IRC logs for #openttd on OFTC at 2023-09-19
⏴ go to previous day
00:42:53 *** johnfranklin has joined #openttd
00:42:53 <johnfranklin> My riding speed cannot exceed 20km/h without a gear box
02:46:13 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:03:51 <Flygon> I'm feeling dumb for only just figuring out how automatic timetable spacing works after 13 years
03:03:55 <Flygon> I'm not the sharpest tool in the shed.
03:04:13 <Flygon> I've had tram networks with literal thousands of trams
03:04:24 <Flygon> And only making loops where I could to try and space them out and avoid clumpage.
03:04:44 <Flygon> I'dve been a good cat litter salesperson.
03:10:46 <emperorjake> Auto separation hasn't even been around that long I think
03:11:19 <emperorjake> Unless you mean the non-JGR automatic timetable
03:36:06 <Flygon> I meant the automatic timetable yeah
03:36:33 <Flygon> There's an auto-separation in jgrpp?..
03:36:41 <Flygon> I am feeling extremely embarassed atm hahaha
04:12:06 <Flygon> I've been one of those people crazy enough to fill the entire map of the 4K Hokkaido scenario
04:12:11 <Flygon> And not know this feature
04:18:42 <Flygon> Oh I did find the auto-separate functuion
04:18:51 <Flygon> I still feel very silly.
04:31:16 <Flygon> My trams carrying soooo much more evenly now.
04:51:08 *** NGC3982_ has joined #openttd
04:52:53 *** NGC3982 has quit IRC (Ping timeout: 480 seconds)
05:24:18 *** keikoz has quit IRC (Ping timeout: 480 seconds)
10:04:32 <andythenorth> is there a list of the acceptable excuses?
10:04:40 <andythenorth> I feel there should be
10:04:46 <andythenorth> is thermonuclear war still valid?
10:32:39 <ahyangyi> How to refuse a thermonuclear war?
10:50:55 <peter1139> Flygon, what do you mean "how [it] works" ... it works!?
11:30:24 <peter1139> Hmm, I'll to bear in mind that #11306 removes bits of docks that were part of my NewGRF docks implementation (bits that previously made sense in vanilla but got changed so are no longer necessary there, but would be with NewGRF docks)
12:02:00 <peter1139> I've tried it and failed miserably every time.
12:20:35 *** D-HUND is now known as debdog
13:02:15 <peter1139> Corrupted savegame is corrupt, shocker?
13:04:44 <_glx_> oh if I remember the discord discussion it loaded fine in 13.0
13:07:04 <peter1139> 27.5MB of NewGRFs required...
13:07:47 <peter1139> And still missing NewGRFs. Oh well.
13:08:43 <peter1139> grfcrawler matches one of the GRFIDs, but it's a different file.
13:10:12 <peter1139> Okay, yes, the savegame is corrupted, and I added a check for invalid road types at some point, which stops the game from loading and then crashing -- now it just fails to load.
13:10:34 <peter1139> The GUI error message only shows one error though.
13:11:50 <peter1139> It doesn't freeze for me though.
13:12:22 <peter1139> Ah, in master you do see all error messages.
13:16:16 <alfagamma7> I definitely should not try my hand on web layout
13:20:46 <andythenorth> peter1139: FlatDocks!
14:06:43 <LordAro> truebrain: i broke my nomad cluster :(
14:09:25 <kamnet> LordAro: I blame solar flares
14:35:41 *** ChanServ sets mode: +v tokai
14:38:20 <peter1139> How am I meant to debug with SDL 2 on Linux?
14:39:22 <peter1139> More than likely, iIf it hits a breakpoint,
14:39:44 <peter1139> More than likely, if it hits a breakpoint, the mouse pointer is still hidden and/or captured...
14:39:45 *** twpol has quit IRC (Read error: Network is unreachable)
14:42:35 <_jgr_> Are you trying to use full screen mode?
14:43:35 <_jgr_> I don't have any problems with the mouse cursor when debugging on Linux
14:44:54 <truebrain> LordAro: What did you do?!
14:56:03 <LordAro> truebrain: upgraded about 5 versions all at once
14:56:13 <LordAro> which, notably, included the raft 3 update
14:56:59 <truebrain> Never get raft in a state it doesn't have consensus
14:57:59 <truebrain> OpenTTD runs on 3 nodes. To upgrade, I scale up 3 nodes first, then scale down 2, wait for new consensus, and scale down the last 😛
14:58:17 <truebrain> Never do 3->6->3 .. as 6 to 3 doesn't give consensus 🙂
14:58:29 <truebrain> Welcome to the world of consensus 😄
15:01:06 <peter1139> Nope. I can comment out the call to SDL_ShowCursor(0) and thusly my cursor appears.
