Config
Log for #openttd on 17th February 2021:
Times are UTC Toggle Colours
00:15:49  *** HerzogDeXtEr has quit IRC
00:16:13  <LordAro> Con_TheGranny: the relevant forum thread, iirc
00:18:38  <FLHerne> Con_TheGranny: Or tell andythenorth when he wakes up
00:18:56  <FLHerne> You missed him by 15 minutes
01:04:36  *** Flygon has joined #openttd
01:28:40  *** Greyfur has quit IRC
02:24:56  <DorpsGek> [OpenTTD/OpenTTD] perezdidac updated pull request #8603: Feature: Object class selection string filtering https://git.io/Jts3i
02:35:44  *** Con_TheGranny has left #openttd
02:36:12  *** con_theGranny has joined #openttd
02:38:05  <con_theGranny> LordAro: I'm new to irc so need more clues on reporting bugs. I guess I will wait to find AndytheNorth
02:39:19  <con_theGranny> I loaded FIRS 3 and it breaks the sub-tropic zone, i.e no water, no water towers, no banks and no diamonds at minimum.
03:18:42  <supermop_Home_> that's not really a bug
03:19:04  <supermop_Home_> ie that's intended behavior
03:24:53  <supermop_Home_> in terms of desert town growth effects, other FIRS cargos fill those behaviors, depending on FIRs economy loaded
03:30:10  *** D-HUND has joined #openttd
03:31:23  *** Wormnest has quit IRC
03:33:32  *** debdog has quit IRC
03:55:54  <DorpsGek> [OpenTTD/OpenTTD] perezdidac approved pull request #8659: Change: "(Business/Chair)man" to "(Business/Chair)person" https://git.io/JtMTq
04:42:43  *** y2kboy23_ is now known as y2kboy23
05:12:15  *** Wuzzy has quit IRC
05:23:03  <DorpsGek> [OpenTTD/OpenTTD] perezdidac opened pull request #8683: Feature: Town name generator for Spanish (Spain) https://git.io/JtMtf
05:25:50  <DorpsGek> [OpenTTD/OpenTTD] perezdidac updated pull request #8683: Feature: Town name generator for Spanish (Spain) https://git.io/JtMtf
05:30:56  <DorpsGek> [OpenTTD/OpenTTD] James103 commented on pull request #8683: Feature: Town name generator for Spanish (Spain) https://git.io/JtMtW
06:01:09  *** WormnestAndroid has quit IRC
06:02:44  *** WormnestAndroid has joined #openttd
06:04:24  *** WormnestAndroid has joined #openttd
06:10:09  *** WormnestAndroid has quit IRC
06:10:22  *** WormnestAndroid has joined #openttd
06:21:39  <DorpsGek> [OpenTTD/OpenTTD] perezdidac commented on pull request #8683: Feature: Town name generator for Spanish (Spain) https://git.io/JtMmk
06:21:45  <DorpsGek> [OpenTTD/OpenTTD] perezdidac closed pull request #8683: Feature: Town name generator for Spanish (Spain) https://git.io/JtMtf
06:21:49  <DorpsGek> [OpenTTD/OpenTTD] perezdidac reopened pull request #8683: Feature: Town name generator for Spanish (Spain) https://git.io/JtMtf
06:41:06  <DorpsGek> [OpenTTD/OpenTTD] perezdidac updated pull request #8683: Feature: Town name generator for Spanish (Spain) https://git.io/JtMtf
07:18:42  *** sla_ro|master has joined #openttd
07:34:36  *** Greyfur has joined #openttd
07:45:31  *** snail_UES_ has quit IRC
08:04:10  *** Wolf01 has joined #openttd
08:08:39  *** HerzogDeXtEr has joined #openttd
08:13:49  *** andythenorth has joined #openttd
08:15:29  <Wolf01> o/
08:24:12  <andythenorth> yo
08:24:27  <andythenorth> why have we removed very flat maps from arctic anyway?
08:24:39  <andythenorth> people are wrong on the internet, but I can't explain to them why
08:24:49  <Eddi|zuHause> because forests
08:24:50  <DorpsGek> [OpenTTD/OpenTTD] auge8472 commented on pull request #8638: Feature: [GS] Allow non-question type windows to have no buttons https://git.io/JtMGW
08:26:05  <andythenorth> I think it's actually because snowline, I think the forests might be misleading
08:26:37  <Eddi|zuHause> well, yes. the forests depend on snowline
08:26:56  <Eddi|zuHause> you need above-snowline areas for forests, and below-snowline areas for farms
08:27:12  <andythenorth> that argument is true, but doesn't carry
08:27:22  <andythenorth> because it can be solved by just putting in ever more red warning dialogues
08:27:29  <andythenorth> to explain how industries work
08:27:36  <andythenorth> which is what the playerbase are being asked to ask for
08:27:48  <Eddi|zuHause> we had those, people didn't understand them (because they weren't very clear)
08:28:07  <andythenorth> it's because they weren't long enough, or detailed enough, and don't fill enough of the screen
08:29:39  * andythenorth is old and should know better than 'internet'
08:29:42  <andythenorth> oops
08:30:28  *** nielsm has joined #openttd
08:30:32  <TrueBrain> I like that people are worried about removing an "OK" button, not seeing that most of the windows you get presented already had none :)
08:31:09  <TrueBrain> it is like complaining there is a small hole in a door, while there are no windows in the house at all
08:31:34  <LordAro> am i remembering correctly that you can close windows with a right-click?
08:31:43  <LordAro> or is it pressing escape?
08:31:46  <LordAro> or both?
08:31:48  <TrueBrain> I believe it is a setting?
08:32:28  <TrueBrain> OpenTTD has so many hidden interactions, I tend to ignore it :P
08:33:46  <Eddi|zuHause> some error windows you can close with a right click
08:33:50  <nielsm> yes transient popups (regular errors and such) close with right click
08:34:11  <nielsm> I think they also close with spacebar (like news windows do)
08:34:37  <TrueBrain> so for me a more interesting question is: do GS windows without a button close on the same interactions
08:34:43  <Eddi|zuHause> andythenorth: the problem with making the popups more detailed is that a) they become too big, and b) you're opening up more and more corner cases that you didn't cover properly.
08:35:31  <DorpsGek> [OpenTTD/OpenTTD] LordAro commented on pull request #8638: Feature: [GS] Allow non-question type windows to have no buttons https://git.io/JtMZC
08:36:57  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8638: Feature: [GS] Allow non-question type windows to have no buttons https://git.io/JtMZ4
08:37:10  <TrueBrain> if they complain that all windows should have an OK button, that it something I can get behind
08:37:23  <TrueBrain> but only complaining GSes can now create one without button .. that is just zooming in on a small detail in the UI :P
08:44:41  <TrueBrain> not enough diskspace to install XCode
08:44:41  <TrueBrain> lol
08:45:27  <TrueBrain> 25GB available
08:45:29  <TrueBrain> how is that not enough?
08:46:26  <TrueBrain> it needs ~70GB
08:46:27  <TrueBrain> lol
08:47:01  <Wolf01> About forests and snowline, it would be cool if on superflat the snowline is capped at max height-1 so you always get snow, ad if you don't want snow then play temperate :P
08:49:37  <andythenorth> funny idea :D
08:51:54  <andythenorth> those terrain names, wtf even are they?
08:52:00  <andythenorth> what is 'very flat' supposed to do?
08:52:40  <TrueBrain> OpenTTD has many windows where someone went overboard by presenting the user with all kinds of toggles ans switches
08:52:52  <TrueBrain> where it feels more like you get access to the developers settings
08:52:56  <TrueBrain> instead of players settings
08:53:02  <TrueBrain> Terrain Generation is one of them :)
08:53:38  <TrueBrain> I still want to try to make a "basic" version, with an "advanced" button, which allows you to configure stuff with more detail, but with less "labels"
08:53:48  <TrueBrain> so instead of "very flat", just say: height modifier 0.2
08:53:49  <TrueBrain> or what-ever
09:03:49  <Wolf01> Look at minecraft worldgen options :P
09:04:30  <TrueBrain> rather, look at Factorio
09:06:02  <andythenorth> look at Warcraft 1
09:06:23  <andythenorth> there aren't any worldgen, it just cycles about 16 maps, and each map has about 4 town locations
09:06:25  <TrueBrain> in my opinion, it is just a wrong idea to give all users so much detailed power over worldgen
09:06:36  <TrueBrain> we really know better how to make good playable maps for most people, especially new players
09:07:17  <andythenorth> there are many awesome heightmaps :P
09:07:23  <andythenorth> but who ever plays a heightmap?
09:07:33  <TrueBrain> more people than I expected, if I believe Discord
09:07:43  <andythenorth> yeah, and everybody has switched to JGRPP
09:07:59  <andythenorth> and is building detailed savegames using precision timetabling of real-world networks
09:08:01  <TrueBrain> there is a pretty active heightmap playerbase, as it turns out
09:08:03  <TrueBrain> which is nice to know
09:08:13  <andythenorth> yeah but how do I start a heightmap game?
09:08:15  <TrueBrain> someone even made a bunch of scripts to automate a lot
09:08:25  <TrueBrain> now you are just trying to be mean ;)
09:08:29  <andythenorth> 'New game' doesn't have any heightmap option.
09:08:30  <TrueBrain> finding a reason to bitch :)
09:08:39  <Wolf01> [10:06:23] <TrueBrain> in my opinion, it is just a wrong idea to give all users so much detailed power over worldgen <- that's why I said to look at minecraft one, based on which preset you select you can specify only some options which that preset let you define
09:08:57  <andythenorth> there are some awesome maps in the discord dev yes
09:08:58  <TrueBrain> Wolf01: and that is why I suggest Factorio, as it is a bit more in between :)
09:09:03  <andythenorth> heightmaps *
09:09:30  <andythenorth> but heightmaps are tacked onto title screen like an afterthought
09:10:08  <Wolf01> Yes, but F sets the values and you can still change everything, which is no different than what is already on OpenTTD
09:10:27  <TrueBrain> Wolf01: I do not agree; their presets are awesome, give you minor control
09:10:35  <TrueBrain> but you can also go all-out and fiddle with every detail if you so like
09:10:45  <TrueBrain> I think that works for OpenTTD, as the first works for 90% of the playerbase
09:10:49  <TrueBrain> and the second works for 10%
09:11:01  <TrueBrain> there will always be a group that wants to say exactly this and that .. and I am fine with that
09:11:10  <TrueBrain> just not by showing it to most of the people :)
09:11:10  <Wolf01> Maybe I should look better at their presets, the last time I generated a map on F it was on beta
09:11:22  <TrueBrain> in that case, yes, you should :)
09:12:54  <TrueBrain> security questions of which the answers cannot be more than 32 chars
09:12:56  <TrueBrain> wtf Apple, wtf
09:13:08  * andythenorth reading https://wiki.factorio.com/Map_generator
09:13:55  <TrueBrain> I would hide the "advanced" stuff by default btw
09:13:57  <TrueBrain> but that is just me
09:14:15  <andythenorth> we ever going to add more climates? (pls say no)
09:14:27  * andythenorth considers the settings being modal to the climate
09:14:49  <TrueBrain> wait, you need to fill in a CC these days for an Apple ID?
