Times are UTC Toggle Colours
00:00:20 <Xaroth> just need to figure out when to exactly sync the copy of the tile map so it doesn't spaz out. 00:04:56 <Xaroth> also figure out why it's constantly seeing type changes on every chunk of the map on bigger maps 00:07:30 *** daspork has quit IRC 00:07:31 *** daspork has joined #openttd 00:09:33 <_dp_> hm, server regularly checks this->HasClientQuit() but does it ever set it? 00:12:09 <Xaroth> hm, frequently changes between MP_TREES and MP_CLEAR 00:12:55 *** ChanServ sets mode: +v orudge 00:12:55 *** ChanServ sets mode: +v Terkhen 00:12:55 *** ChanServ sets mode: +v planetmaker 00:12:55 *** ChanServ sets mode: +v peter1138 00:12:55 *** ChanServ sets mode: +v DorpsGek 00:13:21 <Xaroth> guess we should ignore that 00:13:44 <_dp_> Xaroth, yeah, that sounds normal with tree growth 00:17:21 *** Progman has quit IRC 00:19:28 *** APTX has quit IRC 00:19:30 *** APTX has joined #openttd 00:20:59 *** gelignite has quit IRC 00:22:31 *** jinks has quit IRC 00:22:34 *** jinks has joined #openttd 00:23:02 *** Beerbelott has quit IRC 00:25:42 *** jottyfan has joined #openttd 00:26:19 *** jottyfan has joined #openttd 00:26:47 <Xaroth> heh, there's also a switch between MP_WATER and MP_TREES 00:27:13 <Xaroth> dbg: [map] [UpdateShadowTile] Tile type changed: 8627 6 4 00:27:14 <Xaroth> dbg: [map] [UpdateShadowTile] Tile type changed: 8627 4 6 00:27:19 <Xaroth> water to trees, trees back to water 00:27:51 *** k-man has quit IRC 00:28:13 *** k-man has joined #openttd 00:28:36 *** Sacro has quit IRC 00:28:36 <Xaroth> or rather, other way around 00:28:47 *** Sacro has joined #openttd 00:29:56 <_dp_> that's probably shore tiles 00:36:17 *** roadt_ has quit IRC 00:36:33 *** roadt_ has joined #openttd 00:37:56 *** nnyby has quit IRC 00:37:57 *** nnyby has joined #openttd 00:49:56 *** orudge has quit IRC 00:49:56 *** orudge has joined #openttd 00:52:31 *** mindlesstux has quit IRC 00:52:42 *** mindlesstux has joined #openttd 00:52:57 *** LordArtoo has joined #openttd 00:54:16 *** LordAro has quit IRC 01:03:20 *** glx has quit IRC 01:06:11 <DorpsGek> [OpenTTD/OpenSFX] Wuzzy2 updated pull request #22: Shorten some description texts https://git.io/JtF05 01:06:34 <DorpsGek> [OpenTTD/OpenSFX] Wuzzy2 commented on pull request #22: Shorten some description texts https://git.io/JtbYM 01:07:42 *** luaduck has quit IRC 01:07:42 *** luaduck has joined #openttd 01:08:20 *** innocenat_ has quit IRC 01:08:31 *** innocenat_ has joined #openttd 01:09:56 <DorpsGek> [OpenTTD/OpenSFX] Wuzzy2 commented on pull request #19: Replace Mercurial code with Git code https://git.io/JtbYQ 01:16:31 *** fnutt has quit IRC 01:16:38 *** fnutt has joined #openttd 01:21:08 *** skrzyp has quit IRC 01:21:21 *** skrzyp has joined #openttd 01:22:17 *** iSoSyS has joined #openttd 01:23:00 *** Flygon has joined #openttd 01:24:20 *** m1cr0man has quit IRC 01:24:44 *** m1cr0man has joined #openttd 01:39:54 *** osvaldo[m] has quit IRC 01:39:58 *** osvaldo[m] has joined #openttd 01:42:28 *** murr4y has quit IRC 01:42:41 *** murr4y has joined #openttd 01:46:44 *** JamesRoss[m] has quit IRC 01:46:49 *** JamesRoss[m] has joined #openttd 01:51:19 *** khavik[m] has quit IRC 01:51:27 *** khavik[m] has joined #openttd 01:53:43 *** Tirili has quit IRC 02:00:42 *** shedidthedog[m] has quit IRC 02:00:48 *** shedidthedog[m] has joined #openttd 02:10:02 *** yur3shmukcik[m] has quit IRC 02:10:11 *** yur3shmukcik[m] has joined #openttd 02:10:50 *** karl[m]5 has quit IRC 02:10:55 *** karl[m]5 has joined #openttd 02:11:06 *** Eddi|zuHause has quit IRC 02:12:10 *** natalie[m] has quit IRC 02:12:12 *** natalie[m] has joined #openttd 02:15:08 *** Eddi|zuHause has joined #openttd 02:15:08 *** snail_UES_ has quit IRC 02:15:18 *** snail_UES_ has joined #openttd 02:18:12 *** CornsMcGowan[m] has quit IRC 02:18:17 *** CornsMcGowan[m] has joined #openttd 02:38:40 *** iSoSyS has quit IRC 02:46:55 *** Ttech has quit IRC 02:48:35 *** Ttech has joined #openttd 02:58:00 *** Tirili has joined #openttd 03:03:31 *** Wormnest has quit IRC 03:14:50 *** Heiki[m] has quit IRC 03:14:54 *** Heiki[m] has joined #openttd 03:15:06 *** EmeraldSnorlax[m] has quit IRC 03:15:09 *** EmeraldSnorlax[m] has joined #openttd 03:20:09 *** debdog has joined #openttd 03:23:32 *** D-HUND has quit IRC 03:26:55 *** Tirili has quit IRC 03:29:08 *** Speeder has quit IRC 03:29:31 *** Speeder has joined #openttd 03:32:58 *** muffindrake has joined #openttd 03:34:45 *** muffindrake4 has quit IRC 03:56:55 *** heffer has quit IRC 03:56:56 *** heffer has joined #openttd 04:16:07 *** Gustavo6046 has quit IRC 04:16:09 *** Gustavo6046 has joined #openttd 04:20:04 *** Extrems has quit IRC 04:20:10 *** Extrems has joined #openttd 04:27:43 *** WormnestAndroid has quit IRC 04:27:59 *** WormnestAndroid has joined #openttd 04:56:09 <DorpsGek> [OpenTTD/nml] Andrew350 commented on pull request #173: Fix: Access to persistent storage of towns https://git.io/JtbcM 05:07:48 *** didac has quit IRC 05:08:36 *** lobster has quit IRC 05:12:07 *** y2kboy23 has quit IRC 05:12:10 *** Cursarion has quit IRC 05:12:13 *** Cursarion has joined #openttd 05:12:18 *** y2kboy23 has joined #openttd 05:14:36 *** snail_UES_ is now known as Guest56 05:14:36 *** snail_UES_ has joined #openttd 05:26:34 *** argoneus has quit IRC 05:26:42 *** argoneus has joined #openttd 05:27:00 *** didac has joined #openttd 05:47:46 *** Wuzzy2 has quit IRC 05:58:07 *** andythenorth has joined #openttd 06:01:46 *** blathijs has quit IRC 06:02:20 *** blathijs has joined #openttd 06:17:14 *** Yexo has quit IRC 06:17:30 *** tneo has quit IRC 06:17:46 *** ^Spike^ has quit IRC 06:17:59 *** Yexo has joined #openttd 06:17:59 *** tneo has joined #openttd 06:18:17 *** ^Spike^ has joined #openttd 06:18:20 *** andythenorth has quit IRC 06:24:43 *** andythenorth has joined #openttd 06:36:42 *** cawal[m] has quit IRC 06:36:50 *** cawal[m] has joined #openttd 06:43:06 *** nartir[m] has quit IRC 06:43:08 *** nartir[m] has joined #openttd 06:44:27 *** HerzogDeXtEr has joined #openttd 07:03:43 *** colde has quit IRC 07:03:44 *** colde has joined #openttd 07:22:02 *** jeremy[m] has quit IRC 07:22:07 *** jeremy[m] has joined #openttd 07:22:50 *** einar[m] has quit IRC 07:22:52 *** einar[m] has joined #openttd 07:26:32 *** sla_ro|master has joined #openttd 07:37:34 *** twpol has quit IRC 07:37:45 *** twpol has joined #openttd 08:07:38 *** grag[m] has quit IRC 08:07:47 *** grag[m] has joined #openttd 08:08:08 *** Maarten has quit IRC 08:09:47 *** Wolf01 has joined #openttd 08:11:36 *** crem2 has quit IRC 08:11:47 *** crem2 has joined #openttd 08:13:45 *** LordArtoo is now known as LordAro 08:15:20 *** pothyurf[m] has quit IRC 08:15:25 *** pothyurf[m] has joined #openttd 08:15:36 *** bkilm[m] has quit IRC 08:15:39 *** bkilm[m] has joined #openttd 08:18:24 *** Maarten has joined #openttd 08:27:24 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #8743: Change: Default to a 32bpp blitter. https://git.io/JtbuT 08:29:27 <Wolf01> Ohhh, bigger font is more readable from the bed, I can't undrstand why I didn't figure it out before 08:35:35 <andythenorth> get bigger eyes 08:35:48 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8715: Fix: Vehicle list windows did not update when this year's profit changed https://git.io/JtbuE 08:35:56 <Wolf01> Lol 08:41:30 *** Tulitomaatti has quit IRC 08:41:43 *** Tulitomaatti has joined #openttd 08:46:32 * andythenorth considers wrapping python around the CHIPS C-pre-processor templating 08:47:50 <Wolf01> Why don't you develop an IDE (web based) to create grfs in an user friendly way? 08:47:51 <andythenorth> make -> instantiate python classes for station tiles from config files -> python template .cpp defines -> run the cpp -> nfo -> grfcodec -> grf 08:48:23 <andythenorth> Wolf01 I did have a web-based grf configurator, but I deleted it 08:48:53 <Wolf01> :( 08:49:25 <andythenorth> no big loss, it was quite horrible 09:00:04 <TrueBrain> hmm ... DrawViewport is called 14 times, where I expected it to be called once 09:01:17 *** supermop_Home has quit IRC 09:03:36 *** greeter has quit IRC 09:03:51 *** greeter has joined #openttd 09:03:59 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #8715: Fix: Vehicle list windows did not update when this year's profit changed https://git.io/Jtbgt 09:10:07 <DorpsGek> [OpenTTD/OpenSFX] orudge commented on pull request #19: Replace Mercurial code with Git code https://git.io/Jtbgz 09:11:06 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8715: Fix: Vehicle list windows did not update when this year's profit changed https://git.io/Jtbgw 09:11:12 <DorpsGek> [OpenTTD/OpenSFX] orudge approved pull request #22: Shorten some description texts https://git.io/Jtbgo 09:11:28 <DorpsGek> [OpenTTD/OpenSFX] orudge merged pull request #22: Shorten some description texts https://git.io/JtF05 09:18:32 <TrueBrain> sorry LordAro ; I simply do not understand enough of modern C++ to judge that :) It seems you are absolutely correct, but *shrug* :) 09:20:37 <LordAro> TrueBrain: me either :p 09:21:39 <TrueBrain> :D 09:33:08 <TrueBrain> so our window dirty-rect system is pretty clever, minimizes what needs to be drawn 09:33:13 *** Wolf01 has quit IRC 09:33:14 <TrueBrain> but it makes optimizing Viewport a lot harder :P 09:33:47 *** Wolf01 has joined #openttd 09:33:49 <TrueBrain> but okay, it seems that depending on the map we can offload between 20% and 50% of the drawing time 09:34:27 <TrueBrain> it only does require keeping around sorting data longer than currently :) 09:36:43 <DorpsGek> [OpenTTD/OpenTTD] btzy commented on pull request #8715: Fix: Vehicle list windows did not update when this year's profit changed https://git.io/Jtbaa 09:37:57 <TrueBrain> LordAro: so the first parameter should have a & is what he is saying? 