15:01:40 <peter1139> When the game is stopped on a breakpoint, SDL is of course also stopped, so the built-in stuff that makes the cursor appear when you leave the window does not work.
15:03:45 <peter1139> Okay, just because the cursor stays doesn't mean it works... the window has focus and I can switch to another.
15:04:01 <peter1139> Since when did Gnome have cooperative multitasking :-p
15:11:04 <peter1139> Doesn't look like we set any grabbed modes wither :/
15:16:32 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
15:45:16 <peter1139> _jgr_, okay, if I switch it to SDL v1 it works...
16:04:28 *** HerzogDeXtEr has joined #openttd
16:16:00 *** gelignite has joined #openttd
16:29:49 <peter1139> Welp, it was good knowing you all.
16:32:29 <peter1139> The "Online Safety Bill" passed.
16:53:44 <peter1139> Okay, can I force SDL 1 to be used even with SDL 2 installled...
16:54:22 <pickpacket> The EU appears to be going forward with its Chat Control bill. Consider that the Safe Harbour deal with the US was struck down in court because US laws didn’t respect human rights I don’t see how Chat Control can survive court scrutiny
17:28:07 <andythenorth> time for station grfs?
17:32:03 <belajalilija> andythenorth: Iron Cross?
17:36:03 <belajalilija> Andy train grf should be Iron Paddington
17:37:30 <peter1139> truebrain, replace union with std::variant. Good or bad idea?
17:38:04 <truebrain> Most of the time good, but context matters 😄
17:38:52 <peter1139> `CargoesField::u` :-)
17:39:11 * Rubidium was suggested not to do it for the string parameter stuff
17:40:01 *** Wormnest has joined #openttd
17:42:54 <truebrain> Especially as there is a type field, yes, variant 😄
17:43:04 <truebrain> That is just a self implemented variant 😛
17:43:39 <Rubidium> well, not quite with respect to initialisation (of complex types)
17:44:13 <Rubidium> though, have you considered making a separate sub class for each of the types and not have massive switch blocks on the type?
17:57:45 <peter1139> `std::visit([this, &xpos, &ypos](auto && arg) { this->Draw(xpos, ypos, arg); }, this->u);`
17:57:56 <peter1139> That's kinda ugly but not a switch block.
17:58:45 <peter1139> It feels a bit too loose though, to be honest.
17:59:57 <Rubidium> I meant struct CargoesField { virtual void Draw(int xpos, int ypos); }; struct IndustryCargoesField : CargoesField { void Draw(int xpos, int ypos) { ... } };
18:00:43 <peter1139> Yes, I know what you meant :)
18:00:48 <Rubidium> and instead of CargoesField columns[5] something like std::array<std::unique_ptr<CargoesField>, 5>
18:12:44 *** gelignite has quit IRC (Read error: No route to host)
18:13:04 *** gelignite has joined #openttd
18:17:25 <truebrain> peter1139: It does, compile time, for all variants to have a valid overload on Draw 😉
18:18:31 <truebrain> I was kinda surprised how powerful 'auto' is in visit statements 😛 makes for neat code
18:22:45 <frosch123> you could also just remove the "union"
18:23:01 <frosch123> CargoesField does not exactly look like mass data
18:23:12 <truebrain> BuT tHe MeMoRy FoOtPrInT!!!!1111oneone
18:23:13 <peter1139> frosch123, yes, this is true.
18:23:30 <peter1139> It's temporary when the window is open, no big deal :)
18:23:41 <truebrain> does it increase the savegame filesize?
18:23:46 <truebrain> (don't answer, please, don't )
18:24:25 <frosch123> using a std::variant is probably weird in this case, because there are two enum values with an empty structure
18:24:32 <frosch123> so you would need two empty tag types
18:25:16 <andythenorth> truebrain: not if you use a MBP
18:25:29 <truebrain> let me update my DHCP to reflect that change
18:25:51 <ahyangyi> Andy, how did you manage that 256 industry tile limit in FIRS
18:26:11 <andythenorth> ahyangyi: I just didn't use more than 256? 🙂 😛
18:26:22 <andythenorth> dunno, what problem are you encountering?