09:14:50  <TrueBrain> wtf?
09:14:58  <andythenorth> 'lol Apple'
09:15:11  <andythenorth> 'mistakes were made'
09:16:13  * andythenorth has taken a week off work, normal service will resume next week
09:16:30  <TrueBrain> so how to download xcode in that case .. ugh
09:18:59  <DorpsGek> [OpenTTD/OpenTTD] glx22 commented on pull request #8638: Feature: [GS] Allow non-question type windows to have no buttons https://git.io/JtMcE
09:20:14  <TrueBrain> w00p, turns out I already had an account made years ago .. and that doesn't require all the blabla :P
09:20:32  <TrueBrain> andythenorth: regarding more "climates", no, but more presets would be nice
09:20:40  <TrueBrain> that spices things up :)
09:20:55  <TrueBrain> "small island", "huge hills", etc
09:23:38  <andythenorth> we already do 'slightly modal' controls by turning snowline UI control off for other climates
09:24:44  <andythenorth> if each climate had specific controls, we could offer meaningful presets, and some specific  tuning options
09:25:07  <TrueBrain> I would even go so far that map size is part of the preset .. so you can have a "quick game" (256x256) preset, or a "play with your friends" (1024x1024) preset
09:25:23  <TrueBrain> and I just imagine it very simple: 4 buttons to select climate, single dropdown with presets
09:25:30  <TrueBrain> bottom row reads: "Customize"
09:25:41  <TrueBrain> which brings you to another UI where you can control every aspect :P
09:25:51  <andythenorth> 'control'
09:26:11  <TrueBrain> and presets are per climate, ofc
09:26:41  <andythenorth> 'control' https://www.synthesizers.com/learn/patchanywhere_500.jpg
09:27:02  <TrueBrain> as example, I can image in temperate, you can say: "near the beach" or "in the mountains"
09:27:08  <andythenorth> "The Freedom to Patch Anywhere
09:27:08  <andythenorth> massive patching In a modular synthesizer, you can patch anything to anywhere. Don't worry, nothing will break if you do it wrong. In fact, there is no wrong, but not everything you patch will be musically useful, that's up to you to decide."
09:27:09  <TrueBrain> which gives more water yes/no
09:27:44  <andythenorth> we could even...allow choosing heightmaps from new game window :P
09:27:46  <andythenorth> imagine
09:27:53  * andythenorth crazy talk
09:28:10  <TrueBrain> not constructive way of saying that ;) But yes, with a preview :)
09:28:26  <TrueBrain> anyway, all that is not going to happen for 1.11 :)
09:28:40  <TrueBrain> I do want to experiment with it a bit for 1.12, if I have the time
09:28:58  <TrueBrain> I have more ideas for some UIs, to make them a bit more new-player-friendly
09:29:09  <TrueBrain> but time is a valuable resource
09:29:48  <TrueBrain> and people with pitchforks in their hand, instead of being constructive (not you andythenorth , referring to other people :P), are not contributing to how you divide those resources :D
09:30:58  <andythenorth> time!
09:31:03  <andythenorth> you can't buy more of it
09:32:42  <andythenorth> but you can expend it playing Warcraft 1 from GOG
09:32:53  * andythenorth has been busy click on orcs
09:33:00  <TrueBrain> talk dirty to me
09:33:08  <TrueBrain> I am install XCode
09:33:12  <TrueBrain> not sure what I have to think about that
09:35:15  <andythenorth> you have to think "this will ask for updates often, and (other peoples') security policy says they have to be installed immediately, but they usually break the build, surely that's a problem for all 3 points of the infosec triangle?"
09:35:35  <andythenorth> like...how we will patch emerging CVEs in prod if all dev machines can't build?
09:36:00  <andythenorth> oh wait, your world is not my world :P
09:36:43  <TrueBrain> there are so many things I would like to say about what you just said; instead, I am just going to watch XCode install :)
09:36:54  <andythenorth> better for all of us
09:37:25  <andythenorth> annoyingly GOG doesn't have a mac edition of Warcraft 2
09:37:52  <TrueBrain> I like how my servers-listing "change" seems to be effective .. I should PR this soon, instead of illegally pushing my branch to production ..
09:38:06  <TrueBrain> and yes, I pushed a dev-image to production
09:38:08  <TrueBrain> I DID THAT
09:38:16  <andythenorth> don't tell the sysadmin
09:38:47  <TrueBrain> he was hella pissed, I tell you
09:38:52  <TrueBrain> the CISO also had a word with me
09:39:20  <andythenorth> funny civilisation we've built eh
09:39:28  * andythenorth wonders sometimes if it could all just be better
09:39:31  <andythenorth> but then...people
09:40:00  <TrueBrain> corporate greed ... what can you do!
09:45:17  <TrueBrain> wow, running macOS in a VM drags my host system down too
09:45:19  <TrueBrain> that is annoying
09:50:43  <_dp_> I kinda feel that whole errors/news systems needs an overhaul
09:50:56  <_dp_> not sure how exactly though, still looking for ideas
09:51:02  <_dp_> but current one is pretty much useless
09:51:55  <andythenorth> red dot notifications!
09:51:56  <_dp_> like, for example, wsywg building can get rid of most building errors: https://i.imgur.com/SqNHvQ4.png
09:53:44  <TrueBrain> the current system is very 1994 :)
09:54:34  <andythenorth> it was revolutionary then
09:54:44  <andythenorth> the TTD UI was better than many desktop OSes
09:54:51  <andythenorth> time has not been kind
09:55:43  <TrueBrain> I think it is awesome for what it did
09:55:49  <TrueBrain> but indeed, we nowedays are used to other interactions
09:56:03  <TrueBrain> which makes the question .. do you adjust, or do you "keep the old feeling"
09:59:05  <_dp_> there are better ways to express "feeling" than bad ui :p
10:00:04  <TrueBrain> hmm .. how do I build OpenTTD on macOS
10:00:07  <TrueBrain> it cloned it for me
10:00:12  <TrueBrain> but .. it refuses to open the folder now :D
10:00:32  <LordAro> _dp_: very neat
10:03:07  <andythenorth> refuses? :P
10:03:31  <TrueBrain> yup .. get a nice error if I try to
10:03:56  <TrueBrain> guess XCode doesn't know CMake ..
10:03:58  <TrueBrain> stupid
10:10:48  <TrueBrain> C compiler identification is unknown
10:10:51  <TrueBrain> well, that is not helping
10:11:26  <TrueBrain> andythenorth: how do you do this?
10:12:26  <LordAro> keyboard flailing
10:12:55  <andythenorth> I have a cmake in /usr/local/bin/cmake
10:13:00  <andythenorth> not sure how it got there
10:13:17  <TrueBrain> owh, it was xcode that was being weird
10:13:21  <TrueBrain> brew broke it
10:13:31  <TrueBrain> I unbroke it
10:13:45  <andythenorth> I try to circumvent actual XCode
10:13:55  <andythenorth> but it's the easiest way to get clang etc
10:14:18  <TrueBrain> boy, this is not going to be a quick compile I tjink ...
10:14:20  <TrueBrain> :D
10:14:49  <TrueBrain> I like "cmake --build ." command
10:14:52  <TrueBrain> works on all OSes :)
10:14:59  <TrueBrain> also avoids me understanding MSVC, for example :)
10:23:27  <TrueBrain> okay, my patch seems to work fine on macOS
10:23:42  <TrueBrain> wow, FF is quick ... despite being in a VM
10:23:59  <TrueBrain> damn, the macOS video driver is fast
10:24:26  <TrueBrain> even zoomed-out I hit 60fps just fine (and a simulation rate on FF of 400+)
10:24:29  <TrueBrain> empty map, to be clear
10:25:35  <TrueBrain> right, let's see what michi_cc found as issue with std::chrono .. as I am hitting all the deadlines fine, so I wonder how master looks/feels :D
10:27:16  <TrueBrain> macOS in my VM is quicker in rendering than SDL via WSL ..... lol?
10:28:21  <TrueBrain> its even faster than Win32 video driver
10:29:18  <TrueBrain> graphics rendering takes so much less time
10:29:19  <TrueBrain> interesting
10:30:06  <LordAro> i love how it turns out opengl is entirely unnecessary
10:30:13  <TrueBrain> for macOS, it is
10:30:18  <TrueBrain> for the other two, it is very much needed
10:30:31  <TrueBrain> but I think after that, all 3 OSes are on-par with each other
10:32:30  <TrueBrain> okay, my PR on macOS makes the mouse tail its own tail
10:32:44  <TrueBrain> which is hilarious to see :D
10:33:49  <TrueBrain> ah, no, that is VMWare
10:34:10  *** Flygon has quit IRC
10:34:22  <TrueBrain> macOS already uses OpenGL in the backend
10:34:29  <TrueBrain> at least, I am getting OpenGL errors :P
10:37:01  <TrueBrain> andythenorth: how do you feel about testing some native macOS for me today? :P
10:37:10  <andythenorth> yup
10:37:12  <andythenorth> I feel
10:37:30  <TrueBrain> https://github.com/TrueBrain/OpenTTD/tree/video-performance <- I could really use an opinion if that makes it better or worse on macOS
10:37:46  <TrueBrain> both normal game, how the mouse reacts, but also FF speeds
10:38:05  <TrueBrain> it runs the GUI on a fixed 60fps
10:39:20  <TrueBrain> git fetch https://github.com/TrueBrain/OpenTTD video-performance && git checkout FETCH_HEAD
10:39:25  <TrueBrain> in case you don't know that :)
10:42:39  <andythenorth> saves me knowing that thanks :)
10:42:44  <andythenorth> outsourcing knowing things ftw
10:43:47  <andythenorth> soon it will be time to say 'slow linking is slow' :)
10:44:04  <andythenorth> I think it's just because it doesn't report progress, it _looks_ slow :P
10:44:12  <TrueBrain> if this works, I can abstract the mainloop of all video drivers .. which will be interesting :P
10:45:27  <TrueBrain> allegro driver is faster than SDL
10:45:33  <TrueBrain> some things are just funny :)
10:46:26  <TrueBrain> Allegro + FF == 1 second -> 2 game months
10:46:30  <TrueBrain> (empty map)
10:47:36  <andythenorth> so far so good
10:47:47  <andythenorth> it's hard to judge, but the cursor never seems to lag now
10:48:04  <TrueBrain> good; I was hoping that would fix that ;)
10:48:15  <TrueBrain> how does the framerate window look?
