IRC logs for #openttd on OFTC at 2024-10-13
⏴ go to previous day
00:32:04 *** nielsm has quit IRC (Ping timeout: 480 seconds)
01:56:23 *** gelignite is now known as Guest6166
01:56:26 *** gelignite has joined #openttd
02:03:43 *** Guest6166 has quit IRC (Ping timeout: 480 seconds)
02:09:34 *** godbed_ has joined #openttd
02:12:54 *** D-HUND has quit IRC (Ping timeout: 480 seconds)
02:12:59 *** debdog has quit IRC (Ping timeout: 480 seconds)
02:20:34 *** gelignite has quit IRC (Quit: Stay safe!)
02:58:59 *** gnu_jj_ has quit IRC (Ping timeout: 480 seconds)
03:38:21 *** godbed is now known as debdog
04:46:20 <DorpsGek> - Update: Translations from eints (by translators)
06:26:00 *** herms61 has joined #openttd
10:11:45 <mnhebi> also, unlike people think, writing Assembly, code doesn't look like disassembly here `mov [bp+di],dl`, instead you'd have register names etc. No one sane writes assembly code with no variable etc names.
10:18:36 *** reldred has joined #openttd
10:19:13 <mnhebi> friggin love the design document dissing the graphics of the original Railroad Tycoon btw xD
10:20:24 <mnhebi> then it also throws diss on A-Train and Simcity 2000 and I'm like whaaa
10:22:49 <mnhebi> reldred: don't forget add.
10:24:43 <mnhebi> but yeah, looking at this 179k line disassembly of C.E.O or AIV Networks and I just can't help but feel in awe how Ludde did shit like this all alone to get OpenTTD started..
10:25:41 <reldred> well he also did scumm vm
10:25:44 <mnhebi> gotta see where one function starts, and ends...name it etc.
10:36:08 <peter1138> All it takes is for a commercail product to include some debug information and bam, you've got a whole load of symbols to make your disassembly make much more sense.
10:36:20 <peter1138> It's still assembly though 😄
10:37:42 <mnhebi> yeah, if only all old games included their .pdb's...
10:38:50 <peter1138> Hmm, hopper are so slow 😦
10:48:51 <truebrain> okay, that doesn't work .. let's try this
10:49:55 <truebrain> and now it is missing libpng-dev .. lolz
10:51:24 <truebrain> spam spam spam spam
10:53:18 <truebrain> another repo, being all happy again
11:24:42 <peter1138> YAPF used `m_...` for 'member' variables before we had established that we use `this->...` for those.
11:25:12 <xarick> google chrome killing adblocks, funny
11:25:48 <peter1138> I actually think it might be worth going through removing `m_` from the names and using `this->`, through ALL of YAPF first. With no other changes.
11:27:14 <peter1138> And then also standardising the template parameter naming. All those `_` is not normal 😄
11:27:49 <kuhnovic> I use resharper on my work PC. It really good for renaming things across entire inheritance trees. But I wonder if it's able to track those member names across all the CRTP voodoo.
11:29:03 <peter1138> Curiously recursive template parameters.
11:29:14 <kuhnovic> "magic". But quite cool magic.
11:29:25 <peter1138> Oh, recurring, not recursive.
11:29:33 <peter1138> And pattern not parameters.
11:29:45 <peter1138> Curiously Recurring Template Pattern.
11:29:58 <kuhnovic> Just abbreviate it, you'll never get it wrong 😛
11:29:58 <peter1138> Perhaps Cursed, depending how you feel.
11:30:21 <peter1138> I used something similar to CRTP for dropdown list items, but I think technically that is the wrong way around.
11:30:54 <peter1138> I have no idea what the dropdown list items compile down to, but it makes adding bits and extending them very easy.
11:31:29 <xarick> any hope for reverting 30e1a61? or add some of its functionality as a parameter choice?
11:31:47 <peter1138> You need to properly explain why.
11:32:03 <peter1138> In your PR that does it.
11:32:20 <peter1138> (i.e. don't bother writing it here)
11:34:37 <kuhnovic> I see absolutely no reason to revert that change
11:36:00 <kuhnovic> It removes some strange, incorrect behavior and replaces with correct A* behavior. And no, we're not going to revert things just becase they are inconvenient for your contrived use cases.