18:27:21 <peter1139> Ah, and std::variant doesn't work with type aliases either. Hmm.
18:27:27 <ahyangyi> I'm just looking at the spec and I can't believe FIRS didn't exceed that limit
18:27:53 <truebrain> peter1139: `struct DummyName1` and `struct DummyName2` should work?
18:27:58 <ahyangyi> You have about 80 industries in total and you definitely use more than 3 kinds of tiles per industry
18:28:00 <peter1139> Also the industry tile limit is 512, so...
18:28:13 <ahyangyi> Uh, isn't it 256-per-grf
18:28:21 <frosch123> ahyangyi: a tile can query the industry type and the tile position, so you can actually do most things with a single industry tile per grf
18:28:29 <andythenorth> I definitely don't use more than 3 kinds of tile per industry 🙂
18:28:47 <peter1139> Hmm. Yes. I wonder why, given mixing industry GRFs is a bad idea...
18:28:49 <ahyangyi> So, you used the position cb?
18:28:56 <peter1139> static const IndustryGfx NUM_INDUSTRYTILES_PER_GRF = 255; ///< Maximum number of industry tiles per NewGRF; limited to 255 to allow extending Action3 with an extended byte later on.
18:29:02 <andythenorth> yes, position cb
18:29:17 <ahyangyi> I worry that the position cb doesn't mix well with multiple layouts per industry
18:29:24 <andythenorth> peter1138: we didn't extend it in case we wanted to extend it later
18:29:33 <ahyangyi> But yeah, I don't think FIRS has crazy amounts of different layouts
18:29:42 <peter1139> We limited it to 255 to stop 256 being used when it was not an extended byte.
18:29:45 <andythenorth> no, the position cb is absolutely fine
18:29:55 <peter1139> I'm fairly sure, as it is now an extended byte, that limit can be removed.
18:29:57 <andythenorth> and FIRS does have large numbers of layouts in some cases (auto-generated)
18:30:12 <ahyangyi> That's the trick I want to learn!
18:30:25 <andythenorth> have you got FIRS nml nearby?
18:30:33 *** nielsm has quit IRC (Ping timeout: 480 seconds)
18:30:39 <peter1139> Can you override the original tiles?
18:30:46 <peter1139> Or is the limit of 512 actually 512 - 175?
18:30:47 <frosch123> peter1139: it's in the map array, so you probably do not want to support 64k tiles
18:30:55 <frosch123> i think forcing newgrf to use less tiles is a good thing
18:31:21 <peter1139> frosch123, I was only thinking of allowing NewGRFs to use the full 512 that seems to be available.
18:31:30 <peter1139> So it's still limited, just less limited.
18:31:34 <frosch123> quite possible the 175 are not used unless you use the "override" property
18:31:42 <frosch123> which probably noone uses
18:31:53 <ahyangyi> andythenorth: Thanks!
18:32:03 <andythenorth> I'm looking what you need to search for
18:32:14 <ahyangyi> switch(FEAT_INDUSTRYTILES, SELF, bar_and_section_mill_tile_1_lc_2,
18:32:14 <ahyangyi> (nearby_tile_class( 2, 2) == TILE_CLASS_INDUSTRY) ||
18:32:14 <ahyangyi> (nearby_tile_class( 2, 1) == TILE_CLASS_INDUSTRY) ||
18:32:14 <ahyangyi> (nearby_tile_class( 2, 0) == TILE_CLASS_INDUSTRY) ||
18:32:14 <ahyangyi> (nearby_tile_class( 2, -1) == TILE_CLASS_INDUSTRY) ||
18:32:15 <ahyangyi> (nearby_tile_class( 2, -2) == TILE_CLASS_INDUSTRY) ||
18:32:15 <ahyangyi> (nearby_tile_class( 1, 2) == TILE_CLASS_INDUSTRY) ||
18:32:17 <ahyangyi> (nearby_tile_class( 1, 1) == TILE_CLASS_INDUSTRY) ||
18:32:17 <ahyangyi> (nearby_tile_class( 1, 0) == TILE_CLASS_INDUSTRY) ||
18:32:19 <ahyangyi> (nearby_tile_class( 1, -1) == TILE_CLASS_INDUSTRY) ||