10:48:25  <TrueBrain> are you hitting 33.33 simulation rate, and 60.00 graphics?
10:49:26  <andythenorth> there's a bit of jitter so it's bouncing 1 or 2 decimal places either side of those numbers
10:49:49  <TrueBrain> yeah, like 33.34 etc
10:49:51  <andythenorth> but otherwise yes, and stable
10:49:55  <TrueBrain> good; how is FF?
10:51:04  <andythenorth> very dependent on what's in the viewport
10:51:11  <TrueBrain> ofc, that is normal
10:51:23  <andythenorth> if I go to map corner, full animation off, it's 28-32x
10:51:24  <TrueBrain> but while FF is on, the graphics stays at 60.0-ish? And mouse etc reacts normal?
10:51:47  <TrueBrain> only 30x? Damn, I get more in my VM :P
10:51:57  <andythenorth> yes graphics is 60-ish
10:52:08  <andythenorth> is your VM driving a hi-dpi screen?
10:52:15  <TrueBrain> no :D
10:52:29  <TrueBrain> but drawing should have far less influence on the FF now
10:52:45  <TrueBrain> can you edit your openttd.cfg to read: [gui] <newline> refresh_rate = 30 ?
10:52:50  <andythenorth> ok
10:52:55  <TrueBrain> (it should already have a refresh_rate entry, so search for that)
10:53:29  <andythenorth> ugh this is horrible :)
10:53:41  <TrueBrain> much more laggy, your mouse, I assume?
10:53:56  <andythenorth> yup
10:54:08  <TrueBrain> yeah, so you are really used to 60Hz :D
10:54:13  <andythenorth> it's very hard to define, but the difference is qualitative
10:54:15  <TrueBrain> which, ones you are used to, is hard to get unused to :P
10:54:24  <andythenorth> the 30hz is much more like playing recent vanilla master
10:54:31  <TrueBrain> my screens run on 144Hz .. when you start with that, you don't notice it really
10:54:35  <andythenorth> the mouse runs subtly behind
10:54:38  <TrueBrain> but after a year, doing 60 or 30Hz is ... ugh
10:54:41  <TrueBrain> so I get what you mean :)
10:54:43  <andythenorth> official binaries run more like the 60hz
10:54:54  <TrueBrain> yeah, there is some mouse cheating going on
10:54:59  <TrueBrain> so release builds etc run smoother
10:55:03  <TrueBrain> just completely unpredictable
10:55:11  <TrueBrain> (they do not claim the time, but if it is there, it is given)
10:55:14  <TrueBrain> now it claims the time :)
10:55:22  <andythenorth> I build release, but the recent stack of changes (the one's that improved mac FPS by factor 10) made mouse worse
10:55:33  <TrueBrain> ah, yeah, okay
10:55:37  <TrueBrain> so my PR corrects that again?
10:55:39  <andythenorth> yes
10:55:42  <TrueBrain> good
10:55:43  <andythenorth> whilst retaining factor 10
10:55:53  <andythenorth> possibly slightly improving it even
10:56:06  <andythenorth> might be imagining that, there are no real benchmarks
10:56:13  <TrueBrain> orudge: if you would have some time, I would also love to know if https://github.com/TrueBrain/OpenTTD/tree/video-performance improves things for you :) (just to be sure I don't screw up M1 etc :P)
10:56:23  <TrueBrain> if you want improvements, change refresh_rate to 144 :P
10:56:29  <TrueBrain> or 74
10:56:34  <TrueBrain> depends in your screen honestly
10:57:32  <andythenorth> my mac can't quite hold 144
10:57:42  <andythenorth> it's dropping to 134 or so quite often
10:57:44  <andythenorth> but eh
10:57:51  <TrueBrain> but your mouse should be even smoother :P
10:58:19  <andythenorth> 'smooth'
10:58:28  * andythenorth wonders what the display runs at
10:59:13  <TrueBrain> in Display of configuration you can see this
10:59:35  <TrueBrain> hmm, no, not anymore
10:59:38  <TrueBrain> you used to see this there :P
10:59:41  <andythenorth> probably 60hz
11:00:08  <TrueBrain> I believe most macs are 60Hz indeed
11:00:18  <TrueBrain> I was thinking they already were using 72 or 74Hz
11:00:20  <TrueBrain> but I seem to be wrong
11:00:44  <andythenorth> specs say 60
11:01:04  <andythenorth> let's see what hidpi rendering does with this
11:02:22  <andythenorth> yeah the game can't hold 60fps at hidpi
11:02:25  <andythenorth> nowhere near
11:02:33  <TrueBrain> that is sad
11:03:01  <TrueBrain> I have yet to write code that ensures simulation speed wins from fps, but that means with high dpi that would always kick in :P
11:03:35  <andythenorth> trying to run hidpi on macs would be a bold move
11:03:54  <andythenorth> it drops from 60fps to 30fps during any map movement
11:04:10  <andythenorth> and if full animation is on, with any area of sea in viewport, fps drops to 25
11:04:14  <TrueBrain> loading in sprites, I think
11:04:19  <TrueBrain> oef
11:05:30  <andythenorth> same area of sea, same game, with allow_hidpi = false, 60fps
11:05:44  <TrueBrain> well, it draws 4 times as much
11:05:45  <andythenorth> macs don't run games in hidpi
11:05:49  <TrueBrain> so I can imagine that impacting performance :)
11:06:06  <andythenorth> I am quite -1 to shipping hidpi=true as default for 1.11
11:06:14  <andythenorth> but I won't start a forum poll :)
11:06:19  <TrueBrain> talk it over with orudge and michi_cc I would say :)
11:06:28  <TrueBrain> you three are our OSX experts :D
11:06:35  <andythenorth> "m1, probably fine"
11:06:45  <andythenorth> I still haven't put the build tools on the m1
11:06:53  <andythenorth> not my idea of fun :)
11:07:03  <TrueBrain> so don't :)
11:07:19  <TrueBrain> anyway, tnx for testing andythenorth , good to know this has a similar effect on macOS as the other targets
11:07:24  <TrueBrain> means I can unify the video-driver a tiny bit :D
11:07:33  <andythenorth> so glad to see these improvements :)
11:08:00  <andythenorth> now if we could just find a way to do palette animation, since GPUs dropped it...
11:08:08  <andythenorth> or so I was told about 7 years ago
11:08:31  <milek7> opengl?
11:08:39  <_dp_> lol, tree growth is silly on that uber-ff
11:08:53  <milek7> I think opengl does palette animation in shader?
11:10:10  * andythenorth learning about fragment shaders and other opengl words
11:10:19  <_dp_> TrueBrain, tried your branch as well, is it a known thing that fps seems to cap simulation rate?
11:10:19  <andythenorth> _dp_ see parent topic: trees are silly
11:10:26  <_dp_> like, sim rate changes when I zoom out
11:11:05  <_dp_> on ff ofc
11:13:08  <andythenorth> _dp_ that's not new? :)
11:13:17  <andythenorth> I get same result, but always this way no?
11:13:53  <TrueBrain> _dp_: depends on what you mean with "cap"
11:14:06  <TrueBrain> but yes, if there is a lot to draw, drawing claims time, which means simulation rate goes down
11:14:17  <TrueBrain> this is already the case in master, but now you can see this happening :)
11:14:32  <TrueBrain> and given my branch draws twice as often
11:14:41  <TrueBrain> impacts that sooner
11:14:45  <_dp_> both go from 60 zoomed in to 40
11:14:49  <_dp_> though yeah, makes sense
11:14:56  <_dp_> also it crashed...
11:15:00  <TrueBrain> but, there are two things here: OpenGL will make drawing faster, so that is going to help
11:15:09  <TrueBrain> but also, I want to make code that give priority to simulation
11:15:14  <TrueBrain> so it simply stops drawing as often
11:15:18  <TrueBrain> but that is TODO :)
11:15:57  <TrueBrain> although on FF I am not sure what is best .. possibly on FF the refresh-rate should go down to 30 in that case
11:16:04  <TrueBrain> where in non-FF it should go down to like 10 or so
11:16:05  <TrueBrain> dunno yet
11:16:17  <_dp_> this happended when I tried to unmaximize it: https://imgur.com/a/hYyQpT4
11:16:23  <_dp_> can't replicate though
11:16:37  <_dp_> and no other info, it just closed
11:16:38  <_dp_> weird
11:16:40  <TrueBrain> I haven't seen those errors in a long long long long time :)
11:17:25  <TrueBrain> if you can replicate it, let me know; but possibly this was just X11 barking :)
11:17:33  <andythenorth> this is more fun that writing nml docs eh :)
11:18:01  <_dp_> TrueBrain, yeah, I can, it happens when I unmaximize window on fast forward
11:18:15  <TrueBrain> mind testing with master too?
11:19:06  <milek7> happens on master too
11:19:30  <TrueBrain> as I can maximize and unmaximize all I want, it just keeps on trucking :D
11:19:55  <TrueBrain> just often fails to get in the right resolution
11:19:56  <TrueBrain> that is funny
11:20:10  <TrueBrain> pretty sure I have an idea what is going wrong there
11:20:22  <TrueBrain> well, I am sure I have an idea
11:20:28  <TrueBrain> so that is a really silly sentence :D
11:20:38  * andythenorth tests that on mac for completeness, but eh, completely different subsystem?
11:22:05  <TrueBrain> let me randomly disable a line of code to see if that improves the situation for me ...
11:23:07  <TrueBrain> _dp_: video/sdl2_v.cpp, around line 324, can you disable GetAailableVideoMode line, and try again?
