IRC logs for #openttd on OFTC at 2023-09-06
⏴ go to previous day
00:01:15 *** Guest1952 has quit IRC (Remote host closed the connection)
00:02:12 *** gelignite has quit IRC (Quit: Stay safe!)
01:16:29 *** Wormnest has quit IRC (Quit: Leaving)
02:03:58 *** herms has quit IRC (Quit: bye)
03:00:24 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:22:12 *** D-HUND is now known as debdog
06:00:11 *** Flygon has quit IRC (Remote host closed the connection)
07:26:43 *** gelignite has joined #openttd
07:40:04 <peter1139> Heatmap suggestion helped a lot, so obvious yet I hadn't thought of it.
07:48:59 <peter1139> Although sometimes a bridlepath is just the edge of a field, which is less than ideal on a touring bike.
08:03:46 *** Xaroth has quit IRC (Quit: -)
08:20:50 *** milek7 has quit IRC (Remote host closed the connection)
08:27:56 <LordAro> i've done a few of those
08:28:12 <LordAro> even more fun on a road bike
08:30:28 <peter1139> Well, when I switch to tubeless I might consider that :)
08:31:08 <peter1139> Next planned route is coming up to 100km, which is nice, but probably not one for an evening.
09:23:07 <talltyler> Hey truebrain, I'm finally home and am processing your comments on #11221. I like the approach of a new GameCalendar class for constants, leaving TimerGameCalendar for functions, but am wondering how that will work with inheritance when everything moves to GameCommon and TimerGameCommon. But if a function in TimerGameCommon references a GameCommon type, what happens when that function is inherited
09:23:07 <talltyler> and becomes a TimerGameCalendar function? How do I tell it that it's GameCommon reference now needs to be GameCalendar?
09:23:24 <talltyler> (Sorry that is almost certainly confusing 🙂 )
09:24:17 <truebrain> tttteeeeemmmmmppplllaaaattttteeeessssss 😄
09:24:51 <truebrain> basically, GameCommon will be a template, and GameCalendar / GameEconomy tells GameCommon what "Timer" to use
09:25:10 <LordAro> peter1139: i have thus far kept my exploring loops to around 60k, but i'm running out of things close enough to keep doing that
09:25:18 <talltyler> Okay, so I can deal with that in the next PR, not here?
09:37:17 <peter1139> LordAro, yeah I was pretty lucky I only had a few missing squares to cover last night... and two of those weren't necessary yet.
10:17:42 <alfagamma7> How good is VSCodium folks?
10:17:51 <alfagamma7> Anybody here who has used it?
11:04:37 <peter1139> Hmm, my garden tomatoes are a bit warm...
11:15:10 <peter1139> Unit tests help with the "untested" thing, except of course the unit tests needed to be changed because the ouput got fixed... :)
11:15:29 <truebrain> yeah .. and someone might complain about that: THE SPACES ARE GONE
11:15:44 <peter1139> Then they misunderstand the point of json.
11:16:09 <truebrain> and nobody would be surprised 😛
11:16:15 <peter1139> A quick copy & paste into VS Code, then reformat...
11:16:41 <peter1139> (Or VSCodium, I suppose, if the extensions work...)
11:20:41 <truebrain> VSCodium, for people who trust a random group of people on the internet that "promises to do good" over a company that has actually something to lose when they don't
11:20:48 <truebrain> I don't always understand the logic behind stuff
11:23:15 <peter1139> over a company that has actually demonstrated it's ability to do bad, and keep it hidden?
11:23:58 <truebrain> it is like .. if you don't trust VSCode, use vim or something
11:24:02 <peter1139> One reason I'm trying to use Linux on the desktop again is I'm getting fed up of *adverts* appearing in Windows 11...
11:24:13 <truebrain> but having it repacked by random group of people .. doesn't actually .. fix anything?
11:24:31 <peter1139> So also use Chrome vs Chromium?
11:24:41 <truebrain> same; reason I use Firefox
11:24:56 <peter1139> Linux distros are based on random people packaging other people's stuff.
