IRC logs for #openttd on OFTC at 2021-09-24
⏴ go to previous day
00:11:10 *** b_ has quit IRC (Ping timeout: 480 seconds)
00:17:26 *** Eddi|zuHause has joined #openttd
00:18:35 <Eddi|zuHause> that wasn't as straightforward as i'd liked... (mostly due to the network card driver that was already broken before...)
00:21:44 *** WormnestAndroid has quit IRC (Remote host closed the connection)
00:21:57 *** WormnestAndroid has joined #openttd
00:37:55 *** b_ has quit IRC (Ping timeout: 480 seconds)
00:41:09 *** ChanServ sets mode: +v tokai
00:48:05 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
00:54:30 *** b_ has quit IRC (Ping timeout: 480 seconds)
01:12:18 *** blindvt has joined #openttd
01:12:40 *** blindvt is now known as Guest819
02:06:42 *** Wormnest has quit IRC (Quit: Leaving)
02:10:25 *** debdog has quit IRC (Ping timeout: 480 seconds)
02:47:49 *** _aD has quit IRC (Quit: leaving)
04:32:04 *** tokai|noir has joined #openttd
04:32:04 *** ChanServ sets mode: +v tokai|noir
04:38:40 *** tokai has quit IRC (Ping timeout: 480 seconds)
05:58:14 *** andythenorth has joined #openttd
06:11:26 *** OsteHovel has joined #openttd
06:11:52 *** sla_ro|master has joined #openttd
06:22:55 *** andythenorth has quit IRC (Quit: andythenorth)
06:26:36 *** Guest819 is now known as blindvt
06:33:44 *** OsteHovel has quit IRC (Read error: Connection reset by peer)
06:56:20 *** HerzogDeXtEr has quit IRC (Read error: Connection reset by peer)
07:01:08 *** andythenorth has joined #openttd
07:38:32 *** andythenorth has quit IRC (Quit: andythenorth)
08:37:34 *** andythenorth has joined #openttd
08:38:00 *** andythenorth has quit IRC ()
08:51:54 *** sla_ro|master has quit IRC ()
09:39:21 *** WormnestAndroid has quit IRC (Remote host closed the connection)
09:39:34 *** WormnestAndroid has joined #openttd
09:53:02 *** blindvt has quit IRC (Quit: Leaving)
10:09:01 *** andythenorth has joined #openttd
10:53:41 <Samu> omg, increasing stations touches so many things!
10:55:01 *** andythenorth has left #openttd
11:02:43 <Samu> 2 overloads have similar conversion, I don't know how to fix this
11:03:49 *** andythenorth has joined #openttd
11:25:47 *** andythenorth has quit IRC (Quit: andythenorth)
11:40:25 *** sla_ro|master has joined #openttd
12:13:14 *** virtualrandomnumber has joined #openttd
12:13:55 *** virtualrandomnumber has quit IRC ()
13:26:05 <Samu> i dont know how to make this saveload conversion
14:12:15 <FLHerne> Now I wonder what it would be on those weird architectures where a byte isn't 8 bits :p
14:14:01 <FLHerne> "provided if and only if the implementation directly supports the type" probably rules that out, because it would be insane to support 8-bit-aligned integers and then not have 8-bit bytes
14:14:49 <glx> anyway we won't increase station pool size :)
14:16:13 <FLHerne> I seem to remember actually running out of stations
14:16:22 <FLHerne> it was a pretty stupid game though
14:37:20 <Samu> even cargo packets and linkgraph have StationIDs
15:03:41 *** reldred has quit IRC (Remote host closed the connection)
15:04:04 *** reldred has joined #openttd
15:05:41 *** Wormnest has joined #openttd
15:29:16 <Samu> i just realised I might also have to increase max number of depot indexes
15:29:47 <Samu> how is this depot index working like?
15:30:55 <Samu> will leave depots alone for now
15:44:05 <peter1138> Hmm, wonder if my VR still works.