11:23:13  <TrueBrain> just comment out that line
11:24:19  <_dp_> ok, happens on master but not on beta
11:24:24  <_dp_> I'll check that line now
11:27:30  <TrueBrain> hmm, the problem I was having disappeared now anyway :P
11:27:48  <TrueBrain> owh, wait, I was testing SDL1, oops
11:27:53  <TrueBrain> found a bug in SDL1 :P
11:28:17  <_dp_> nope, still happens with that line commented
11:28:25  <TrueBrain> and if you comment out line 328, the SetWindowSize?
11:30:24  <LordAro> _dp_: i've seen error messages like that if i've updated the graphics driver but not rebooted
11:30:49  <LordAro> though if it's actually initialised itself, probably not the cause here
11:30:57  <_dp_> I just rebooted today so shoudn't be an issue :p
11:31:46  <_dp_> still crashes withouot SetWindowSize
11:32:00  <TrueBrain> too bad; those lines of code in the SDL driver are a bit weird :P
11:32:12  <TrueBrain> will see if I can reproduce it in my Linux VM
11:33:59  <TrueBrain> yippie, I can
11:34:12  <TrueBrain> funny, means that the Windows X11 server allows what-ever is causing this just fine :D
11:34:34  <milek7> it might just not use XShm
11:35:37  <TrueBrain> blegh, I hate it when small changes cause a full rebuild of OpenTTD
11:35:40  <TrueBrain> takes for-ever
11:35:46  <_dp_> ok, it happens on normal speed as well, but not if game is paused
11:36:35  <_dp_> doesn't seem to happen on title screen though %)
11:37:10  <milek7> given it only happens on ffwd, it maybe tries to submit too large image after window is resized but before resize event is processed?
11:38:03  <_dp_> not only ffwd, it can happen on normal speed as well, just mb less likely
11:38:38  <TrueBrain> I will bisect and find the cause, no worries :)
11:39:11  <_dp_> though for small games only on ffwd
11:40:28  <TrueBrain> possibly I introduced this with the SDL cleanup, and some bit is now falling over differently .. I can reproduce it, so it will be fine :)
11:40:45  <TrueBrain> just bisects takes for ever it is insists on recompiling every time :P
11:48:33  <TrueBrain> not really sure what triggers a full rebuild ..
11:48:53  <TrueBrain> feels like we have some dependency a bit wrong in CMake
11:50:00  <_dp_> just lang changes are probably enough to force full rebuild
11:50:36  <TrueBrain> hmm, I guess that is the cause I am experiencing, yes
11:50:39  <TrueBrain> annoying, yes
11:50:40  <TrueBrain> :D
11:52:19  <TrueBrain> bisecting blames an OSX patch :P
11:52:58  <_dp_> haha, everything is Apple's fault :p
11:54:32  <TrueBrain> and of course I first need to recompile everything
11:54:52  <TrueBrain> but yeah, a change that was hidden in an OSX-only PR
11:54:59  <TrueBrain> makes changes for all drivers :D
11:55:09  <TrueBrain> who approved that PR anyway?!
11:55:20  <TrueBrain> what a shitty developer is that, just approving that without testing it on other drivers ..
11:56:56  <TrueBrain> https://github.com/OpenTTD/OpenTTD/commit/8906e9e0fd6dc7bd64f9ad2c633da6f110dbf921 seems to be the cause
11:57:08  <TrueBrain> (and I reviewed it, to be clear, that is why I can call myself shit :P)
11:57:17  <TrueBrain> well, lunch first, looking at why this breaks it after :P
11:57:42  <TrueBrain> owh, no, it crashes before that too
11:57:52  <TrueBrain> lol .. back to bisecting ...
12:12:32  <orudge> TrueBrain: tested it on Intel MBP, seems fine in that I am able to get 60fps most of the time - if zoomed out fully, then it does drop (HiDPI here)
12:12:45  <orudge> Will test on M1 (which is not HiDPI, as I only have a Mac Mini plugged into a 1080p screen)
12:12:58  <TrueBrain> does it feel better or worse?
12:13:04  <TrueBrain> (lovely non-factual question :D)
12:16:06  <orudge> I'll be honest, I can't really answer that one way or the other. It seems much the same either way. Obviously the sim rate is considerably faster in ffwd compared to the previous version.
12:17:56  <orudge> In an empty game, I'm hitting 9999fps sim rate using the Intel build on M1 (because I haven't built it on my M1 natively), with 60fps graphics rate - no hidpi here
12:19:16  <orudge> Sorry, possibly not the most helpful :) But it doesn't seem to make things worse for me
12:19:27  <TrueBrain> that is very helpful :)
12:19:38  <TrueBrain> not a regression, check
12:19:41  <TrueBrain> improves FF speed, check
12:19:47  <TrueBrain> no mention of weird mouse behaviour, check
12:20:05  <TrueBrain> tnx a lot orudge :)
12:20:51  <TrueBrain> it was reported std::chrono made things worse for macOS, but both our users don't notice it, so I guess we are good :)
12:35:12  <TrueBrain> _dp_: found it! https://github.com/OpenTTD/OpenTTD/commit/52317bb7df6d153ec332e5403ef75cb7bc86b8f6 :)
12:35:17  <TrueBrain> nobody understood what it was doing
12:35:22  <TrueBrain> but it prevents a crash with X11 :P
12:39:08  <LordAro> lol
12:41:18  <TrueBrain> I wonder what happens here exactly .. as that line still feels like a hack
12:42:09  <supermop_Home_> yo
12:42:22  <TrueBrain> ah, random forum threads might give a clue ... X11 doesn't like it if you update rects outside the window
12:43:07  <TrueBrain> so that is why that line prevents the problem
12:43:17  <TrueBrain> explains why unmaximize is the issue
12:43:21  <TrueBrain> and also why not always
12:47:10  *** roadt_ has joined #openttd
12:50:28  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #8684: Fix 52317bb7: [SDL2] ensure we don't try to blit out of bounds https://git.io/JtMzR
12:50:42  <TrueBrain> _dp_: happy you found it before 1.11 :D
12:51:07  <TrueBrain> and now it has a proper comment explaining why it is there :) Totally makes sense ... could not come up with that based on the old comment :P
12:52:39  *** WormnestAndroid has quit IRC
12:52:52  *** WormnestAndroid has joined #openttd
12:54:05  *** roadt__ has quit IRC
12:59:27  *** snail_UES_ has joined #openttd
13:00:00  <_dp_> lol
13:03:06  <DorpsGek> [OpenTTD/OpenTTD] orudge approved pull request #8684: Fix 52317bb7: [SDL2] ensure we don't try to blit out of bounds https://git.io/JtMgk
13:05:36  *** WormnestAndroid has quit IRC
13:05:49  *** WormnestAndroid has joined #openttd
13:06:15  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8684: Fix 52317bb7: [SDL2] ensure we don't try to blit out of bounds https://git.io/JtMzR
13:14:05  *** Samu has joined #openttd
13:26:52  <TrueBrain> if you exit the game during landscape generation, the generation is not aborted :D
13:26:56  <TrueBrain> but screen drawing stops
13:27:00  <TrueBrain> so you have to wait till it finishes :P
14:05:54  <andythenorth> I refuse!
14:05:58  <andythenorth> I want to draw soone!
14:06:05  <andythenorth> sooner*
14:22:11  *** con_theGranny has quit IRC
14:35:16  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8680: Several improvements to SDL2 video driver https://git.io/JtMrm
14:40:28  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8680: Improvements for all video drivers https://git.io/Jt1ur
14:44:52  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8680: Feature: configurable refresh-rate and default to 60fps https://git.io/Jt1ur
14:44:52  <TrueBrain> found a better name for the PR :)
14:45:25  <TrueBrain> oops, a word fell off :P
14:45:40  <TrueBrain> okay, the result of this PR is much better than I could have hoped for
14:46:04  <TrueBrain> did not expect such improvements were possible on a 17 year old game with such little code changes :D
14:54:45  <milek7> > Mostly, on SDL it takes 5ms to draw a frame.
14:54:48  <milek7> wait, where?
14:54:50  <milek7> on win32?
14:55:43  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8680: Feature: configurable refresh-rate and change default to 60fps https://git.io/Jt1ur
15:05:02  <TrueBrain> okay, tested SDL2, SDL1, Allegro, Cocoa, Win32, dedicated server and emscripten; they all work as I would expect \o/ :D
15:05:08  <LordAro> \o/
15:05:13  <TrueBrain> the only thing I notice more, but that is not a regression of my PR: there is visual tearing
15:05:22  <TrueBrain> most video drivers are not vsync'd it seems
15:05:29  <LordAro> /o\
15:05:32  <TrueBrain> I already knew that, as the trailer also has this
15:05:51  <nielsm> at least in olden times, vsync was not possible in windowed applications on windows
15:05:53  <TrueBrain> but that is something for another PR :)
15:05:57  <nielsm> only in exclusive fullscreen modes
15:06:07  <TrueBrain> I know with OpenGL we are vsync'd
15:07:08  <TrueBrain> so it appears that works for windows too
15:07:15  <TrueBrain> but I really do not know more about that stuff :D
15:07:25  <TrueBrain> all I know is that I can see that OpenTTD currently is not :P
15:07:34  <TrueBrain> and if you watch the trailer closely, you see this happening a lot too :)
15:07:48  <TrueBrain> but okay, nobody reported it .. ever?
15:07:50  <TrueBrain> so I guess we are good :P
15:08:10  <TrueBrain> but seriously, 60fps ... map scrolling is so much nicer to the eye :D
15:08:15  <TrueBrain> it feels ... like silk
15:09:25  <andythenorth> \o/
15:09:42  <TrueBrain> let me create a preview, as even in the browser that is noticeable (to me)
15:10:46  <TrueBrain> nielsm: as most games run in borderless windowed mode these days, I guess that is something they solved by now? I dunno :)
15:11:13  <TrueBrain> might really be OpenGL / Direct3D only stuff :D
15:11:40  <nielsm> yeah it's a good while since they fixed so you can vsync in windowed apps under DWM
15:12:04  <nielsm> and yes I think you need to use an ogl or d3d surface to do it
15:12:18  <TrueBrain> especially on SDL, which is the slowest video driver for me, tearing is noticeable :)
15:12:37  <TrueBrain> but if OpenGL lands, that is all a problem of the past, so I am not that worried honestly
15:13:38  <TrueBrain> this is the only PR in weeks I tested on so many different targets :P It is nice to see that is not really difficult in 2021
15:13:52  <TrueBrain> not fun to do, but also not difficult
15:14:05  <TrueBrain> right, I guess I should add a FF window to set game speeds or something ...