11:25:23 <truebrain> yup; but this is repackaging stuff from a vendor you say "you don't trust"
11:25:28 <truebrain> I fail to see how that actually solved anything
11:25:46 <truebrain> it is still the same source
11:25:51 <peter1139> They trust the code, they don't trust the changes made in the released package.,
11:25:51 <truebrain> just with slightly different config
11:26:06 <truebrain> yeah, and I would argue, if you can't trust the release process, you can't trust the code either
11:26:21 <peter1139> Telemetry is often a big thing for people, see the fuss over Audacity.
11:26:59 <_glx_> And then you have people typing random commands found on internet to install stuff
11:29:10 *** milek7 has quit IRC (Remote host closed the connection)
11:31:22 <andythenorth> it's on stack overflow though
11:31:28 <andythenorth> so just pase it into shell
11:37:46 <alfagamma7> Sticking to nano then
11:44:29 *** locosage has joined #openttd
11:44:29 <locosage> what's the deal with TileIndexDiff now? subtracting tiles produces TileIndex for some reason
11:44:55 <locosage> though I guess putting diff into TileIndex would simplify stuff a lot
11:45:24 <locosage> add strong type to put a completely different thing there xD
11:49:44 <locosage> TileIndex is unsigned though, is it even supposed to store diffs?
11:50:11 <locosage> guess it will probably work with overflows but still...
12:01:14 <peter1139> Trying to come up with a good off-road route is a pain :(
12:02:02 <peter1139> Some many little connecting bits are marked as only a footpath, so the routing goes a very long way around, often the way you want to come back :p
12:03:14 <peter1139> (And sometimes it routes me on things that actually are only suitable for walkers.
12:05:11 <truebrain> hmm .. I am trying to find a nice way to encode/decode our network protcol in Rust, but that is challenging .. we made some weird choices 🙂
12:05:26 <truebrain> not all packets have a version indicator, and it is derived from other fields
12:07:50 *** milek7 has quit IRC (Remote host closed the connection)
12:11:49 <andythenorth> time for a new prototocol? 😛
12:11:59 <truebrain> if I could rewrite history ...
12:15:36 <andythenorth> just have 2 competing standards
12:34:29 <_glx_> It seems preview workflow is broken
13:11:51 <LordAro> truebrain: I can't private message you, can I?
13:12:03 <truebrain> nope; you would have to join Discord, sorry.
13:13:30 *** TrueBrain_ii has joined #openttd
13:13:37 <TrueBrain_ii> or I can join IRC .. not sure which is easier :P
13:13:50 <truebrain> wow ... I still had HexChat installed ..
13:14:03 <truebrain> (I need it to test the bridge from time to time ..)
13:14:26 <truebrain> You can PM TrueBrain_ii now LordAro 😛
13:15:41 <peter1139> Could be a fake scam account!
13:16:00 <truebrain> it could .. you could be fake too
13:16:07 <truebrain> Peterodium I would trust more
13:28:19 <peter1139> I wouldn't trust anyone claiming to be me. Especially me.
13:29:00 <peter1139> Can I be bothered to reply to the open-source project that doesn't see a problem with crypto-currency donations...
13:49:11 <andythenorth> repeat for most "can I...?"
13:50:16 <peter1139> If you don't know by now that it's a scam...
13:50:20 <peter1139> You just don't care.
13:50:40 <andythenorth> it's digital trust
13:50:52 <andythenorth> no indpendent accountable authority
13:50:55 <andythenorth> so clearly much better
13:51:02 <andythenorth> if you're a libertarian fuckwit
13:51:32 <andythenorth> /me is actually a libertarian, just you can't have liberty without authority sadly
13:51:44 <peter1139> Damn, map 31 of FreeDoom is a gimmick which I have failed on :(
14:15:56 <truebrain> even our length indicators aren't always the same .. some are u8, others u16 ..
14:18:12 <truebrain> every byte counts, or something silly
14:20:05 <_glx_> it should be size_t but can be an issue with 32/64
14:20:28 <truebrain> or just, you know, pick one!
