IRC logs for #openttd on OFTC at 2024-02-15
⏴ go to previous day
00:00:44 <Eddi|zuHause> the original TT came on 2 diskettes, so less than 3MB
00:01:55 <Eddi|zuHause> (i think installed disk space was a little bit more)
00:02:43 <reldred> Our desktop PC in 1996 had a 2GB hdd, except we later found it was actually 3GB and IBM formatted it by default without support for bigger than 2GB partitions, and their recovery software formatted the drive as such as well. Yet lo and behold when it finally got replaced and we stripped the PC for giggles we found it had a whole-ass 50% extra space...
00:03:17 <reldred> Back in those days we had to choose whether at any given time we were going to have The Sims or Diablo 2 installed. Not both.
00:03:35 <reldred> There were a lot of experimentations with ZIP Drives, but god damn were they slow.
00:03:43 <reldred> Satisfyingly noisy though.
00:13:10 <Eddi|zuHause> reldred: 2GB was probably a BIOS limitation
00:13:50 <reldred> Also I'm really smart, I just realised that if you want SNMP to work on both routers in a redundancy pair you really should make sure SNMP instance isn't trying to use the innactive VRRP interface.
00:14:32 <reldred> Glorious! We're getting data now!
00:14:54 <peter1138[d]> Loopback interfaces.
00:15:15 <reldred> Mikrotik have only *just* started exposing loopback interfaces. I'm still on 7.8
00:15:44 <reldred> I'll move up to a newer version when I convert the whole setup over to MLAG's and layer 2 leaf and spine
00:16:36 <peter1138[d]> On Mikrotik, just create a bridge with no ports. That'll be a loopback interface.
00:17:25 <reldred> I already had a handy interface for such management purposes available so I just bound it to that
00:17:52 <reldred> I really wanna play with MVRP in Mikrotik's as well, it's coming in 7.14
00:18:10 <reldred> Will make distributing customer vlanid's in our datacentre significantly easier
00:20:58 <reldred> Mine's just a vlanid on the main bridge. This network is still using a mixture of old vlaning and new style vlan filters on bridges depending where it is and what it's doing, but I'll be doing everything with vlan bridge filters on the new architecture.
00:21:42 <peter1138[d]> Dunno what font this is using but that 9 is pretty stylized.
00:21:54 <reldred> unfortunately it means one of my two core switches is sitting at 40-50%cpu which is disgusting but that switch has to do a couple things that are impossible with the new hardware accelerated bridge vlan filters.
00:22:17 <reldred> but those tasks will switch to the two routers which will become the border gateway leaf.
00:22:29 <reldred> which can hapilly do it all night long
00:22:34 <reldred> without eating their CPU's
00:22:48 <reldred> yeah, joys of interfacing with customers BYO carriers,
00:23:34 <reldred> and a fucky ENNI with some services that were ordered incorrectly which then resulted in some poor spot-decisions on how to configure that boxed me into a corner later on.
00:23:58 <peter1138[d]> Winbox is missing chunky bevels 😦
00:24:15 <reldred> yeah it could do with chunky bevels
00:26:00 <reldred> but yes the next goal is kill RSTP (well, minimize it) and roll out mlag top to bottom. Atm. it takes between 30s and 3m for RSTP to settle on topology changes which is admittedly better than my predecessors design (everything just fails because there were no redundant paths), but I can do a lot better now.
03:00:45 *** Wormnest has quit IRC (Quit: Leaving)
03:11:09 *** debdog has quit IRC (Ping timeout: 480 seconds)
03:14:37 *** gnu_jj_ has joined #openttd
03:18:05 *** gnu_jj has quit IRC (Ping timeout: 480 seconds)
03:35:11 *** D-HUND is now known as debdog
08:29:26 *** keikoz has quit IRC (Ping timeout: 480 seconds)
11:30:14 <peter1138[d]> Bah at std::distance not working with const vs non-const.
11:32:17 <andythenorth> is it for grf docks? 😄
12:07:35 *** Ox7C5 has quit IRC (Ping timeout: 480 seconds)
12:11:35 <xarick> I have a problem with Visual Studio 2022... every 7 days of usage, I have like 200 GB of temporary files written to disk...