11:39:29 <kuhnovic> And optionally enabling/disabling this behavior is even worse. Adding modes and options turns testing into a nightmare.
11:51:05 <peter1138> Hmm, I wonder best to do threaded sound loading.
11:52:37 <peter1138> I think I'd need to read the binary data into memory first, and then handle it.
11:53:01 <peter1138> Otherwise it's very likely to have multiple sounds playing together trying to read from a file at the same time.
11:54:06 <peter1138> And if the sound is in NewGRF, possibly trying to read sound data and sprite data at the same time, which means that using a queue and only loading one sound at a time is not useful.
11:55:52 <peter1138> It is mostly moot. Sound loading does not take much time, but with Opus compression and Soxr resampling it can encroach into noticeable stutter (on first load, the sounds are only loaded once)
12:05:17 *** akimoto has joined #openttd
12:28:00 *** akimoto has quit IRC (Remote host closed the connection)
13:21:24 <LordAro> peter1138: protip: all-out effort in 5C makes your throat hurt
13:21:54 <peter1138> I am going to struggle getting back into it.
13:25:09 <Borg> howdy, any GRF guruz around? espacially industry and tiles
13:25:39 <Borg> I have problem changing IndustryTiles properties.. I mean I can change them, but sounds and tile animation defaults are wiped.
13:25:49 <Borg> is there a way to prevent it.. or at least rebuild them?
13:27:28 <Borg> I didnt not yet done RTFS :)
13:42:05 <peter1138> LordAro, just fatigued after the last bout of possibly-that-virus-but-not-sure.
13:45:56 <johnfranklin> I have a running nose for a few days
13:48:10 <_glx_> oh breaking emscripten is so rare
14:00:07 <Borg> okey I found it myself, doing RTFS:
14:00:09 <Borg> A copied tile should not have the animation infos copied too.
14:00:46 <Borg> so now question is.. Can I easly recreate default tile animation and sounds somehow? possible?
14:15:56 <peter1138> Ah, my defailt search excludes .txt files, because lang files... 😄
14:49:08 <xarick> no asserts so far, I'm quite pleased
14:49:56 <xarick> but trackdir is noticeable slower, even with the change to 8x8
14:57:28 <xarick> i'mma test 4x4 for the lols
15:02:14 <xarick> wow, watching this stuff effortessly scale is magic
15:03:44 <truebrain> /me awaits the bug reports for changelog.txt (similar to known-bugs.txt 😛 )
15:04:03 <xarick> seems to be doing 25 lookaheads
15:13:31 <xarick> very big performance impact with 2x2... turned worse than i expected
15:17:05 <xarick> 1x1 returns path not found 😛
15:19:17 <xarick> 32x32 not possible to test atm
15:27:01 <_glx_> of course when testing CI stuff on my fork it always picks 22.04
15:44:01 <peter1138> truebrain: Can if you want.
15:44:11 <peter1138> Or rather, if you don't object 😉
15:46:48 <truebrain> I assume it shows the same issues 😄
15:46:59 <truebrain> But might be time to make that file markdown too, imo
15:59:23 <peter1138> Okay, a bit more to it.
16:01:15 <_glx_> the structure is present
16:01:34 <peter1138> Yeah, I had to disable some bits.
16:01:50 <peter1138> By default we only show the first 20 entries.
16:02:11 <peter1138> And changing the format breaks that test, so I'm changing that too 😄
16:03:41 <Borg> ^- oh lols :) now wonder I cannot recreate Power Plant Smoke
16:07:52 <peter1138> Yeah, there does not seem to be a way to use the default animation handlers for custom graphics.
16:08:18 <peter1138> Chimney smoke only works for industry tile 8, and anything added is not 8.
16:08:47 <Borg> you could add some hack in property.. if industry tile was copied from industry tile 8
16:08:55 <Borg> but I guess its not worth the effort
16:09:05 <Borg> the same problem w/ bzz effect.. I managed to get animation.. no sound tho
16:09:13 <peter1138> Well, no, if you're adding a flag, you just create smoke if the flag is set.
16:09:33 <Borg> yeah.. thats other optio.
16:09:36 <peter1138> You can use base sounds effectsfor sure.
16:09:59 <Borg> maybe I can live w/o smoke then :) If I will have bzz :D
16:11:59 <peter1138> A general way to add effect-vehicle smoke for industries (and perhaps even houses or objects) might be nice though.