14:20:36 <truebrain> receive_PACKET_CONTENT_CLIENT_INFO_ID -> u16
14:20:41 <truebrain> receive_PACKET_CONTENT_CLIENT_INFO_EXTID -> u8
14:21:27 <_glx_> ID can be 1, 2 or 3 bytes
14:22:16 <_jgr_> Having length fields be an integer number of bytes is luxury
14:23:02 <_jgr_> Network packet formats typically have length fields of all sorts of strange bit sizes
14:23:21 <truebrain> well, we are TCP here, not a serial connection 😄
14:31:22 *** TrueBrain_ii has quit IRC (Quit: Leaving)
14:34:54 <truebrain> `worker: ClientInfoList { content_type: BaseMusic, openttd_version: 4294967295, branches: Some([ContentClientInfoListBranch { branch: "vanilla", version: "14.0" }]) }`
14:35:01 <truebrain> I love how easy Rust can make (de)serialization
14:35:09 <truebrain> it just took a bit to learn him our weird protocol 😄
14:37:12 <truebrain> never noticed, but on Linux loading a game hides folders which are hidden
14:37:19 <truebrain> but the default is ~/.local/share/openttd/save
14:37:21 <truebrain> so if you go up enough
14:37:23 <truebrain> you can't get back in
14:38:06 <_jgr_> You can use the home button to get back to the start
14:38:18 <truebrain> yes, but as that was in another path in my case
14:38:24 <truebrain> I had no way to navigate to the XDG home 😄
14:38:54 <_jgr_> On my setup at least, ~/.openttd is shown even though the other dotfiles are hidden
14:39:03 <andythenorth> truebrain: doesn't GPT know our protocol?
14:39:16 <truebrain> might be that we made an exception for the old home folder .. but the XDG one isn't showing up for me 😄
14:40:16 <truebrain> now .... how am I going to tell `serde` when a list has an u8 counter and when a u16 counter .. hmm
15:07:41 <_glx_> hmm fixing preview build won't be easy
15:22:07 <_glx_> ok I have a potential workaround
15:49:37 <peter1139> Ah, non-repeatable builds because the build environment got updates...
15:50:07 <peter1139> Technically it should all be using specific versions?
15:50:18 <peter1139> But maintaining that... urgh.
15:50:26 <_glx_> well build env need update but we assumed latest was always fine
15:51:14 <_glx_> and upgrading emscripten image won't help here
15:56:48 <_glx_> of course CI doesn't use the version in the PR
16:00:35 <LordAro> mildly surprised npm install doesn't automatically know the maximum version it can install is 9.x
16:05:35 <truebrain> lol, it still failed 😄
16:06:19 <truebrain> it uses the `master` version
16:06:24 <truebrain> so preview will not work till you merge 🙂
16:09:03 <_glx_> yeah I always forget that 🙂
16:09:25 *** Wormnest has joined #openttd
16:12:04 <_glx_> annotations in preview are weird
16:13:19 <_glx_> would mean WITH_ASSERT is defined but assert is not
16:39:40 <kamnet[d]> |I remember that the cost values for building and for landscaping could be set separately in former versions of OpenTTD, but now they seem to have been merged into one. Is this feature still hidden somewhere in the game, so that I can access it through the openttd.cfg file?
16:39:40 <kamnet[d]> Was this ever a thing in OpenTTD? I don't remember.
17:25:40 *** ahyangyi has joined #openttd
17:25:40 <ahyangyi> peter1139: Wow, I didn't hear that in 2021 but thankfully the distro takes care of it
17:37:40 <Eddi|zuHause> kamnet[d]: my initial reaction to that post was "you must be thinking of a patchpack or something"
18:08:34 *** HerzogDeXtEr has joined #openttd
18:36:40 <michi_cc[d]> kamnet[d]: Maybe confusion with some basecost NewGRF?
18:47:58 <andythenorth> how many open PRs do we have? 🙂
18:48:02 * andythenorth hasn't looked for a long time
18:48:28 <andythenorth> do we get a prize or badge for 100?
18:48:44 <Eddi|zuHause> i was just typing to suggest reaching 100 :p
19:20:45 <pickpacket> I played today! Been a couple of months
19:21:11 <andythenorth> slow macOS linker is slow
19:22:33 <andythenorth> maybe I should buy a faster laptop
19:26:00 <andythenorth> truebrain: querystring is used where?