12:11:55 <xarick> considering this is an SSD... I'm worried
12:12:24 <xarick> already got alerts from windows about low disk space
12:12:33 <xarick> this didn't use to happen, until some recent updates
12:13:07 <xarick> windows disk utilities don't clear these temp files
12:13:30 <xarick> it says they do, but when I refresh, they're not deleted, they're there again
12:13:54 <xarick> I only use Visual Studio for OpenTTD
12:14:39 <xarick> I solved both alerts about low disk space by manually deleting the files myself
12:14:53 <xarick> both times, 200 GB of temp data from Visual Studio 2022
12:26:15 <xarick> they're forcing me to buy a 2nd SSD
12:26:57 <xarick> a 4 TB SSD is expensive 😦
12:28:27 <xarick> I wish I wouldn't be forced to go back to HDDs
12:28:58 <belajalilija> I run my PC on an SSD but have HDD for storage
12:29:16 <belajalilija> It’s plenty fast enough imo
12:29:36 <xarick> this temp data is written to C:
12:29:59 <johnfranklin> 10 mins to launch Cities Skylines
12:30:13 <belajalilija> It’s only a few for me
12:30:33 <xarick> I must have enabled some stupid beta setting on visual studio when I was asked
12:30:45 <xarick> and now I'm paying the price
12:31:02 <xarick> can't remember what I enabled..
12:31:17 <xarick> it didn't use to do this
12:31:23 <belajalilija> belajalilija: But i also have a CPU from only like 2021 and 128gb of ram
12:32:13 <belajalilija> I inherited some money
12:32:33 <peter1138[d]> xarick: maybe set up a RAM disk.
12:32:42 <belajalilija> My pc needed upgrade
12:32:43 <peter1138[d]> (Not sure how, so you'll need to look that up)
12:32:57 <belajalilija> It was running on 2015 hardware
12:33:17 <peter1138[d]> It'll fill up RAM instead so might cause other issues, but...
12:34:34 <_glx_> VS tried to make me activate ReFS
12:35:33 <johnfranklin> Is upgrading CPU possible?
12:45:12 <peter1138[d]> Okay, is there such a thing as a small gaming PC these days...
12:45:31 <peter1138[d]> Not super high end, but... uh, better than whatever andy's kids bought :p
12:46:03 <peter1138[d]> Not that small...
12:47:45 <peter1138[d]> And "PC" tends to mean "desktop PC", otherwise I'd've said laptop or portable.
12:51:39 <_glx_> xarick: My guess is it's the PDB files
12:52:03 <peter1138[d]> I found several GB in my /tmp caused by clang crashing, using on fmt:: stuff.
12:52:29 <peter1138[d]> `&*classes.cbegin()` is quite ugly 😦
12:53:15 <peter1138[d]> Get index in vector: `this - classes.data()` -> `std::distance(&*classes.cbegin(), this)`
12:53:34 <peter1138[d]> Wrong dereference 🙂
12:54:37 <peter1138[d]> > Per-window zoom levels - Adjust the zoom level for each window independently.
12:54:56 <peter1138[d]> > Hey Code voice command - Start a chat session with a voice command.
12:55:06 <pickpacket> I just realised something funny
12:55:09 <peter1138[d]> I'm pretty sure that is not something I want in my code editor 😦
12:55:20 <pickpacket> you can build tunnels under mines and oil wells
13:03:11 *** tony_pixel has joined #openttd
13:03:11 <tony_pixel> Holy shit you're right
13:03:17 *** snarkyerica has joined #openttd
13:03:17 <snarkyerica> peter1138[d]: i mean... my rig fits inside an ikea Kallax slot
13:03:29 <snarkyerica> it's pretty small as far as PCs go lol
13:04:45 <peter1138[d]> MINISFORUM stuff might fit the bill, but GamersNexus had some issues 🤷
13:07:38 <peter1138[d]> Hmm, maybe something MiniITX based but with enough space for a proper GPU.