09:39:09 <LordAro> i'm really not sure 09:39:20 <LordAro> i can't find any change that makes any difference to the asm output, anyway 09:39:33 <TrueBrain> indeed 09:39:56 <TrueBrain> it is optimized to fully use registers 09:40:28 <LordAro> again, might be a consequence of the minimal code, but i don't think so 09:44:14 <TrueBrain> gcc is even worse 09:44:24 <TrueBrain> it just makes GetDisplayProfitThisYear2 jump to GetDisplayProfitThisYear 09:44:27 <TrueBrain> like IN YOUR FACE 09:44:28 <TrueBrain> :P 09:44:52 <LordAro> haha 09:45:14 <TrueBrain> so somehow it picks up the Money class is a primitive, I guess 09:45:26 <TrueBrain> well, more likely, that it fits in a register 09:45:36 <TrueBrain> after that, all this std::move() becomes irrelevant 09:46:32 *** gregdek has quit IRC 09:46:33 *** gregdek has joined #openttd 09:52:58 *** Exec has quit IRC 09:53:00 *** Exec has joined #openttd 09:58:16 *** dvim has quit IRC 09:58:27 *** dvim has joined #openttd 10:01:45 <TrueBrain> okay, so there really is a lot to gain for big games if we can split ViewportDoDraw() in a part that influences the game state and a part that does not 10:01:52 <TrueBrain> about 50% on most games is spend in either one 10:02:21 <TrueBrain> the heavy NewGRF games, like one with Timberwolf's, are less impacted (as most time is spent in parts that requires a lock on the game-state) 10:03:04 *** robert[m]2 has quit IRC 10:03:07 *** robert[m]2 has joined #openttd 10:03:12 <TrueBrain> but it will cost more memory, mostly vectors 10:05:00 <TrueBrain> is there an easy way to get the memory size of a vector, hmm 10:05:02 <TrueBrain> or a class 10:11:15 <LordAro> sizeof() 10:11:15 <LordAro> :p 10:12:29 <TrueBrain> yeah 10:12:33 <TrueBrain> I was hoping for something more clever :D 10:13:19 <TrueBrain> okay, it is about 500KB of data, the vectors combined 10:13:34 <TrueBrain> well, 50KB in normal zoom 10:13:42 <TrueBrain> 750KB in max zoomout 10:14:04 <TrueBrain> and 3MB in extreme games like ProGame5 10:14:51 <TrueBrain> 15MB if I go 2k 10:17:18 <peter1138> Are we still trying to keep within the CPU memory cache? :p 10:17:23 <peter1138> FUCK 10:17:25 <TrueBrain> so yeah, in high resolutions with huge games, memory goes up from 1KB to 15MB in that way 10:17:28 <peter1138> Is OpenTTD still trying to keep within the CPU memory cache? :p 10:17:53 <TrueBrain> I can imagine for 4k resolution this goes up to 60MB 10:17:57 <peter1138> Thats only 15000%, nothing to worry about. 10:18:01 <TrueBrain> and for HighRes 4k to .. euh .. 120MB 10:19:24 <TrueBrain> currently the game uses 220MB for me, so on my resolution it would go from 220MB to 235MB 10:19:32 <TrueBrain> I am curious how much memory a 4k resolution takes 10:19:56 <_dp_> what data are you looking at? I don't think gui memory usage should be affected by screen resolution 10:20:03 <_dp_> except for viewport drawing ofc 10:20:29 <TrueBrain> ... so yes, it is influenced by it ;) 10:22:21 <LordAro> TrueBrain: how about with something like zbase? 10:22:28 <TrueBrain> owh, do I dare .... 10:22:30 <TrueBrain> good question :D 10:24:08 <TrueBrain> owh, wait, it currently uses 40KB, not 1KB .. so from 40KB to 15MB for ProGame5 zoomed out :P 10:25:11 <TrueBrain> zbase consumes 400MB of RAM for me 10:25:17 <TrueBrain> so an increase of 15MB is even less there :P 10:25:27 <TrueBrain> either way, I don't like this solution as it scales really poorly with resolution :D 10:26:07 <_dp_> to draw the screen you basically just need some strings and a list of sprites 10:26:10 <_dp_> can't take that much 10:26:27 *** gelignite has joined #openttd 10:26:28 <TrueBrain> I was just watching a youtube movie about armchair developers 10:27:13 <andythenorth> were you in an armchair? 10:28:38 <TrueBrain> its funny to see how overly optimized OpenTTD's drawing is :) 10:28:48 <TrueBrain> clearly build in a time where people still cared about that stuff 10:30:20 <TrueBrain> okay, the current code can already burst to the same values, when the full screen is redrawn in a single iteration 10:30:43 <TrueBrain> that basically only happens during zoom-out 10:31:01 <TrueBrain> possibly also the reason that is so sluggish, as the sorter will have a hard time sorting that all at once 10:31:55 <peter1138> Yeah, see, if there was a *proper* OpenGL driver... *runs away very fast* 10:33:15 <_dp_> hey, my sorter is quite fast :p 10:37:49 *** rptr has quit IRC 10:49:44 *** virtualrandomnumber has joined #openttd 10:49:55 *** virtualrandomnumber has quit IRC 10:57:39 <TrueBrain> okay, made a very crude implementation of splitting the Viewport code 10:57:52 <TrueBrain> this game that was running on 13 simulation fps, is now doing 15 10:58:09 <TrueBrain> with 10ms for drawing now done in parallel with the gametick 11:00:24 <TrueBrain> funny, simulation can now run at, for this game, 20fps, but drawing is down to 5fps 11:00:32 <TrueBrain> simply because I am fully zoomed out and it cannot keep up :D 11:00:37 <TrueBrain> (instead of them both being at 5fps) 11:06:16 *** reldred has quit IRC 11:06:27 *** reldred has joined #openttd 11:07:13 <TrueBrain> hitting 9999 simulation fps on a fully zoomed-out map is weird 11:07:50 <TrueBrain> https://pasteboard.co/JQ8bmwZ.png 11:09:55 <TrueBrain> owh, and this is a debug build 11:09:57 <TrueBrain> lol 11:11:30 <TrueBrain> and frosch123 said it couldn't be done in a week, pffft 11:11:41 <TrueBrain> well, it breaks sometimes, as too many things assume what is happening shouldn't happen 11:11:47 <TrueBrain> mostly, fonts sometimes crash the game :P 11:12:13 <TrueBrain> https://github.com/OpenTTD/OpenTTD/compare/master...TrueBrain:split-viewport-drawing <- last patch in the series .. it is ugly code :D 11:13:52 <TrueBrain> it can be done smarter by using a pool of threads 11:14:02 <TrueBrain> and assigning each a "vd" to be drawn 11:14:25 <TrueBrain> but that would make OpenTTD really become multithreaded, I am not ready for that 11:15:06 <peter1138> So the graphics framerate is dropped because so much is updating, but the sim can still run flat out... 11:15:14 <TrueBrain> exactly 11:15:27 <TrueBrain> so what is on your screen has 50% less influence on the simulation speed 11:15:34 <peter1138> That is ... interesting. 11:16:35 <TrueBrain> slowly ... removing what needs a lock on the game-state, and what can run without, basically :) 11:17:05 *** Samu has joined #openttd 11:17:15 <peter1138> What's the frame rate if it's not running flat out? 11:17:21 <TrueBrain> but the more proper solution here is moving a lot of these Viewport _vd functions into a single class, so it is more clear it is isolated 11:17:41 <TrueBrain> 33.33fps for simulation and 60fps for graphics 11:18:17 <peter1138> Yeah, ideally you'd throttle the sim rate to not run so fast as to cause the framerate to drop. Heurrghghghghgghistics.. 