15:14:14  <TrueBrain> as I am sure LordAro will complain about FF now :P
15:14:41  *** JGR has joined #openttd
15:15:14  <LordAro> a window? i was imagining a dropdown
15:16:02  <TrueBrain> something
15:16:23  <TrueBrain> nielsm: btw, a layered window system solves another issue: the responsiveness during map generation
15:16:46  <TrueBrain> I was looking at that earlier today, as it still annoys the fuck out of me .. but in the current setup that isn't really possible to make more responsive for most parts of the generation
15:16:53  <JGR> On screen tearing, I had quite bad issues with this on Linux, these changes to SDL seemed to resolve the issue: https://github.com/JGRennison/OpenTTD-patches/commit/51477334be06e2a91541f397b7cbc227ab392bfa https://github.com/JGRennison/OpenTTD-patches/commit/1298f49be69b9cf0ba0936b4a353258ae7875ac0
15:16:55  <TrueBrain> a layered window system like you explained would
15:17:02  <JGR> They may no longer be relevant with the recent changes though, I've not caught up on them all yet.
15:17:34  <TrueBrain> the first for sure is interesting
15:17:55  <TrueBrain> both are
15:18:02  <TrueBrain> none of the recent changes do anything with that
15:18:17  <TrueBrain> so yeah, I can imagine those resolve the issue honestly :)
15:18:56  <TrueBrain> I guess those loops should be merged :P
15:19:03  <TrueBrain> but in general, I wonder if we shouldn't just track "biggest rect"
15:19:03  <nielsm> when you're really cpu or memory constrained, a clip region based window system makes sense, but when you get spare cpu and memory a layered one is much more flexible, and easier to work with overall
15:19:07  <TrueBrain> I believe OpenGL will do that too
15:19:28  <TrueBrain> tnx JGR , I will look at that :D
15:19:33  <JGR> No problem
15:19:36  <TrueBrain> wasn't planning too, but if a solution is presented ...
15:19:39  <TrueBrain> might as well :P
15:22:27  <TrueBrain> okay, for shit and giggles, open (I suggest Chrome) https://preview.openttd.org/pr8386/ and https://preview.openttd.org/pr8680/
15:22:30  <TrueBrain> create a new game in both
15:22:37  <TrueBrain> right click your mouse, and move around
15:22:39  <TrueBrain> slowly
15:22:39  <TrueBrain> fast
15:22:41  <TrueBrain> pick one
15:22:45  <TrueBrain> can you spot the difference? :D
15:27:49  <Timberwolf> 8680 is nice (for bonus points, I'm trying it on a baby MBP touchbar)
15:30:38  <andythenorth> called it, one is 60
15:30:51  * andythenorth didn't check the fps until after testing, nor read the PR
15:31:01  <TrueBrain> :D
15:31:05  <andythenorth> it's quite obvious
15:31:09  <andythenorth> can't explain why
15:31:26  <TrueBrain> funny aint it? :D
15:31:50  <andythenorth> I think it's the map scroll that does it
15:36:00  <Timberwolf> Is there anything else in it? 8386 this Mac is struggling to keep the simulation rate above 30fps, whereas 8680 it's comfortably sat at 1.00x
15:36:13  <TrueBrain> JGR: with your patches applied, the tearing changes for me, but for sure it has a positive effect here too :)
15:36:43  <TrueBrain> Timberwolf: yeah, 8680 does more things to allow everything to hit deadlnes a bit better
15:37:36  <TrueBrain> for emscripten we depend a bit on the refresh rate of your screen, as browsers calls us on that frequency
15:37:43  <TrueBrain> the new code should align a bit better there
15:38:44  <LordAro> TrueBrain: oh wow
15:38:50  <LordAro> that is so very silky
15:39:14  <TrueBrain> cool aint it? And that for a game from 1994/1995 :P
15:39:51  <Timberwolf> I have to admit, the first thing I did was install Timberwolf's Trains and go through a subset of my "things which break dirty rectangling" list.
15:39:56  <Timberwolf> No problems noted :)
15:40:02  <TrueBrain> \o/
15:40:07  <TrueBrain> that is honestly good to hear
15:40:13  <TrueBrain> as ... well .. I am not 100% sure it is safe
15:40:14  <TrueBrain> :P
15:40:46  <LordAro> TrueBrain: why is "Video output" always coloured yellow?
15:40:50  <LordAro> (emscripten)
15:41:24  <nielsm> when I try the emscripten previews in firefox my mouse cursor is constantly drifting downwards
15:41:52  <TrueBrain> LordAro: because of how emscripten works, basically .. it acts like it takes for-ever to draw
15:41:54  <TrueBrain> which is a lie
15:41:59  <TrueBrain> but it is when the browser takes back control
15:42:03  <TrueBrain> I might be able to fake that
15:42:29  <LordAro> nielsm: no such issues here
15:42:32  <TrueBrain> nielsm: huh? That is odd :D But in general, FireFox isn't the best for these kind of things
15:42:40  <TrueBrain> I have like 50% of the frames I have with Chrome
15:42:51  <nielsm> might be related to me running 150% dpi scaling on that display
15:48:08  <orudge> [15:13:39] <TrueBrain> this is the only PR in weeks I tested on so many different targets :P It is nice to see that is not really difficult in 2021 <-- I still need to finish the OS/2 update... :D
15:48:37  <TrueBrain> you .. silly ... mister
15:48:38  <TrueBrain> :D
15:50:10  <orudge> Could potentially even build for OS/2 via GitHub Actions via an SSH relay to a VM... :P Reliability could be an issue there though
15:50:42  <TrueBrain> yeah .... no :P
15:50:58  <TrueBrain> isn't there a Docker to cross-compile? :)
15:51:47  <orudge> Not last time I checked, but you never know...
16:02:34  <FLHerne> Haiku?
16:05:20  <TrueBrain> JGR: I guess doing the 80% is no longer needed with the other change you made? At least, checking the SDL code, it should behave nearly identical :)
16:21:01  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain opened pull request #8685: Fix: [SDL2] simplify what to redraw to prevent tearing https://git.io/JtM1T
16:21:26  <TrueBrain> similar solution than JGR showed, but with the eye on OpenGL collapsed the rects to a single one in a different place :)
16:21:44  <JGR> Quite probably yes. I made the 80% change first and the main motivation was performance. I think that it would be reasonable to prefer the the single rectange change.
16:21:45  <TrueBrain> it could some more Linux people testing it
16:22:33  <TrueBrain> I stole code from michi_cc 's branch :D
16:23:42  *** Progman has joined #openttd
16:30:15  *** Wormnest has joined #openttd
17:03:50  *** JGR has quit IRC
17:03:51  *** Tulitoma1tti has joined #openttd
17:03:59  *** mirrorbird has joined #openttd
17:05:14  <DorpsGek> [OpenTTD/OpenTTD] perezdidac updated pull request #8683: Feature: Town name generator for Spanish (Spain) https://git.io/JtMtf
17:05:20  *** APTX has quit IRC
17:05:38  *** Tulitomaatti has quit IRC
17:14:07  <DorpsGek> [OpenTTD/OpenTTD] perezdidac closed pull request #8552: Feature: Prevent vehicles to unload cargo in implicit stops in cargodist mode if the cargo is not accepted in such station https://git.io/JLhF7
17:18:00  <DorpsGek> [OpenTTD/OpenTTD] EratoNysiad opened issue #8686: Unable to change townnames in Scenario editor after pr8566 https://git.io/JtMyZ
17:29:41  *** Progman has quit IRC
17:32:18  *** Wuzzy has joined #openttd
18:20:05  *** frosch123 has joined #openttd
18:21:21  <andythenorth> quak
18:24:37  *** gelignite has joined #openttd
18:26:28  <Xaroth> if it talks like a duck...
18:28:50  <frosch123> moo
18:29:02  <frosch123> no ducks allowed
18:29:28  <TrueBrain> frosch123: would you be so kind and try out https://github.com/OpenTTD/OpenTTD/pull/8685 for me? See if it changes anything for you?
18:29:35  <TrueBrain> I think it changes absolutely nothing for you :)
18:29:49  <TrueBrain> (which is a good thing)
18:54:16  <DorpsGek> [OpenTTD/OpenTTD] DorpsGek pushed 1 commits to master https://git.io/JtM7G
18:54:17  <DorpsGek>   - Update: Translations from eints (by translators)
18:58:15  <frosch123> TrueBrain: i don't see a difference. i also see no issue with master
18:58:26  <TrueBrain> frosch123: that is a positive thing :)
18:58:40  <TrueBrain> and yeah, I couldn't reproduce it on all machines
18:58:43  <TrueBrain> just a few that have this issue
18:58:46  <TrueBrain> and when they have it, it is terrible
18:59:00  <TrueBrain> tnx for testing!
18:59:04  <DorpsGek> [OpenTTD/OpenTTD] spnda opened issue #8687: CMake include error when building new project. https://git.io/JtM7E
19:00:34  <LordAro> do we support building in paths containing spaces?
19:00:37  <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #8685: Fix: [SDL2] simplify what to redraw to prevent tearing https://git.io/JtM7o
19:00:55  <TrueBrain> LordAro: I dont see a reason why not
19:01:02  <TrueBrain> as if we don't, it means we are not escaping something
19:01:09  <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #8685: Fix: [SDL2] simplify what to redraw to prevent tearing https://git.io/JtM76
19:01:24  <LordAro> potentially
19:01:36  <LordAro> makefiles are pretty terrible at dealing with paths with spaces
19:01:59  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8685: Fix: [SDL2] simplify what to redraw to prevent tearing https://git.io/JtM7X
19:02:14  <TrueBrain> LordAro: the initial CMake I created, I created with a space in my path
19:02:17  <TrueBrain> to avoid this mistake
19:02:24  <TrueBrain> it is not difficult, just easy to mess up :D
19:02:46  <LordAro> :)
19:03:56  <TrueBrain> really cannot believe CMake doesn't protect that better honestly :)
19:04:18  <LordAro> add a space into the CI build path :p
19:04:37  <TrueBrain> well, that is a good way, honestly
19:04:42  <TrueBrain> write that in the issue plz :D
19:06:23  *** _2TallTyler has joined #openttd
19:14:22  <DorpsGek> [OpenTTD/OpenSFX] Wuzzy2 updated pull request #14: Change readme, license, descriptions  https://git.io/JtrZp
19:14:24  <TrueBrain> sorry frosch123 , really don't know what you mean with your comment :(
19:14:51  <frosch123> what about the first rectangle?
