IRC logs for #openttd on OFTC at 2023-12-05
⏴ go to previous day
00:12:00 *** igor[m] has joined #openttd
01:05:15 *** leward[m] has joined #openttd
01:05:20 *** magdalena[m] has joined #openttd
01:39:52 *** Bilb[m] has joined #openttd
02:39:19 *** yubvin[m] has joined #openttd
02:46:37 *** JamesRoss[m] has joined #openttd
02:59:22 *** Wormnest has quit IRC (Quit: Leaving)
03:03:04 *** nolep[m] has joined #openttd
03:09:09 *** godbed has quit IRC (Ping timeout: 480 seconds)
03:09:19 *** Zathras has quit IRC (Ping timeout: 480 seconds)
04:03:45 *** Zathras has joined #openttd
04:06:49 *** debdog has quit IRC (Ping timeout: 480 seconds)
04:07:04 *** D-HUND has quit IRC (Ping timeout: 480 seconds)
04:23:04 *** zzy2357[m] has joined #openttd
04:52:04 *** virtualrandomnumber has joined #openttd
04:52:11 *** virtualrandomnumber has quit IRC ()
06:13:47 *** hamstonkid[m] has joined #openttd
06:46:11 *** joey[m] has joined #openttd
07:09:04 *** einar[m] has joined #openttd
07:25:27 *** patricia[m]1 has joined #openttd
07:30:02 *** tokai|noir has joined #openttd
07:30:02 *** ChanServ sets mode: +v tokai|noir
07:36:35 *** tokai has quit IRC (Ping timeout: 480 seconds)
08:17:52 *** CornsMcGowan[m] has joined #openttd
08:21:53 *** Zathras is now known as debdog
08:26:32 <_zephyris> I'm getting a weird effect. Test builds using the updated make file for NewGRFs works perfectly for clean and rebuilds, gives the exact expected `orig_extra.grf` NewGRF, identical `.obj` files etc. But, the `.grf.hash` files are picked up as modified by git, even though the hash itself looks unchanged...
08:35:24 *** menelaos[m] has joined #openttd
08:38:06 <Eddi|zuHause> checked difference in line endings?
09:01:16 <peter1138> I did originally trim the has file so it had no line endings at all, but the git CI complained about that.
09:11:44 <_zephyris> Do they need line endings then?
09:12:26 <_zephyris> *imagines a PR to add two lube ending*
09:13:00 *** FelixActually[m] has joined #openttd
09:13:41 <peter1138> They have line endings, the problem now is, presumably, the difference between Windows/DOS and Linux line endings.
09:34:52 *** throwawayaccount[m] has joined #openttd
10:12:32 *** Smedles_ has quit IRC (Remote host closed the connection)
10:14:05 *** Smedles has joined #openttd
10:17:54 *** Smedles has quit IRC (Remote host closed the connection)
10:19:25 *** Smedles has joined #openttd
10:26:05 <Eddi|zuHause> when did that split occur? 50 years ago?
10:26:49 <Eddi|zuHause> maybe closer to 40?
10:29:19 <Eddi|zuHause> even google isn't really sure
10:32:49 <peter1138> Mac used to use just CR, I believe.
10:33:45 <peter1138> And the Internet generally uses CRLF.
10:43:44 <Eddi|zuHause> and why was there never a split between CRLF and LFCR?
10:48:02 <peter1138> Because CRLF comes from typewriters, where the line feed would happen once the carriage was returned to the end stop.
10:49:07 <peter1138> The force to move the carriage is less than the force to turn the lever.
10:51:42 *** shedidthedog[m] has joined #openttd
11:10:08 <andythenorth> what shall we do today?
11:10:22 <LordAro> same thing we do every day
11:40:07 <xarick> I need someone to convince me to buy a 5950X over a 5800X3D
11:40:30 <andythenorth> what do these words mean?