11:18:53 <TrueBrain> the balancing act there is a matter of opinion, I found out 11:19:00 <TrueBrain> some want the simulation to be as quick, at the cost of drawing 11:19:11 <TrueBrain> others want drawing to hit 60fps as much as possible, at the cost of simulation 11:19:30 <Eddi|zuHause> speaking of framerate, i was looking for the setting yesterday, but couldn't find it 11:19:31 <peter1138> I think that's reasonable if you're hitting framerate and simrate limits < 1x. 11:19:41 <TrueBrain> Eddi|zuHause: its an ini-file-only setting 11:20:04 <Eddi|zuHause> i would have expected it in game setting, where resolution and stuff is 11:20:27 <TrueBrain> currently, I think nobody should change it unless they know what they are doing :) 11:22:18 <peter1138> Can Windows explorer mix directories and files together? I should know this... 11:22:33 <TrueBrain> yes, it can 11:23:00 <TrueBrain> hmm .. if I put this work in another thread, it can already start with sorting and drawing while the next rect on the screen is started .. making this even faster 11:23:50 <Eddi|zuHause> well, i don't think i know what's happening, but my fps seemed to be 30, not 60 11:24:15 <TrueBrain> did you change your openttd.cfg to test a while back? :D 11:24:30 <Eddi|zuHause> i don't think so 11:26:33 <Eddi|zuHause> "display_hz = 0" <-- this? 11:26:41 <TrueBrain> no, that is an old setting and removed now 11:26:45 <TrueBrain> refresh_rate is what you are looking for 11:28:02 <Eddi|zuHause> yeah, that's set to 30. but i don't think i ever touched that 11:28:17 <TrueBrain> well, the game sets it to 60 by default shrug 11:28:56 *** jgx has quit IRC 11:29:18 *** jgx has joined #openttd 11:41:26 <Eddi|zuHause> i don't think i'll find out when this was set. maybe some earlier test version set it to 30 by default 11:46:12 <LordAro> it would be nice to add it to the options window, i think 12:11:47 <dihedral> hey hey 12:19:48 <TrueBrain> LordAro: my main issue is, and the reason for not being in favour of adding it to the options window, is that increasing the refresh_rate has an impact on simulation speed, something you wouldn't expect (and most games don't have :P) 12:19:59 <TrueBrain> so I think we should first make sure that the impact of that is far less than it currently is 12:20:05 <TrueBrain> before adding it for anyone to change :) 12:20:30 <TrueBrain> it was why I originally wanted to keep it a const, but I can understand that some people will have to change it to 30 or so in order to play the games they can play now 12:20:33 <TrueBrain> for the same reason :) 12:20:54 <TrueBrain> but a dropdown with <30> <60> as only values, would result in more questions (and complaints) of users than simply not mentioning it 12:21:03 <TrueBrain> at least, that was my reasoning for doing it this way :) 12:21:52 <TrueBrain> sometimes you have to protect the user from himself, basically :D 12:22:48 <Eddi|zuHause> well, you can try reducing fps when simulation speed dips below x1? 12:30:43 <Eddi|zuHause> so the behaviour would be ideal conditions: simulation = x1, fps = user setting, under stress: simulation = x1, fps = reduced, and under high stress (fps < 33), also drop simulation speed, so simulation ticks and refresh rate are roughly equal 12:45:58 <TrueBrain> there are many things we can do to improve a lot of this, but it requires time and testing :) 12:46:24 <TrueBrain> and as I mentioned earlier, it also turns out people expect different things from the game when there is a resource shortage 12:46:34 <TrueBrain> so this takes more time to figure out 12:47:52 <TrueBrain> owh boy, we have 1 place in the code that does misused _realtime_tick 12:48:04 <TrueBrain> it uses it to check if there has been a draw-frame 12:49:21 <FLHerne> dihedral: Ahey! Seems this is not-seen-for-ages people week. 12:49:38 <dihedral> :-) 12:50:00 <TrueBrain> FLHerne: be careful, before you now it ludde drops by :) 12:50:05 <TrueBrain> but indeed: hi :) 12:50:31 <dihedral> the nice thing must be, that one cannot stay away for too long ;-) 12:50:46 <TrueBrain> some try more than others :P 12:51:01 <dihedral> i recently sent a message to darkvater - wow was he surprised :-D 12:51:29 <dihedral> "recently" .... in relation to my "recent" visits to this channel may not habe been all that recent 12:51:42 <TrueBrain> so 4 years ago, you mean :P 12:51:56 <dihedral> nah - i believe it was late summer 2020 12:52:05 <TrueBrain> recent enough :) 12:53:14 <dihedral> uh - december even 12:53:29 <dihedral> (and summer 2020) :-P 12:53:41 <TrueBrain> stalker :P 12:54:09 <dihedral> na - that's what I concider "keeping in touch" 12:56:37 <TrueBrain> tomato tomato :P 12:56:42 <TrueBrain> what have you been up to lately? 12:58:19 <dihedral> busy - quote busy 12:58:44 <dihedral> have a company in karlsruhe and am working on my ppl 12:59:37 <dihedral> how about yourself? 13:02:02 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8747: Codechange: replace _realtime_tick with std::chrono https://git.io/JtFpt 13:02:13 <TrueBrain> was bored with my last job, quit, and have been working on OpenTTD instead :P 13:02:26 <dihedral> financed how? :-P 13:02:36 <TrueBrain> unrelated: w00p! No more _realtime_tick, no more race-condition with my other PR \o/ 13:02:37 <dihedral> i did notice that you have been busy with openttd alot 13:02:42 <TrueBrain> dihedral: a bank accont, ofc :) 13:02:53 <dihedral> Ha! that's what they all say :-P 13:02:59 <LordAro> TrueBrain: nice 13:03:28 <TrueBrain> cannot believe it was defined in debug.h .. 13:05:17 <LordAro> TrueBrain: newgrf_debug_sprite_picker stuff feels real hacky 13:05:25 <LordAro> oh, it was in window.cpp already 13:05:37 <LordAro> no more hacky than it already was then :p 13:05:38 <TrueBrain> that doesn't change it is a hack :) 13:05:40 <TrueBrain> :D 13:05:51 <TrueBrain> it is SLIGHTLY better, as it no longer abuses a variable for something unrelated 13:05:59 <LordAro> mm 13:06:15 *** supermop_Home has joined #openttd 13:06:45 <LordAro> i suppose i could argue that creating lots of static variables is just creating lots of "global" variables rather than just using a single one 13:06:47 <supermop_Home> hello 13:06:48 <LordAro> but eh 13:07:34 <TrueBrain> LordAro: did I add any static variables? 13:08:03 <TrueBrain> I should have only removed a few 13:08:26 <LordAro> oh, i can't read properly 13:08:37 <LordAro> yes, on the whole you've removed a couple 13:08:46 <TrueBrain> pfew :D Otherwise I screwed up :P 13:09:03 <dihedral> @TrueBrain, since when? 13:09:22 <dihedral> i mean - when did you quit? 13:09:23 <TrueBrain> that I screwed up? According to the forums, a few weeks ago :P 13:09:25 <TrueBrain> :D 13:09:28 <TrueBrain> somewhere last year 13:09:41 <dihedral> and what did you not like about your job? 13:09:46 <TrueBrain> the working part 13:09:49 <dihedral> ... 13:09:52 <TrueBrain> :D 13:09:54 <dihedral> har har 13:09:56 <LordAro> lol 13:10:01 <TrueBrain> I am not even kidding :D 13:10:32 <dihedral> yeah, can understand - but more specifically? :-P 13:10:36 <andythenorth> so is Belguas visiting next? 13:10:44 <andythenorth> Belugas * 13:10:48 <LordAro> @seen Bjarni 13:10:48 <DorpsGek> LordAro: Bjarni was last seen in #openttd 9 years, 20 weeks, 5 days, 12 hours, 51 minutes, and 41 seconds ago: <Bjarni> heh 13:10:59 <TrueBrain> @seen ludde 13:11:00 <DorpsGek> TrueBrain: I have not seen ludde. 13:11:01 <dihedral> uh - that was the guz 13:11:10 <TrueBrain> bad DorpsGek , bad bad DorpsGek 13:11:11 <TrueBrain> @seen Yexo 13:11:12 <DorpsGek> TrueBrain: I have not seen Yexo. 13:11:12 <andythenorth> TrueBrain we could just patreon you :P 13:11:16 <TrueBrain> he keeps forgetting people 13:11:26 <TrueBrain> andythenorth: meh; means people have expectations 13:11:30 <andythenorth> yes 13:11:33 <andythenorth> I would expect reports 13:11:35 <LordAro> :D 13:11:40 <andythenorth> and we would want to review you quarterly 13:11:44 <dihedral> patreon can be unconditional 13:11:48 <andythenorth> and we'd have to have a HR policy 13:11:56 <dihedral> as it is voluntary anyway 13:12:07 <andythenorth> TrueBrain we could construct it so you have to review yourself! 13:12:13 <andythenorth> on livestream 13:12:13 <dihedral> andythenorth, you want to patreon TrueBrain not openttd 13:12:16 <TrueBrain> https://gist.github.com/TrueBrain/25a5a6f98f761ef5be27b493b1b6ada5 there you go andythenorth 13:12:26 <andythenorth> yes 13:12:31 <andythenorth> then you can sign it off, yourself 13:12:40 * andythenorth has one sponsor, andythenorth 13:12:54 <andythenorth> I just rely on doing work to fund it 13:13:09 <andythenorth> interestingly, getting paid for newgrf commissions is now a thing 13:13:13 <andythenorth> not for me, but for other people 13:13:16 <TrueBrain> I am pretty sure only a very few people can make a living out of Open Source work via patreon :) 13:13:34 <TrueBrain> haha, who is doing that andythenorth ? :D That is pretty funny :) 13:15:04 <dihedral> TrueBrain, Next report could be a link to itself :-D 13:15:28 <TrueBrain> dihedral: fixed 13:15:48 <dihedral> perfect 13:16:45 <TrueBrain> okay, with the _realtime_tick removed, I can thread drawing and gameloop for all platforms safely 13:16:47 <TrueBrain> so that is nice 13:17:18 <TrueBrain> less platform-specific code \o/ 13:17:49 <TrueBrain> next would be to refactor viewport.cpp into more class-based system 13:21:11 <TrueBrain> ugh, still avoiding working on the GameSpeed window ... lets pick an issue that has been unresolved for a while .. 13:32:13 <supermop_Home> oof all that messing around to get pretty vehicle cargoes and i forgot to add plant fibers 13:32:27 <TrueBrain> tssk 13:33:36 <FLHerne> Yexo reappeared pretty recently for a while? 