19:15:02  <frosch123> doesn't it always include (0,0) into the rectangle?
19:15:04  <DorpsGek> [OpenTTD/OpenSFX] Wuzzy2 commented on pull request #14: Change readme, license, descriptions  https://git.io/JtM5Z
19:15:21  <TrueBrain> ah, you mean that
19:15:26  <Wuzzy> I have just removed the outdatd translations in OpenSFX PR
19:15:27  <TrueBrain> means you also found a bug in the OpenGL branch :P
19:15:44  <Wuzzy> because they are no longer correct (description changed due to license)
19:16:06  <Wuzzy> care to review OpenSFX #14, please?
19:16:24  <Wuzzy> https://github.com/OpenTTD/OpenSFX/pull/14
19:17:10  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain commented on pull request #8685: Fix: [SDL2] simplify what to redraw to prevent tearing https://git.io/JtM58
19:17:36  <TrueBrain> michi_cc: what would be a better way, check in BoundingRect() if the Rect was empty, or do it in MakeDirty, or did I copy/paste your code wrong ^^ :)
19:20:10  *** gelignite has quit IRC
19:24:42  <_2TallTyler> One of my default setting improvements in 8463 doesn't seem to have worked in 1.11.0-beta1, to use non-stop orders by default. The settings menu lists is as default-off. Can anyone confirm?
19:25:00  <supermop_Home_> andythenorth would it be obnoxious to open an issue about the opengfx hotel?
19:26:39  <michi_cc> TrueBrain: Let me read my code. I'm not sure I actually thought about that problem :)
19:26:59  <TrueBrain> frosch123 is giving you a free early review :D
19:27:05  <TrueBrain> he just doesn't know :P
19:31:14  <michi_cc> Okay, the way I implemented it is wrong :) I would actually tend to change BoundingRect (i.e. by ignore rects that have width&&height==0) just in case the function is used somewhere else in the future.
19:31:54  <TrueBrain> k, let me fix that in my PR already :D
19:33:05  <michi_cc> It actually not that surprising, BTW, that the new layered drawing on OSX is very fast.
19:33:22  <TrueBrain> it uses OpenGL in the backend, I noticed :P
19:33:35  <andythenorth> supermop_Home_ no it would not!
19:33:38  <michi_cc> What the OpenGL PR does is basically two things: improve moving pixels to screen and improve palette animation by offloading it to the GPU.
19:33:43  <andythenorth> I have been not doing the PR
19:33:43  <LordAro> _2TallTyler: confirmed, doesn't appear to have worked
19:33:54  <andythenorth> because then I don't know who will approve the PR
19:33:56  <Wuzzy> seems like everyone ignores me now. šŸ™
19:33:58  <TrueBrain> you do have to test my performance PR btw michi_cc ; I couldn't find any reason why std::chrono would be of issue
19:34:06  <andythenorth> and I was in no mind for debating 2 hotel sprites until we are all dead
19:34:21  <andythenorth> if someone else does the PR, I will approve it :P
19:34:24  <Timberwolf> Wuzzy: There's just a lot on, PRs take a while. e.g. I have one that's a few weeks old, it'll happen when it happens.
19:35:19  <frosch123> _2TallTyler: yep, you changed the wrong setting
19:35:25  <michi_cc> Now on OSX a CALayer is basically a thing wrapper around a GPU surface. Giving the system a bitmap reference to upload to the GPU is excatly what part 1 of OpenGL PR does, except of course that OSX can use the most optimal implementation.
19:35:53  <TrueBrain> michi_cc: makes sense
19:35:55  <michi_cc> So that only leaves the improved palette animation, which as it seems currently is not enough speedup to compensate the less optimal GPU upload.
19:35:55  <frosch123> "gui.sg_new_nonstop" is wrong, "gui.new_nonstop" would have been the right thing
19:36:04  <_2TallTyler> Are those SLV_22 and SLV_93 lines to denote an obsolete setting?
19:36:20  <frosch123> it used to be a game setting instead of a client setting
19:36:26  <frosch123> so it was stored in savegames
19:36:32  <TrueBrain> michi_cc: so we can further improve the OpenGL implementation, you say :D :D :D
19:36:36  <TrueBrain> (not for the first iteration!)
19:36:41  <frosch123> you actually need to revert that change, since it may break old savegames
19:36:50  <frosch123> *very old* savegames only though :p
19:37:29  <michi_cc> TrueBrain: Improving the OpenGL implementation might translate to "write a Metal backend" for OSX :D
19:37:44  <TrueBrain> michi_cc: haha :D
19:38:13  <_2TallTyler> Can I revert just that change in a new PR which also changes the correct setting, or do I need to revert the entirety of #8463? Never reverted a PR before.
19:38:56  <frosch123> just make a new PR, which changes the two defaults
19:39:08  <_2TallTyler> OK, will do. Thanks.
19:39:15  *** jottyfan has joined #openttd
19:39:21  <frosch123> put the hash of the old commit into the "Fix" commit message
19:39:32  *** jottyfan has quit IRC
19:39:34  <_2TallTyler> Righto
19:39:41  <frosch123> though we can also add that on squash :)
19:39:54  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8685: Fix: [SDL2] simplify what to redraw to prevent tearing https://git.io/JtM1T
19:39:55  <TrueBrain> michi_cc: something like this what you had in mind too ^^?
19:41:01  <michi_cc> Yes, though I might actually have checked X and Y to allow "line" extensions.
19:43:23  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8685: Fix: [SDL2] simplify what to redraw to prevent tearing https://git.io/JtM1T
19:43:34  <TrueBrain> fair enough
19:43:42  <TrueBrain> depends on what you define as rect
19:43:47  <TrueBrain> is top=0, bottom=0 a line?
19:43:49  <TrueBrain> or is it nothing?
19:44:03  <TrueBrain> :D
19:44:17  <TrueBrain> currently, I was assuming it was nothing btw
19:44:36  <TrueBrain> which is fine, as what-ever you pick, do it consistent for your usage, and you are fine
19:44:59  <michi_cc> "top | bottom | left | right == 0" :D
19:45:47  <frosch123> how about -1 then? or std::optional<rectangle> :p
19:46:37  <TrueBrain> okay, you guys fight out what the best C++ way of doing this is, I will write the code, k? :P
19:47:37  <michi_cc> Bitwise or is definitely the C way, not the C++ way :)
19:49:40  <michi_cc> The most C++ way is probably "std::optional<Rect> _dirty_rect;"
19:50:19  *** argoneus4 has joined #openttd
19:50:38  <michi_cc> But I think a normal if is okay for us :
19:50:39  <michi_cc> )
19:50:44  *** lastmikoi has quit IRC
19:50:44  *** argoneus has quit IRC
19:50:44  *** dih has quit IRC
19:50:44  *** argoneus4 is now known as argoneus
19:50:54  *** dihedral has joined #openttd
19:51:40  <TrueBrain> so something like this :P
19:51:43  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8685: Fix: [SDL2] simplify what to redraw to prevent tearing https://git.io/JtM1T
19:51:44  *** lastmikoi has joined #openttd
19:51:58  <TrueBrain> I still have tearing btw, but now it is vsync tearing .. without this PR, I have a very weird tearing (vertical, instead of horizontal)
19:52:53  *** berndj-blackout has joined #openttd
19:52:54  <michi_cc> It's future proof, and nicely uncommon C :)
19:53:08  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8685: Fix: [SDL2] simplify what to redraw to prevent tearing https://git.io/JtM1T
19:53:15  *** iSoSyS has joined #openttd
19:53:22  *** milek7_ has joined #openttd
19:54:00  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain updated pull request #8685: Fix: [SDL2] simplify what to redraw to prevent tearing https://git.io/JtM1T
19:54:01  <TrueBrain> now with doxygen comments
19:54:16  *** Tulitomaatti has joined #openttd
19:54:27  *** Timberwo1f has joined #openttd
19:54:39  <DorpsGek> [OpenTTD/OpenTTD] michicc approved pull request #8685: Fix: [SDL2] simplify what to redraw to prevent tearing https://git.io/JtMdd
19:54:40  <michi_cc> frosch might have a different opinion, but meh ^^
19:54:41  *** avdg has quit IRC
19:54:41  *** Hazzard has quit IRC
19:54:41  *** berndj has quit IRC
19:54:41  *** Greyfur has quit IRC
19:56:01  *** blathijs has quit IRC
19:56:05  *** Tulitoma1tti has quit IRC
19:56:20  *** Timberwolf has quit IRC
19:56:25  *** Guest12698 has quit IRC
19:56:25  *** fonsinchen has quit IRC
19:56:45  *** tneo has quit IRC
19:56:50  *** milek7 has quit IRC
19:56:55  <Wolf01> People are dying!
19:57:23  *** m811 has joined #openttd
19:57:52  *** avdg has joined #openttd
19:58:11  *** iSoSyS has quit IRC
19:58:56  <frosch123> michi_cc: did you already fight with glx? :p
19:59:30  <michi_cc> I think there's a slight tendency for my PR.
19:59:41  *** Hazzard has joined #openttd
19:59:54  <TrueBrain> that is not fighting :(
20:00:21  <michi_cc> I can't approve myself. I could close glx' PR though :D
20:00:40  *** tneo has joined #openttd
20:00:46  <DorpsGek> [OpenTTD/OpenTTD] frosch123 approved pull request #8678: Fix #8676: GUI-visible settings may not be part of misc settings. https://git.io/JtMFf
20:00:46  *** moll has quit IRC
20:01:01  <frosch123> reject it as "this has already been fixed"?
