IRC logs for #openttd on OFTC at 2012-02-12
⏴ go to previous day
00:00:05 <Nat_as> i hate how you have to make an internatonal airport if you want more than 3 terminal places
00:00:14 <Nat_as> what if I have cargo planes waiting to load 100%
00:01:16 <Rhamphoryncus> yup. Switching to intercontinental is something you just do everywhere ASAP
00:01:27 <Rhamphoryncus> Planes need a new mechanic to balance them better
00:01:38 <Nat_as> man, looking at the Newgrif tutorial, I can say that's one thing where we have more advanced code than Simutrans.
00:01:47 <Nat_as> I sugested airplanes being able to mix cargo types
00:01:57 <Nat_as> they trade flexibility for inefficiency.
00:02:10 <Nat_as> they can move anything anywhere, but cost lots of money.
00:02:12 <Rhamphoryncus> I mean better than the chance to crash
00:02:35 <Nat_as> really airplanes are more safe than trains or cars IRL
00:02:55 <Nat_as> High speed trains are really dangerous.
00:03:38 <Rhamphoryncus> I was just thinking, YACD with different pricing brackets depending on time. Some passengers may really want a fast flight and pay more for it
00:04:17 <Nat_as> i also suggested differentiating between commuters and travelers.
00:05:18 <Rhamphoryncus> The game would have to estimate travel times to only put those passengers on the plane though
00:06:01 <Nat_as> I'd rather see massive improvements to cargo desitnation handling before that.
00:07:41 <Nat_as> oh, commuters don't care about being stuffed into things and will travel both ways
00:08:11 <Nat_as> Travelers do care and will want to go somewhere as quick as possible and stay there
00:08:23 <Nat_as> that might be oversimple though.
00:08:35 <Nat_as> Tourism industries would be cool
00:10:11 <Rhamphoryncus> That could be simplified in to comfort levels
00:10:52 <Nat_as> The "Passingers go both ways" thing however could be applied to all passingers
00:11:08 <Nat_as> balanced by simply reducing the amount of passingers created by half
00:11:35 <Nat_as> giant trains that move people from a big city to a small city and return empty are silly.
00:12:05 <Rhamphoryncus> I'd put that on cargodest again
00:12:05 <Nat_as> reminds me of the "Fema death camp" that turned out to be an Amtrack facility
00:12:23 <Nat_as> OH NO, AMTRACK IS IN ON THE CONSPIRACY!
00:12:28 <Nat_as> DON'T GET IN THE TRAINS!
00:13:12 <Elukka> my favorite part of conspiracy theorists is how they ignore real conspiracies
00:15:38 <Rhamphoryncus> the passingers, they're everywhere!
00:15:49 <Rhamphoryncus> They're what happen when you feed a passenger after midnight
00:15:59 <Elukka> they are much like the sheeple but a bit less dangerous
00:16:44 <frosch123> DorpsGek is our sheeple
00:17:56 <Elukka> the passingers skulk through abandoned stations and derelict sidings at night
00:34:01 <Nat_as> I made a cute little psudo subway system
00:34:05 <Nat_as> but I had to cheat to afford it
00:48:32 *** supermop has joined #openttd
00:49:25 <Nat_as> hmm, are signals strictly necessary exiting stations?
00:49:53 <Nat_as> my subway layout seems to sugest they are not, but that might just be because the trains are as short as possible and get out of the way quickly
00:50:08 <Nat_as> I have a two platform set up leading into a double track,
00:50:25 <Nat_as> immediatly outside the station is an X allowing trains to switch to either platform when arriving
00:50:40 <Nat_as> and then there are route signals seperating the left and right raills into coming and going
00:51:02 <Nat_as> I used to put signals between the X and the stations but I left them out this time to save room and I don't notice any issue
00:52:14 <Rhamphoryncus> entering or exiting the station?
00:52:31 <Rhamphoryncus> hrm. I'd have to see the layout
00:52:45 <Nat_as> I have Double tracks, Route signals, a cross, and the station itself
00:53:03 <Nat_as> it used to be double tracks, route signals, a cross, normal signals, and the station itself
00:53:28 <Rhamphoryncus> But the big caveat on most of this is normal block signals treat all the track as a single block, but if you're using path signals then the train can switch in to path mode and may not need signals
00:56:50 <Rhamphoryncus> That's a terminus station? It's the end of the line?
00:57:28 <Nat_as> it was, but i added another
00:57:33 <Nat_as> and it's still working fine
00:58:39 <Nat_as> it could be that the trains are too short and too fast to hit each other
00:58:52 <Nat_as> I want to know if I could replicate this building cargo routes
00:59:48 <Rhamphoryncus> I'm experimenting now
01:00:12 <Rhamphoryncus> For terminus it definitely works
01:01:26 <Rhamphoryncus> This is all 1-way? Or two-way?
01:05:12 <Rhamphoryncus> Okay, what happens if you have path signal, cross, station, cross, normal signal, and this is all 1-way (they continue through the station to the next station), when they're at the path signal they grab a path up until the next stopping point
01:05:37 <Rhamphoryncus> For a terminus station this is the end of the line. For a non-terminus station it's whatever signal is after
01:06:21 <Rhamphoryncus> Beware, with a cross after the station they may choose to swap, then sit there and load for a while, preventing any other train from leaving (because they took the path)
01:07:33 <Rhamphoryncus> There is something to help you with this. Click on the gear at the top left, advanced settings, interface, display options, set Show reserved tracks to on
01:12:28 <Nat_as> it seems that tracks are reserved when a train exits the station
01:12:35 <Nat_as> they count as signal blocks
01:13:33 <Nat_as> I really wish subways and underwater tunnels were possible
01:13:43 <Rhamphoryncus> they're reserved from before the train enters the station
01:13:58 <Rhamphoryncus> A train won't go unless it has a stopping point
01:16:01 <Nat_as> no i mean it automaticly reserves as the train exits
01:17:50 <Nat_as> or rather they become reserved as the train enters the station and stay that way till it leaves the block
01:18:09 <Rhamphoryncus> oh, that is what I said
01:19:12 <Rhamphoryncus> I notice that have a cross immediately after the station irks the pathfinder. It takes a second or two after the train is clear before another train will enter. It probably has a special case for signals immediately after the station
01:20:14 <Rhamphoryncus> It's much better to have a cross before the station, then signals immediately after
01:21:24 <Nat_as> yes but that doubles the lenght of the station
01:21:49 <Nat_as> which is bad if you are trying to build in a city
01:22:17 <Nat_as> how hard would it be to make OTTD maps layerd?
01:22:28 <Rhamphoryncus> Hard. I still want to do it though :)
01:22:38 <Nat_as> they would have to change the memory allocation of each tile right?
01:23:20 <Nat_as> but it would allow cool things like subways, elivated raill, and airplanes crashing into tall buildings because you put an airport in the center of a city.
01:23:22 <Rhamphoryncus> Redesign that, the save format, the path finding, the tools for building
01:24:10 <Nat_as> isn't that a function of the route shape? it should always be discribable as a web right?