15:44:15 <LordAro> Samu: just to be clear, whatever you're modifying now is entirely for your ownh use - if you try PRing this, i'm going to be annoyed
15:45:11 <peter1138> More stations? Haha
15:45:23 <peter1138> LordAro, got my road bike back :D
15:50:07 <peter1138> And ordered replacement mudguards for the tourer...
15:50:38 <peter1138> And solved 2 layers of a Rubik's cube, stuck on the 3rd now.
15:55:10 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
15:55:33 *** WormnestAndroid has joined #openttd
16:00:45 <LordAro> peter1138: i need to order some proper mudguards for my now-winter-and-commuting bike
16:01:15 <peter1138> I went for SKS Longboards as it's a full-on tourer.
16:01:30 <peter1138> I've got clip on things for the road bike but I don't intend to use it in the rain any more.
16:01:42 <peter1138> Well, light shower while out is okay.
16:01:55 <LordAro> i've seen what the clipon ones do to the frame, i'm not putting them on my shiny new bike
16:01:58 <peter1138> Pissing down from the start, less so :D
16:02:50 <peter1138> My road bike does actually have custom mounting points, and I have the mudguards originally but the rear snapped, so... annoying.
16:02:52 <LordAro> in theory the ones from the manufacturer should fit a bit better
16:03:20 <LordAro> £40 does seem a bit steep though
16:03:50 <peter1138> Worse is the mounting point on the seat stays had a thread-lock bold and the head got mashed trying to remove it.
16:04:32 <peter1138> chain stays, not seat stays.
16:04:40 <LordAro> i had to get 2 of my bottle cage mounts drilled out after they completely rounded out a couple months ago
16:04:45 <peter1138> So there's a little bit of the old mudguard still stuck in there, heh.
16:05:19 <peter1138> Hmm, £32 for a new set. Maybe it's worth trying to get that bolt out...
16:06:42 <peter1138> Maybe I should rebuild my old bikes...
16:07:56 <peter1138> The MBK tourer I had was a pretty shit frame and had excessive flex when pedalling. Tyre would rub both chainstays depending on the stroke :p
16:21:21 <peter1138> I went for SKS Longboard as the "normal" length SKS mudguards I got for another bike were disappointingly short.
16:22:22 <peter1138> There's a reflector on those but it's in the wrong position really. Partly due to framesize, the bottom bracket is a bit lower than the axle.
16:31:19 <nielsm> but it turns out my GfxFillPolygon function is _very_ slow, so framerate gets pretty terrible
16:31:36 <nielsm> it's not great for this use, at least not right now
16:32:45 <LordAro> luckily it's jsut a debugging feature, where frames are largely irrelevant, right? :p
16:33:03 <nielsm> well it should be fast enough to not drive yourself mad
16:33:11 <LordAro> ok, that is quite slow
16:34:20 <glx> but this debugging feature helped me to see forcing trackid keyed nodes may not be necessary
16:35:45 <glx> when an open node is updated to a lower cost, the trackid is also updated
16:36:00 <glx> *trackdir in both previous lines
16:37:24 <nielsm> I'll push that in a bit
16:37:37 <peter1138> What are the colours?
16:37:44 <nielsm> (mainly to properly show when both overlapping trackdirs are there)
16:37:53 <nielsm> the cost relative to max cost
16:38:19 <nielsm> I'm not sure why it makes pale yellow for the lowest
16:38:56 <peter1138> Palette colours or rgba?
16:39:14 <peter1138> Then either that's just the scheme, or an obiwan,.