11:40:54 <andythenorth> my children want gaming PCs, words like '3050' are being said
11:41:21 <xarick> that's a graphics card
11:41:52 <xarick> Nvidia GeForce RTX 3050, a poor one
11:42:52 <LordAro> i mean, it's not top end, but it's not like it's bad
11:42:58 <ahyangyi> Add the first number and the third number and you get a rough estimation of the card's power. For example, 3050 and 2060 are both "power 8" cards, but 4090 is "power 13" 😛
11:44:15 <andythenorth> it's probably a lot faster than the M1 GPU
11:44:32 <andythenorth> it has 128 cores vs. 8 according to spec I found
11:45:39 <andythenorth> no wait, that's wrong
11:46:21 <xarick> the more to the left upper corner, the better
11:46:48 <andythenorth> I assume the M1 GPUs are too low to chart?
11:49:02 <xarick> I don't know anything Apple... They're not known for gaming
11:50:46 <peter1138> I don't think those cores are directly comparable.
11:51:34 <peter1138> Sad that my 2080 Ti is only half way up...
11:51:51 <andythenorth> it's just for coin mining though?
11:51:53 <peter1138> On the other hand, doesn't make much difference for OpenTTD.
11:51:55 <andythenorth> and playing Doom
11:52:17 <peter1138> MS FS 2020 was nice, but as I'm in Linux these days... not so much.
11:53:47 <_glx_> _zephyris: Usually happens when building with Linux from a windows clone, or with windows from a Linux clone
11:55:12 <peter1138> Hmm, yeah, git has built in crlf conversion, but may not necessarily use sensible defaults.
11:56:02 <_zephyris> IIRC I ran WSL through Visual Studio Code instead of WSL terminal, maybe that was it
11:56:55 <peter1138> Was your clone made in WSL or Windows?
11:57:25 <peter1138> Also, should our server list window be green?
11:57:35 <_zephyris> Oh, maybe I cloned with the github GUI instead...
11:58:25 <_glx_> Hmm a workaround could be to use `configure_file` instead of `file` maybe
11:59:27 <_zephyris> Haven't seen that window in a long time. I think I last played multiplayer in 1996...
12:01:47 <peter1138[d]> Should I tweak drop down padding to match TTD, and if so, which style? 😉
12:04:29 <peter1138> (Old OpenTTD has two different drop down list windows, not related to each other at all)
12:04:53 <peter1138> At some point we merged them but the spacing became "a bit of both".
12:05:35 <Eddi|zuHause> does that include the menu popping up when pressing main taskbar buttons?
12:06:30 <peter1138> That's one of the two different windows.
12:08:04 <peter1138> Oh actually there were 3.
12:08:23 <peter1138> The list with the company icons on it was a separate window handler to the other plain text list.
12:32:32 <xarick> I can solve the long stalls, but not the rivers being built disconnected
12:33:11 <peter1138> I think the solution probably involves terraforming.
12:37:43 <Eddi|zuHause> all my fiddlings in rivers ended with "we need rivers on every possible slope"
12:41:12 <peter1138> River foundations? :D
13:26:47 <xarick> How is AyStar initiated?
13:26:56 <xarick> How are the default values applied
13:27:22 <xarick> the max_search_nodes is initiated as 0, and I wonder why
13:37:07 <_glx_> hmm after checking documentation, using `configure_file` won't help as it discards newlines from source file and replaces them with native (or specific, if the command says so) on write, but there's no way to keep newline style from source
13:50:16 <_pruple> hooray, people who don't have a ctrl key can now build Class 20 pairs in Sunshine 🙂
13:52:00 <peter1138> Magical pointless callback that didn't save many duplicated sprites in Iron Horse after all?
13:53:52 <_pruple> having things like cloned 37s being randomly a/b end forward is nice though. 🙂
13:55:25 <_glx_> I think the only viable option will be `configure_file` with specified `NEWLINE_STYLE`, and .gitattributes forcing `*.hash` eol, or maybe just no eol in `*.hash` and .gitattributes saying they are binary
13:56:08 <andythenorth> peter1138: cut 3s average off the compile time
13:56:24 <peter1138> _glx_, isn't it just the usual git-cloned-in-wrong-format issue?
13:56:39 <peter1138> Any of the source files could have the same issue.