01:24:12 <Rhamphoryncus> huh, that gives me an interesting thought: when building a custom airport you could "build" the flight path it uses to land. This would prevent buildings from grabbing that space
01:24:14 <Nat_as> even if routes overlap
01:24:25 <Nat_as> yeah that would be cool
01:24:31 <Rhamphoryncus> No, it understands the route as a graph and figures out the optimal way to get there
01:25:51 <Nat_as> yeah that's what I thought
01:26:02 <Nat_as> making a map 3d should not effect that
01:26:19 <Nat_as> anyways, 'building' air routes would be cool.
01:26:42 <Rhamphoryncus> The algorithms are usually pretty abstract, but the actual code will need some changes to handle it
01:26:49 <Nat_as> maybe small airports don't need it, but large ones become dangerous if you don't.
01:27:02 <Nat_as> the biggest problem is changing the memory
01:27:08 <Rhamphoryncus> AI's would need work
01:27:10 <Arafangion> Nat_as: Aren't they like boats... No chance of crashes in air?
01:27:19 <Arafangion> Even if they overlap.
01:27:21 <Rhamphoryncus> No, memory is easy :)
01:27:29 <Nat_as> Well this sugestion would allow air crashes >:3
01:27:41 <Nat_as> so it's just three easy problems/
01:27:53 <Arafangion> Nat_as: Realistic air traffic management is very tricky.
01:28:14 <Rhamphoryncus> Nat_as: if you build a landing path then there's nothing to crash in to
01:28:20 <Arafangion> You have corridors, and modern planes monitor nearby planes and ensure that they keep the right distance, etc.
01:29:09 <Arafangion> Mid-air crashes are relatively rare, except for airports.
01:29:49 <Nat_as> I used to be in the air force learning to be an ATC
01:29:56 <Nat_as> hence why I am here now
01:30:00 <Rhamphoryncus> If you're arguing realism then compared to trains they essentially never happen. That's not the issue though. They need a better balancing factor
01:30:16 <Arafangion> Right now planes ruin the fun of the game.
01:30:37 <Nat_as> but it would be cool to 'build' takeoff and landing coridoors, and holding fixes and such
01:30:55 <Nat_as> I'm imagining giant transparent radar display symbols that are layered over the map
01:30:57 <Arafangion> We also need more helicopters.
01:31:16 <Arafangion> As the game progresses, the options in planes and airports become severely restricted.
01:31:35 <Nat_as> Boats should have bigger docks, Airplanes should have custom layouts, and trains should have subways
01:32:01 <Rhamphoryncus> You forgot trucks
01:32:17 <Arafangion> What about trucks?
01:32:30 <Nat_as> I'm here for the trains
01:32:45 <Arafangion> Rhamphoryncus: What about trucks? :)
01:32:57 <Arafangion> I think the trucks are all fine, but maintaining them is a bitch .
01:33:01 *** theholyduck has joined #openttd
01:33:03 <Rhamphoryncus> Their stations need work too
01:33:09 <Nat_as> although i supose trucks would benifit from tunnels and sky bridges the same way trains would
01:34:06 <Arafangion> Rhamphoryncus: They have lots of options with stations, though.
01:34:24 <Rhamphoryncus> garbage and not garbage?
01:34:28 <Nat_as> stations have lots of options, what they need are a coherent slice of options.
01:34:39 <Nat_as> take the best ones and put them into a pack
01:34:45 <Nat_as> we need more automatic stations.
01:34:49 <Nat_as> and more non station tiles.
01:34:59 <Arafangion> The main issue I have with trucks, is that they usually just go for the closest station, rather than the most 'ideal' station.
01:35:16 <Nat_as> also non station buildings that do shit
01:35:21 <Nat_as> like hotels would be nice.
01:36:02 <Nat_as> Raillroad Tycoon3 had hotels and bars that gave you money for passengers leaving and arriving at a station.
01:36:19 *** theholyduck has joined #openttd
01:37:16 <Rhamphoryncus> Arafangion: they *have* code to pick empty stations, it just fails fairly often
01:38:10 <Rhamphoryncus> Actually, that's probably one of the best things to really work on. It's totally achievable without significant consequences. Unlike everything else we've mentioned here.
01:38:37 <Nat_as> even though nobody cares about them
01:39:24 <Rhamphoryncus> I've actually found trucks are often the most profitable
01:39:51 <Arafangion> Trucks require the least setup, and somehow, don't seem to be as sensitive to slow cargo delivery.
01:40:04 <Arafangion> But they're a PITA to manage.
01:43:26 *** theholyduck has joined #openttd
01:43:53 <Nat_as> yes I hate trucks because of how hard route management is in this game
01:44:11 <Rhamphoryncus> Do you use shared order lists?
01:44:20 <Nat_as> another feature I wish this game had from simutrans is route handing separate from vehicle handling.
01:44:44 <Nat_as> you mean like when you click on another train to copy it's orders?
01:45:06 <Nat_as> or is there a better way to quickly change several vehicle's orders at once?
01:46:06 <Arafangion> Rhamphoryncus: Managing the shared order lists is also PITA. :)
01:46:12 <Arafangion> Rhamphoryncus: How do you keep track of those lists?
01:46:22 <Rhamphoryncus> Make a group for each list
01:46:29 <Nat_as> where are shared order lists?
01:46:36 <Rhamphoryncus> Put one truck in to it then use "add shared vehicles
01:46:37 <Arafangion> I must be using an old OTTD.
01:46:48 <Rhamphoryncus> naw, the features are well hidden :P
01:46:58 <Rhamphoryncus> Nat_as: ctrl-click when duplicating or cloning
01:47:29 <Rhamphoryncus> or select an existing vehicle, click "go to", then ctrl-click on a vehicle you want to join with
01:47:46 <Nat_as> oh I know how to copy orders
01:48:08 <Nat_as> I was wondering if there were sets of orders stored independently of veichiles that you could quickly switch into and out of
01:48:09 <Rhamphoryncus> It's not a copy. Using ctrl makes it shared
01:48:59 <Rhamphoryncus> You can do switching like that *if* you have a dummy truck somewhere with those orders.
01:49:57 <Nat_as> maybe if there was an administrative building that stored sets of orders
01:50:10 <Rhamphoryncus> Also, route management is what I'm working on. It will do this
01:50:24 <Rhamphoryncus> Create a route first, then create vehicles to use it
01:50:29 <Nat_as> so you don't have to get annoying popups of "train is waiting in the depot.
01:50:45 <Nat_as> and it would be an excuse for pretty non track buildings
01:50:50 <Nat_as> which are always nice.
01:51:11 <Rhamphoryncus> Anyway, afk for a few hours at least
02:08:53 *** tparker has joined #openttd
02:12:27 *** theholyduck has joined #openttd
02:36:27 *** theholyduck has joined #openttd
02:37:30 *** tparker has joined #openttd
02:48:36 *** Biolunar_ has joined #openttd
03:08:48 *** TheStick has joined #openttd
03:09:14 <TheStick> Hey, I've got a question...
03:09:43 <TheStick> ...are there binaries for a "dedicated server" version of OpenTTD on a generic Linux machine?
03:13:37 <Arafangion> TheStick: Is that different from the regular binary?
03:13:55 <TheStick> ...in that it only works as a server for a multiplayer game...