13:29:14 <LordAro> i bought my new computer bits the other day
13:29:23 <LordAro> 7800X3D & associated MB & RAM
13:29:58 <peter1138[d]> "Gaming Processer" 😮
13:30:17 <LordAro> it's as high spec as i can reasonably justify
13:30:31 <LordAro> given i barely play games these days anyway
13:30:59 <LordAro> but i've got to wait for Noctua to ship me a fan mount adapter first
13:31:13 <LordAro> because AM5 was not a thing when i bought it in 2016
13:36:09 <peter1138[d]> My partner wants a new PC. Been using a laptop for years, and the old desktop is... a Phenom. Back when that was the AMD's path.
13:36:34 <LordAro> the first computer i built was a Phenom
13:36:39 <peter1138[d]> About the only thing worth keeping is the SSD I guess
13:39:40 <LordAro> NVMes aren't horrendous these days
13:41:55 <peter1138[d]> Worth keeping as in I can stick it in something else and use it for /tmp without worrying about it killing the main SSD 😉
13:42:31 <LordAro> though you have to try really hard to actually kill an SSD
13:42:58 <LordAro> unless you're writing >100GB a day i wouldn't worry about it
13:49:25 *** Xaroth9232 has joined #openttd
13:56:35 *** Xaroth923 has quit IRC (Ping timeout: 480 seconds)
13:58:02 <peter1138[d]> We both know that, I was paraphrasing from earlier 🙂
14:01:04 <LordAro> i can't be expected to read all that
14:11:58 <xarick> you can create an empty list with `local list = AIList();`, but nowhere in the documentation I could find that you could do that.
14:15:36 <LordAro> is the intention to disallow creating a bare list?
14:36:25 <_glx_> AIList is the bas for other lists, no need to document the constructor
14:50:36 <xarick> I'm creating a `GetTileOfAirport`
14:53:05 <peter1138[d]> As in, what problem are you trying to solve?
14:53:51 <xarick> I want to get a non-hangar tile to use as order destination
14:56:16 <peter1138[d]> I think you are trying to work around instead of solving a problem.
15:02:12 <peter1138[d]> Hmm, okay, maybe changing that bad design is too much effort (breaks too much)
15:03:03 <andythenorth> in case you were wondering
15:03:20 <peter1138[d]> I was wondering.
15:05:58 <xarick> all this time I thought OpenGFX+ Airports 0.5.0 was letting AIs to place rotated airports
15:08:09 <xarick> why are some airports repetitions...
15:15:48 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
15:16:29 <xarick> this was not what I expected
15:23:59 <peter1138[d]> What's the issue?
15:25:16 <peter1138[d]> ScriptAirport::BuildAirport doesn't have a parameter to change rotation.
15:27:08 <xarick> I thought each type was a rotated version of an airport
15:28:37 <xarick> is there a reason why there are repeats?
15:31:53 <xarick> 0/10, 1/15, 2/19, 3/16, 4/17, 5/13, 6/20, 7/18, 8/22
15:32:59 <xarick> seems to me the first 9 airports were supposed to be not available for AIs
15:33:52 <xarick> but doing that is guaranteed AIs won't work
15:36:37 <peter1138[d]> Having the default list of airport types available as an enum is a misfeature.
15:36:42 <peter1138[d]> Should not exist.
15:43:01 *** HerzogDeXtEr has joined #openttd
15:44:11 <xarick> we can remove the enum
15:46:34 <peter1138[d]> There are 14 airports with OpenGFX+ airports loaded.
15:51:48 <peter1138[d]> Yeah, this is quite badly designed 😄
15:52:18 <peter1138[d]> airports have a class-based system, but also have an id override manager sytem.
15:52:24 <peter1138[d]> The UI uses the class-based system only.
15:52:33 <peter1138[d]> But for scripts it uses the override manager.
15:53:14 <peter1138[d]> So players can only see the active airports, but scripts (and the game, internally) can see all the old IDs as well. But they are mapped to a substitute within the class-based system.
15:53:42 <peter1138[d]> This override manager system is quite flawed, I think.
15:54:16 <peter1138[d]> Probablky a hold-over from TTDPatch design -- it allows some things to fallback to defaults if a NewGRF is removed, but they aren't going to work anyway, so what's the point.