13:56:46 <_glx_> it is, but on windows it's frequent
13:57:09 <_glx_> you clone on windows then run wsl on the clone
13:57:23 <peter1138> If you do that then you've done it wrong, tbh.
13:58:03 <peter1138> I don't think the build system should work around other peoples' git configurations
13:58:15 <andythenorth> it was about 1.5 MB of duplicated sprites in the grf also
13:58:58 <andythenorth> not quite sure how the size of the input PNGs relates to the grf size
13:59:19 <andythenorth> input graphics dir is ```12,702,231 bytes (18.9 MB on disk)```
13:59:19 <_glx_> anyway if clone on windows, and file generated by wsl, you usually do the commit on windows, so it will fix eol
13:59:35 <andythenorth> grf is ```27,849,074 bytes (27.9 MB on disk)```
13:59:51 <peter1138> PNG is compressed, GRF is not.
14:00:28 <peter1138> Compression would not work very well, as each sprite is an individual image.
14:00:47 <peter1138> Well, not as effectively.
14:00:52 <andythenorth> $somebody looked at sprite atlas or something?
14:01:17 <peter1138> I was investigating creating it dynamically, which would have no effect here.
14:02:05 <_glx_> with MSVC wsl can't touch files in source because it firsts rsync to wsl but never rsync from wsl
14:02:33 <_glx_> the issue is most likely only with vscode, running wsl from windows folder
14:04:29 <_glx_> which is also super slow (reading/writing host files from wsl is very slow)
14:05:57 *** gelignite has joined #openttd
14:18:55 *** gelignite has quit IRC (Quit: Stay safe!)
14:25:12 <_glx_> as a test I edited a .hash from CRLF to LF and github desktop just warns me it will change it back to CRLF
14:30:57 <_glx_> so if it's only an eol missmatch it's noop
14:32:24 <_glx_> nothing to worry about, usually default git install has proper eol settings
14:32:53 <_glx_> not like svn in the old time 🙂
14:53:44 *** gelignite has joined #openttd
15:17:32 <xarick> i regret buying a HDR monitor 😦
15:17:53 <xarick> desktop capture becomes with grayed out colours
15:20:57 <peter1138> That's not the fault of the monitor.
15:25:00 <_glx_> all the chain should support HDR
15:26:30 <xarick> river gen long stall vs a possible fix
15:33:35 <_glx_> bitrate seems too low, but the colours look correct
15:59:42 *** virtualrandomnumber has joined #openttd
16:02:34 *** virtualrandomnumber has quit IRC ()
16:05:15 *** Wormnest has joined #openttd
16:34:00 <xarick> Do these colors look correct?
16:36:57 *** HerzogDeXtEr has joined #openttd
16:54:30 <xarick> dang, hardware acceleration sucks
16:54:57 <xarick> AMF can't even encode correctly, it has its own stalls
17:00:18 <peter1138> If you using HW acceleration you should probably enable vsync, and/or make sure the refresh is not too high.
17:04:34 *** merni has quit IRC (Quit: User went offline on Discord a while ago)
17:11:21 <xarick> no, this is just the case of AMD sucking
17:15:13 <peter1138> Oh, so it's a 165Hz monitor. Nice.
17:17:02 <locosage> hm, if end year is configurable does it even store highscores separately for each?
17:18:30 *** virtualrandomnumber has joined #openttd
17:19:00 *** virtualrandomnumber has quit IRC ()
17:24:10 <peter1138> No, only the name and score is saved.
17:29:31 <_zephyris> Sounds like a hack rather than a proper solution tbh
17:30:04 <_zephyris> Would be fun to fix the table, might encourage some speed run type players
17:36:27 <locosage> maybe not include the end year at all in that message then?
17:36:44 <locosage> also, performance score isn't that great of a goal to speed run
17:38:20 <talltyler> Yeah, it could just be `Top companies` 🙂
17:41:42 <_zephyris> Top companies, game length in years, flag for +/- newgrfs
17:42:48 <_zephyris> locosage: Yeah, I was actually pondering race to a score as an excuse endgame criterion
17:57:17 <peter1138> Well, feel free to redesign the mostly useless highscore system, I was just fixing a simple-to-fix bug.