03:14:05 <TheStick> ...that is, zero graphics, sero gameplay on the machine itself.
03:14:33 <TheStick> Because I'm getting this error:
03:14:35 <TheStick> error while loading shared libraries: libSDL-1.2.so.0: cannot open shared object file: No such file or directory
03:15:05 <glx> you have to build dedicated version yourself IIRC
03:15:33 <TheStick> ...how would I do that on a Windows machine...?
03:15:59 <glx> on windows you just get the normal build
03:16:20 <glx> it has dedicated included
03:16:33 <TheStick> ...really, I have two options. Compile on this Win7, or sftp and compile on the generic Linux.
03:16:40 <TheStick> Cuz I want it to run off the Linux.
03:16:57 <glx> (same for linux except if you need a full dedicated without X)
03:17:13 <TheStick> Which is what I need.
03:18:14 <TheStick> Lessee. Get gzip arc of the source...
03:19:27 <TheStick> ...something tells me this is gonna get messy.
03:21:14 <TheStick> ...libs from where...?
03:24:28 <Arafangion> TheStick: What distro are you using, incidentally?
03:24:59 <Arafangion> TheStick: That's usually what the question applies to, yes. :)
03:25:14 <TheStick> ...long story short, I'm repurposing a MineOS CRUX distro.
03:25:27 <TheStick> ...a distro primarily intended to host a Minecraft server.
03:25:39 <Arafangion> TheStick: Then you'll probably have to figure things out yourself. :)
03:25:52 * Arafangion is biased towards debian.
03:26:11 <TheStick> ...does anyone know which libs I'm gonna need to see if I can't snag 'em from somewhere...?
03:26:48 <TheStick> ...oh, yeah. That would help...;
03:26:56 * TheStick is obviously flying blind here...;
03:27:45 <glx> of course this page is not for dedicated servers
03:28:10 <Arafangion> And SDL isn't neccessarily something that depends on X122
03:28:35 <glx> sdl usually requires a window manager
03:29:27 <glx> basically you start with ./configure --enable-dedicated
03:30:37 <TheStick> Error: configure: error: no liblzo2 detected
03:30:50 <glx> and the important libs for a server are zlib, liblzma and liblzo
03:31:02 <TheStick> ...let's see if I can't find that anywhere...
03:32:14 <TheStick> ...uhm. What's the typical extension of a lib that Linux can run...?
03:33:02 <glx> you need the dev versions
03:33:21 <glx> and the corresponding headers
03:33:57 <TheStick> (I'm gonna use the 1.1.5 version of OpenTTD, do I really need the dev version?)
03:34:32 <glx> you always need the dev version of libs to compile
03:34:54 <Rhamphoryncus> The -dev *packages* include what you need to compile a program that uses them
03:34:58 <TheStick> So I'm looking for liblzo2.a, right...?
03:35:27 <Rhamphoryncus> Surely that distro has some sort of package system available
03:37:46 <TheStick> prt-get: no command given. try prt-get help for more information
03:38:40 <Rhamphoryncus> Check their website
03:38:53 <Rhamphoryncus> It's possible to grab upstream tarballs and install yourself, but doing this repeatedly will cause debris to collect and in the long run will be MUCH more work
03:38:59 <Rhamphoryncus> And require much more learning
03:39:06 <glx> seems they have xz and zlib
03:39:18 <glx> but I don't really see dev versions
03:41:59 <TheStick> ...so then where would I need to run prt-get depinst openttd?
03:44:12 <TheStick> ...also, I'm getting a worrying error:
03:44:22 <TheStick> [Config error: can't access /usr/ports/*]
03:44:37 <TheStick> Where * is one of "core", "opt", "contrib"
03:48:05 <TheStick> ...ugh. I gotta go do something. Be right back.
04:08:57 <TheStick> They don't have a liblzo2 port / package.
04:36:14 *** supermop has joined #openttd
04:47:38 <TheStick> Ok, so I tried skipping liblzo2 entirely, since I can't seem to find it...
04:47:45 <TheStick> ...and I'm trying to run the Makefile...
04:48:02 <TheStick> ...and I'm smacked with Syntax erros.
04:50:39 <TheStick> Is everyone here simply idling...?
04:57:48 *** Markavian` has joined #openttd
04:58:10 <supermop> most of the people are in europe, so they are asleep
04:58:50 <TheStick> I'm trying to compile OpenTTD, dedicated-mode, on a Linux Box.
04:59:08 <TheStick> ...unfortunately, I'm getting lovely syntax errors.
05:00:57 <TheStick> ...any possibility I could get a pinch of assistance...?
05:02:14 <TheStick> (Running a CRUX distro, btw.)
05:02:16 <supermop> i dont know anything about it, but i'd suggest posting in the forum, and waiting until morning
05:02:51 <supermop> or seeing if someone has written something about it in the development subforum
05:03:09 <TheStick> ...but, then... if it's a syntax error thing, then wouldn't that be a bug...?
05:03:42 <supermop> i actually dont know really any programming so i am nut sure
05:04:29 *** Markavian has joined #openttd
05:35:31 <Rhamphoryncus> TheStick: methinks you'd be better off installing a different distro
05:35:56 <TheStick> Is there anything that can run console-only...?
05:36:28 <Rhamphoryncus> There's swaths of linux servers quietly powering the world
05:36:45 <Rhamphoryncus> And probably ever major distro has a server version
05:37:07 <TheStick> ...course, I picked CRUX since it (supposedly) has a small memroy footprint, leaving more for Minecraft...
05:37:43 <TheStick> I've only got 0.75 MiB to work with...;
05:38:03 <Arafangion> TheStick: You're not getting linux on that, then.
05:38:03 <Rhamphoryncus> of ram? Not even a meg?
05:38:16 <Rhamphoryncus> Well you CAN, but it won't be any easy, major distro
05:38:28 <Arafangion> Rhamphoryncus: Not sure you could, even, with tomsrtbt.
05:38:30 <Rhamphoryncus> that's not so bad
05:38:41 <TheStick> Screwed that one up.
05:40:25 <Rhamphoryncus> tomsrtbt is designed for small disks, not small ram
05:41:28 <Arafangion> Rhamphoryncus: Even so, it's a 2.0.x kernel.
05:41:47 <Arafangion> Rhamphoryncus: And the whole kernel has to be put into RAM.
05:42:37 <Rhamphoryncus> TheStick: Either way I know debian or ubuntu server versions would run in that
05:42:47 <Rhamphoryncus> I'd be far more concerned about openttd itself than linux
05:43:21 <Rhamphoryncus> Not having X, gnome/kde, a graphical browser, or any of that makes a HUGE difference
05:43:45 <TheStick> ...anyways, I found a syntax error with the Makefile...;
05:46:32 <TheStick> ...and around how much would OpenTTD want...?
05:47:43 <Arafangion> TheStick: Considerably more. :)
05:47:56 <Arafangion> Also, linux needs more than 2MB as well.
05:48:17 <TheStick> ...would I be able to run OpenTTD with 0.75 GiB of RAM?
05:49:27 <Arafangion> Referring to RAM as "0.75 GiB of RAM" is an exceededly odd way of saying it, mind you.