13:35:07 *** dvim has quit IRC 13:35:07 *** gregdek has quit IRC 13:35:08 *** greeter has quit IRC 13:35:08 *** Maarten has quit IRC 13:35:08 *** crem2 has quit IRC 13:35:08 *** y2kboy23 has quit IRC 13:35:08 *** WormnestAndroid has quit IRC 13:35:08 *** Extrems has quit IRC 13:35:08 *** Speeder has quit IRC 13:35:08 *** EmeraldSnorlax[m] has quit IRC 13:35:08 *** Ttech has quit IRC 13:35:08 *** fnutt has quit IRC 13:35:08 *** innocenat_ has quit IRC 13:35:08 *** mindlesstux has quit IRC 13:35:08 *** k-man has quit IRC 13:35:08 *** Smedles has quit IRC 13:35:08 *** Vadtec has quit IRC 13:35:08 *** dwfreed has quit IRC 13:35:08 *** mikegrb has quit IRC 13:35:13 *** mikegrb has joined #openttd 13:35:19 <TrueBrain> yeah, and DorpsGek didn't track that, which annoys me :P 13:35:38 <andythenorth> I suspect yexo got another job :P 13:35:42 <andythenorth> oof 13:35:44 *** k-man has joined #openttd 13:36:01 <LordAro> did you look to see if the dorpsgek issue was an irc bot issue, or a database issue? 13:36:09 *** dvim has joined #openttd 13:36:09 *** gregdek has joined #openttd 13:36:09 *** greeter has joined #openttd 13:36:09 *** crem2 has joined #openttd 13:36:09 *** y2kboy23 has joined #openttd 13:36:09 *** WormnestAndroid has joined #openttd 13:36:09 *** Extrems has joined #openttd 13:36:09 *** Speeder has joined #openttd 13:36:09 *** EmeraldSnorlax[m] has joined #openttd 13:36:09 *** fnutt has joined #openttd 13:36:09 *** innocenat_ has joined #openttd 13:36:09 *** mindlesstux has joined #openttd 13:36:09 *** Smedles has joined #openttd 13:36:09 *** Vadtec has joined #openttd 13:36:09 *** dwfreed has joined #openttd 13:36:34 <TrueBrain> I did not 13:36:40 <TrueBrain> as I did not care enough, I have to admit :) 13:36:49 <LordAro> fair enough :p 13:37:09 <TrueBrain> it is somewhere between too-lazy and if-I-find-the-problem-it-means-I-have-to-fix-it :P 13:37:12 <TrueBrain> btw: certbot! 13:37:24 <LordAro> aaah 13:38:17 *** Ttech has joined #openttd 13:39:08 *** Maarten has joined #openttd 13:44:02 <TrueBrain> I have this ticket, that is invalid, because it talks about a solution, not the real problem 13:44:09 <TrueBrain> but the real problem, might still be a thing 13:44:14 <TrueBrain> tempted to close the ticket 13:44:19 <TrueBrain> as I am too lazy to check if the real problem is a thing :D 13:44:26 <TrueBrain> (or also, possibly, invalid) 14:00:02 *** nielsm has joined #openttd 14:04:11 * andythenorth looks on reddit 14:05:02 <andythenorth> WHY 14:05:04 <andythenorth> WHY 14:05:05 <andythenorth> WHY 14:05:05 <andythenorth> https://www.reddit.com/r/openttd/comments/lstczw/why/ 14:05:42 <TrueBrain> best reply evah 14:05:48 <TrueBrain> (the one reply) 14:05:51 <supermop_Home> those ginzu a4s giving me some real 1995 vibes 14:06:03 <andythenorth> diesel A4? 14:06:07 <andythenorth> A4 with pans? 14:06:11 <andythenorth> hmm 14:09:33 <DorpsGek> [OpenTTD/OpenTTD] EratoNysiad commented on pull request #8604: Feature: allow limiting zoom level of NewGRF-provided sprites https://git.io/Jtb7A 14:12:11 *** gelignite has quit IRC 14:18:15 <TrueBrain> so if 5 commands use an ugly hack 1 way, and 1 command an ugly hack the other 14:18:24 <TrueBrain> but the 1 command is slightly less ugly 14:18:28 <TrueBrain> should I change all 5, or the 1 :P 14:18:30 <TrueBrain> :D 14:24:51 *** Samu has quit IRC 14:25:49 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on issue #8045: AI estimated costs are 0 for commands that fail while returning a cost https://git.io/JvSaL 14:26:40 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #8748: Fix 02e770ff: allow estimating CloneVehicle if short on money https://git.io/Jtbdq 14:26:41 <TrueBrain> 14 year old bug ^^ :D 14:27:29 <TrueBrain> and a nice example of: the issue starter might say A, but means B, and tried to solve it in a way that breaks more than it solves :D 14:29:35 <Eddi|zuHause> that's a typical samu 14:29:47 <TrueBrain> "reading between the lines" skill leveled up \o/ 14:29:50 <TrueBrain> Eddi|zuHause: yup 14:30:20 <TrueBrain> changing something for ALL commands while ONE command was at fault :D 14:30:50 <Eddi|zuHause> and then he gets mad if someone proposes an alternate solution 14:32:56 <TrueBrain> I spend 50% of the time working on this finding out if and which commands return a partial cost when erroring out :P 14:33:36 <Eddi|zuHause> command return values are a fractal on their own :p 14:39:11 <_dp_> vehicle cloning may not be the only command to behave that way... 14:39:49 <TrueBrain> I checked; it is 14:39:56 <TrueBrain> I wrote that in my comment even 14:43:48 <DorpsGek> [OpenTTD/OpenTTD] 2TallTyler commented on pull request #8604: Feature: allow limiting zoom level of NewGRF-provided sprites https://git.io/JtbFB 14:47:09 <_dp_> well, mb not for money but terra limit seems to affect estimations as well 14:47:15 <_dp_> which is kinda questionable 14:49:07 <_dp_> pretty much a bug I'd say even as it's not that different from money 14:52:20 <LordAro> TrueBrain: "... hopefully for good this time" 14:53:00 <TrueBrain> LordAro: sorry, context? :D 14:53:12 <nielsm> I really want to get JGR's route restriction signals and realistic braking in 14:53:17 <LordAro> 02e770f commit message 14:53:32 <nielsm> haven't tried them but in principle they sound really good 14:53:34 <TrueBrain> LordAro: ghehe, yeah :D 14:54:29 <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #8748: Fix 02e770ff: allow estimating CloneVehicle if short on money https://git.io/JtbFp 14:55:08 <_dp_> oh, and terraform estimations are wrong anyway so it's not really unaffected there :p 14:57:05 <TrueBrain> LordAro: "AIVehicle::CloneVehicle now correctly returns estimate when short on cash" <- good enough? 14:57:13 <LordAro> yeah 14:57:26 <_dp_> nielsm, doesn't realistic braking only work with pbs? 14:57:42 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8748: Fix 02e770ff: allow estimating CloneVehicle if short on money https://git.io/Jtbdq 14:57:50 <TrueBrain> tnx LordAro :) 14:58:02 <nielsm> _dp_ yes, I don't see any problem with that 14:59:44 <_dp_> well, I guess that would be another option never to enable in a goal game :p 15:08:22 <_dp_> interesting way to implement realistic braking would be to have fake head of the train always a braking distance ahead of the real one 15:08:33 <_dp_> but probably would be quite hard to do in openttd 15:09:30 <LordAro> how very "shadows" 15:10:30 <TrueBrain> OpenTTD wouldn't be OpenTTD if trains cannot stop instantly 15:10:38 <TrueBrain> makes me giggle with Factorio too 15:10:40 <TrueBrain> so much fun :D 15:11:27 <Eddi|zuHause> _dp_: the main problem with that idea is that braking distance changes with speed, so you need this pseudo-vehicle to be of variable length 15:11:54 <_dp_> Eddi|zuHause, I think variable length is the least problem here :p 15:12:27 <Eddi|zuHause> _dp_: what i would rather do is check the length of the path reservation 15:12:47 <Eddi|zuHause> _dp_: if less than braking distance, try extending it 15:12:54 <Eddi|zuHause> _dp_: if extending fails, brake 15:14:36 <_dp_> well, reservation isn't that different from a fake head anyway 15:17:35 <_dp_> btw, I don't think #6503 was ever fixed :p 15:20:04 <LordAro> clearly not 15:22:19 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #8748: Fix 02e770ff: allow estimating CloneVehicle if short on money https://git.io/JtbNX 15:22:47 <DorpsGek> [OpenTTD/OpenTTD] LordAro reopened issue #6503: Abnormalities in train subtile coordinates when reversing at the end of line https://git.io/fhZIc 15:22:49 <LordAro> mwhaha 15:23:01 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain closed issue #8045: AI estimated costs are 0 for commands that fail while returning a cost https://git.io/JvSaL 15:23:04 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8748: Fix 02e770ff: allow estimating CloneVehicle if short on money https://git.io/Jtbdq 15:23:47 <Eddi|zuHause> LordAro: so you're volunteering? :p 15:24:25 <LordAro> nope! 15:25:52 <_dp_> and I don't think I ever reported similar stuff for road vehicles because it's such a mess that I couldn't figure out a logical way to plot subtile movement for them xD 15:26:39 <Eddi|zuHause> movement is also a fractal... 