19:26:20 <truebrain> somewhere you can type text 😛
19:27:55 <truebrain> the console also seems like a place that has changed
19:41:36 <andythenorth> talltyler: reading
19:41:44 <andythenorth> first thought was 'do it in GS' 😛
19:42:15 <talltyler> But GS can change settings, so it can control this feature
19:42:52 <andythenorth> I just think we should do everything in GS 😛
19:42:59 <talltyler> Oh, and if anyone feels like UI design (which nobody does), #11170 needs some design ideas 😛
19:43:14 <andythenorth> ok the PR is very clear, but everything about daylength breaks my brain, so I can't give a useful answer
19:43:30 <talltyler> Fair enough, I guess
19:44:00 <talltyler> It will need some sort of communication to GRF so industry texts can be accurate, I think
19:44:58 <andythenorth> I would consider a scaled {STRING_FORMAT}
19:45:31 <andythenorth> if I've understood correctly, it's very unlikely I'd check a var to scale numbers myself
19:45:39 <andythenorth> text stack is hard enough already
19:45:55 <andythenorth> I try to never touch anything to do with string formatting in grf
19:46:00 <talltyler> lol, I should read my own PR. It already provides the percentage to NewGRF 🙂
19:46:16 <andythenorth> touch the text stack once every 5 years and walk away
19:47:19 <talltyler> I think the only thing in FIRS that breaks is the helptext for scrap yards, which says "Not more than 2048 units" or something
19:47:47 <talltyler> Boosted production texts don't give actual cargo quantities
19:47:58 <andythenorth> how often is primary cargo produced (non-callback)
19:48:03 <andythenorth> is it also on 256 ticks?
19:48:15 *** Deep3D has quit IRC (Quit: znc)
19:49:18 <talltyler> Non-callback production keeps the usual frequency and the quantity is scaled. Callback production has the frequency adjusted.
19:50:08 <talltyler> I think most of the helptexts I break are my own 😛 (IOTC and Improved Town Industries)
19:51:40 <andythenorth> I don't have a really strong opinion
19:52:00 <andythenorth> it seems to reopen the long-standing settled policy of 'do it in grf' though
19:52:21 <talltyler> Yes, I'm aware of that 🙂
19:52:22 <andythenorth> 'do it in grf' is much less work
19:52:36 <andythenorth> what else do we scale?
19:52:53 <andythenorth> the consensus has been against OpenTTD being able to scale vehicle capacity, speed or costs
19:53:06 <andythenorth> so lots of grf authors have to provide their own parameters
19:53:06 <talltyler> "do it in GRF" is only less work if you have access to GRF source code and the skill to modify it
19:53:23 <andythenorth> but when I wanted to scale time in grf, the channel consensus was 'lol you have lost your mind'
19:53:23 <talltyler> We do scale costs though, in basecosts
19:53:31 <andythenorth> seems 2+2=4 or 5 depending on mood
19:53:50 <talltyler> I am not part of the "do it in grf" consensus 😛
19:54:32 <talltyler> There's a middle ground obviously
19:54:33 <andythenorth> experience with JGRPP shows that players will break grfs like FIRS with settings
19:54:43 <andythenorth> and then they will report it
19:54:47 <andythenorth> and then I will ignore it
19:54:52 <andythenorth> after a small period of rage
19:55:19 <andythenorth> ignoring player reports seems like the way forward though
19:55:26 <_glx_> just skip the rage period, it's not needed 🙂
19:55:52 *** belajalilija has joined #openttd
19:56:31 <belajalilija> I’ve messed with a server map and found that firs caused problems with rukts and the sbb set
19:56:47 <belajalilija> But i chalked it up to versions of things being old
19:57:23 <talltyler> I am willing to bet real money that RUKTS and the SBB set do silly things with cargos, and they are to blame 😉
19:58:49 <andythenorth> I am looking for the bug reports to see
19:58:54 <andythenorth> but now I think I hallucinated them
19:59:03 <andythenorth> worrying how much of my life might be wholly imaginary
19:59:22 <andythenorth> TL;DR "why are cargo amounts so low" etc