05:49:45 <TheStick> Just trying to be specific...;
05:52:02 <Arafangion> TheStick: Do keep in mind that measuring RAM precisely is more involved than you realise.
06:13:39 *** kkb110__ has joined #openttd
06:18:09 *** kkb110_ has joined #openttd
06:57:31 *** andythenorth has joined #openttd
07:01:57 *** theholyduck has joined #openttd
07:03:41 *** sla_ro|master has joined #openttd
07:06:23 *** Markavian has joined #openttd
07:16:51 *** andythenorth has joined #openttd
07:33:10 *** andythenorth has joined #openttd
07:43:49 <Rhamphoryncus> Hmm, I think I've put together a robust timetabling algorithm. Anybody interested in the details?
08:02:14 *** cmircea has joined #openttd
08:27:36 *** Firartix has joined #openttd
08:28:13 *** Chris_Booth has joined #openttd
08:28:57 *** Devroush has joined #openttd
08:34:36 *** Progman has joined #openttd
08:37:26 *** DayDreamer has joined #openttd
08:40:06 *** lmergen has joined #openttd
08:44:01 *** TWerkhoven has joined #openttd
08:46:03 *** tokai|mdlx has joined #openttd
08:52:16 *** ChanServ sets mode: +v tokai
08:55:41 *** tokai|noir has joined #openttd
08:55:41 *** ChanServ sets mode: +v tokai|noir
09:00:43 *** tokai|mdlx has joined #openttd
09:13:13 *** tokai|noir has joined #openttd
09:13:13 *** ChanServ sets mode: +v tokai|noir
09:20:12 *** Alberth has joined #openttd
09:20:12 *** ChanServ sets mode: +o Alberth
09:21:58 <Rhamphoryncus> Can anybody give me some quick pointers on how to save/load a byte array? Technically I could get away with 2 or 3 bits for each value
09:23:10 <Rhamphoryncus> Hmm, SL_ARR looks promising
09:23:26 <TinoDidriksen> Then you can define that you're using 3 per value and just write that out.
09:24:44 <TinoDidriksen> But it's a lot of code to pack the bits. Much easier to just dump a byte array to a stream.
09:25:28 <Rhamphoryncus> Yeah, perfectly easy for me to treat it as nibbles
09:25:47 <Rhamphoryncus> Or.. a uint32 array and 3-bits 10 times
09:26:54 <TinoDidriksen> That's going to be some fiddling code to work with...
09:27:20 <TinoDidriksen> I'd go with nibbles if space is that needed.
09:27:21 <Rhamphoryncus> Nah, I'll just throw together a macro to do bitshifting for get/set
09:27:50 <TinoDidriksen> Macro, ew... inline template, ftw.
09:28:02 <Rhamphoryncus> heh. Not that proficient in C++
09:28:56 <andythenorth> what's wrong with macros? :)
09:29:05 <andythenorth> genuine question
09:29:24 <Rhamphoryncus> A C macro is purely a textual substition
09:29:36 <TinoDidriksen> They're unsafe. inline templates are type checked at compile time; macros can be abused.
09:29:56 <Rhamphoryncus> So you can have something like this: #define FOO &#(*#}J{ #*$#(
09:30:01 <TinoDidriksen> Same reason you should prefer const globals instead of #define constants.
09:30:56 <Rhamphoryncus> constants aren't nearly as bad.. but I can still see a const global being better
09:42:36 <Rhamphoryncus> TinoDidriksen: so.. could you actually give me a hint?
09:43:12 <TinoDidriksen> Sounded like you had it under control in your own way...
09:43:27 <Rhamphoryncus> I'm using grep and browsing, trying to learn saveload
09:44:10 <Rhamphoryncus> I'm sure it does arrays, I've even found one or two mentions, but I haven't found an example yet
09:46:32 <Terkhen> Rhamphoryncus: I suggest checking trunk revisions that made changes to saveload to get an idea of how it works
09:46:42 <Terkhen> it is a bit confusing :)
09:47:36 <Rhamphoryncus> That's an understatement
09:51:16 <Terkhen> I understood it briefly while creating the persistent storage pool codechanges, now I barely remember it :P
09:51:35 * Rhamphoryncus finds openttd.svg and finds out that it contains.. vectorized pixel art.
10:01:53 * andythenorth wonders if nmlc -> nfo -> grfcodec -> grf is faster than nmlc -> grf ?
10:03:48 <Rhamphoryncus> Hmm.. SlArray.. Load_ORDL, Save_ORDL. More promising stuff.
10:05:45 <Terkhen> andythenorth: probably not
10:06:05 <Rubidium> Terkhen: it actually might be
10:06:43 <Rubidium> nml tries both compression formats, grfcodec only the specified one
10:09:15 * andythenorth wonders how to test
10:11:31 <Rhamphoryncus> what does SLE stand for? Save load encoding? Save load entry?
10:11:38 *** sla_ro|master has joined #openttd
10:14:09 <Alberth> and do that several times, to eliminate disk caching effects
10:15:02 *** JVassie has joined #openttd
10:15:32 *** TGYoshi has joined #openttd
10:15:40 <Rubidium> or use a ramdisk and disable swap
10:15:44 *** cmircea has joined #openttd
10:15:44 *** supermop has joined #openttd
10:15:44 *** Belugas has joined #openttd
10:15:44 *** Arafangion has joined #openttd
10:15:44 *** ccfreak2k has joined #openttd
10:15:44 *** oxygen.oftc.net sets mode: +o Belugas
10:15:51 *** ChanServ sets mode: +v Belugas
10:18:33 *** peteris has joined #openttd
10:19:25 <andythenorth> grfcodec will barf on my directory structure :|
10:19:32 * andythenorth temporarily hacks that
10:20:35 <Alberth> always nice, programs that enforce a particular directory structure :p
10:22:49 <andythenorth> ok, averages, rounding down because it's not that significant:
10:23:18 <andythenorth> nmlc -> grf: 13s
10:23:46 <andythenorth> grfcodec -> grf: 0.3s (oops, can't round that down) :O
10:24:03 * andythenorth wonders how to edit makefile to change this :)
10:24:29 <andythenorth> silly old grfcodec wants my nfo in sprites/
10:24:42 <andythenorth> but that's not tragic
10:25:11 * andythenorth still hasn't figured out makefile editing yet
10:25:36 <andythenorth> and I'm not replacing it with python - makefile has three years work and all those targets and edge cases taken care of...