15:55:24 <peter1138[d]> Meanwhile, especially for industries and houses, you waste many IDs as the original IDs aren't reused.
15:59:31 <peter1138[d]> Do AIs get to choose newgrf station / roadstop type?
16:00:08 <xarick> there's some airport that has holes in it
16:00:32 <xarick> it's not airport tile, it's not hangar tile, but st->airport.Contains it
16:01:08 <peter1138[d]> I recommend implementing a GetAirportTerminal() call, and only list the first terminal, or GetAirportTerminals() and listing all terminals.
16:01:30 <peter1138[d]> For orders, you only care about hangar vs terminal.
16:01:38 <peter1138[d]> There's no reason to list all tiles.
16:02:47 <peter1138[d]> `GetTerminalOfAirport()` would be fine, and match the equivalent existing `GetHangarOfAirport()`
16:03:26 <peter1138[d]> `GetTerminalsOfAirport()` could potentially allow selecting a specific terminal, although the game doesn't care about that.
16:04:39 <peter1138[d]> Same could be said about a possible `GetHangarsOfAirport()` though, and in fact for hangars the game *does* care which hangar tile it is.
16:06:57 <peter1138[d]> Can we make AIs look at preview images and, compare them, and make them pick aesthetically pleasing choices? 😄
16:07:34 <xarick> maybe if it could read names
16:08:50 <peter1138[d]> In terms of AI 'understanding', text is just as useless as images.
16:11:22 <xarick> it would be interesting to choose which terminal group to pick
16:11:44 <xarick> make the aircraft forcefully land on one side or the other
16:20:52 *** Wormnest has joined #openttd
16:35:06 <xarick> technically not looking for the terminal tile, but...
16:46:59 <peter1138[d]> Hmm, I guess maybe terminals are part of the state machine rather than the tile layout.
17:02:29 <xarick> GetNonHangarOfAirport?
17:02:52 <_glx_> isn't helistation the most problematic one ?
17:03:23 <_glx_> st->xy points to the hangar
17:03:47 <xarick> st->xy as I learned, could even point to bare land
17:04:06 <_glx_> ah yes there's one layout with a hole there
17:04:42 <xarick> it's a problem I found with ships
17:04:53 <xarick> but now i'm on aircraft
17:06:25 <xarick> static TileIndex GetNonHangarOfAirport(TileIndex tile);
17:06:25 <LordAro> the assert seems redundant
17:07:22 <_glx_> yeah t being in st->airport is kinda mandatory
17:07:41 <_glx_> as you're looping over st->airport
17:08:12 <xarick> holes are still part of st->airport though
17:08:23 <xarick> kinda messed up, but okay
17:08:31 <_glx_> yes it's just an area IIRC
17:16:37 *** edchapter has joined #openttd
17:16:37 <edchapter> Hi, I was doing translation recently, and I want to check all my strings before they commit to the main branch. Is there a way to do this?
17:23:18 <xarick> I need a good name: GetNonHangarOfAirport, GetTileOfAirport, GetDestinationOfAirport, GetStationOfAirport,...
17:24:18 <_glx_> edchapter: The only way is to clone the source, dowload the lang file from translator and build
17:24:42 <_glx_> or you can wait for the nightly build then you can check them
17:25:15 <_glx_> and fix if needed for the next one
17:26:32 <edchapter> _glx_: Ok. I will consider doing this, thanks.
17:26:56 <_glx_> you can get it from steam
17:34:30 <xarick> GetNonHangarOfAirport in an unelegant name, but right to the point
17:38:00 <LordAro> i think i'd rather "GetTileOfAirport" with a note in the docs to say that it's guaranteed not to be a hangar
17:42:51 <xarick> * Get a tile of the airport guaranteed not to be hangar.
17:47:15 <edchapter> What is {P 0 "" s} referring to?
18:08:50 <xarick> next and probably last function I wanna add
18:34:53 <DorpsGek> - Update: Translations from eints (by translators)
18:38:53 <peter1138[d]> Forums are great. "I set up a 4kx4k map, and AIs have 3000-5000 vehicles each, why is it laggy?"