15:26:55 <Eddi|zuHause> we have this length and speed distortion on diagonals 15:27:06 <Eddi|zuHause> we have jumpy movement on x4 zoom 15:27:46 <_dp_> diagonals aren't that much of a problem for movement 15:28:30 <_dp_> but sorting out subtile stuff will help the 4x cause for sure 15:28:39 <Eddi|zuHause> https://github.com/OpenTTD/OpenTTD/issues/1063 15:30:32 <_dp_> Eddi|zuHause, yeah, but it's more of a ui/gamplay issue, movement code doesn't really care about it 15:32:09 <_dp_> it's unreasonable to expect smooth train graphics when rails do 45 deg corner :p 15:32:32 <Eddi|zuHause> "smooth" and "inconsistent" are very different 15:35:11 <supermop_Home> maybe instead of worrying about bounding boxes, it would be best to just always force rear catenary sprites to be drawn before any vehicle etc? 15:35:26 <_dp_> that inconsistency comes from the grid-based nature of the game 15:37:03 <_dp_> you can't change graphics without changing speed but changing speed raises balance issues 15:38:43 <Eddi|zuHause> _dp_: i disagree. you can not *completely* fix it with integers, but you could make it more consistent. the main problem (as discussed in the old ticket) is that this breaks all kinds of backwards compatibility with graphics 15:45:30 <_dp_> integers are kinda irrelevant here but, yeah, graphics compatibility is a significant issue here as well 15:45:49 <_dp_> but what I'm saying that inconsistency didn't come out of nowhere, it has a reason to exist 15:46:18 <_dp_> which cased graphics to be made the way they are 15:46:27 <Eddi|zuHause> the inconsistency comes exactly from a quick-and-dirty integer implementation that was never properly fixed 15:47:01 <Eddi|zuHause> like rounding 1/sqrt(2) to 0.5 15:47:11 <Eddi|zuHause> or to 1 15:47:23 <_dp_> but game isn't rounding it to 0 or 1 15:47:27 <Eddi|zuHause> in different parts of the code 15:47:28 <_dp_> it's rounding to 128 and 196 15:47:41 <_dp_> which could easily be more precise 15:48:02 <_dp_> *192 15:49:04 <Eddi|zuHause> the main problem isn't how exact the rounding is. the problem is that it's rounded differently for speed and length 15:49:42 <_dp_> if it was different trains would have gaps :p 15:49:56 <Eddi|zuHause> so on making a turn, the train changes length, so for a while, front and tail move at different speeds 15:50:33 <_dp_> yeah, it's not your typical euclidean space :P 15:53:13 *** Progman has joined #openttd 15:58:09 <_dp_> but I think it is a metric space nonetheless 15:58:58 *** Wuzzy has joined #openttd 15:59:43 <_dp_> so in a way it doesn't change length and speed, it just uses non-euclidean length 16:00:32 <Wuzzy> so is there any way i can help with work on <https://github.com/OpenTTD/OpenSFX/pull/19> now? i am a bit overwhelmed atm 16:00:35 <Eddi|zuHause> i don't think you understood what i was trying to say 16:07:50 *** Flygon has quit IRC 16:10:04 <_dp_> neither do you :p 16:10:38 <_dp_> basically, if game used euclidean space there would be no point in ever using diagonal tracks 16:11:06 <Eddi|zuHause> what? 16:11:11 <Eddi|zuHause> you're making no sense 16:16:57 *** gelignite has joined #openttd 16:22:46 <_dp_> euclidean distance makes all directions equal. but on a grid building diagonal stuff is much harder 16:23:00 <_dp_> there would be no point going diagonal if there was no benefit to it 16:23:38 <_dp_> so game has to compromise 16:28:54 *** Samu has joined #openttd 16:32:17 <Wolf01> _dp_: we should switch to hex grid 16:33:39 <_dp_> if anything it would be better to have no grid at all 16:33:51 <_dp_> but that would be a completely different game 16:34:03 <_dp_> like nimby for example 17:04:01 *** supermop_Home has quit IRC 17:06:11 *** Wormnest has joined #openttd 17:30:27 *** supermop_Home has joined #openttd 17:41:23 *** rptr has joined #openttd 17:59:15 *** Wormnest has quit IRC 17:59:15 *** supermop_Home has quit IRC 17:59:15 *** Ttech has quit IRC 17:59:15 *** Vadtec has quit IRC 17:59:15 *** Smedles has quit IRC 17:59:15 *** fnutt has quit IRC 17:59:15 *** EmeraldSnorlax[m] has quit IRC 17:59:15 *** Speeder has quit IRC 17:59:15 *** WormnestAndroid has quit IRC 17:59:15 *** y2kboy23 has quit IRC 17:59:15 *** crem2 has quit IRC 17:59:15 *** gregdek has quit IRC 17:59:15 *** mindlesstux has quit IRC 17:59:15 *** Extrems has quit IRC 17:59:15 *** dwfreed has quit IRC 17:59:15 *** innocenat_ has quit IRC 17:59:15 *** greeter has quit IRC 17:59:15 *** dvim has quit IRC 17:59:15 *** Maarten has quit IRC 17:59:15 *** mikegrb has quit IRC 17:59:15 *** robert[m]2 has quit IRC 17:59:15 *** roadt_ has quit IRC 17:59:15 *** APTX has quit IRC 17:59:15 *** daspork has quit IRC 17:59:15 *** k-man has quit IRC 17:59:15 *** jgx has quit IRC 17:59:15 *** reldred has quit IRC 17:59:15 *** twpol has quit IRC 17:59:15 *** colde has quit IRC 17:59:15 *** snail_UES_ has quit IRC 17:59:15 *** Gustavo6046 has quit IRC 17:59:16 *** heffer has quit IRC 17:59:16 *** muffindrake has quit IRC 17:59:16 *** murr4y has quit IRC 17:59:16 *** nnyby has quit IRC 17:59:16 *** jinks has quit IRC 17:59:16 *** DorpsGek has quit IRC 17:59:16 *** ST2 has quit IRC 17:59:22 *** supermop_Home has joined #openttd 17:59:22 *** Wormnest has joined #openttd 17:59:22 *** Maarten has joined #openttd 17:59:22 *** Ttech has joined #openttd 17:59:22 *** dwfreed has joined #openttd 17:59:22 *** Vadtec has joined #openttd 17:59:22 *** Smedles has joined #openttd 17:59:22 *** mindlesstux has joined #openttd 17:59:22 *** innocenat_ has joined #openttd 17:59:22 *** fnutt has joined #openttd 17:59:22 *** EmeraldSnorlax[m] has joined #openttd 17:59:22 *** Speeder has joined #openttd 17:59:22 *** Extrems has joined #openttd 17:59:22 *** WormnestAndroid has joined #openttd 17:59:22 *** y2kboy23 has joined #openttd 17:59:22 *** crem2 has joined #openttd 17:59:22 *** greeter has joined #openttd 17:59:22 *** gregdek has joined #openttd 17:59:22 *** dvim has joined #openttd 17:59:22 *** k-man has joined #openttd 17:59:22 *** mikegrb has joined #openttd 17:59:22 *** jgx has joined #openttd 17:59:22 *** reldred has joined #openttd 17:59:22 *** robert[m]2 has joined #openttd 17:59:22 *** twpol has joined #openttd 17:59:22 *** murr4y has joined #openttd 17:59:22 *** nnyby has joined #openttd 17:59:22 *** roadt_ has joined #openttd 17:59:22 *** jinks has joined #openttd 17:59:22 *** APTX has joined #openttd 17:59:22 *** daspork has joined #openttd 17:59:22 *** ST2 has joined #openttd 17:59:22 *** DorpsGek has joined #openttd 17:59:22 *** muffindrake has joined #openttd 17:59:22 *** heffer has joined #openttd 17:59:22 *** Gustavo6046 has joined #openttd 17:59:22 *** snail_UES_ has joined #openttd 17:59:22 *** colde has joined #openttd 17:59:22 *** synthon.oftc.net sets mode: +ov DorpsGek DorpsGek 18:01:53 *** frosch123 has joined #openttd 18:10:25 *** Wuzzy2 has joined #openttd 18:11:47 *** Wuzzy has quit IRC 18:22:20 *** dvim has quit IRC 18:22:20 *** gregdek has quit IRC 18:22:20 *** greeter has quit IRC 18:22:20 *** crem2 has quit IRC 18:22:20 *** y2kboy23 has quit IRC 18:22:20 *** WormnestAndroid has quit IRC 18:22:20 *** Extrems has quit IRC 18:22:20 *** Speeder has quit IRC 18:22:20 *** EmeraldSnorlax[m] has quit IRC 18:22:20 *** fnutt has quit IRC 18:22:20 *** innocenat_ has quit IRC 18:22:20 *** mindlesstux has quit IRC 18:22:20 *** Smedles has quit IRC 18:22:20 *** Vadtec has quit IRC 18:22:20 *** Wormnest has quit IRC 18:22:20 *** supermop_Home has quit IRC 18:22:20 *** dwfreed has quit IRC 18:22:20 *** Ttech has quit IRC 18:22:35 *** supermop_Home has joined #openttd 18:22:35 *** Wormnest has joined #openttd 18:22:35 *** Ttech has joined #openttd 18:22:35 *** dvim has joined #openttd 18:22:35 *** gregdek has joined #openttd 18:22:35 *** greeter has joined #openttd 18:22:35 *** crem2 has joined #openttd 18:22:35 *** y2kboy23 has joined #openttd 18:22:35 *** WormnestAndroid has joined #openttd 18:22:35 *** Extrems has joined #openttd 18:22:35 *** Speeder has joined #openttd 18:22:35 *** EmeraldSnorlax[m] has joined #openttd 18:22:35 *** fnutt has joined #openttd 18:22:35 *** innocenat_ has joined #openttd 18:22:35 *** mindlesstux has joined #openttd 18:22:35 *** Smedles has joined #openttd 18:22:35 *** Vadtec has joined #openttd 18:22:35 *** dwfreed has joined #openttd 18:26:58 <DorpsGek> [OpenTTD/OpenTTD] michicc merged pull request #8743: Change: Default to a 32bpp blitter. https://git.io/Jtdoj 18:27:27 <DorpsGek> [OpenTTD/OpenTTD] 2TallTyler opened pull request #8749: WIP Feature: Setting to loop the same year forever https://git.io/JtNkN 18:28:46 <andythenorth> :o 18:28:47 <andythenorth> :) 18:32:21 *** Strakaty has joined #openttd 18:37:00 <DorpsGek> [OpenTTD/OpenTTD] James103 commented on pull request #8749: WIP Feature: Setting to loop the same year forever https://git.io/JtNI2 18:38:20 *** robert[m]2 has quit IRC 18:38:20 *** Maarten has quit IRC 18:38:20 *** mikegrb has quit IRC 18:38:50 *** Maarten has joined #openttd 18:38:50 *** mikegrb has joined #openttd 18:38:50 *** robert[m]2 has joined #openttd 18:41:05 *** colde has quit IRC 18:41:05 *** snail_UES_ has quit IRC 18:41:05 *** Gustavo6046 has quit IRC 18:41:05 *** heffer has quit IRC 18:41:05 *** muffindrake has quit IRC 18:41:05 *** jinks has quit IRC 18:41:05 *** nnyby has quit IRC 18:41:05 *** murr4y has quit IRC 18:41:05 *** reldred has quit IRC 18:41:05 *** jgx has quit IRC 18:41:05 *** ST2 has quit IRC 18:41:05 *** twpol has quit IRC 18:41:05 *** DorpsGek has quit IRC 18:41:19 *** jgx has joined #openttd 18:41:19 *** reldred has joined #openttd 18:41:19 *** twpol has joined #openttd 18:41:19 *** colde has joined #openttd 18:41:19 *** snail_UES_ has joined #openttd 18:41:19 *** Gustavo6046 has joined #openttd 18:41:19 *** heffer has joined #openttd 18:41:19 *** muffindrake has joined #openttd 18:41:19 *** murr4y has joined #openttd 18:41:19 *** nnyby has joined #openttd 18:41:19 *** jinks has joined #openttd 18:41:19 *** DorpsGek has joined #openttd 18:41:19 *** ST2 has joined #openttd 18:41:19 *** reflection.oftc.net sets mode: +ov DorpsGek DorpsGek 18:43:33 *** rptr_ has joined #openttd 18:44:10 <frosch123> so did dih*dral succeed in hiring Tru*Brain? that part is unclear from the log 18:44:29 <TrueBrain> was he attempted to do so? Oops .. might have missed that :P 18:45:58 <frosch123> the meetup is some years ago, but back then he had an IT company 18:46:34 <andythenorth> we could all hire each other? 