10:26:15 * andythenorth ponders shell script
10:26:40 <andythenorth> maybe I just tell the makefile it's an nfo project :)
10:27:14 <Terkhen> yeah, that should work
10:29:22 *** tokai|mdlx has joined #openttd
10:32:53 <CIA-1> OpenTTD: rubidium * r23932 /trunk/ (17 files in 7 dirs): -Codechange: split the NewGRF text window into its own source files
10:34:29 <andythenorth> bash alias works for me
10:34:37 <andythenorth> alias makebandit='nmlc bandit.nml --nfo=sprites/bandit.nfo && grfcodec -e bandit.grf && mv bandit.grf /Users/andy/Documents/OpenTTD/data'
10:34:57 <andythenorth> leave the makefile alone
10:35:02 *** ChanServ sets mode: +v tokai
10:35:03 <andythenorth> don't break what I can't fix :)
10:35:17 <CIA-1> OpenTTD: rubidium * r23933 /trunk/src/ (lang/english.txt newgrf_gui.cpp): -Codechange: make the text file window strings more generic (LordAro)
10:35:55 *** JVassie_ has joined #openttd
10:36:17 <Alberth> andythenorth: I often put such sequences in a real script file, less likely to get lost :)
10:36:36 * andythenorth never read the manual on how to do that
10:37:13 *** lmergen has joined #openttd
10:38:24 *** lmergen has joined #openttd
10:39:54 *** tokai|mdlx has joined #openttd
10:41:03 <Alberth> put that in a file (makebandit or so), chmod u+x makebandit and run it with ./makebandit
10:42:10 *** theholyduck has joined #openttd
10:42:52 <CIA-1> OpenTTD: rubidium * r23934 /trunk/src/lang/ (43 files in 2 dirs): -Update: other language files w.r.t. the previous change
10:47:26 <andythenorth> 8s less frustration :)
10:51:22 <Alberth> you're working around the natural barrier which is designed to keep newgrfs small :p
10:51:44 <CIA-1> OpenTTD: rubidium * r23935 /trunk/src/ (newgrf_config.cpp textfile_gui.cpp textfile_gui.h): -Codechange: generalise GetTextfile
10:54:16 <Zuu> In one network timeout string the computer of a user is blamed that it takes too long to join. In another the user is blamed because he/she took too long to download the map. I can understand that we blame a user to be too slow to type the password, but shouldn't his/her computer be blamed for slow download speed rather than the user him/herself?
10:54:39 <Zuu> STR_NETWORK_ERROR_TIMEOUT_MAP <-- string that could be changed from user to computer.
10:55:14 *** TWerkhoven has joined #openttd
10:56:25 <Rhamphoryncus> Zuu: I agree. I'd go with "Client took too long to download the map"
10:58:20 <CIA-1> OpenTTD: rubidium * r23936 /trunk/src/ (7 files in 6 dirs): -Feature [FS#5047]: readme/licence/changelog viewer for AI and game scripts (LordAro)
11:01:19 <Zuu> Yep that is what I had in mind.
11:02:03 <Zuu> You could of course starting to say that his ISP is to slow etc. but that is taking things too far as the slowness of download can't easily be told what the cause is.
11:02:39 *** Chris_Booth_ has joined #openttd
11:07:07 *** Chris_Booth_ is now known as Chris_Booth
11:10:54 <andythenorth> Alberth: CETS compile time is ~4 mins :D
11:11:24 <Alberth> good for making some coffee :)
11:19:37 *** tokai|noir has joined #openttd
11:19:37 *** ChanServ sets mode: +v tokai|noir
11:20:39 <Rubidium> with grf container version 2 you might do some tricks to prevent nml from re-encoding the graphics each time. Then it might become significantly faster, although dep checking will be a nightmare
11:21:24 <andythenorth> dep checks seem to be a headache
11:28:15 <andythenorth> anyone ever written procedural graphics generator?
11:28:24 * andythenorth is reluctant to draw
11:28:32 <andythenorth> anyone / anyone here /s
11:29:37 <Alberth> all renderer programs?
11:30:09 <andythenorth> this would be pure 2d bitmap composition
11:30:40 <Rhamphoryncus> andythenorth: I can't help but think that's a bad idea
11:30:44 <andythenorth> with alpha / mask / genlock support
11:30:51 <andythenorth> drawing is boring
11:30:55 <andythenorth> writing code is interesting
11:31:14 <Rhamphoryncus> But you're not going to end up with much variety doing that
11:31:15 <andythenorth> if I draw truck components, does anyone want to lay them out according to simple rules?
11:31:26 * andythenorth considers amazon mechanical turk
11:31:41 <Rhamphoryncus> You'll be like the face configurator thing
11:32:06 <andythenorth> trucks are pretty vanilla
11:33:21 <Rhamphoryncus> Mix-and-match trailers I could see
11:33:57 *** frosch123 has joined #openttd
11:34:23 <Rhamphoryncus> How many distinct cabs do you actually see on there?
11:34:31 <andythenorth> there are 4 or so
11:35:00 <andythenorth> the actual graphics will need maybe 12 distinct cab styles
11:35:53 <andythenorth> they're on a standard grid of 4px units
11:36:01 <andythenorth> so composing them would be possible - in theory
11:36:08 <Rhamphoryncus> Ignoring the wheels then yeah, it's 4
11:36:18 <andythenorth> wheels can be comped on top
11:36:46 <Rhamphoryncus> You'd honestly spend more time programming than you would doing it in gimp
11:36:51 <Rhamphoryncus> and you'd get a better result in gimp
11:37:25 <andythenorth> the sad time comes when I improve a truck cab
11:37:35 <andythenorth> then have to copy it into n photoshop files
11:37:48 <andythenorth> maintaining graphics is a PITA
11:38:01 <andythenorth> I could actually do this with Adobe's Flash IDE
11:38:13 <andythenorth> which supports symbols that can be comped
11:38:29 <andythenorth> and easy procedural positioning of sprites
11:38:38 <andythenorth> but the flash IDE sucks
11:38:41 <andythenorth> that's my old life
11:38:53 <andythenorth> After Effects can do comps of symbols too
11:39:31 <CIA-1> OpenTTD: rubidium * r23937 /trunk/src/lang/english.txt: -Change: improve the wording of some of the timeout related "error" messages
11:41:57 <andythenorth> photoshop smart objects
11:42:01 <andythenorth> instances of symbols
11:51:43 <andythenorth> that's not very future proof
12:02:45 *** Shiners has joined #openttd
12:26:26 *** Phoenix_the_II has joined #openttd
12:28:10 *** lmergen has joined #openttd
12:39:46 * andythenorth figures out a better way to handle calculating run cost etc
12:40:07 <andythenorth> if no value is set, calculate the value
12:40:13 <andythenorth> if a value is set, use the value
12:40:27 <andythenorth> no magic happens to a value that is manually set
12:40:35 <andythenorth> hence magic xOR no-magic
12:40:57 <Rhamphoryncus> Sanity is a rare commodity ;)
12:41:14 *** micasousa99 has joined #openttd
13:48:35 *** micasousa99 has joined #openttd
14:12:41 <andythenorth> defining a function that's only called once per class...
14:12:49 <andythenorth> ...seems tidy, but actually makes for worse code?
14:13:05 <andythenorth> more pointless indirection, harder to read
14:14:14 <Rhamphoryncus> not necessarily
14:14:21 <Zuu> If you can abstract away some code by puting it in a function with a good descriptive name, the code that calls the function could become easier to read.
14:14:25 <Rhamphoryncus> It's a question of abstraction
14:14:44 <Rhamphoryncus> Potentially MUCH easier to read
14:15:21 <andythenorth> could be set_graphics_file
14:15:33 <andythenorth> similar to modify_capacities_fifth_wheel_trucks
14:15:51 <andythenorth> seems over-engineered to use functions for these though
14:16:00 <andythenorth> even though I personally find it much easier to read
14:16:15 <Rhamphoryncus> yeah.. it's only a small dent but I think I'd do it too
14:16:15 <Zuu> 140? the paste doesn't contain that many lines.