18:48:08 <peter1138[d]> There is no "++" operator on that type.
18:53:13 <nanapipirara> It would be nice if you could click on an industry in the _Industry chain_ window, and it would open the _Industries_ window with filters set so that all industries of the type you clicked would show up in a list. 👀
18:57:49 <frosch123> there is a button to link it to the small map
18:58:04 <_glx_> where they are filtered
18:58:22 <xarick> `(::AirportTypes)0` this is ugly looking, magic number... etc
18:58:29 <locosage> peter1138[d]: there's more in topic than just that
19:00:59 <nanapipirara> frosch123: Yeah that's nice too, but the Industries window allows for clicking to position the camera at the industry location.
19:01:54 <frosch123> so does the smallmap ?
19:02:49 <nanapipirara> I guess. Maybe it's a _me_ problem, but I find the mini map very cumbersome to navigate
19:03:36 <nanapipirara> Maybe it's because I play on a laptop screen most of the time
19:04:45 <nanapipirara> When I hook up to a big screen it's easier to use the mini map
19:05:03 <nanapipirara> Just a thought :p
19:08:02 <peter1138[d]> Also it's "annoying" that "find missing content online" for a savegame will only look for NewGRFs, not AI/GS.
19:09:14 <locosage> peter1138[d]: according to the player it worked fine until they added a few trains while on pause
19:09:42 <locosage> I've no idea how it managed to work fine but something doesn't add up there
19:11:06 <locosage> also, it's only a 6 year game so like 1.5 hours, kinda qualifies as "just started"
19:12:01 <locosage> it's not players fault that ai managed to lag it into death so fast
19:13:23 <locosage> well, opcode limit aside
19:17:57 <locosage> might be something about that ai that made it suddenly start to lag so much
19:18:56 <peter1138[d]> Would be nice if debugging squirrel was feasible.
19:20:25 <_jgr_> The code looks like it just spins, using as many opcodes as allocated to it
19:21:28 <_jgr_> There is no sleep in the main loop
19:22:58 <peter1138[d]> Anyway, something is up with settings because I can't change #opcodes.
19:24:04 <_jgr_> Vanilla does not support changing those in a nexisting game
19:25:47 <peter1138[d]> Oh, so the error is wrong then.
19:50:28 <peter1138[d]> Nice, fixed it. But I had to change the API, and the AI...
19:51:42 <peter1138[d]> Oh actually, it doesn't need an API change 😄
19:52:56 <peter1138[d]> // local company_vehicles_count = AIVehicleList().Count();
19:52:56 <peter1138[d]> local company_vehicles_count = AIGroup.GetNumVehicle(AIGroup.GROUP_ALL, AIVehicle.VT_ROAD);
19:55:13 <peter1138[d]> Hmm, my API-modded method is slightly faster, but this is still much faster than without the change.
19:58:41 <truebrain> Lol, building a list to get the count .. yes, that is expensive 😄
19:58:55 <peter1138[d]> It would be fine with the normal limits.
19:59:41 <_jgr_> The Squirrel language encourages using lists for everything
20:00:56 <_jgr_> Script developers can't be expected to understand what that actually entails from a performance point of view
20:01:25 <truebrain> Don't think anyone expected them to 🙂 hence my suggestion for a JIT 🙂
20:07:51 <peter1138[d]> Well, I have no idea how such a thing would be implemented.
20:08:14 <andythenorth> Squirrel is nuts eh
20:08:15 <peter1138[d]> And of course the squirrel code for that is nothing.
20:09:01 <peter1138[d]> Optimizing that isn't really JIT either.
20:10:05 <andythenorth> Truebrain made a modern gameplay modding spec once 😛
20:10:19 <andythenorth> those were great days 🙂
20:10:22 <truebrain> wow, translators have been busy ... many languages are done now 🙂
20:10:31 <truebrain> 17 are fully translated
20:10:50 <truebrain> (47 are not, but still, great progress)
20:13:45 <Rubidium> talking about translations, what are we going to do with the number format stuff? 14 or 15 or never?