20:01:07  <TrueBrain> :D
20:01:25  <DorpsGek> [OpenTTD/OpenTTD] michicc merged pull request #8678: Fix #8676: GUI-visible settings may not be part of misc settings. https://git.io/JtX9t
20:01:28  <DorpsGek> [OpenTTD/OpenTTD] michicc closed issue #8676: Crash when opening settings window https://git.io/JtXiF
20:01:47  <DorpsGek> [OpenTTD/OpenTTD] michicc commented on pull request #8677: Fix #8676: Revert a2c3197f4 and unconditionally load settings "early" https://git.io/JtMFJ
20:01:50  <DorpsGek> [OpenTTD/OpenTTD] michicc closed pull request #8677: Fix #8676: Revert a2c3197f4 and unconditionally load settings "early" https://git.io/JtXSK
20:01:52  *** Terkhen has joined #openttd
20:02:13  *** blathijs has joined #openttd
20:02:22  *** fonsinchen has joined #openttd
20:04:30  <DorpsGek> [OpenTTD/OpenTTD] frosch123 commented on pull request #8683: Feature: Town name generator for Spanish (Spain) https://git.io/JtMFt
20:04:33  <DorpsGek> [OpenTTD/OpenTTD] frosch123 closed pull request #8683: Feature: Town name generator for Spanish (Spain) https://git.io/JtMtf
20:05:44  *** glx has joined #openttd
20:05:44  *** ChanServ sets mode: +v glx
20:06:50  <DorpsGek> [OpenTTD/OpenTTD] frosch123 approved pull request #8659: Change: "(Business/Chair)man" to "(Business/Chair)person" https://git.io/JtMFG
20:07:55  *** andythenorth_ has joined #openttd
20:12:27  *** nielsm has quit IRC
20:13:50  *** andythenorth has quit IRC
20:19:36  <DorpsGek> [OpenTTD/OpenTTD] TrueBrain merged pull request #8685: Fix: [SDL2] simplify what to redraw to prevent tearing https://git.io/JtM1T
20:34:34  <DorpsGek> [OpenTTD/OpenTTD] glx22 closed issue #8601: Female CEOs might get the title of ā€œBusinessmanā€ or ā€œChairmanā€ https://git.io/Jt365
20:34:37  <DorpsGek> [OpenTTD/OpenTTD] glx22 merged pull request #8659: Change: "(Business/Chair)man" to "(Business/Chair)person" https://git.io/Jtrqd
20:40:32  *** jottyfan has joined #openttd
20:44:36  *** jottyfan has quit IRC
20:56:18  *** Progman has joined #openttd
21:02:49  *** gelignite has joined #openttd
21:04:00  *** NGC3982 has joined #openttd
21:04:02  <NGC3982> oh hai
21:05:16  <_2TallTyler> Does anyone have a comprehensive proof that path signals are not more computationally-intensive than block signals? I see this opinion stated several times a week on Discord and would like a better argument against it than "Timberwolf said so in one of his videos"
21:05:40  <supermop_Home_> so i am dumb, i don't really get how a PR works for a sprite. Like for a base set, if i want to add new sprites, how does that work?
21:06:11  <_2TallTyler> I'm working on a new implementation of #7504 so I expect a lot of people to come out of the woodwork with that argument :)
21:06:51  <Eddi|zuHause> supermop_Home_: look at #8556
21:07:19  <frosch123> _2TallTyler: noone ever measured it. the theoretical discussions all compare oranges with apples
21:07:55  <Xaroth> andythenorth_: https://i.redd.it/njkbuir592i61.jpg
21:08:11  <supermop_Home_> Eddi|zuHause is that an opengfx issue?
21:08:12  <Xaroth> every time I read that sentence I have to think of that.
21:08:50  <Eddi|zuHause> supermop_Home_: no, for openttd
21:08:54  <frosch123> _2TallTyler: for example: with block signals, the signals update everytime a train passes, and the pathfinder of the train runs at every junction. for path signals: updating signals and pathfinding is mostly the same thing. the path finder runs when the signal is passed, and then follows the reservation
21:09:45  <frosch123> _2TallTyler: so, if you have a big junction and only one path signal in front, then the pathfinder will run only once. but with a block signal the pathfinder would run at every intermediate junction
21:10:25  <frosch123> _2TallTyler: most comparisons claim path signals are more expensive because they have to run the pathfinder. but then forget that the train won't have to run it later.
21:11:05  <Eddi|zuHause> supermop_Home_: it depends on what exactly you want to add
21:11:43  <frosch123> supermop_Home_: the "extra" grf of a baseset is a regular newgrf, it works the same as any other newgrf
21:12:11  <_2TallTyler> Thanks for the explanation, frosch123
21:12:23  <supermop_Home_> just replace the arctic hotel sprite
21:12:38  <supermop_Home_> i can't see that issue eddi
21:12:43  <glx> extra is basically a list of action 5 IIRC
21:13:03  <supermop_Home_> oh its a pr
21:14:11  <frosch123> _2TallTyler: ah, i forgot one item: some people make the distinction that on straight track without any junctions, path signals would be expensive because they run a pathfinder when no path needs finding
21:14:55  <frosch123> but i don't know either whether it really searches for a path, or whether it stops when it notices there is just straight track
21:15:44  <andythenorth_> Xaroth lolz
21:15:47  <andythenorth_> what a come down
21:16:13  <frosch123> Xaroth: everyone knows andy would never ride a horse. he likes ponies
21:16:55  <andythenorth_> supermop_Home_ just replace the sprites, there are probably 4 or 6, climate specific?
21:17:00  <andythenorth_> then PR
21:17:55  <Eddi|zuHause> supermop_Home_: if you just want to change the existing sprite, you just modify the file and pr that
21:18:08  <glx> finding path in long straight track is trivial I think, so probably not that expensive
21:18:54  <frosch123> glx: i don't know whether it only searches up to the next signal, or until the destination
21:19:12  <supermop_Home_> andythenorth_ you misunderstand how little i understand about this. so gh is like what branches are you comparing, so one branch is a .png, and the other branch is opengfx master?
21:19:25  <andythenorth_> supermop_Home_ maybe Eddi|zuHause will do it :)
21:19:28  <andythenorth_> then I can approve?
21:20:17  <Eddi|zuHause> supermop_Home_: just read the git explanation of how to make a PR.
21:20:44  <Eddi|zuHause> it works the same way no matter if you modify a code file or an image
21:27:11  <andythenorth_> oops, /me clicked 'settings' in a recent pull of master :)
21:27:15  <andythenorth_> oh dear
21:27:39  <TrueBrain> not recent enough, clearly
21:27:45  <andythenorth_> 'recent'
21:27:50  <TrueBrain> :D
21:28:02  <andythenorth_> supermop_Home_ TBH I can do it, I just fear it will sit for months because nobody else will want to answer to players "ZOMG you deleted the beautiful rendered OGFX hotel sprite"
21:28:16  <andythenorth_> so the PR might not move
21:28:25  <andythenorth_> let's be optimistic!
21:28:48  <andythenorth_> where is my opengfx checkout :P
21:28:50  *** Progman has quit IRC
21:28:53  <TrueBrain> I answered to players today!
21:28:56  <TrueBrain> well, not about OpenGFX :P
21:29:10  <supermop_Home_> andythenorth_ tell them it's my fault
21:29:27  <frosch123> i read that as "prayers" :p
21:29:27  <andythenorth_> someone still has to approve :P
21:30:19  <frosch123> we had an ogfx dev, but they only played with original graphics, so they quit
21:30:39  <andythenorth_> I think I kept my approval rights :P
21:30:59  <TrueBrain> blind approval? I can do that too! :P
21:31:20  <frosch123> supermop_Home_: anyway, we need to add the two new gui sprites to ogfx in the next month. and then there will be a new release
21:31:40  <frosch123> so, if you can PR before that, you are less likely to have to wait a year for the next release :p
21:31:53  <glx> oh if you touch opengfx images andythenorth_, there's a glitch with some tunnel tiles
21:32:33  <glx> and the opengfx is "broken" too, for some accented chars
21:33:45  <andythenorth_> do we know what needs fixed?
21:33:51  <glx> oh supermop_Home_, we can see issue #44 ;)
21:35:49  <andythenorth_> now it's the game called 'hunt the openfx sprite' :D
21:36:06  <andythenorth_> no, I'm not proposing reorganising the spritesheets either, they rarely need changed :P
21:36:11  <glx> well for fonts there's an open issue
21:36:42  <andythenorth_> rad arctichotel.png is a thing
21:36:44  <glx> for tunnel glitch I don't see an open issue, but I can try to remember how to spot it ingame
21:36:49  <andythenorth_> unusually easy this one
21:37:26  <andythenorth_> ogfx in windows palette? :o
21:38:56  *** Progman has joined #openttd
21:39:20  <andythenorth_> oh we have some ANIM warnings for high score :D
21:39:28  * andythenorth_ must have missed some last year
21:41:40  <frosch123> no, it was not missed
21:41:54  <frosch123> we made a lof of fun about the sprite being completely broken and noone ever reporting it
21:42:02  <frosch123> it was probably too much work to fix it
21:44:53  <supermop_Home_> are any new gui needed to draw?
21:45:56  <supermop_Home_> glx yay
21:47:59  <andythenorth_> supermop_Home_ FIRS picks up the ogfx sprite https://gist.github.com/andythenorth/311126e1d2cb2ad3c31bc86bc02a26d4#gistcomment-3635022
21:48:13  <andythenorth_> I haven't found one in a town yet :P
21:48:22  <andythenorth_> so glad I don't make houses, testing them must be horrible
21:49:10  <supermop_Home_> andythenorth_ do you get the house version with firs loaded?
21:49:22  <andythenorth_> yes
21:49:27  <andythenorth_> I don't nerf the town version
21:49:32  <andythenorth_> found one, it's date sensitive
21:49:44  <supermop_Home_> also lmk if you want all the assorted other bits for in firs
21:49:57  <supermop_Home_> or / direction
21:50:15  <andythenorth_> https://gist.github.com/andythenorth/311126e1d2cb2ad3c31bc86bc02a26d4#gistcomment-3635029
21:50:23  <supermop_Home_> i didn't draw / yet but shouldn't take more than a couple hours
21:50:45  <michi_cc> _2TallTyler: I've never measured the PBS code I wrote, but from a theoretical POV: Path signals are better if multiple junction tiles are in one signal block due to less pathfinder calls; straight line no difference.
21:51:01  <supermop_Home_> well it looks better than the rendered one at least
21:51:20  <andythenorth_> I _think_ it has converted to windows palette ok
21:51:28  <andythenorth_> there are fewer greys, it might have lost some shading
21:51:31  <michi_cc> The one thing path signals are worse is when trains wait at a red/blocked signals as they can't passively wait for the signal to turn red but have to periodically query for a free path.