14:17:05 <andythenorth> when setting props, I don't really want to read "if foo, blah, else: other blah"
14:17:18 <andythenorth> I hate if-else blocks in the middle of otherwise simple code
14:17:48 <Zuu> Sounds like you have good arguments for defining a function.
14:18:00 <andythenorth> except....violates the 'do it once, do it long hand'
14:18:08 <andythenorth> 'do it twice, write a function'
14:20:04 *** Wolf01 is now known as Guest2341
14:20:05 *** Wolf03 is now known as Wolf01
14:30:56 *** kkb110_ has joined #openttd
14:35:34 * andythenorth did the functions anyway :)
14:37:59 *** sla_ro|vista has joined #openttd
14:40:11 <Rhamphoryncus> what's this "do it once, do it long hand" you speak of?
14:40:40 <Rhamphoryncus> IMO, it's a myth. Functions are NOT to prevent duplicate code. They're to provide abstraction
14:40:58 <Rhamphoryncus> goto is to prevent duplicate code ;)
14:45:27 <Rubidium> abstractions prevent duplicate code
14:46:57 <Rhamphoryncus> Abstractions can reduce code size. Duplicate code is a consequence, not a cause
14:47:06 <andythenorth> functions move your code away
14:47:16 <andythenorth> that can be bad when you're trying to figure out 'wtf?'
14:47:37 <andythenorth> function in another file calling another function in another file is a PITA
14:47:42 <Rhamphoryncus> Can be. Abstraction isn't a silver bullet
14:47:47 <andythenorth> lots of project-grep = dull
14:48:03 <andythenorth> ideally all the code you need *right now* fits in the viewport
14:48:19 <andythenorth> if it doesn't fit in the viewport, reduce the scope of your feature :)
14:48:27 * andythenorth has a 13" screen
14:48:36 <andythenorth> other people I work with have 22" dual screen
14:48:47 <Rhamphoryncus> But the reality is I use grep as my IDE ;)
14:49:03 <andythenorth> viewport may be a flawed measurement :)
14:49:19 <andythenorth> ho. a new truck graphic
14:49:27 * andythenorth has started actually...drawing
14:54:42 *** sla_ro|master has joined #openttd
14:54:49 * andythenorth ponders curling bandit.cfg from web
14:58:31 <andythenorth> slows down building
14:59:16 *** lofejndif has joined #openttd
14:59:51 *** JVassie has joined #openttd
15:13:01 *** FLHerne has joined #openttd
15:35:27 *** Biolunar has joined #openttd
15:49:08 *** valhallasw has joined #openttd
16:36:22 *** Chris_Booth has joined #openttd
17:04:38 * Rhamphoryncus lurks behind andythenorth
17:29:59 <CIA-1> OpenTTD: smatz * r23938 /trunk/src/script/script_scanner.hpp: -Codechange: declare ScriptScanner::Initialize() and make it abstract, make the other overloaded Initialize() protected
17:45:43 *** Chris_Booth has joined #openttd
17:52:32 *** kmichael has joined #openttd
17:56:00 <andythenorth> is it exam season?
17:56:07 <andythenorth> can't be the weather luring people out doors
17:56:26 <andythenorth> maybe the european interwebs has iced over
17:57:49 <Rubidium> andythenorth: well, the weather is quite nice for ice skating
17:58:56 <Rubidium> though the last week's weather was very unpredictable (according to the once running the trains)
17:59:24 <andythenorth> anybody want to reimplement eGRVTS in nml? I think it's about 1 day's work + 1 day finding mistakes :P
18:00:03 <Rubidium> i.e. 70-80% of the time the sun could shine last week it did shine, no rain, relatively stable temperatures...
18:01:53 * Rubidium wonders how long it takes until the first real 32bpp or ez NewGRF is released
18:02:15 <andythenorth> eGRVTS would be a good candidate, zeph has EZ sprites for a number of vehicles
18:02:33 <andythenorth> I have a truck set [nearly] framework
18:03:12 <andythenorth> which I am itching to test on some other hapless coder
18:03:40 <andythenorth> as that will uncover strangeness in my thinking
18:14:20 <Hirundo> Is there a way to extract all info (at least action0) from eGRVTS into some tracking table?
18:16:52 <Hirundo> Then the conversion to NML would be quite doable, but manually setting a few dozen properties for more than a few dozen vehicles would feel like a fabulous waste of time
18:17:48 <Hirundo> and then there's still a lot of sprites and loading stages, articulation, etc to fix
18:18:12 <andythenorth> Hirundo: zeph was pretty good about templating his sprites
18:18:25 <andythenorth> eGRVTS is very...modular
18:18:42 <Hirundo> sprites yes, but acc. his statements he writes his NFO in plain hex
18:19:11 <andythenorth> if $someone could parse the nfo into a sane data structure...I could write the other side to parse it into a python config file
18:20:05 <andythenorth> we might never write any nml at all :o
18:20:05 <andythenorth> dict per vehicle would be favourite, or JSON or similar
18:20:11 <andythenorth> or straight into config file
18:21:00 <andythenorth> that's why I haven't maintenance-patched it for various bugs
18:21:12 <andythenorth> I am not going to spend time with my finger on the screen counting bytes
18:21:18 <Hirundo> I'm looking into grf2html output now
18:21:26 <andythenorth> FIRS nfo->nml conversion script was pretty robust
18:23:37 *** JVassie has joined #openttd
18:24:57 <andythenorth> could parse the dom of the grf2html output
18:26:03 *** andythenorth is now known as Guest2364
18:26:04 *** andythenorth has joined #openttd
18:30:24 <andythenorth> I'm quite interested in learning to write python config files as well as read them
18:30:33 <andythenorth> so I would be happy to figure that out
18:31:28 <Hirundo> as far as i can see, the nfo is quite easy
18:31:43 <Hirundo> there's no action 6/7/9/D
18:31:58 <Hirundo> only actions 0-4, 8 and one action 11 (sound)
18:34:19 <andythenorth> plausible conversion then
18:34:31 <Hirundo> varaction2 are mostly for articulation only
18:35:11 <Hirundo> the few graphics switches I could find are for trailers, which seem to share IDs
18:35:32 <andythenorth> I kind of know the eGRVTS codebase, as we copied it for HEQS
18:35:38 <andythenorth> zeph coded v1 of HEQS
18:35:40 <andythenorth> then I took over
18:35:45 <andythenorth> I commented it ;P
18:36:05 <andythenorth> and added formatting :P
18:37:34 <andythenorth> I don't share trailer IDs anymore
18:37:39 <andythenorth> we have plenty spare ;)
18:39:44 <Hirundo> No wonder, I tended to get lost in the P-list
18:39:58 <andythenorth> I wouldn't mind manually redoing all the properties
18:40:02 <andythenorth> tedious, but not hard
18:41:12 * Hirundo would mind tediousness
18:41:29 <andythenorth> but I really don't want to sort out the pngs + realsprites
18:41:31 <andythenorth> I hate doing that
18:41:56 <andythenorth> i.e. rebuilding the templated offsets etc
18:42:25 <andythenorth> BANDIT codebase uses the NML template system btw
18:43:09 *** sla_ro|master has joined #openttd
18:43:17 <Hirundo> you could ask zeph, if he has any useful template info for the sprites
18:43:37 *** supermop has joined #openttd
18:44:24 <Hirundo> then some python / PIL script should work
18:44:34 <andythenorth> I could also ask him if he minds the conversion :)
18:44:40 <andythenorth> if he likes nfo, this might annoy him
18:45:03 <Hirundo> sounds like a good thing (tm)
18:45:06 <andythenorth> could be the first grfcontainer2 grf though?