16:40:38 <nielsm> pushed the arrow heads
16:41:06 <nielsm> some nice integer vector rotation and shit
16:41:28 <glx> the scheme uses colours 42 to 48 in .\docs\palettes\palette_key.png
16:41:56 *** Eddi|zuHause has quit IRC ()
16:42:10 *** Eddi|zuHause has joined #openttd
16:53:12 *** Progman has joined #openttd
16:59:14 <glx> ok it's the 197 (first value in all color gradiants)
17:00:54 *** WormnestAndroid has quit IRC (Read error: No route to host)
17:00:58 <nielsm> glx yeah this looks better: const int colour = 42 + cost * 8 / this->maxcost;
17:01:29 *** WormnestAndroid has joined #openttd
17:01:55 <glx> unzoomed arrows are ugly ;)
17:02:41 <nielsm> I didn't even test unzoomed
17:12:58 <glx> ok with the arrows (and trackdir nodes) I can clearly see paths going backwards to start point
17:13:17 <glx> no wonder there's too many closed nodes
17:16:33 <glx> visually I see no errors in my test map
17:19:27 <nielsm> just for sanity checking, maybe check in ViewportPfOverlay::AddTile if it's adding the same tile*td multiple times
17:34:04 <glx> added an assert for safety
17:35:04 <nielsm> pushed some slightly better arrows
17:41:40 *** andythenorth has joined #openttd
17:42:30 <glx> oops playing with zoom triggered an division by zero
17:43:59 <glx> 0, -1 and 0,0 vectors are not very nice :)
17:44:20 <nielsm> those really shouldn't happen, I think
17:45:02 <glx> well the 3 most unzoom breaks the drawing
17:47:17 <andythenorth> pls entertain me
17:49:26 <glx> pta and ptb are next to eachother
18:09:38 <andythenorth> occasionally I wish there was a step through newgrf debugger
18:09:42 <andythenorth> but it wouldn't be practical :P
18:09:59 <andythenorth> if I really cared I could just learn the C++ debugger I guess
18:10:10 <andythenorth> I have logic not doing what I expect :P
18:12:58 *** Flygon has quit IRC (Quit: A toaster's basically a soldering iron designed to toast bread)
18:48:23 <DorpsGek> - Update: Translations from eints (by translators)
18:51:57 *** Beer has quit IRC (Quit: Leaving)
18:56:07 <Samu> I think I found an unused global variable in the saveload code
19:05:36 <FLHerne> Yeah, but if you remove it I think you'll blow up loading old savegames
19:09:49 <peter1138> It could be replaced with a CONDNULL.
19:17:02 *** Gustavo6046 has quit IRC (Ping timeout: 480 seconds)
19:30:14 <Samu> from what I understand, cargo_paid_for was a thing in Vehicle struct, and was moved to paid:for in CargoPacket
19:30:50 <Samu> sometime between that commit and today, paid_for disappeared
19:31:21 *** Gustavo6046 has joined #openttd
19:31:41 <Samu> how do I check history in reversed?
19:31:55 <Samu> see the commits after which make things disappear
19:33:37 <glx> even in this commit the global was unused
19:35:34 <glx> should have been COND_NULL(..., 45, 67) there
19:42:53 <Samu> well, i still can't load savegames, this conversion is tough stuff
19:55:04 <Samu> current_order of disaster vehicles... DestinationID aren't even real stations, they're states
20:29:20 <Samu> i can finally load savegames! hurray
20:29:32 <Samu> now I'm getting asserts after a tick :)
20:30:08 <Samu> the old invalid station id has to be updated everywhere
20:30:43 <glx> that's the job of AfterLoad
20:37:12 <Samu> in CargoPacket, how do i make the saveload change something that is private? do I just move to public?
20:37:29 <Samu> StationID needs conversion but it's private
20:39:27 <Samu> 'CargoPacket::source': cannot access private member declared in class 'CargoPacket'
20:41:14 <glx> nielsm: both dpi->zoom used in your last commit should be vp->zoom (dpi->zoom is constant ZOOM_LVL_NORMAL)
20:42:24 <glx> Samu: you need to be friend, it's done in many place in saveload code
20:45:38 <nielsm> oh, does that perhaps happen when it goes up/down locks?
20:46:36 <peter1138> "Fix: Changing raw industry funding method redraws GUI (#9572)"
20:46:49 <nielsm> because I haven't gotten anything like it when zoomed all the way out
20:47:05 <peter1138> Dodgy commit messages... makes it sound like it did but wasn't meant to.