20:17:22 <wensimehrp> I was also thinking of that
20:20:10 <truebrain> Rubidium: when approved -> merge, I would say. And hopefully before we branch, if you ask me.
20:20:15 <peter1138[d]> Merge and do beta4?
20:21:39 <peter1138[d]> It's changed a bit since I last looked, it was language-file only.
20:25:37 <xarick> Disaster_Zeppeliner_Init is targeting airport.type AT_SMALL or AT_LARGE
20:25:51 <xarick> what about the other types and NewGRF airport types?
20:26:13 <xarick> some are complete copies
20:31:15 <xarick> {Disaster_Zeppeliner_Init, 1930, 1955}, // zeppeliner heh
20:32:52 <xarick> AS(city, 6, 6, 1955, CalendarTime::MAX_YEAR, 5, 5, 24, ATP_TTDP_LARGE, APC_LARGE, STR_AIRPORT_CITY, SPR_AIRPORT_PREVIEW_LARGE),
20:33:14 <xarick> 1955, interesting, there's a tiny chance the zeppeliner can crash onto a large airport
20:39:16 <locosage> that's some error I've never seen before xD `Error: Okay... something went horribly wrong. I couldn't load the fallback sprite. What should I do?`
20:41:06 <kuhnovic> Does anyone know why the tile number is displayed in hex? I find it rather annoying, especially during a debugging session where you get regular base-10 numbers (in Visual Studio that is)
20:41:33 <truebrain> that sounds like a VS issue more than the debugging 😉
20:43:48 <_jgr_> The tile index is inherently in two parts
20:44:20 <locosage> I can't remember any case where hex there was useful
20:44:27 <locosage> but can remember plenty where dec would be
20:44:56 <_jgr_> I find that a bit weird TBH, does the MSVC debugger not understand the 0x prefix?
20:45:15 <truebrain> kuhnovic: What is fun about these lines of thinking, that if you would change it to decimal, and someone with a debugger outputing hex could go: Why isn't it hex? My debugger does it in hex! And changes it back 😉 Sorry, I had an internal laugh about that flip/flop situation 🙂
20:46:04 <Rubidium> truebrain ought to know why it's in hex, he committed it ;) :D
20:46:08 <truebrain> `Co-ordinates` .. why is there a `-`? 😄
20:46:09 <kuhnovic> The TileIndex is just a uint32_t under the hood, why would any debugger show that in hex? (this is coming from a very window/VS-focused dev...)
20:46:27 <truebrain> kuhnovic: because why would it show decimal? 🙂
20:46:36 <kuhnovic> Because I don't speak robot 😛
20:46:37 <truebrain> why is one representation superior to the other, really? 😄
20:46:56 <truebrain> open regedit, enjoy the robot speak 😛 😄
20:47:33 <truebrain> anyway, more serious, it is most likely from pre-mapsize area, as the first 2 nibbles are the Y, the other 2 are the X, in hex
20:48:38 <kuhnovic> ... which doesn't really work with maps larger than 256x256 😛
20:48:48 <truebrain> well, not completely true btw
20:48:53 <truebrain> in hex bits are easiest understood
20:49:01 <truebrain> in a decimal number you are not going to see which bits are active
20:49:04 <truebrain> in hex that is pretty trivial
20:49:40 <kuhnovic> It's clear I haven't had my fair share of bit-wrestling yet 😛
20:49:41 <truebrain> think about it like this: at least it is prefixed with `0x` 🙂
20:49:58 <truebrain> the other argument is that it is less typing
20:50:23 <Rubidium> I doubt there will be many that would get mad if you were to make a PR that changes it
20:50:26 <truebrain> the arguments for decimal are fewer ... 😛
20:50:36 <kuhnovic> I'm not against having it if it has a use. I was just wondering what the usecase was. And personally I'd like to have the decimal equivalent next to it, as that can help a lot with debugging.