18:45:18 <andythenorth> or the first with EZ anyway
18:45:45 <Hirundo> I should be working on NML
18:46:04 <CIA-1> OpenTTD: translators * r23939 /trunk/src/lang/ (8 files): (log message trimmed)
18:46:04 <CIA-1> OpenTTD: -Update from WebTranslator v3.0:
18:46:04 <CIA-1> OpenTTD: croatian - 4 changes by VoyagerOne
18:46:04 <CIA-1> OpenTTD: english_US - 4 changes by Rubidium
18:46:04 <CIA-1> OpenTTD: finnish - 2 changes by jpx_
18:46:06 <CIA-1> OpenTTD: german - 1 changes by NG
18:46:06 <CIA-1> OpenTTD: italian - 1 changes by lorenzodv
18:46:17 <Hirundo> instead of discussing, which tends to *not* produce working grfs out of thin air
18:46:32 <Hirundo> Fixing NML doesn't either, but might help ;-)
18:46:39 * andythenorth finds monologue necessary prelude to coding ;)
18:46:45 <andythenorth> dialogue is even better
18:59:59 <Elukka> what station set is that on the left?
19:02:56 *** andythenorth has joined #openttd
19:09:35 *** HerzogDeXtEr has joined #openttd
19:45:51 *** lmergen has joined #openttd
19:46:50 <CIA-1> OpenTTD: smatz * r23940 /trunk/src/ (3 files in 2 dirs): -Codechange: remove superfluous semicolons
19:54:28 *** lofejndif has joined #openttd
20:32:44 <Hirundo> Yexo: Doing #3571 could cause grf breakage in some corner cases, is that acceptable?
20:33:33 <Hirundo> 'corner cases' being AFAIK only grfs with vehicles that have capacity but are not refittable
20:47:57 *** amews_aj has joined #openttd
20:48:59 <amews_aj> Hi, After upgrading 1.1.3 -> 1.1.5, the fonts are all weird. Tried resetting cfg file, but still fonts have changed. Wrong size, nonbold etc. Is this on purpose, or is something wrong?
20:50:30 <SmatZ> does that happen when you switch the game to English?
20:50:58 <SmatZ> what could have changed, is that the translation in 1.1.5 now includes some characters that are not included in the base font
20:51:50 <amews_aj> where do I change language?
20:52:27 <SmatZ> (not Advanced Settings)
20:52:59 <amews_aj> Can't I force a the old look on the translated version ?
20:53:42 <Rubidium> maybe an update of OpenGFX helps
20:54:08 <amews_aj> Rubidium, when installing 1.1.5 I chose to download opengfx
20:54:14 <amews_aj> shouldn't that cause it to be updated?
20:55:25 <SmatZ> in trunk, Danish looks fine... but it might be because of FS#5055
20:55:53 <SmatZ> which added letters like "OE"
20:56:17 <Rubidium> SmatZ: only to the original
20:56:19 <amews_aj> I think the letter "" was included in translations of previous versions as well
20:57:06 <Rubidium> amews_aj: is 1.1.4 okay?
20:59:45 <planetmaker> Hirundo: the question is: what's a good default cargo type?
21:00:01 <planetmaker> or should that rather be done explicitly by the player? IMHO yes
21:00:11 <planetmaker> then probably as a separate "property"
21:00:24 <planetmaker> though that's... difficult. It'll have to be a list property
21:00:29 <Rubidium> amews_aj: not needed, no different langauge updates there
21:00:43 <Rubidium> amews_aj: does 1.1.3 (with opengfx 0.4.2) still show the bold characters?
21:00:44 <Hirundo> planetmaker: why? you can specify only one, right?
21:00:45 <amews_aj> Well I just did, same thing
21:00:52 <amews_aj> Rubidium, trying that next
21:01:06 <planetmaker> Hirundo: yes-ish. But what if that cargo doesn't exist?
21:02:19 <Hirundo> planetmaker: then it picks first refittable, if possible using the order in the CTT
21:02:45 <SmatZ> hmm 1.1 looks fine for me @ Danish
21:02:52 <Hirundo> unless the vehicle is not refittable at all, in which case the default cargo is the only cargo and refitting is not possible (like the default vehs)
21:02:58 <Rubidium> SmatZ: which opengfx are you using?
21:02:59 <amews_aj> Rubidium, no, 1.1.3 is also bad now
21:03:52 <__ln__> Danish shouldn't be using any unusual characters (non-Latin-1) so it's a bit weird.
21:03:53 <SmatZ> Rubidium: ok, not with OpenGFX
21:04:07 <SmatZ> it looks fine with original graphics, but not with opengfx ...
21:04:42 <amews_aj> so what to do to solve it? Bugfix? Something I can do temporarily ?
21:04:50 *** Chris_Booth_ has joined #openttd
21:04:50 <Rubidium> amews_aj: did you ever use the graphics of Transport Tycoon Deluxe?
21:05:34 <amews_aj> I don't have the game, and using it without a copy of the game would not be legal I suppose
21:06:01 <Rubidium> amews_aj: section 9 of the readme has a short bit about how to override the font
21:06:46 <Rubidium> planetmaker: amews_aj "wants" to report a regression in OpenGFX w.r.t. glyphs gone missing for Danish in OpenTTD 1.1.x
21:06:56 <amews_aj> Rubidium, I did that, but I cannot find the default fonts
21:07:31 <Rubidium> amews_aj: the "default" font is in opengfx, but when characters (glyphs) are missing it asks the operating system for the best font
21:08:48 <amews_aj> Rubidium, so I can't get the "default" font back by manual override? It's not a standard font ?
21:08:52 <Rubidium> planetmaker: opengfx 0.4.2, Danish translation: in OpenTTD 1.1.x misses glyphs, in 1.2.x is does miss glyphs
21:09:14 <planetmaker> in both versions? hm
21:09:28 <amews_aj> Rubidium, Would it work downgrading opengfx, or only if I downgrade to openttd 1.1.3 as well?
21:09:32 <Rubidium> planetmaker: in 1.2.x it does NOT miss glyphs ;)
21:09:45 <Rubidium> amews_aj: you could downgrade opengfx
21:10:02 <planetmaker> it might as well be that I made the checks to strict.
21:10:11 <Rubidium> but you need to manually find the older version
21:10:24 <amews_aj> Rubidium, not on a build server ?
21:10:33 *** Chris_Booth_ is now known as Chris_Booth
21:10:35 <planetmaker> amews_aj:: Do you know which OpenGFX you used before?