21:51:37  <frosch123> supermop_Home_: i think the sprites from openttd.grf are good enough. we only had one pencil icon, and the location icon also won clearly
21:51:56  <michi_cc> I.e. path signals are generally better except if your network is mostly a traffic jam :)
21:52:03  <_2TallTyler> michi_cc I haven't seen any argument about path signals at junctions, just whether they're worse than block signals on junction-less track
21:52:37  <andythenorth_> how much pathfinding could it even be doing?
21:52:42  <andythenorth_> there's one node to consider
21:52:45  <michi_cc> They aren't. Pathfinder calls only ever occur on junction tiles.
21:52:45  <andythenorth_> and one edge
21:53:03  <glx> ok for at least rail tunnels in / direction there's a missing pixel, spottable when moving a window over it
21:53:50  <michi_cc> And putting a path reservation on the tile will always be done, even for block signals, just a bit later when the train enters the tile instead of when it enters the signal block. Still exactly the same amount of work.
21:54:02  <andythenorth_> supermop_Home_ https://gist.github.com/andythenorth/311126e1d2cb2ad3c31bc86bc02a26d4#gistcomment-3635031
21:54:32  <michi_cc> I wrote that shit so I should know it. But I do have to admit I never did any in-depth profiling.
21:54:38  <andythenorth_> I think people confuse what looks like work (drawing the reservation) with actual work
21:54:55  <andythenorth_> like I think nmlc is slow to do lang, because it sometimes prints a lot of warning lines :P
21:55:05  <andythenorth_> not empirically true, but my brain knows otherwise
21:55:16  <supermop_Home_> andythenorth_ now i feel like i need to fix that church
21:55:29  <andythenorth_> supermop_Home_ this is why I don't touch opengfx
21:55:36  <andythenorth_> it's not that I think it's shit and want to throw it away
21:55:51  <andythenorth_> it's just that it's 80% and could be brought up to an even standard
21:55:59  <andythenorth_> but so much...work
21:56:29  <glx> and there's no real consistency in interface gfx (for me)
21:56:31  *** Samu has quit IRC
21:56:57  <andythenorth_> whenever I enable it, I get tempted to...tweak
21:58:18  <FLHerne> andythenorth_: Printing to the console can actually be surprisingly slow :p
21:58:34  <andythenorth_> so I AM RIGHT!
21:58:35  <andythenorth_> fancy that
21:59:31  <supermop_Home_> andythenorth_ yeah, and its hard to set a scope for the tweaking... limit yourself to just fixing the snow, or maybe the whole roof, or redraw the whole thin?
21:59:40  <andythenorth_> supermop_Home_ https://gist.github.com/andythenorth/311126e1d2cb2ad3c31bc86bc02a26d4#gistcomment-3635036 alternative guest room
21:59:43  <andythenorth_> and below, snow version
22:00:00  <andythenorth_> think I prefer that second version, shape is easier to parse
22:00:21  <supermop_Home_> and once the most egregious thing is fixed, the next thing that looked ok before now sticks out as the new worst
22:00:48  <supermop_Home_> the little gables stand out better at scale
22:01:05  <andythenorth_> yeah I was just looking up what they are technically called
22:01:13  <andythenorth_> I call them gables, but I thought that might be wrong :P
22:01:24  <andythenorth_> they tie the two tiles together
22:01:28  <supermop_Home_> the one without  (A) works better maybe as a stand alone one tile cheap ski town motel
22:01:42  <andythenorth_> very banff
22:01:48  <andythenorth_> there are flashing pixels on the end wall btw
22:01:51  <andythenorth_> easy fix but eh
22:02:30  <supermop_Home_> lake placid has both quaint tiny 8 room lodges like that, and big grand hotels with a big atrium
22:02:41  <supermop_Home_> which end wall?
22:02:59  <supermop_Home_> thought i'd got all of those
22:03:13  <andythenorth_> nearest SE end
22:03:29  <andythenorth_> colour is b06c54 in hex, and there's a an L-shaped group of it amongst other pixels
22:03:37  <supermop_Home_> ugh
22:03:38  <andythenorth_> it's a brick colour, really common issue this one
22:04:31  <andythenorth_> colour 230 here https://github.com/andythenorth/iron-horse/blob/master/palette_key.png
22:04:51  <andythenorth_> you want 78 or so
22:05:16  <andythenorth_> supermop_Home_ fancy making the snow more similar to rest of opengfx?
22:05:32  <andythenorth_> shading at edges seems to be the convention
22:07:04  <supermop_Home_> i tried to make it look 'thick' so it has a brighter face along SE edge, as if it was vertical for 2px there. i can add a little more color around corners tho
22:08:55  <supermop_Home_> just got to find my psd real quick
22:09:59  <supermop_Home_> 78 are the bad ones?
22:10:42  <supermop_Home_> or 227-231 are bad?
22:17:14  *** sla_ro|master has quit IRC
22:17:16  <andythenorth_> 227-231 are animated
22:23:44  <supermop_Home_> okie dokie
22:28:01  *** mirrorbird has quit IRC
22:31:03  <andythenorth_> hmm we broke Arctic Farms
22:31:14  <andythenorth_> that makes me look like a right twat in the forums?
22:35:01  <frosch123> at work there is an internal "forum"
22:35:32  <frosch123> one forum is about IT, but mostly about complaining about IT
22:36:13  <frosch123> today there was a thread, where people complained that they have to click a window away after signing wia the VPN
22:36:29  <frosch123> apparently it's very annoying, and the extra click (after entering your PW) is a lot of work
22:36:46  <frosch123> they some random guy posted a shell script, to log in to the VPN without dialogs
22:37:16  <frosch123> but that script also skipped the password dialog, by storing the raw password in the script
22:37:25  <supermop_Home_> nice
22:37:38  <TrueBrain> Hahaha
22:37:44  <frosch123> which funnily gaines a lot of praised, until IT told them to never ever do that
22:37:55  <frosch123> so they spend more work to add an console input for the PW
22:38:02  <frosch123> all about removing a single click?
22:38:44  <frosch123> anyway, I am not IT, but that forum always shows me that not only our users are like that :)
22:39:09  <frosch123> people come up with the weirdest suggestions to solve the weirdest non-problems
22:39:28  <TrueBrain> Lolz @ reply on our forums
22:39:34  <TrueBrain> Holy crap, that dude is lost
22:40:06  <TrueBrain> He took a wrong turn somewhere and is now somewhere in the forest
22:40:50  <frosch123> well, flat arctic maps don't work :) either you have no forests or no farms
22:40:58  <TrueBrain> I love that it turned into demanding answers .. as if we are on trial :D
22:42:37  <frosch123> is there some rule 73 on the internet: if you start with argueing about "your freedom", you have already lost the conversation?
22:42:53  <TrueBrain> I love the complaints about the screenshot of the minimaps .. he is aware those are heightmaps, right?
22:44:21  <TrueBrain> But yeah, a conversation where someone has a pitchfork in his hand is a lost conversation before it started :(
22:45:46  <supermop_Home_> haha what topic?
22:48:20  <TrueBrain> In general
22:48:32  <TrueBrain> Can't miss it :p
22:48:38  <andythenorth_> I did a reply and deleted
22:48:42  <andythenorth_> I did another one, better
22:49:15  <andythenorth_> dunno why this pisses me off, I think it's that I already told him in the GH issue that this is happening
22:49:38  <andythenorth_> then he goes to forums and starts a poll, as though this is a democracy, and not some kind of clusterfuck un-nameable political structure
22:50:01  <andythenorth_> the walls of text are just...oof
22:50:26  <TrueBrain> Just a typical "I have been wronged" thread .. we all made them, and we have seen many more :)
22:50:48  <andythenorth_> I think I stopped early
22:50:57  <andythenorth_> I joined internet culture via usenet
22:51:14  <andythenorth_> usenet was both very prone to numbered nitpicks
22:51:18  <TrueBrain> And I am no longer allowed to solve that in my favorite way :p
22:51:18  <andythenorth_> and very prone to flaming them
22:56:23  <_2TallTyler> I'm attempting to change the "Enable the signal GUI" to a drop-down, enabling the user to select Path signals only, All signals, or None (original off behavior). I've changed it from bool to uint8, but the selector is a left-right arrow, not a drop-down. What am I missing?
22:57:00  <frosch123> andythenorth_: no need for more replies, erato already posted the best reply "8. ugly grass" :)
22:57:13  <andythenorth_> oofsie
22:57:25  <_2TallTyler> https://pasteboard.co/JOPWZrQ.png
22:57:26  <frosch123> i like that reply :)
22:57:57  <andythenorth_> supermop_Home_ I won't do more ogfx tonight, going to sleep soon-ish :)
22:58:11  <andythenorth_> but send me any tweaks, I might get them in tomorrow
23:00:03  <frosch123> _2TallTyler: guiflags, SGF_MULTISTRING i think
23:02:09  <_2TallTyler> I thought I had that, but had a typo of "guiflats" ... :p
23:02:14  <_2TallTyler> Works now
23:02:31  <supermop_Home_> ok
23:02:56  <supermop_Home_> sorry i was talking to my super about fixing a water hammer in the pipes from upstairs
23:03:03  <supermop_Home_> want me to PM it?
23:03:28  <supermop_Home_> i fixed the animated pixels and tried to shade the snow more
23:04:21  <frosch123> _2TallTyler: not sure whether a setting in the setting tree is the best option
23:04:44  <frosch123> the signal gui could itself have expand buttons to hide block, path and semaphore singals
23:05:59  <frosch123> oh, it's already tomorrow...
23:06:00  <frosch123> night
23:06:04  *** frosch123 has quit IRC
23:08:11  *** Progman has quit IRC
23:08:33  <LordAro> ooh, forum drama
23:08:53  <LordAro> they're Zorg on GH, aren't they?
23:09:02  <supermop_Home_> ??
23:09:10  <TrueBrain> Yes LordAro
23:09:10  <_2TallTyler> Yes, same person
23:10:19  <andythenorth_> standard forum drama, it's about almost nothing
23:10:57  <supermop_Home_> andythenorth_ sent pm
23:11:32  <andythenorth_> got it cheers  :)
23:17:17  * andythenorth_ must to sleep
23:17:21  *** andythenorth_ has quit IRC
23:21:47  <supermop_Home_> man
23:22:10  <supermop_Home_> where do they get the energy
23:35:17  <TrueBrain> Who does?
23:38:01  *** gelignite has quit IRC
23:48:20  *** Wolf01 has quit IRC
23:54:58  *** glx has quit IRC
23:59:34  *** _2TallTyler has quit IRC

Powered by YARRSTE version: svn-trunk