18:32:19 <ahyangyi> (nearby_tile_class( 1, -2) == TILE_CLASS_INDUSTRY) ||
18:32:21 <ahyangyi> (nearby_tile_class( 0, 2) == TILE_CLASS_INDUSTRY) ||
18:32:21 <ahyangyi> (nearby_tile_class( 0, 1) == TILE_CLASS_INDUSTRY) ||
18:32:23 <ahyangyi> (nearby_tile_class( 0, 0) == TILE_CLASS_INDUSTRY) ||
18:32:23 <ahyangyi> (nearby_tile_class( 0, -1) == TILE_CLASS_INDUSTRY) ||
18:32:25 <ahyangyi> (nearby_tile_class( 0, -2) == TILE_CLASS_INDUSTRY) ||
18:32:25 <ahyangyi> (nearby_tile_class( -1, 2) == TILE_CLASS_INDUSTRY) ||
18:32:25 <frosch123> peter1139: yes, but it would allow one grf to take all, so you never get small grfs with hotels or stuff
18:32:27 <ahyangyi> (nearby_tile_class( -1, 1) == TILE_CLASS_INDUSTRY) ||
18:32:27 <ahyangyi> (nearby_tile_class( -1, 0) == TILE_CLASS_INDUSTRY) ||
18:32:29 <ahyangyi> (nearby_tile_class( -1, -1) == TILE_CLASS_INDUSTRY) ||
18:32:29 <ahyangyi> (nearby_tile_class( -1, -2) == TILE_CLASS_INDUSTRY) ||
18:32:31 <ahyangyi> (nearby_tile_class( -2, 2) == TILE_CLASS_INDUSTRY) ||
18:32:31 <ahyangyi> (nearby_tile_class( -2, 1) == TILE_CLASS_INDUSTRY) ||
18:32:33 <ahyangyi> (nearby_tile_class( -2, 0) == TILE_CLASS_INDUSTRY) ||
18:32:33 <ahyangyi> (nearby_tile_class( -2, -1) == TILE_CLASS_INDUSTRY) ||
18:32:35 <ahyangyi> (nearby_tile_class( -2, -2) == TILE_CLASS_INDUSTRY)
18:32:37 <ahyangyi> I think I already understand what you did 😛
18:33:03 <andythenorth> because of the way FIRS auto-generates nml, it's quite trivial
18:33:13 <peter1139> ahyangyi, please don't paste so much junk into the channel.
18:33:21 <andythenorth> unless I run out of IDs, it would be more work to give each tile a unique ID
18:33:31 <peter1139> (That's andythenorth's job)
18:33:52 <andythenorth> I used 183 IDs so far eh
18:34:25 <ahyangyi> peter1139: Sorry, I edited the message now; though I don't think that edit can proprgate to IRC 😮
18:34:40 <peter1139> truebrain, I decided to start small, and deal with IndustryCargoesWindow::ind_cargo, which is either an IndustryType, or a CargoID - NUM_INDUSTRYTYPES.
18:35:02 <peter1139> But they're byte/uint8_t :-)
18:37:03 <peter1139> (Although the reason I am looking at this is because I make CargoID a strong type, and that makes that union fail...)
18:38:53 <DorpsGek> - Update: Translations from eints (by translators)
18:39:26 <ahyangyi> I see that the industry tile switch can use the PARENT (the industry?)'s layout_num
18:40:15 <ahyangyi> time to study nmlc code I guess...
18:43:11 <ahyangyi> ( hmm, it's called `related_scope` in grf-py. I think I know how to do it now... )
18:48:26 <truebrain> now to test it properly for aircrafts 😄
18:48:55 <andythenorth> ahyangyi: wonder if it was already built into grf-py? Or is that what you're doing?
18:52:28 <ahyangyi> AFAIK grf-py doesn't support anything related to industry layouts at this moment
18:53:05 <peter1139> truebrain, would you expect anything different for aircraft?
18:53:08 <ahyangyi> And... I don't know what I am doing
18:53:21 <truebrain> peter1139: nothing; but I didn't test it, so it is good to do so
18:54:40 <truebrain> for aircrafts it was, rightfully, pointed out that `v->tile` is 0 when in the air
18:54:46 <truebrain> so I assume this is already updated when cargo is handled
18:54:52 <truebrain> but .. I just want to be 100% sure 🙂
18:55:16 <truebrain> especially during savegame upgrades 🙂
18:55:42 <LordAro> truebrain: hmm, consul
18:55:51 <truebrain> LordAro: I never bothered
18:56:17 <LordAro> oh, how did you do service detection then?