18:03:43 <locosage> making useful highscore systems is basically what I've been doing for 10 years xD
18:03:56 <locosage> but there is no simple one-fits-all goal that can be just added
18:04:12 <locosage> game settings are adjusted for each goal individually, not the other way around
18:05:31 <locosage> and I'm even taking it beyond what vanilla settings can do
18:07:26 <peter1138> Does your system work for vanilla games?
18:17:01 <locosage> hardly, only reasonable vanilla goal I can think of is bank balance
18:17:24 <locosage> and even then it's kind of meh especially when you can do silly exploits like selling vehicles on the other end
18:20:05 <locosage> and there are plenty of settings that make moneymaking goal straight up stupid
18:20:19 <locosage> cargodist, town growth, freight multiplier etc
18:21:39 <peter1138> So the default system is 10 different statistics, which are combined into a single score.
18:22:53 <peter1138> Can you give me an example of a GameScript that uses the new system?
18:23:01 <peter1138> (That I can download from the content system)
18:23:42 <locosage> other 9 statistics are even worse than money
18:23:56 <locosage> you mean league table gs? no, only the one I had in pr as an example
18:24:10 <locosage> but not many gamescripts were even updated since
18:24:15 <locosage> and even less goal scripts
18:25:57 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
18:29:43 <talltyler> Huh, so we show a different table in multiplayer games, but you can’t access it from the main toolbar… maybe it only shows when the game ends??
18:31:40 <talltyler> I’m looking at removing the year from the string, but there are two strings…one that shows up when you open the high score screen from the menu or a single player game, and the other in networking games. But I’m unsure how to open that screen in a networking game
18:34:04 <peter1138> When the game ends...
18:34:18 <talltyler> That’s what I thought 🙂
18:38:07 <talltyler> Okay, so it shows the newspaper first and then the table
18:38:40 <DorpsGek> - Update: Translations from eints (by translators)
18:39:59 <andythenorth> peter1138: Not many GS in the world 😛
18:41:51 <locosage> put it on bananas so peter1138 can download it xD
18:41:52 <xarick> I think it's up to date still
18:42:15 <xarick> well, it's not really finished, but... fine
18:42:36 <peter1138> It feels like we added a whole new complex feature, that conflicts with a built-in feature, and... one person uses it... and then trying to make a minor fix to the built-in feature starts a stupid argument about how the built-in feature is useless...
18:43:34 <xarick> it doesn't have a readme
18:43:51 <locosage> that argument isn't really about your fix
18:44:11 <locosage> built-in system exists for historical reasons, you can fix whatever, it won't make that score any more meaningful
18:46:56 <xarick> oh, the upload feature on bananas doesn't let me select folders
18:47:45 <talltyler> You need to make a .tar, I think
18:51:32 <andythenorth> Or certain zips are accepted
18:51:38 <xarick> minimum requirements, I'm not exactly sure, I guess 14.0?
18:52:21 <andythenorth> Not in bananasas though
18:52:35 <andythenorth> Tends to require a specific FIRS version
18:52:46 <andythenorth> Also tends to crash hard
18:52:56 <andythenorth> And lag the game
18:53:45 <xarick> oh, i guess 13. something
18:53:53 <xarick> when was this feature introduced
18:56:59 <LordAro> xarick: please don't add unfinished things to bananas
18:59:03 <xarick> must be failing on info.nut, it says it's 0 kbytes
19:03:13 <xarick> i'll do this later, i dont remember details how I was uploading before
19:46:43 <truebrain> does Year 0 actually exist?
19:47:26 <truebrain> so I think you are fixing the wrong bug here 😛
19:49:13 <peter1138> As an ending year, 0 means never.
19:52:16 <talltyler> Right, it’s the special value to disable the end year popup
19:53:30 <talltyler> Also year 0 does exist, you can start a game in that year 🙂
19:53:42 <truebrain> I would like to file a bug report
19:53:50 <truebrain> this game does something illegal
19:55:36 <talltyler> I mean, there are no vehicles available so you can’t do much… 😛
19:55:47 <truebrain> as the year doesn't exist! The map should be one big void!