20:49:04 *** Gustavo6046_ has joined #openttd
20:49:14 <Samu> i dont know how to make friend StationID
20:49:46 <Samu> need an example to follow
20:52:39 <glx> oh you already have CargoPacket::AfterLoad(), fix the id there
20:53:38 *** Gustavo6046 has quit IRC (Ping timeout: 480 seconds)
20:53:38 *** Gustavo6046_ is now known as Gustavo6046
20:56:31 <Samu> oh, nice find, I didn't look too clearly
20:57:52 <Samu> cp->source is being used in the functions above, i guess it needs to be the first thing to change
21:04:08 <glx> nielsm: no it's a plain water tile
21:09:29 <Samu> im surprised savegames are loading without crashing, seems too good to be true
21:09:30 <glx> and of course it doesn't always fail
21:11:07 <Samu> linkgraph_sl.cpp is chinese to me
21:11:19 <nielsm> only line I can think of that could theoretically div by 0 would be this:
21:11:19 <nielsm> int rot_vec_len = std::max(1u, IntSqrt(rot_vec.x * rot_vec.x + rot_vec.y * rot_vec.y));
21:11:42 <nielsm> well, only source of a zero that could be divided by
21:12:18 <Samu> will continue with this tomorrow, cyas goodnight
21:12:52 <glx> yes that's the line (well the next instruction is the breaking one) but I can't find why it seems to happen randomly
21:13:49 <nielsm> I wonder if IntSqrt is degenerate in some cases and ends up zero when it shouldn't
21:15:05 <glx> rev_vec is {0, -1} and rot_vec is {0, 0}
21:16:46 <nielsm> seriously, no tile should be that small...
21:18:17 <glx> that's one of the short trackdir
21:20:21 *** Samu has quit IRC (Ping timeout: 480 seconds)
21:28:17 *** andythenorth has quit IRC (Quit: andythenorth)
21:28:18 *** Speedy` has quit IRC (Read error: Connection reset by peer)
21:28:57 <glx> another tile, still LEFT_S, but this time I was also moving the viewport
21:29:18 *** andythenorth has joined #openttd
21:31:20 *** Speedy` has joined #openttd
21:32:58 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
21:33:04 <glx> oh maybe it is, I see a pattern in the failing tiles 0x1A99, 0x1B19, 0x1B99
21:41:08 *** WormnestAndroid has quit IRC (Ping timeout: 480 seconds)
21:41:30 *** WormnestAndroid has joined #openttd
21:43:55 <glx> anyway with std::max it should be safe
22:00:24 *** andythenorth has quit IRC (Quit: andythenorth)
22:01:10 *** Wormnest has joined #openttd
22:05:16 *** HerzogDeXtEr has joined #openttd
22:19:16 *** nielsm has quit IRC (Ping timeout: 480 seconds)
22:25:07 *** Gustavo6046_ has joined #openttd
22:25:08 *** Gustavo6046 has quit IRC (Remote host closed the connection)
22:25:12 *** Gustavo6046_ is now known as Gustavo6046
22:36:27 *** sla_ro|master has quit IRC ()
22:41:09 *** Wolf01 has quit IRC (Quit: Once again the world is quick to bury me.)
22:53:43 *** Wormnest has quit IRC (Ping timeout: 480 seconds)
23:13:13 *** Gustavo6046 has quit IRC (Ping timeout: 480 seconds)
23:13:36 *** Gustavo6046 has joined #openttd
23:23:05 *** ChanServ sets mode: +v tokai
23:24:01 *** Progman has quit IRC (Remote host closed the connection)
23:24:25 *** Gustavo6046 has quit IRC ()
23:24:33 *** Gustavo6046 has joined #openttd
23:29:56 *** tokai|noir has quit IRC (Ping timeout: 480 seconds)
23:31:20 *** Wormnest has joined #openttd
continue to next day ⏵