18:56:29 <truebrain> since Nomad 1.3? they have a built-in variant
18:56:35 <truebrain> it is less powerful, but more than sufficient for OpenTTD
18:56:47 <truebrain> it is a bit hidden in their documentation that it exists these days
18:56:52 <truebrain> as well .. they make money with Consul 😛
18:57:22 <LordAro> i was trying to spin up a test nginx job earlier and it just shouted at me about missing consul
18:57:43 <truebrain> yeah, you need to explicitly tell it to use nomad service discovery, instead of consul
18:58:39 <peter1139> Hmm, if I just remove that union, the six arrays could be cut down to three. But naming may not be ideal...
19:00:20 <truebrain> LordAro: very specially: `provider = "nomad"`
19:00:32 <truebrain> it has limitations, but for most goals it works "good enough"
19:00:59 <truebrain> ironically, today I learnt that in my dayjob they are using proxmox. So I actually knew what it was this time 😛
19:03:00 <LordAro> truebrain: yup, that appears to work!
19:03:07 <LordAro> well, the job runs, at any rate
19:03:19 *** esselfe has quit IRC (Remote host closed the connection)
19:03:21 <LordAro> i've gone with podman rather than docker, because it sounded better
19:04:10 <truebrain> Docker is "reinventing" itself to stay relevant, but is doing a lot less innovation on docker itself
19:04:46 <truebrain> (their reinventing is a good call in my opinion btw; just to be clear 😄 )
19:05:02 <truebrain> podman on the other hand is going full-in on the containerization, and is doing some nice things
19:05:11 <truebrain> like rootless containers, which is really nice for local setups
19:05:24 <truebrain> no more files that are owned by `root` because a container created them, and shit 😛
19:06:06 <LordAro> success, i appear to have successfully hosted an nginx container
19:06:18 <truebrain> babies first steps!
19:06:37 <frosch123> to encourage reviewers of #11306: only the first commit is horrible, the others are readable 🙂
19:06:53 <LordAro> tomorrow's problem is getting the networking to make it accessible at :8080 as requested, rather than :23847
19:07:06 *** esselfe has joined #openttd
19:07:30 <truebrain> LordAro: that will be a fun thing to read up on .. you have like 20 ways of doing that 😛
19:07:41 <truebrain> feel free to check OpenTTD/infra to see what route I went with 😛
19:07:50 <LordAro> and also convincing the marketing team that adding inline custom css/js to every vaguely custom drupal page is not how drupal is Supposed To Work
19:07:56 <truebrain> but it is important to understand a bit how you can setup networking in Nomad
19:07:59 <LordAro> truebrain: i've seen traefik mentioned in a few places
19:08:21 <truebrain> You could; but that is mostly to handle edge-traffic, mostly when you want https
19:08:32 <truebrain> for OpenTTD I only run nginx, and have cloudflare do the https part
19:08:47 <LordAro> yeah, i don't have cloudflare :p
19:08:59 <truebrain> no, but do you need https for internal website, is a question you have to answer 🙂
19:09:08 <truebrain> as https requires a domain, means your domain gets on the transparancy logs ..
19:09:23 <truebrain> it is not a good / bad question; just a question that needs an answer 😛
19:09:41 <truebrain> if you do https, traefik makes it easier, as it does letsencrypt (or possibly others) for you
19:09:55 <LordAro> we already have a few wildcard certs, that's not particularly a problem
19:10:16 <truebrain> wildcard certs could be done by nginx too 🙂 But traefik is a nice piece of software, makes routing rather easy
19:10:28 <truebrain> I believe it even has nomad service discovery, so it can be done fully automated
19:10:54 <truebrain> frosch123: the first commit is a perfect place to add a trojan to the code
19:10:57 <truebrain> nobody is going to notice
19:12:24 *** gelignite has quit IRC (Quit: Stay safe!)