19:56:33 <truebrain> nah, you have to fix it for me!
19:56:48 <talltyler> `Feature: Delete the entire map if the user chooses an invalid year`
19:59:02 <truebrain> "It was a dark place"
20:07:01 <michi_cc[d]> `Feature: Big bang now properly modeled`
20:07:44 <andythenorth> Quantum Redstone Entanglement Bits
20:07:46 <talltyler> I'm sure Peter has a patch for it
20:18:05 <peter1138> No I'm busy with other pointless patches.
20:24:41 <peter1138> But speaking of Big bang... hows cargodest...
20:33:35 *** Smedles has quit IRC (Ping timeout: 480 seconds)
20:39:04 *** Smedles has joined #openttd
20:42:51 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
20:48:09 <andythenorth> I have a local implementation
20:48:22 <andythenorth> It exists only in the savegame
20:50:13 <talltyler> “It doesn’t work on my machine”
20:50:56 *** gelignite has quit IRC (Quit: Stay safe!)
20:51:37 <andythenorth> “Patched live in production”
21:19:55 <andythenorth> tab to all buttons? 😮
21:26:40 *** nielsm has quit IRC (Ping timeout: 480 seconds)
21:38:00 <peter1138> What shall I replace that magic number with?
21:40:12 <ahyangyi> truebrain: Depending on how do you represent 1 BC
21:40:20 <ahyangyi> After all, 1 BC doesn't have to be "-1 AD"
21:42:20 <andythenorth> didn't we patch the buy menu to optionally show wagon intro dates?
21:42:24 <andythenorth> or was that JGRPP?
21:43:29 <xarick> D:\OpenTTD\OpenTTD GitHub\OpenTTD\src\vehicle_base.h(788): warning C4100: 'limited': unreferenced formal parameter
21:43:42 <xarick> what is "unreferenced formal parameter"...
21:45:31 <peter1138> It means you have a parameter which is not used.
21:46:59 <xarick> ah... well that's the intention
21:47:34 <peter1138> You can just remove the name then.
21:47:48 <peter1138> e.g. "bool limited" -> "bool"
21:47:52 <xarick> okay my bad, it is used only in some overrides
21:48:54 <peter1138> You can keep the name in the overrides where it is used.
21:54:39 <xarick> nop, it did not like it
21:56:16 *** keikoz has quit IRC (Ping timeout: 480 seconds)
21:57:21 <peter1138> Did you remove more than just the name?
21:59:21 <andythenorth> /me must to sleep
21:59:26 <andythenorth> such day-long hangover
22:12:37 <xarick> `virtual ClosestDepot FindClosestDepot(bool limited = false) { return {}; }`
22:12:37 <xarick> > D:\OpenTTD\OpenTTD GitHub\OpenTTD\src\vehicle_base.h(788): warning C4100: 'limited': unreferenced formal parameter
22:25:21 *** wallaby2 has joined #openttd
22:26:09 <xarick> well the code builds, and works, but that warning annoys me 😦
22:26:52 <_glx_> we have [[maybe_unused]] for that
22:28:35 <xarick> erm, nope, still warning me. Maybe I'm doing it wrong
22:30:05 *** wallabra has quit IRC (Ping timeout: 480 seconds)
22:32:57 <_glx_> yes it's a marker for the parameter, just search for it in the code, we used it in many places
22:34:13 <xarick> I see, there was an example 5 lines above, and I didn't notice
22:36:49 <xarick> but I think it's a bit overkill for something that is "user error"
22:38:13 <xarick> a better approach would be to call the pathfinder upon reaching a segment, but this one happens in ProcessOrders instead and not VehicleController
22:38:36 <xarick> ProcessOrders is done every tick 😦
22:39:33 <_glx_> of course it's done every tick, it needs to find the destination
22:41:45 <_glx_> once the depot is found it's handled by the controller
22:59:44 <xarick> I got a feeling my fix is gonna backfire
23:09:58 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
23:26:25 <peter1138> Heh, back when grfcodec did PCX files
continue to next day ⏵