18:46:46 <TrueBrain> pyramid scheme? 18:46:54 <andythenorth> if we can just keep the money moving around on 30 day invoicing, I think it works 18:47:24 <frosch123> we could make some VAT triangle trade 18:47:36 <andythenorth> there is some illegal form of that 18:47:46 <TrueBrain> UK is now out of the EU; not sure if business there still is against 0% VAT now :P 18:47:46 * andythenorth won't be indulging 18:48:15 <TrueBrain> annnyyywwaaayyy, hello to you too frosch123 :) 18:48:38 <frosch123> i read you prepared a challenge for me to break something 18:48:49 <TrueBrain> breaking it is easy, honestly 18:49:09 <TrueBrain> as I missed at least one place that changes _vd out of the scope I thought it had 18:49:17 <TrueBrain> but what do you think about such solution? 18:49:27 *** rptr has quit IRC 18:49:28 <TrueBrain> a thread should constantly be reading _vd_queue of course 18:49:36 <frosch123> i did not look at the code yet 18:50:05 <TrueBrain> its funny how the NewGRFs you have loaded make such change from worth the time to barely useful :P 18:50:48 <TrueBrain> but viewport.cpp is such ugly code :P It needs some refactoring before such patch would be doable to get accepted :) 18:51:01 <TrueBrain> mainly _vd's access needs to be more .. in contorl 18:51:26 <frosch123> well, when you are done with thread-pooling the viewport drawing, i can hint you how to thread-pool the newgrf stuff :p 18:51:48 <frosch123> i will set the estimate for multi-thread newgrf to two week, just so you take the challenge bait :) 18:52:27 <TrueBrain> that would be safe to do so? Now that would make it interesting :) 18:52:27 <TrueBrain> frosch123: hahaha :D 18:52:33 <TrueBrain> well, this was a few hours work instead of a week, so ... :P 18:52:45 <TrueBrain> nah, a real PR would take a week or so 18:53:27 <frosch123> resolving vehicle sprites in paralell is only limited by using too many global variables 18:53:43 <frosch123> but they access the gamestate read-only 18:53:47 <TrueBrain> OpenTTD does have that issues, yes ... too many global state :D 18:53:58 <TrueBrain> cool :D 18:54:17 <frosch123> about the game-state itself. you can shuffle the vehicle ticks around, so they do the acceleration-computation in paralle, and then advance sequentially 18:54:19 <TrueBrain> I found another optimization that would remove some of the stutter when doing things quickly ... never draw the whole screen all at once :P :P 18:54:46 <TrueBrain> normally the buffer to sort is like 10k items 18:54:50 <TrueBrain> but in peek it is 8M items 18:54:54 <TrueBrain> for some reason, that takes longer 18:55:53 <TrueBrain> but the dirty-rect in the viewport is crazy complex 18:55:57 <TrueBrain> a line-scanner of some sorts 18:56:04 <TrueBrain> couldn't figure it out within 10 minutes, so just ignored it 18:56:04 <frosch123> yes, the spritesorter is quadratic or cubic 19:01:40 <DorpsGek> [OpenTTD/OpenTTD] DorpsGek pushed 1 commits to master https://git.io/JtNtR 19:01:41 <DorpsGek> - Update: Translations from eints (by translators) 19:02:32 <TrueBrain> also a funny thing, viewport locations are updated after they are drawn 19:02:34 <TrueBrain> which feels funky 19:02:55 <TrueBrain> (also means they always lag behind 1 tick) 19:04:14 *** Alkel_U3 has quit IRC 19:04:28 *** Alkel_U3 has joined #openttd 19:07:22 <_dp_> sprite sorter is not that quadratic anymore, it's nearing qsort territory now 19:07:23 <peter1138> Does game time backwards yet? 19:08:02 <_dp_> you sure can make a quadratic worst case, but should be near-linear for a real game 19:08:15 <TrueBrain> https://github.com/OpenTTD/OpenTTD/commit/c693463c29f98f763cc4dd6346527262b34ff6ea <- what is this for anti-feature? :P 19:08:17 <andythenorth> peter1138 if we go fast enough, does it overflow into signed? 19:08:34 <_dp_> well, loglinear I guess since it uses std::sort 19:08:42 <peter1138> Don't look at me... 19:10:12 <peter1138> Have you undone it? 19:10:26 <TrueBrain> no, it is rightfully there; just the commit message is funny :D 19:10:33 <peter1138> Oh. 19:10:49 <peter1138> Yeah, it describes the what, not the why. 19:10:56 <peter1138> Well, sort of. 19:11:29 <TrueBrain> it is part of the PR that made my work possible, so yeah ;) 19:11:47 <peter1138> It is amusing to watch vehicles teleport across the map because the game didn't bother drawing the inbetween states. 19:12:10 <peter1138> I was useful after all? Damn. 19:12:25 <TrueBrain> yeah, I do apologize for pointing that out 19:12:47 <TrueBrain> okay, in r1 it was already that first the screen was redrawn, than the viewport locations were updated 19:12:48 <TrueBrain> very odd 19:12:56 <TrueBrain> than? then 19:13:00 <TrueBrain> lol ... typing hard 19:13:27 <TrueBrain> GitHub interface has a lovely addition to blame functionality .. no clue how long it is in there 19:13:37 <TrueBrain> but on each line you can click: show me the file from before this commit 19:13:41 <TrueBrain> so it makes delving down really easy 19:14:06 <TrueBrain> also: https://github.blog/changelog/2021-02-23-github-discussions-graphql-api-public-beta/ 19:14:13 <TrueBrain> means we can make the bot announce those events I guess 19:15:02 <TrueBrain> no webhook event yet, bah 19:17:40 <TrueBrain> ah, viewport is moved after drawing, as otherwise it doesn't know what pixels to redraw .. 19:17:42 <TrueBrain> lazy :P 19:18:34 <supermop_Home> is it worth suing my landlord? 19:23:20 <TrueBrain> holy crap, I enabled the overlay for cargo flow 19:32:06 *** ericnoan has quit IRC 19:32:27 <peter1138> uh oh 19:32:40 *** ericnoan has joined #openttd 19:36:56 *** jgr has joined #openttd 19:43:12 *** rptr_ has quit IRC 19:49:17 *** otetede has joined #openttd 19:50:59 <michi_cc> And with that TrueBrain vanished and was never to be seen again :D 19:51:12 *** Artea has quit IRC 19:51:16 *** Artea has joined #openttd 20:00:50 <TrueBrain> in horror, at best :P 20:00:51 <TrueBrain> :D 20:01:40 <TrueBrain> does anyone mind if I mention the Discord we are on for OpenTTD on our webpage for Support question? 20:02:10 <TrueBrain> no doubt Steam will give an influx of people with stupid questions, and they are kinda setup of dealing with it, and doing a good job at it 20:02:29 <TrueBrain> would take that burden away from us :P 20:03:11 <TrueBrain> and very specific, I mean mentioning here: https://www.openttd.org/contact.html 20:03:19 <TrueBrain> (and on the Steam page) 20:04:16 <frosch123> i would vote for removing all personal emails from that page, add links to forums, reddit, discord 20:04:31 <frosch123> and make sure that info@ is only used for serious stuff 20:04:33 <TrueBrain> I was working on something similar :P 20:04:39 <frosch123> never figure out a wording for that though 20:04:55 <TrueBrain> but it does require making the Discord (and possibly reddit) "official" 20:05:22 <TrueBrain> I will cook something up, so we can see a bit what is going on :) 20:07:52 <TrueBrain> in return, I am guessing the Discord owner will ask if we can make the Discord official, by going through the validation process that we, openttd.org, have at least 1 owner in that Discord. I assume that isn't a real problem for anyone? 20:08:15 <TrueBrain> I would make DorpsGek at info@ that owner 20:08:44 <frosch123> nah, they don't need to be official 20:08:54 <TrueBrain> quid-pro-quo 20:09:18 <frosch123> we just need to find a wording, so that people understand to ask other people :) 20:09:47 <TrueBrain> I mean, is there a reason not to make it a validated Discord server? 20:09:59 <TrueBrain> wording is easy, I think :D 20:13:31 <_dp_> is there any good way for slowing down map download on localhost (for testing)? 20:14:02 <frosch123> wasn't that asked just yesterday? 20:14:28 <_dp_> idk, I may have missed that 20:16:30 <frosch123> 20:27:21 <LordAro> https://stackoverflow.com/q/3536249/995325 20:16:31 <_dp_> oh, nvm I forgot to make 4k map 20:16:40 <_dp_> that solves it nicely xD 20:17:55 <michi_cc> It's a well run Discord, I wouldn't mind if anyone put a stamp on it. 20:18:24 *** Speedy` has quit IRC 20:18:36 *** Speedy` has joined #openttd 20:19:10 *** glx has joined #openttd 20:19:10 *** ChanServ sets mode: +v glx 20:21:36 *** Markk has quit IRC 20:21:41 *** Markk has joined #openttd 20:35:41 <_dp_> I finally got the desync I've been hunting for a year 20:35:50 <_dp_> but a repro involves patched server and a python script 20:35:56 <_dp_> fan-freaking-tastic :/ 20:37:19 <andythenorth> ha 20:41:16 <_dp_> lol, I have progress, unpatched server but 2 python scripts now xD 20:42:08 <LordAro> it counts! 