19:14:56 <truebrain> I am thinking here about the people writing the changelogs frosch123 🙂
19:15:45 <truebrain> for that I always kinda blindly remove codechange and cleanup. To have these "fix" commits in there, which don't actually fix anything, is just tedious to filter out 😛
19:15:55 <truebrain> there is another "fix" commit in the series, which should just be a codechange tbh
19:17:01 <frosch123> ok, so "fix warning" should be just "codechange"
19:17:27 <truebrain> I don't always remember this myself, but I tend to only use `Fix` for fixing something the user noticed
19:17:48 <truebrain> it is also a "fix" to change C-style stuff to C++-style 😛
19:19:07 <truebrain> `Codechange, Cleanup: Changes without intentional change of behaviour from user-point-of-view. Difference between "Codechange" and "Cleanup" is somewhat subjective.`
19:19:11 <truebrain> I am shocked, we even wrote that down
19:19:18 <truebrain> although ... `Fix, Revert: Fixing stuff.`
19:19:22 <truebrain> is as vague as it gets 😛
19:20:03 <truebrain> anyway, otherwise it looks fine to me frosch123 ; CI is happy, so I think it is fine anyway 😛
19:20:13 <peter1139> Sometimes my Codechanges are accidental Fixes too.
19:20:37 <truebrain> `/w34189 # 'identifier' : local variable is initialized but not referenced` <- ideally I would have a bit more information why we ignore that warning, but it is MSVC ... I don't think anyone actually cares 😛
19:21:15 <truebrain> `/w34100 # 'identifier' : unreferenced formal parameter` <- I don't even know what a "formal" parameter is ... do we have informal parameters?!
19:21:53 <frosch123> those warnings are enabled, not disabled
19:22:00 <truebrain> peter1139: I think that is why the "intential" got added to the sentence 😄
19:22:15 <truebrain> frosch123: huh? Lol .. I assumed `/w3` meant they are already in the 3-series
19:22:34 <truebrain> okay, fuck MSVC, I just don't care anymore 😛
19:23:24 <truebrain> "intential"? I can't even copy/paste anymore .. I will give up on English too, okay?
19:38:02 <frosch123> only rewording, no diff
19:38:52 <frosch123> and yes, i do not want to squash commits 2..N into commit 1 :p
19:38:53 <peter1139> Ah that's better, I no longer have missing engines.
19:44:18 <peter1139> Okay so aircraft no longer carry waste... that's a change. Hmm.
19:44:30 <peter1139> It's correct, but I wonder how that affects savegames. Hmm.
19:46:06 <blindvt> meh. i was wondering why regression_{regression,station} timeout on me after 1500 (!) seconds. write(2, "Error: Failed to find a graphics"..., and indeed, no graphic set on that pristine box. Would have been nicer if `make test` would fail faster in this situation ;)
19:46:39 <locosage> ahyangyi: It supports them on grf/nfo level, there just aren't any fancy classes like for vehicles yet
19:48:12 <_glx_> truebrain: we're in /W3, but thos 2 warnings are from /W4 so we move them to /W3 because /W4 warns too much
19:49:28 <blindvt> Rubidium: `bool my_cmd` can be removed completely. Are you sure?
19:50:23 <blindvt> Rubidium: there's a comment that this is somewhere used to distinguish user- and non-user stuff; didn't look if that comment was stale, too. You certainly know that better
19:52:19 <peter1139> Hmm, okay, existing aircraft can still carry the wrong cargo.
19:53:31 <frosch123> hmm, so even more stuff to remove
19:57:52 <truebrain> _glx_: My point was that they start with /w3, so one would assume it is in /w3 ....
19:58:28 <_glx_> yeah syntax is weird, but /w3nnnn is to change level
19:59:50 <peter1139> "Can't buy aircraft" well that's...
19:59:53 <frosch123> oh, it does not even enable them? it just moves them to a level, which happens to be enabled?
19:59:56 <peter1139> Feature: Remove aircraft.