20:50:39 <truebrain> "because MSVC does show values in decimal" is the best argument there 😛
20:50:58 <frosch123> kuhnovic: iirc there is a console command "scrollto", which requires this format
20:51:11 <Rubidium> scrollto doesn't require hex
20:51:13 <frosch123> or because you want to index the map array in a debugger
20:51:22 <_jgr_> kuhnovic: The first log_2(x size) bits are the x, this is the case even if it isn't 256 x 256
20:51:49 <truebrain> frosch123: it supports both; 0x work, and decimals work
20:51:57 <truebrain> someone was actually paying attention when implementing that command, it seems 😛
20:52:02 <Rubidium> scrollto defaults to decimal, unless you prefix it with 0x..
20:52:19 <frosch123> ok, fine, change it to decimal then 🙂
20:52:25 <truebrain> that is more typing!
20:52:31 <truebrain> and my debugger shows it in hex!
20:52:38 <truebrain> and <insert other bullshit reasoning here>
20:52:40 <Rubidium> not for relatively small maps ;)
20:52:41 <truebrain> (goes both ways 😛 )
20:52:47 <kuhnovic> Spacebar heating! Pitchforks!
20:52:50 <truebrain> Rubidium: no map small enough!
20:52:52 <locosage> you can show them in hex but that's just unnecessary extra step that needs to be remembered
20:53:07 <truebrain> kuhnovic: yes! Nu ben je ingeburgerd!
20:53:15 <Rubidium> 0x0001 is way more than 1, 0xFFFF is more characters than 65535
20:53:43 <truebrain> Rubidium: pfff, 0x aren't characters! I don't need to remember those! 😛
20:53:47 <_jgr_> Hexadecimal really shouldn't be difficult to remember given the amount of development that you've done
20:54:47 <truebrain> kuhnovic: still can't get over `Co-ordinates`, but I don't see that in any of our language files, so not sure what you did there 😛
20:55:11 <_jgr_> In the spirit of bike-shedding it'd make more sense to remove the tile index entirely unless the developer setting is turned on
20:55:11 <Rubidium> truebrain: it's in US English
20:55:12 <kuhnovic> Don't look at me, I didn't put that there 😛
20:55:22 <locosage> quite a common case that I had is that I have tileindex in dec and can't find that tile in the game
20:55:27 <truebrain> Rubidium: ah .... always forget that even exists 😛
20:55:32 <locosage> because scrollto doesn't show exact tile and ? shows hex
20:56:24 <truebrain> _jgr_: should that whole window even be available if developer settings are turned off?
20:56:59 <_jgr_> Yes, it tells you what type of tile it is, whoe owns it, the rail/road type name, etc
20:57:01 <kuhnovic> I guess it's useful to find out what town a tile belongs to
20:57:12 <truebrain> always found it odd you have to discover it like that
20:57:38 <truebrain> feels like it is compensating 😄
20:57:51 <_jgr_> How else would you discover it?
20:58:13 <truebrain> the only times I used it in actual games, is to find out why I can't destroy a road, as it belongs to another town than I expected
20:58:57 <locosage> there is a lot of weirdness when two towns intersect
20:59:05 <locosage> town can't grow on roads of another town, etc
20:59:08 <truebrain> anyway, always wondered about the actual usefulness of that window, and whether there are not better ways to get there 🙂
20:59:23 <truebrain> the internal coordinates of something feels "developer" to me at least 🙂
20:59:49 <kuhnovic> I like it a lot as a debugging tool. Other than that I never use it.
20:59:50 <locosage> that window is so useful I have it on hover instead of click in cmclient
21:00:14 <locosage> and that's not because of debugging
21:00:22 <locosage> for debugging I don't use it that often
21:00:31 <_jgr_> If you use fancy objects, GRF stations, etc it is handy to see what things are
21:01:02 <locosage> it's invaluable for server moderation
21:01:08 <locosage> to see what company is being naughty
21:01:22 <truebrain> _jgr_: so to bikeshed some more, I would argue you can remove the whole position without developer setting on 😛
21:01:32 <locosage> in cb to see house population and so on
21:02:06 <peter1138[d]> TileIndex as a number isn't very useful there anyway.
21:02:26 <locosage> for development it is useful
21:03:03 <truebrain> so kuhnovic , next time you wonder about something, just keep it to yourself would you 😛 😛 😛 (sorry, long day, so I am just being a troll at this point)
21:03:14 <peter1138[d]> Anyway, 22ms per tick instead of 650ms per tick is somewhat better for that 4kx4k savegame.