20:42:17 <_dp_> and 2 bugs in network gui as a bonus xD 20:44:23 <_dp_> there is openttd server called mashinky oO 20:44:59 <frosch123> call your's trainfever? 20:45:19 *** _2TallTyler has joined #openttd 20:45:20 <TrueBrain> and for a moment, enjoy the moment of missing certain other servers from the list :) 20:45:23 <TrueBrain> I should still PR that :P 20:46:26 <frosch123> i don't know what you mean :p 20:50:48 <TrueBrain> working on the website reminds me how much the layout sucks for 2021 :P 20:51:19 <TrueBrain> https://pasteboard.co/JQc0gBR.png 20:51:32 <TrueBrain> some text still needs work, but how about something like this on our Contact page? 20:52:00 <frosch123> not feature requests on gh please :) 20:52:26 <TrueBrain> that shouldn't be there, no 20:52:39 <TrueBrain> notice how info@ is now at the bottom? :D 20:53:22 <_2TallTyler> Maybe provide a link to the manual instead of sending people to ask signal questions on Discord/Reddit? 20:53:39 <_2TallTyler> I don't know what you have in the menu above, maybe you already have a link 20:53:42 <TrueBrain> there is already a Manual button, yes 20:53:50 <frosch123> and a community button 20:53:57 <TrueBrain> so if people got here, they already went past that :P 20:54:00 <frosch123> it's just to avoid adding a captcha to the emails 20:54:37 <TrueBrain> so, improvement? 20:55:04 <frosch123> some category "how to complain about other players on the server i am playing on" 20:55:11 <frosch123> not sure where to link to though :p 20:55:14 <TrueBrain> exactly :P 20:55:27 <TrueBrain> reddit seems the current target :P 20:55:29 <TrueBrain> but that is not nice :D 20:55:38 <_2TallTyler> An improvement though, definitely 20:55:49 <_dp_> frosch123, /dev/null :p 20:56:27 <frosch123> can we display a category "cat pictures" on april 1st? 20:56:48 *** michi_cc has quit IRC 20:56:48 <frosch123> by forum pm to andy, he'll never figure it out 20:56:51 *** michi_cc has joined #openttd 20:56:51 *** ChanServ sets mode: +v michi_cc 20:57:03 *** supermop_Home has quit IRC 20:58:24 <frosch123> TrueBrain: do the forum links link to the specific subforums? 20:58:39 <frosch123> problems / feature requests 20:58:39 <TrueBrain> not yet, will do 20:59:14 <_2TallTyler> Question: If I have a CompanyID, how can I get the Company associated with it? I know how to go the other way with company->index, just not the reverse. 20:59:32 <TrueBrain> GetCompanyById or something? There is some function like that 20:59:32 <frosch123> Company::Get or Company::GetIfValid 20:59:33 <glx> Company::Get 20:59:46 <TrueBrain> bah, I was waaayyyy off :P 20:59:53 <LordAro> Get::Company 21:00:00 <frosch123> one asserts for invaild ids, the other returns nullptr 21:00:17 <TrueBrain> frosch123: do you have a problem with it if we validate the Discord server? (to approach it from the different angle) 21:00:51 <frosch123> i have no idea what that means :) but as long as i do not have to be there, i don't care :) 21:00:57 <TrueBrain> you don't 21:00:58 <glx> we don't own it (same as forum), but it can be helpful 21:01:08 <TrueBrain> it just adds a green checkmark, indicating openttd.org validated the Discord as OpenTTD's ;) 21:01:23 <TrueBrain> that is literally everything :P 21:01:46 <glx> and I think it's the same with reddit 21:01:50 <TrueBrain> I guess I should also ask the reddit owners if they don't mind if we link to them .. I doubt they will 21:01:53 <TrueBrain> but always better to ask 21:02:08 *** gnu_jj has quit IRC 21:02:13 *** gnu_jj has joined #openttd 21:02:45 <frosch123> i thought they were the same 21:03:01 <TrueBrain> I believe the owners are 2 different people, but I will check :) 21:05:43 <_2TallTyler> `Company::GetIfValid(company)` says incomplete type is not allowed. Am I doing something wrong? 21:06:12 <frosch123> that means you miss the #include 21:06:33 <frosch123> in general: when c++ gives you a weird error message, it's always a missing #include 21:07:09 <frosch123> msvc usually says "missing type specifier, assuming int" 21:07:15 <frosch123> other compilers are more clever 21:08:17 <_dp_> there are plenty of weird compiler errors 21:08:18 <_2TallTyler> So I'm missing the #include for the while which contains GetIfValid? Or Company? 21:08:24 <_dp_> try forgetting } somewhere 21:08:27 <_2TallTyler> *file, not while 21:08:29 <_dp_> or messing with templates :p 21:09:02 <DorpsGek> [OpenTTD/website] TrueBrain opened pull request #192: Add: rework our "Contact" page https://git.io/JtNn3 21:09:05 <glx> company_type.hpp ? 21:09:14 <_dp_> _2TallTyler, company_base.h i think 21:10:02 <_2TallTyler> Aha, that did it. Thanks! 21:12:08 <DorpsGek> [OpenTTD/website] TrueBrain updated pull request #192: Add: rework our "Contact" page https://git.io/JtNn3 21:12:56 <_dp_> well, that's one weird error... https://i.imgur.com/buGBREo.png 21:13:11 <TrueBrain> especially that HUGE cursor :P 21:13:26 <TrueBrain> and I see my PR is still not merged :D 21:13:43 <TrueBrain> owh, it is not reviewed yet :P 21:13:46 <TrueBrain> such slackers here :D 21:14:10 <TrueBrain> frosch123 / michi_cc : C++ wizards, what is your verdict on https://github.com/OpenTTD/OpenTTD/pull/8715#discussion_r583467886 ? 21:16:10 <michi_cc> Move semantics are only relevant if copying is expensive. For a type that is not larger than a pointer and has no constructor/destructor, it has no practical meaning. 21:16:39 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8715: Fix: Vehicle list windows did not update when this year's profit changed https://git.io/JtNnF 21:16:46 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain approved pull request #8715: Fix: Vehicle list windows did not update when this year's profit changed https://git.io/JtNnN 21:16:51 <TrueBrain> cheers :) 21:16:58 <frosch123> yes, no std::move for trivial types. it only slows down non-inlined debug builds 21:17:26 *** supermop_Home has joined #openttd 21:18:29 <frosch123> but ther GetOldestVehicleAge is broken 21:18:44 <LordAro> oh no 21:18:58 <TrueBrain> do tell? 21:19:05 <frosch123> i need to login 21:19:11 <frosch123> but can the list be empty? 21:19:13 <TrueBrain> you are not by default? :o 21:22:42 <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #8715: Fix: Vehicle list windows did not update when this year's profit changed https://git.io/JtNcu 21:23:05 <TrueBrain> lol .. lovely details :D 21:24:01 <frosch123> btw. you can also use std::accumulate with std::max 21:25:30 <frosch123> just in case you want to become a haskell programmer in the future 21:33:40 <andythenorth> that's probably the most epic gravedig ever https://www.tt-forums.net/viewtopic.php?f=29&t=6566 21:34:06 <_dp_> you can do functional programming in python! 21:34:22 <_dp_> coworkers were very happy to see my 2 screen long function call :p 21:35:14 <frosch123> andythenorth: what's that game? i don't recognise the title in the window bar 21:35:23 <andythenorth> me neither 21:35:31 <andythenorth> 17 years though 21:36:28 <_dp_> haha, BUSTED! :p 21:36:44 <frosch123> _dp_: every modern language can do functional programming, lambdas and coroutines 21:36:54 <frosch123> they like to talk about that all the time 21:38:49 <frosch123> you can recognise the real functional programmers when they complain that ohter languages are not self-modifying... 21:40:46 <DorpsGek> [OpenTTD/OpenSFX] glx22 commented on pull request #19: Replace Mercurial code with Git code https://git.io/JtNCA 21:41:47 <glx> if you want to do functional go Caml and O'Caml 21:41:52 *** ookfof[m] has quit IRC 21:41:56 *** ookfof[m] has joined #openttd 21:42:04 <DorpsGek> [OpenTTD/OpenTTD] 2TallTyler opened issue #8750: Graph point connector lines are transparent https://git.io/JtNWv 21:43:41 <frosch123> i only hear about clojure 21:43:54 <frosch123> is caml newer? 21:43:56 <_dp_> looks like opengl has thinner lines 21:44:02 <glx> caml is old 21:44:17 <glx> learned it around 2000 21:46:47 <LordAro> looks like the lines are transparent 21:48:12 <frosch123> anti-aliasing? 21:49:49 <TrueBrain> I blame _2TallTyler , he worked last on the graphs! 21:50:12 <_dp_> TrueBrain, it's fine without opengl :p 21:50:16 <frosch123> oh, silly me. we draw our lines in our own blitters... i thought we had opengl lines :p 21:50:20 <michi_cc> Without having checked this issue, in general there are some corner cases where the 8bpp only looked at the palette value and the 32bpp blitters only look at rgb, while the 40bpp blitter keeps both. 21:50:23 <TrueBrain> irrelevant _dp_ , we need someone to blame, so that is it 21:50:37 <_2TallTyler> I thought I had broken it at first! 