16:12:30 <peter1138> Could save on the awkward tile animation that FIRS etc do.
16:13:20 <Borg> peter1138: looks like easy to do... just add extra bit to special flags (Property 12)
16:14:52 <Borg> or no so easy.. because X,Y is hard calculated for now
16:14:54 <peter1138> Kind-of. We then fall foul of "I want the smoke to look different" 🙂
16:15:43 <truebrain> _glx_: or just force 24.04 for now, and we can change that into "latest" in a month or so
16:15:53 <peter1138> Limited to 20 versions like before...
16:40:28 <andythenorth[d]> peter1138: “Pls subscribe me to your list”
17:03:23 <peter1138> Hardcoding to 24.04 temporarily is probably better than making it more complex to support 22.04 briefly?
17:08:25 *** gelignite has joined #openttd
17:17:55 <peter1138> Assuming it all compiles with newer clang.
17:33:09 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
17:34:43 <xarick> for some reason ship is not pathfinding correctly though
17:40:14 <xarick> it's setting intermediate destination
17:40:24 <xarick> and there's only 1 region
17:41:44 <xarick> "ship is already there problem"
17:57:58 <xarick> test: ship finding nearest depot on a 4096x4096 map from right corner to left corner
18:02:49 <peter1138> High-level pathfinders, all the way down.
18:21:22 *** akimoto has joined #openttd
18:46:05 *** akimoto has quit IRC (Remote host closed the connection)
18:46:49 *** merni has quit IRC (Quit: User went offline on Discord a while ago)
18:51:33 <xarick> I'm learning something
18:51:48 <xarick> the smaller the area, the more accurate it is
18:54:45 <xarick> 8x8 - happens to pick the best depot choice again
18:56:22 <xarick> 16x16 - not the best depot choice anymore, but also not the worst
18:57:39 <xarick> 32x32 - the worst depot choice
19:03:57 *** Borg has quit IRC (Quit: leaving)
19:31:54 <_glx_> xarick: surprisingly I'm quite sure it has been said here last week 🙂
19:32:45 <_glx_> because low level run is bound to patches found by high level
19:38:01 *** akimoto has joined #openttd
19:41:13 <kuhnovic> Bigger water regions also have a higher chance to be invalidated, simply because they contain more tiles. And updating large water regions takes a lot longer than smaller ones.
19:41:24 <kuhnovic> So I think you are beginning to see that it's all a big trade-off
19:49:08 *** gelignite has quit IRC (Quit: Stay safe!)
20:11:43 <soylent_cow[m]> OMG lol this game has TRAM!!!
20:12:16 <soylent_cow[m]> i am hella digging the Czech one, it's purdy at 32bit
20:15:45 <soylent_cow[m]> ❓️ is it advisable to raise the 500 train limit in server config?
20:15:55 <soylent_cow[m]> what are the safe higher values?
20:38:31 <peter1138> Have you reached it as a limit?
20:54:42 <andythenorth[d]> lunch was not worth discussing today
21:19:26 <soylent_cow[m]> <peter1138> "Have you reached it as a limit?" <- ahaha no, one of the players did though
21:19:42 <soylent_cow[m]> i am at like 100 trains though, i can see how it's very possible
21:19:52 <soylent_cow[m]> playing 2k x 2k map
21:19:55 <peter1138> Isn't that a yes then?
21:27:23 <soylent_cow[m]> hmmm giant screenshot option not there...
21:30:38 *** akimoto has quit IRC (Remote host closed the connection)
21:35:23 <xarick> peter1138: what do you mean by that?
21:36:12 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
21:36:21 <xarick> after 2 days of testing, no more asserts happening
21:36:50 <xarick> on my 70k ships tests :p
21:51:05 *** tokai|noir has joined #openttd
21:51:05 *** ChanServ sets mode: +v tokai|noir
21:57:59 *** tokai has quit IRC (Ping timeout: 480 seconds)
22:13:29 <soylent_cow[m]> <peter1138> "Isn't that a yes then?" <- yes 😹
22:17:43 <johnfranklin> What is Peter’s Avatar?
22:32:03 *** nielsm has quit IRC (Ping timeout: 480 seconds)
23:45:08 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
continue to next day ⏵