21:03:16 <locosage> kuhnovic: _DEBUG != development :p
21:03:44 <kuhnovic> truebrain: Hell no, I love throwing some fuel on the fire 😄
21:03:51 <kuhnovic> locosage: Fair enough
21:04:06 <peter1138[d]> Show all of binary, decimal, octal and hexadecimal formats.
21:04:40 <kuhnovic> And draw prime numbers in purple
21:06:08 <locosage> peter1138[d]: you forgot fibonacci coding :p
21:09:27 <Rubidium> also... bear in mind that with _DEBUG builds (and -d misc=1) the tile's location is dumped into the Debug output which is even easier to put into your debugger ;)
21:11:57 <kuhnovic> Rubidium: True true, but the land info too works in release as well. Pro's and con's to everything.
21:12:27 <kuhnovic> rubidium42viaGitHub: Speak of the devil...
21:16:04 <Rubidium> kuhnovic: the Debug dumping works in release builds as well, you just need to up the debug level misc a bit. That could be seen as the source of the hexadecimal representation, with a decimal representation in the UI (for when someone wants to scrollto or so)
21:27:16 <frosch123> in some cases there are multiple valid abbreviations in german :/
21:27:30 <_glx_> in MSVC you can use hex for numbers, but it's all or nothing
21:27:32 <frosch123> the larger the number, the fewer there are
21:31:03 <_glx_> yeah I just paused randomly so it was somewhere in video driver
21:35:53 <xarick> enough PRs from me, got 9 open
21:46:45 <xarick> AirportCompatibility hmm
21:50:20 <peter1138[d]> Windows-only bah
22:05:04 *** Ox7C5 has quit IRC (Quit: Lost terminal)
22:16:29 <kuhnovic> xarick: that bug is so much weirder than it looks
22:22:54 <peter1138[d]> uint vs default enum storage type 😮
22:31:14 <reldred> PeterNviaGitHub: fridaemon there we go, fixed in OpenTTD. If a roadstops grf only has drive through stops it now completely hides the other four buttons.
22:42:33 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
22:43:38 <xarick> oh crap, i forgot something
22:48:45 *** nielsm has quit IRC (Ping timeout: 480 seconds)
22:50:23 <xarick> CanPlaneUseAirport maybe needs another name now
22:50:33 <xarick> GetAirportCompatibilityWithPlane ?
22:51:21 <xarick> GetAirportPlaneCompatibility?
22:51:35 <xarick> GetAirportCompatibility
22:52:06 <xarick> help me with a name talltyler
22:55:27 <talltyler> xarick: That sounds good. Making things is always hard. 🙂
22:57:28 <xarick> the word compatibility is quite large
23:00:33 <xarick> > Sure, here are a few shorter, but still meaningful, suggestions for the function name:
23:00:36 <xarick> > Remember, the best function names are concise, yet accurately describe what the function does. It’s also good practice to follow the naming conventions established in the rest of your codebase. Happy coding! 🚀
23:01:37 <peter1138[d]> Don't write code with "AI" tools, they are trained off stolen copyright content
23:12:21 <xarick> `/* static */ ScriptAirport::AirportPlaneCompatibility ScriptAirport::GetAirportPlaneCompatibility(AirportType airport_type, PlaneType plane_type)`
23:12:39 <xarick> but it's gonna be this
23:14:26 <peter1138[d]> AIAirport::...Airport... is sort of redundant, but it's done elsewhere.
23:23:59 <xarick> shouldn't be returning -1 for helicopters 😐
23:26:16 *** keikoz has quit IRC (Ping timeout: 480 seconds)
23:28:00 <xarick> IsValidStation accounts for oilrigs
23:35:49 <xarick> ah, right, gotta rename PR
23:38:44 <xarick> get airportnumhelipads, i wonder what it returns for oilrig
23:42:35 *** Deep3D has quit IRC (Ping timeout: 480 seconds)
23:44:12 *** ChanServ sets mode: +v tokai
23:51:10 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
continue to next day ⏵