21:51:02 *** Tirili has joined #openttd 21:51:21 <_2TallTyler> But you reviewed and approved my code so it would be partially your fault too ;) 21:51:28 <TrueBrain> darnit! 21:51:31 <TrueBrain> remind me to stop reviewing :P 21:51:32 <TrueBrain> :D :D 21:52:08 <_dp_> blame CS, everything is his fault.... partially :p 21:55:34 *** Samu has quit IRC 21:59:21 *** frosch123 has quit IRC 21:59:28 *** ircer[m] has quit IRC 21:59:39 *** ircer[m] has joined #openttd 22:00:27 *** otetede has quit IRC 22:02:18 <nielsm> https://0x0.st/-KQY.webm I got a thing kinda working, and needed a test file 22:02:24 *** iarp[m] has quit IRC 22:02:29 *** iarp[m] has joined #openttd 22:03:19 <glx> haha encoding issues ;) 22:06:36 <andythenorth> nielsm is livestream? :P 22:07:00 <nielsm> I don't have a proper driver for it at hand, but I know it's supposed to be capable of grapgics 22:07:14 <nielsm> because I can see the shadow of a windows 98 logo on the paper roller 22:07:26 <nielsm> (printer test page printed without paper) 22:07:42 <glx> I think our amstrad dmp 2160 should still work (just needs new ribbon probably) 22:08:22 <nielsm> you're more likely to rejuvenate the old ribbon than find a new one that fits 22:08:29 <glx> and it used to work on at least win95 (I'm remember I tried it) 22:08:55 <glx> compatible with epson driver 22:09:24 *** sla_ro|master has quit IRC 22:17:06 <DorpsGek> [OpenTTD/OpenTTD] ldpl opened issue #8751: Disconnecting clients affect other clients in the queue https://git.io/JtN8c 22:23:23 <DorpsGek> [OpenTTD/website] nielsmh commented on pull request #192: Add: rework our "Contact" page https://git.io/JtN8P 22:24:18 <TrueBrain> nielsm: we follow the forums? :D 22:24:25 <TrueBrain> I will add something in that vibe, good idea :) 22:25:28 <nielsm> because I know I've seen talk in here before mentioning long threads on reddit/discord about bugs and inconveniences that then never reach anyone here until much much later, after everyone is already angry nobody is doing anything about it 22:26:10 <glx> well bugs should go on github first 22:26:30 <nielsm> yes exactly, that's the point :) 22:26:45 <nielsm> if you discuss it in a thread on reddit, chance is nobody here will see it 22:26:49 <_dp_> some feature requests are better stay on reddit though :p 22:34:02 <_dp_> oh, got desync repro without python as well, updated the bug 22:34:40 *** _2TallTyler has quit IRC 22:35:07 <_dp_> now I'm pretty sure that accounts for at least some of the desyncs plaguing the servers 22:35:38 <_dp_> and mb some other weird errors as well 22:48:07 <DorpsGek> [OpenTTD/website] TrueBrain updated pull request #192: Add: rework our "Contact" page https://git.io/JtNn3 22:48:35 <DorpsGek> [OpenTTD/website] TrueBrain commented on pull request #192: Add: rework our "Contact" page https://git.io/JtNBC 22:51:15 <DorpsGek> [OpenTTD/website] TrueBrain updated pull request #192: Add: rework our "Contact" page https://git.io/JtNn3 22:52:43 *** gelignite has quit IRC 23:04:10 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on issue #8751: Disconnecting clients affect other clients in the queue https://git.io/JtN8c 23:04:58 *** jottyfan has joined #openttd 23:05:43 *** jottyfan has quit IRC 23:07:01 <_dp_> oh, and desync is probably because clients are not queued 23:07:23 <_dp_> so when second one tries to catch up client is already gone 23:07:25 <andythenorth> bed 23:07:26 *** andythenorth has quit IRC 23:07:46 <_dp_> kinda makes it a separate bug though if that's true 23:08:37 *** nielsm has quit IRC 23:15:30 *** Tirili has quit IRC 23:18:40 <DorpsGek> [OpenTTD/OpenTTD] ldpl commented on issue #8751: Disconnecting clients affect other clients in the queue https://git.io/JtN8c 23:25:52 *** Wolf01 has quit IRC 23:27:07 <TrueBrain> lol, a bug in our state machine .. another 14 year old bug fix incoming tomorrow? :D 23:29:57 <LordAro> tis the season for it 23:31:57 <TrueBrain> one still needs a review :D 23:32:25 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #8746: Fix bddfcaef: don't tell twice that a client left because of a timeout etc https://git.io/JtNE8 23:32:42 <TrueBrain> \o/ 23:32:55 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8746: Fix bddfcaef: don't tell twice that a client left because of a timeout etc https://git.io/JtFxq 23:32:59 <TrueBrain> cheers 23:33:16 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #8747: Codechange: replace _realtime_tick with std::chrono https://git.io/JtNE0 23:33:17 <LordAro> oh, i did already look through this one 23:33:32 <LordAro> let's find out if chrono is stable enough 23:33:44 <DorpsGek> [OpenTTD/OpenTTD] JGRennison commented on issue #8751: Disconnecting clients affect other clients in the queue https://git.io/JtN8c 23:33:54 <DorpsGek> [OpenTTD/OpenTTD] LordAro approved pull request #8745: Fix: don't link to OpenGL with SDL2 as backend; SDL2 dynamically loads it https://git.io/JtNEu 23:34:45 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8745: Fix: don't link to OpenGL with SDL2 as backend; SDL2 dynamically loads it https://git.io/JtFQ4 23:36:09 <TrueBrain> LordAro: our video drivers depend on it; I sure hope it is :D 23:36:19 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8747: Codechange: replace _realtime_tick with std::chrono https://git.io/JtFpt 23:36:40 <LordAro> true enough :) 23:38:58 <_dp_> feels like overkill to create a bug for it but can someone with better english fix this message? https://github.com/OpenTTD/OpenTTD/blob/master/src/network/network_server.cpp#L1108 23:39:23 <_dp_> smth along the lines "spectator ({client_id}) issued non-spectator command {cmd}" 23:39:59 <Xaroth> then create a PR for it? :P 23:40:19 <LordAro> _dp_: as you say would be fine 23:40:48 <LordAro> s/issuing command/issued non-spectator command {cmd}/ 23:44:15 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8741: Move GameLoop into a thread (and no longer run Paint in a thread) https://git.io/JtdUz 23:44:25 <TrueBrain> I think ^^ is now ready for review, but I need some sleep first to be sure :P 23:44:32 <_dp_> hm... and what's the correct format for command id? xD 23:46:22 <_dp_> though I guess those messages don't particularly care about correct format strings 23:47:07 <_dp_> pretty sure %d isn't quite the right one for uint32 23:48:56 <LordAro> %u 23:49:33 <_dp_> not %"PRIu32" ? 23:49:41 <LordAro> well, technically speaking, yes 23:49:45 <LordAro> but urgh 23:49:52 <LordAro> can you imagine having to write that every time? :p 23:50:20 <_dp_> no, so I use fmt::* :p 23:50:40 <_dp_> or %d and cast everything to int xDD 23:50:52 <TrueBrain> or just, you know, %u :) 23:52:30 <_dp_> does %u actually work fine on 64 bit archs? 23:53:06 <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8741: Move GameLoop into a thread (and no longer run Paint in a thread) https://git.io/JtdUz 23:53:09 <TrueBrain> okay, so this is ready for review ^^ :) 23:53:25 <TrueBrain> _dp_: grep for it in our code, and get your answer 23:53:28 <LordAro> dejavu 23:53:45 <TrueBrain> nah, now I am sure :) 23:54:00 <TrueBrain> would really appreciate it if other people could test it for a bit 23:54:02 <TrueBrain> see if they can break it 23:54:15 <_dp_> TrueBrain, the first thing I did is I grepped PRIu32, 0 hits :p 23:54:19 <TrueBrain> looking at the code, I see no way it can deadlock, but .. 23:54:25 <TrueBrain> _dp_: and now grep %u ... 23:54:36 <TrueBrain> some questions are already answered by just looking at how other code does it, really :) 23:54:45 <_dp_> I don't need to grep to see %d used for uint32 :p 23:55:05 <TrueBrain> or, of course, you can create problems where there are now :) That is fine too :) 23:55:08 <TrueBrain> now = none 23:55:09 <_dp_> question is should I try to make it correct or just do like other code does :p 23:55:34 <TrueBrain> lot of code does %u, so I am not sure what you hope to gain from this reasoning :) 23:55:55 <TrueBrain> I am just waiting for the "OpenTTD sucks" sentence :D 23:56:16 <_dp_> mostly I'm just curious how %u manages to work fine even on 64 architectures xD 23:56:31 <TrueBrain> what do you expect that breaks? 23:56:38 <_dp_> I'd expect it to crash or print garbage 23:56:46 <TrueBrain> why? 23:56:50 <_dp_> since it should expect 64 value 23:56:55 <_dp_> but gets 32 23:57:01 <_dp_> * bit 23:57:14 <TrueBrain> pretty sure %lu is for 64bit? 23:57:29 <TrueBrain> %u does unsigned int 23:57:43 <TrueBrain> (by specs) 23:58:23 <_dp_> but is int 4 bytes? 23:59:03 <_dp_> oh, int is <= 32 23:59:05 <TrueBrain> src/stdafx.h: typedef unsigned int uint; 23:59:08 <_dp_> so it only breaks on 16 bit :p 23:59:12 <TrueBrain> for OpenTTD, "uint" == %u 23:59:35 <_dp_> it's lu that will break on 32/64 :p