19:59:40 <andythenorth> then after bug hunting someone points out it's some daylength or cargo scaling factor
20:01:51 * andythenorth stops reading annoying tt-forums FIRS thread
20:03:11 <talltyler> I need to continue my internet detox. No Reddit or Railway Preservation News, and less forums and real-world news 🙂
20:03:27 <talltyler> There's just so much negativity that I don't need in my life
20:04:04 <andythenorth> sometimes I stop reading all of it
20:04:10 <andythenorth> then the dopamine drags me back
20:05:19 <talltyler> Yeah, the dopamine is tough to shake
20:05:57 <talltyler> I wish you made it to the OpenTTD meetup, it was a great time 🙂
20:06:21 <talltyler> I'm trying to nurture more human connections for my dopamine instead of consuming content
20:06:54 <andythenorth> ach I've somehow broken my build, it has lost nlohmann_json
20:07:02 <andythenorth> that's disappointing
20:07:16 <belajalilija> talltyler: I removed every grf in that scenario and messed around with them, firs was the only problem causer, was like firs3, really old map xd
20:08:08 <belajalilija> It was causing vehicles to not be available
20:08:12 <talltyler> Removing GRFs is even more like to cause problems 😄
20:08:25 <belajalilija> The problem was there
20:08:34 <belajalilija> I removed things to find the cause
20:08:43 <belajalilija> Expecting it to be a track thing
20:08:51 <andythenorth> `The following packages are already installed:
20:08:51 <andythenorth> nlohmann-json[core]:arm64-osx -> 3.11.2
20:08:51 <andythenorth> nlohmann-json:arm64-osx is already installed`
20:08:58 <belajalilija> But it was firs xd
20:09:02 <andythenorth> ok what lolz have I done?
20:09:46 <andythenorth> belajalilija: FIRS can't and doesn't cause vehicles to not be available, unless they don't have a valid cargo
20:10:17 <belajalilija> I can find you the scenario if you like
20:10:26 <belajalilija> It was very unusual
20:10:35 <belajalilija> But it was firs that was the cause
20:11:05 <andythenorth> it's totally inspectable in grf code
20:11:10 <andythenorth> FIRS was not the cause
20:11:18 <talltyler> > ignoring player reports seems like the way forward though
20:11:42 <talltyler> Odds are that the scenario creator broke it before uploading it
20:11:59 <andythenorth> you removed every grf, so what vehicles could be not available?
20:12:24 <andythenorth> I'm sure I was told how to fix this already, but I don't have notes, sorry 😦
20:12:55 <Timberwolf> I had a stupid like that the other day. I had a wagon disappearing only in FIRS games. Turned out I'd reused ID 95 (default game maglev grain wagon), not defined a cargo, and it was one of the FIRS economies which doesn't have that cargo.
20:14:38 <belajalilija> andythenorth: Removing firs fixed it, removing other grfs didn’t
20:15:33 <andythenorth> FIRS can't be expected to fix broken vehicle grfs 🙂
20:15:36 <andythenorth> that's out of scope
20:20:21 <belajalilija> andythenorth: Check Discord channel #multiplayer-jgr when you have a minute, you’ll understand better what i did then xd one of my messages links to the map file and the other to my finding out the cause
20:20:50 <belajalilija> I don’t understand why it did it then and i still don’t now
20:20:55 <belajalilija> It’s just amusing
20:24:07 <andythenorth> cause will be missing cargos
20:24:21 <andythenorth> as FIRS has literally no grf actions for FEAT_TRAINS
20:25:24 <andythenorth> so in engineering terms, either it's the known issue, or you found an OpenTTD bug
20:25:49 *** _pruple has joined #openttd
20:26:58 <belajalilija> andythenorth: Wouldn’t surprise me if it’s either
20:27:13 <belajalilija> I do have a habit of finding bugs
20:27:19 <andythenorth> truebrain: sorry for ping, but my gist above, you will have told me before how to fix this, but I forgot 😛
20:28:05 <_glx_> andythenorth: so vcpkg has it, did you forgot `-DCMAKE_TOOLCHAIN_FILE` ?
20:28:23 <truebrain> Was about to suggest: rerun cmake?