20:02:23 <truebrain> frosch123: just please don't lose sight of time; I rather have this PR as-is with follow-up PRs to make it even better, than to have this PR drag on for weeks 🙂
20:03:19 <frosch123> i won't work on it today, maybe tomorrow, definitely not thursday, so maybe friday
20:03:28 <frosch123> time enough for you to just merge it first :p
20:03:47 <truebrain> well, I rest my case, yes 😛
20:04:11 <truebrain> But a week is fine .. just not that it ends up like the "make scrollbar position storage bigger" PR 😄
20:04:38 <truebrain> that PR is 3.5 years old now 😄
20:05:03 <frosch123> i may include #11316 then
20:05:21 <frosch123> the two PRs conflict anyway
20:05:26 <truebrain> I am also fine if you say, and Rubidium agrees ofc, let's take this PR as-is, and let's fix the "we can remove those parameters" after-the-fact
20:05:43 <truebrain> we might also want to check up on existing `maybe_unused` tags
20:08:02 <frosch123> aldot/blindvt also posted some other ununsed parameters last night, which are not detected by any compiler
20:08:10 <Rubidium> I've seen nothing shocking when reviewing it, just some minor remarks. I'm fine with pushing that to a follow-up PR
20:08:13 <frosch123> so these warnings are kind of weird
20:08:56 <truebrain> these kind of things are an endless battle after all 🙂
20:09:08 <truebrain> those `flags` I commented on are also unused, but in another way than you expect .. 😛
20:09:17 <truebrain> which requires a refactor of another level 😄
20:09:57 <truebrain> Anyway, I should also have addressed all things you mentioned for #11283 🙂
20:10:29 <frosch123> yeah, i was also unsure what to do with the flags 🙂
20:10:44 <truebrain> just too strict OO being applied 😛
20:10:46 <truebrain> that is all that is 😄
20:12:06 <truebrain> I might fix that one properly one of these days, if I don't forget .. as it is just silly, and I think I wrote it 😄
20:12:56 <andythenorth> so CHIPS station grf has 4 different ground tile types
20:13:09 <andythenorth> to match different FIRS industries / different roleplay
20:13:31 <andythenorth> kinda weird if I just duplicate every station tile to have each ground type?
20:13:39 *** muffindrake has joined #openttd
20:15:00 <truebrain> `Realtime distributed OLAP datastore, designed to answer OLAP queries with low latency` .. always nice, when a site explains nowhere wth OLAP is ... 😛
20:15:01 <brickblock19280> there isn't really a better alternative unless we have bgt
20:15:16 <truebrain> `online analytical processing` is the answer
20:15:23 <brickblock19280> the tiles with track could relay on industrial trackset
20:16:59 <frosch123> auto-merge scared me, i did not press that button :p
20:17:08 <truebrain> we all have that issue 😄
20:17:16 <truebrain> it should show a bit better what is actually happens, tbh
20:18:10 *** muffindrake has quit IRC ()
20:19:43 <peter1139> Ah right, when you de-overload enum values things may stop working :D
20:19:52 <peter1139> Maybe I should not do that.
20:22:37 *** muffindrake has joined #openttd
20:24:17 <peter1139> Damn it, maybe I should just remove libsdl2 dev.
20:25:34 <_glx_> andythenorth: these are magic tiles, they can have buffers, houses, cranes, ... depending on position
20:38:07 <andythenorth> I might do all the building tiles with different ground also
20:38:41 <andythenorth> it's cheap to generate, but adds clutter for the player
20:52:19 <frosch123> haha, just noticed there is a user "param", who i highlighted in a comment. i wonder how often that happens, people commenting on doxygen `@param`
20:55:48 <andythenorth> so....station cargo subtype refits? 😛
20:55:55 <andythenorth> but use them to set ground texture?
20:56:39 <frosch123> did you try the dutch road funiture objects?
20:57:36 <frosch123> you build them next to the road, and they overlap the road sprite
20:58:11 <frosch123> so you could use on station tile for the ground, and the tile behind/in-front for stuff on-top
20:59:46 <andythenorth> is it the best thing ever? 🙂
21:03:13 <frosch123> i am already happy you did not suggest ctrl+clicking station tiles to cycle graphics
21:03:31 *** keikoz has quit IRC (Ping timeout: 480 seconds)
21:05:01 <andythenorth> doesn't ctrl already do something? PP
21:05:13 <andythenorth> oh there is an overbuild cycle I think for stations?
21:05:23 <andythenorth> some sets do some magic on repeated building
21:21:24 <andythenorth> I wonder if organising station tiles by class is the thing
21:21:44 <andythenorth> what if they were flat, but the text search was more powerful?
21:21:59 <andythenorth> and maybe they had taxonomies (tags or so)
21:24:43 *** muffindrake has quit IRC (Quit: muffindrake)
21:25:57 *** muffindrake has joined #openttd
21:25:59 *** muffindrake has left #openttd
21:31:44 <andythenorth> I never quite know where to look in ISR, even though it has clear logic for classes 🙂
21:32:09 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
21:34:25 <peter1139> I have a patch for searching but it's a bit overwhelming...
21:34:45 <peter1139> (And if you have enough station NewGRFs loaded it trips up on the scrollbar limit :D)
22:18:14 *** tokai|noir has joined #openttd
22:18:15 *** ChanServ sets mode: +v tokai|noir
22:24:55 *** tokai has quit IRC (Ping timeout: 480 seconds)
continue to next day ⏵