21:10:56 <Rubidium> amews_aj: yes, but using the 1.1.3 installer will not get the older opengfx
21:11:01 <amews_aj> the one that was installed when openttd 1.1.3 was out
21:11:09 <Rubidium> planetmaker: I'm assuming it's monospace related
21:11:38 <planetmaker> well... I specifically rewrote that part already once ;-)
21:11:49 <planetmaker> and have a special section for monospace
21:12:01 <planetmaker> which simply is not included for openttd 1.1.x
21:12:14 <amews_aj> Rubidium, Ok. Maybe I'll just go with english UI at the moment.
21:12:27 <planetmaker> amews_aj: in order to fix: which strings do not work anymore?
21:12:45 <amews_aj> planetmaker, well all of them. The strings are there, but with wrong font
21:12:45 <planetmaker> I *need* something to pinpoint in order to address this bug
21:13:01 <planetmaker> what is "wrong" font?
21:13:19 <amews_aj> Don't know what font it is... Want a screenshot?
21:13:23 <planetmaker> can you show me (screenshot)?
21:13:43 <planetmaker> post them to imagebin.com
21:17:11 <planetmaker> and... that's how it should or should not look?
21:17:19 <planetmaker> (It's not the game-supplied font)
21:17:36 <CIA-1> OpenTTD: smatz * r23941 /trunk/ (7 files in 5 dirs): -Add: support for clang
21:17:48 <Rubidium> planetmaker: that's showing the opengfx font isn't chosen
21:17:55 <Rubidium> whereas in 1.2.0 it is chosen
21:18:11 <Rubidium> and according to amews_aj the opengfx font was chosen in an earlier version of opengfx as well
21:19:17 <Rubidium> but I'll try bisecting
21:21:49 <amews_aj> Rubidium, planetmaker, Can a bugfix be expected anytime soon ?
21:22:15 <SmatZ> amews_aj: 1.2.0-beta4 should work for you ;)
21:22:30 <amews_aj> SmatZ, But I cannot play on a lot of servers with 1.2.0
21:22:43 <Rubidium> amews_aj: depends on what you think is soon
21:23:21 <Rubidium> if soon is in astrological or paleotological time scales, then yes
21:23:26 <planetmaker> amews_aj: with 1.2.0 you cannot play on any :-P
21:23:52 <amews_aj> Of course :) But any guesses? (I won't complain if it isn't correct)
21:24:00 <SmatZ> #openttdcoop.stable allows you to play beta4 ;)
21:24:29 <planetmaker> amews_aj: you'll probably have to estimate my time budget
21:24:35 <planetmaker> Even I myself fail at that
21:24:44 <Rubidium> darn it... opengfx doesn't compile anymore with my image boundary patch :(
21:25:07 <planetmaker> there will be one before April
21:25:27 <amews_aj> I am a developer myself. Is it an easy fix that I will be able to do myself ?
21:26:00 <planetmaker> the difficult thing here probably is building OpenGFX. And knowing the NewGRF language
21:26:04 <Rubidium> depends on what the bug is
21:26:11 <amews_aj> oh, it's a custom language?
21:26:13 <planetmaker> Technically it's probably not that complicated. But what Rubidium says
21:26:26 <planetmaker> It first needs to be found where it is
21:26:28 <amews_aj> Rubidium, just sounded like you thought to know what caused the bug ?
21:27:35 <Rubidium> my statement was in the order of: "last time this car was in the garage you did mess with the motor management, so the malfunction of the motor might have to do with that"
21:28:11 <Rubidium> and bisecting isn't going as I hoped
21:28:26 <Rubidium> hit already two non-working revs
21:28:46 <amews_aj> Oh well, english is fine. Is there any release plan for 1.2.0 yet? I mean, maybe it's not worth looking for the bug in 1.1.5 now ?
21:29:04 <Rubidium> amews_aj: it's not a bug in OpenTTD
21:29:24 <Rubidium> it's a bug in OpenGFX, which is a separate project with a separate release schedule
21:29:33 <Yexo> amews_aj: there is no official release plan, but look at the release dates for 1.1 and 1.0 to get some idea
21:29:38 <amews_aj> Rubidium, well no, but it's a combination. Works with 1.2.0 you said
21:30:10 <Rubidium> then ~7-ish weeks if nothing major comes up
21:31:12 <planetmaker> That looks disgusting, SmatZ :-(
21:31:15 <Rubidium> SmatZ: look at fontcache.cpp
21:31:26 <Rubidium> I've got the feeling the font is bad though
21:31:49 <amews_aj> I noticed the "X"'s on the windows are misplaced as well here
21:32:03 <amews_aj> but maybe just due to wrong font
21:32:41 <Rubidium> SmatZ: I guess it's r23038
21:33:16 <DorpsGek> SmatZ: Commit by peter1138 :: r23038 trunk/src/fontcache.cpp (2011-10-18 17:57:42 UTC)
21:33:17 <DorpsGek> SmatZ: -Fix: Check that the selected font size is valid the font face in use and choose the nearest size to that selected if not. Font metrics should then just work.
21:35:55 <SmatZ> Rubidium: indeed, reverting r23038 on 1.1 fixes the misalignment
21:39:18 <Rubidium> planetmaker: I suspect 858 (1139-1150)
21:40:25 <planetmaker> Thanks for testing, Rubi. I'll check it this week
21:41:28 <Rubidium> though when bisection is done I'll know if it's really that revision
21:47:10 <Rubidium> it might be that 1.2.0 is smarter about missing glyphs
21:47:50 <planetmaker> IIRC we took out a few special glyphs from the checks
21:49:16 <Rubidium> nope... doesn't seem to be smater
21:53:58 <Rubidium> r23582 'fixes' it in trunk
21:54:38 <Rubidium> so whatever characters were removed due to that in opengfx should remain in the branch for 1.1.x and earlier
22:03:40 <Rubidium> amews_aj: my initial guess about the bug in opengfx was wrong and I fear a proper review to fix the complete bug requires quite some checking. So I doubt it'll be a quick fix
22:05:14 <Rubidium> nevertheless replacing opengfx with 0.3.7 should be a good short term fix
22:06:50 <amews_aj> There are no important bugfixes that I will miss then ?
22:07:30 <Rubidium> depends what you perceive to be important
22:07:49 <amews_aj> Guess I'll just take a look at the changelog
22:08:06 <Rubidium> it's mostly minor graphics issues and improved graphics you'll be missing out
23:08:17 <planetmaker> good night here, too
23:08:43 *** nobody_else has joined #openttd
23:18:31 <Nat_AS> is there any newgrifs that add a tourism industry?
23:19:05 <Nat_AS> factories that just have a high passenger rating and maybe a bit of goods/food/mail
23:24:56 <Hirundo> ECS has 'tourists' as a separate cargo
23:26:27 <Nat_AS> I could understand them as a separate subcategory of passengers. But requiring them to use different cars is silly.
23:29:10 <Nat_AS> it would be cool if the game recognized different classes of passenger.
23:29:20 <Nat_AS> and maybe different classes of goods.
23:29:41 <Nat_AS> they could be stored in the same cars, BUT go to different destinations.
23:50:22 *** Phoenix_the_II has quit IRC
continue to next day ⏵