20:29:04 <_glx_> check how ci-build.yml does it (lines 247-249)
20:29:21 <truebrain> Should be in his bash history 😛
20:30:41 <andythenorth> nah, I only have `cmake ..`
20:30:53 <andythenorth> I have saved the command locally
20:31:23 <andythenorth> `cmake .. -DCMAKE_TOOLCHAIN_FILE=../../vcpkg/scripts/buildsystems/vcpkg.cmake -DCMAKE_OSX_ARCHITECTURES=arm64 -DCMAKE_BUILD_TYPE=Release ..`
20:31:29 <andythenorth> is what I have in my old notes
20:31:54 <_glx_> if you don't tell cmake to use vcpkg it won't
20:33:02 <andythenorth> `cmake .. -DCMAKE_TOOLCHAIN_FILE=../../vcpkg/scripts/buildsystems/vcpkg.cmake` doesn't work, same issue
20:33:07 <andythenorth> what have I typoed
20:33:40 * andythenorth looking for ci-build.yml
20:33:48 <_glx_> I recommand using full path
20:35:14 <truebrain> Still think we should vendor nlohmann .. so much easier, and it is a header only anyway
20:35:21 <truebrain> Might fiddle with that soon
20:38:30 <andythenorth> seems I don't have permissions to run vcpkg.cmake
20:38:56 <andythenorth> it's always quite concerning when permissions change for no reason, especially on deps from remote sources
20:40:09 <_glx_> but you don't run vcpkg.cmake
20:40:25 <_glx_> cmake is just reading it
20:41:11 <andythenorth> oh it's not a script?
20:42:21 <_glx_> cmake usually wants fullpath (even source files are actually handled with full path)
20:43:45 <andythenorth> I've given it full path, and also tried with user dir as path root
20:43:50 <andythenorth> same issue with both
20:43:58 <andythenorth> do I need to delete vcpkg and reinstall?
20:44:37 <andythenorth> is this not just underscores vs em-dash?
20:45:23 <andythenorth> my cmake doesn't know any package with nlohmann_json
20:45:33 <andythenorth> and say it's an invalid package name format
20:48:23 <_glx_> but vcpkg has it, cmake just need to know it has to use vcpkg
20:51:07 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
20:52:11 <andythenorth> not sure how I can do that 🙂
20:52:29 <andythenorth> I assume it's a typo in my cmake commands, but I can't see it
20:53:55 <andythenorth> ok I've now broken cmake
20:54:05 <andythenorth> I rm-ed my build dir, which is standard practice
20:54:38 <_glx_> `-DCMAKE_TOOLCHAIN_FILE=/full/path/to/vcpkg.cmake` should be a good start, but maybe it also needs `-DCMAKE_OSX_ARCHITECTURES=arm64` and `-DVCPKG_TARGET_TRIPLET=arm64-osx`
20:54:55 <andythenorth> ok found the typo 🙂
20:55:13 <andythenorth> and the previous build dir had cached files in it
20:55:17 <andythenorth> which is why I rmed it
20:55:30 <andythenorth> they were masking the typo in the path to the toolchain file
20:55:31 <_glx_> yeah cmake cache can be annoying
20:57:30 <andythenorth> one of the commands I pasted has `..` twice 😛
20:58:12 <_glx_> ah yes, but should not be an issue 🙂
20:58:44 <andythenorth> build works, I've kept notes here, thanks
21:15:55 *** Deep3D{ has joined #openttd
21:17:54 <peter1139> LordAro, 35.9 is... brisk :o
21:18:33 *** Deep3D has quit IRC (Ping timeout: 480 seconds)
21:18:33 *** Deep3D{ is now known as Deep3D
21:23:19 <andythenorth> lost my test save
21:23:29 <andythenorth> any way I can fix this?
21:24:53 <andythenorth> it's rare that this occurs
21:25:46 <andythenorth> ach it's cached action 14
21:26:28 <andythenorth> still catches me out after so many years 😛
21:27:29 <andythenorth> I changed min. compatible version in action 14
21:27:47 <andythenorth> but rescan files won't pick that up
21:28:04 <andythenorth> fixed by OpenTTD restart 😛
21:29:35 <andythenorth> wonder how many 'bugs' are just caching, in the whole world
21:31:23 <andythenorth> lost all my railtypes
21:32:55 <andythenorth> existing tiles have all reverted to RAIL when the save loaded
21:33:24 <andythenorth> I've recently rebased to upstream master, save was made with a branch
21:36:29 <andythenorth> oof that was a big complicated savegame to test GS
21:36:31 <andythenorth> I think it's lost
21:37:51 <ketsuban[d]> andythenorth: There are two hard problems in computer science: cache invalidation, naming things, off-by-one errors and buffer overflows.
