think it's one per map
20:46:16 <glx> and 1 max per town, if the setting says so
20:46:23 <andythenorth> forcing it by town (it's mod of town ID) doesn't work to cluster industries
20:46:33 <andythenorth> pure distance check is better, I just worry about performance
20:46:49 <andythenorth> mod(town ID) *must* be faster than measuring distance to nearby industries ?
20:47:04 <andythenorth> newgrf performance is such a dark art
20:47:59 <glx> but town ID distribution is random usually
20:48:24 <andythenorth> that was a positive feature :)
20:48:34 <andythenorth> if we had regions, I think this would work
20:49:30 <andythenorth> OpenTTD fork when?
20:51:42 <andythenorth> regions would be agglomerations of towns I think
20:52:15 <andythenorth> just give each town a parent
20:52:21 <andythenorth> regional capital
20:52:33 <andythenorth> but what's the regional capital for a regional capital!
20:52:47 <peter1138> Have you considered writing a document about what you want to do?
20:53:11 *** Samu has quit IRC (Quit: Leaving)
20:58:01 <andythenorth> I am trying things in current spec first, but then I should
20:58:17 <andythenorth> I suspect anything I write is less good than these 2
20:58:28 <andythenorth> but 3 might be the charm eh
20:59:28 <andythenorth> ha those two are both GS
20:59:36 <andythenorth> that's never going to work eh :D
21:01:50 <andythenorth> TL:DR; clay pits in 'cornwall', cider farms in 'somerset', sheep farms in 'powys' and steel mills in 'swansea'
21:03:46 <andythenorth> I could just do it on map xy, but then...map sizes :P
21:23:56 <DorpsGek> [OpenTTD/bananas-frontend-web] James103 commented on issue #51: "manager" URLs not blocked for content that doesn't belong to you https://git.io/JLM3j
21:27:51 <Xaroth> I'm missing whiskey distilleries up north, andythenorth.
22:16:01 <Xaroth> More like north of Glasgow
22:43:27 *** andythenorth has quit IRC (Quit: andythenorth)
23:06:14 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
23:23:16 *** Progman has quit IRC (Remote host closed the connection)
23:35:38 *** gelignite has quit IRC (Quit: Stay safe!)
continue to next day ⏵