21:39:31 <andythenorth> I'm really puzzled how I've done this
21:39:36 <andythenorth> there are many moving parts
21:40:09 <andythenorth> I've changed my build back to the branch the save was made with
21:44:24 <andythenorth> not sure why it just loses the grf
21:44:31 <andythenorth> there will be a reason...but finding it
21:44:49 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
21:45:12 <andythenorth> `(py311) OTTD_graphics/Iron_Horse/iron-horse (main|✚2) % python src/polar_fox/git_info.py
21:45:12 <andythenorth> 6651 horse/3.2.0-12-gbba79ea1 undefined horse 3.2.0-12-gbba79ea1`
21:48:01 <andythenorth> I imagine the MD5 doesn't match, but that's normal?
21:48:56 <_glx_> grf not compatible with the one used in savegame ?
21:49:18 <andythenorth> I think it's that
21:49:23 <andythenorth> I bumped min. compatible version
21:49:29 <andythenorth> due to changed vehicle IDs
21:49:48 <andythenorth> but I have been reloading the save constantly for the last week or 2
21:49:55 <andythenorth> to test GS changes
21:50:29 <andythenorth> probably I didn't restart OpenTTD though
21:50:52 <andythenorth> this is the thing I most hate about GS dev
21:51:03 <andythenorth> that's a 100 year whole-map savegame to test behaviour
21:51:05 <andythenorth> now I have to do it again
21:51:56 <andythenorth> cause of the broken is grf though
21:52:39 <andythenorth> if I adjust min. compatible version I can get the savegame to load
21:53:02 <andythenorth> but all railtypes are reverted to RAIL, all ELRL trains are stuck and the railtypes construction menu is borked
21:56:35 <_jgr_> You could just build an old version of the GRF and put that in the GRF folder as well
21:57:02 <LordAro> peter1139: yup. quite pleased with how i felt during that. hard, but not overly so
21:57:16 <LordAro> complete lack of any wind was helpful
21:57:41 <LordAro> would've bren even faster if we hadn't hit every light on the way back in
21:57:55 <andythenorth> _jgr_: I will try
21:58:01 <andythenorth> although that will be broken in other ways
21:58:18 <andythenorth> it's not that I expect this to work 🙂
21:58:47 <andythenorth> it's just that I've been caught out here because action 14 caching hides the problem until restart
21:59:07 <andythenorth> it would be better if it broke immediately after the compatibility change 😛
22:05:03 <andythenorth> 'rescan files' is a bit misnamed eh
22:05:06 <andythenorth> as it doesn't really
22:05:33 <andythenorth> only a restart actually rescans the newgrf dir
22:24:29 * andythenorth directs andythenorth to the github 'new issue' button
22:43:20 <truebrain> So a title '"rescan files' doesn't invalidate action14 cache" would be more to the point?
22:43:35 <andythenorth> I don't know if that's factual yet
22:43:49 <andythenorth> 'action14 is cached' might be another piece of nonsense I've imagined
22:44:32 <truebrain> It is just that such bug titles are confusing 😛
22:44:48 <andythenorth> confusion all round, double helping
22:45:13 <andythenorth> I'm not sure what "rescan files" is supposed to do
22:45:51 <andythenorth> the button here appears to have different behaviour
22:46:03 <andythenorth> but I haven't read the source to see if it's the same method being called
22:46:36 <andythenorth> I wanted to go to bed an hour ago 🙂 but then peter can catch me out on not filing bug reports, again 😛
22:47:29 <andythenorth> shit, the time 😛
22:47:51 <andythenorth> alarm clock will be ringing too soon, zzzz time
23:30:49 *** gelignite has quit IRC (Quit: Stay safe!)
continue to next day ⏵
