Back to Home / #openttd / 2016 / 02 / Prev Day | Next Day
#openttd IRC Logs for 2016-02-20

---Logopened Sat Feb 20 00:00:08 2016
01:05-!-Snail [] has quit [Quit: Snail]
01:54-!-_johannes [] has joined #openttd
02:32-!-andythenorth [] has joined #openttd
02:44-!-ConductorCat [] has joined #openttd
02:54-!-roidal [] has joined #openttd
02:59-!-Alberth [~alberth@2001:981:c6c5:1:be5f:f4ff:feac:e11] has joined #openttd
02:59-!-mode/#openttd [+o Alberth] by ChanServ
03:00-!-AdmiralKew [~pcc31@] has joined #openttd
03:10*andythenorth needs a faster horse
03:10<@Alberth>this a firs bug?
03:11<andythenorth>nah don’t think so
03:12<andythenorth>it’s related something like the removal of gridlines :)
03:13-!-smoke_fumus [~smoke_fum@] has joined #openttd
03:13<@Alberth>just couple more horse for more speed?
03:13<andythenorth>faster compile :P
03:14<@Alberth>ha :)
03:15<andythenorth>my ‘go faster’ approach with incremental compiles is *very* fast if only a few things changed
03:15<andythenorth>but the overhead for a compile from clean is huge
03:19*andythenorth might remove all the nml-nfo stuff
03:19<andythenorth>plain nml, unified compile
03:20<andythenorth>seems to take about 15 seconds per roster, scales linearly with number of rosters
03:21<andythenorth>there is unfinished support for compiling just one roster
03:27-!-Ketsuban [~ketsuban@2a02:c7d:a33e:4100:685d:5ddd:f62c:a680] has quit [Quit: Leaving]
03:28<@Alberth>just disable a few others while testing one
03:28-!-sla_ro|master [slamaster@] has joined #openttd
03:29<andythenorth>with the incremental compile overhead, it’s 2 mins for first compile
03:29<andythenorth>versus 30s without
03:30<andythenorth>need to compile 4 times without mistakes to make it worth it :P
03:30<andythenorth>every time I make a mistake, the 30s compile wins
03:32<@Alberth>so make lots of mistakes? :D
03:32<andythenorth>don’t even need to try :P
03:38-!-sim-al2 [] has quit [Ping timeout: 480 seconds]
03:40-!-AdmiralKewl [~pcc31@] has joined #openttd
03:43-!-andythenorth [] has quit [Quit: andythenorth]
03:46-!-AdmiralKew [~pcc31@] has quit [Ping timeout: 480 seconds]
04:11-!-Progman [] has joined #openttd
04:13-!-TrueBrain_ [] has joined #openttd
04:14-!-TrueBrain is now known as Guest4754
04:14-!-TrueBrain_ is now known as TrueBrain
04:17-!-Guest4754 [] has quit [Ping timeout: 480 seconds]
04:19<_johannes>I'm still on that rail net graph exporter... The difficult thing is that most YAPF function require a train as argument
04:19<_johannes>so, if I'm trying to find the route from B->C, but my train is still at A, I'll need to somehow get the train to B
04:20<_johannes>would you suggest to let the train drive the whole suggested path, or instead manipulating its location each time I need it to be in another station?
04:21-!-Quatroking [] has joined #openttd
04:24<@Alberth>considered implementing your own A*?
04:26<_johannes>Alberth: don't know if that's a good idea :P
04:26<@Alberth>or maybe add a second way to start up the search?
04:26<_johannes>possible, with some modifications in Yapf
04:26<@Alberth>I don't know what properties of the train are used while searching
04:27<@Alberth>well, the puzzle is that speed of existing A* is crucial
04:27<_johannes>one could try to pass a "pseudo train" or similar
04:28<_johannes>with crucial, you mean "bad", or "relevant"?
04:28<@Alberth>making it faster is ok, making it slower is not
04:28<@Alberth>it's the #1 bottleneck in moving vehicles around at the map
04:28<@Alberth>so immediately related to the max number of vehicles in the game
04:29<@Alberth>although I am not convinced the current implementation is the best you can have :p
04:31<@Alberth>the entire template stuff looks like an elaborate way to prevent any compiler optimizations, to me
04:31<@Alberth>at the time, it was no doubt faster, but several years of compiler development have passed since then
04:35<_johannes>you mean virtual functions would increase speed here?
04:36<@Alberth>I don't know, but the current implementation just locks any option out
04:36<@Alberth>compilers tend to get optimized for the common code case
04:37<@Alberth>a class with a base class and virtual functions is quite common :)
04:38<@Alberth>the more convoluted code you hack around things, the less likely it is that the compiler understands what you're doing
04:39<@Alberth>so you totally rely on your own coding ability for optimizing
04:39<@Alberth>even if the author could do that X years ago, it won't be optimal today
04:40<@Alberth>processors, memory, and compilers change continuously
04:44<+michi_cc>_johannes: You need your own replacement for CYapfFollowRailT::ChooseRailTrack anyway, what prevents you from choosing yoour origin differently than the existing implementation?
04:46<sim642_>Alberth, templates don't prevent compiler optimizations because templates generate non-template code in the background which gets optimized just as any other code
04:47<@Alberth>you looked at the yapf code?
04:47<sim642_>furthermore, the optimizations might be different for different types the template is used with to provide even greater level of optimization for speficic use
04:47<sim642_>no, not really
04:47-!-sim642_ is now known as sim642
04:50<sim642>the suboptimal aspect of templates would be the produced binary size if the same template is used with a huge number of different arguments since that causes duplicate functionality code but I'm not sure if it'd cause much performance impact
04:51<_johannes>michi_cc: good point! looks like one can simply pass NULL as the vehicle in FindPath
04:51<_johannes>and yes, the origin and destination could be found otherwise
04:51<+michi_cc>_johannes: I wouldn't do that unless you want to ignore railtypes.
04:52<_johannes>michi_cc: oh, yes, I see it
04:53<_johannes>michi_cc: maybe just replacing all occurences of Vehicle* by something that meets the minimum requirements the pathfinder needs?
04:53<_johannes>like: FindPath(UsedVehicleDate* v)
04:53<roidal>what high-resolution graphic sets would you suggest?
04:54<+michi_cc>What'S wrong simply using the train you are graphing?
04:54<roidal>for landmarks, and vehicles
04:54<_johannes>michi_cc: if my train has the order list a->b->c->b, then to find the track from b to c, I'll need to move it to b first, correct?
04:58<_johannes>ah, you mean: I pass the origin to my st... function, use that to get the origin (instead of getting it from the train), but then still pass the train to FindPath()?
04:58<@Alberth>check how it uses orders?
05:01<_johannes>already tried, but all these function only check the next order
05:02<_johannes>ok, I think I got it...
05:05<roidal>hm, can the basesets GFX use the newGFX format too?
05:08<@Alberth>basesets and newgrf are quite separate worlds, afaik
05:08<@Alberth>but both support high resolution graphics, and full colour
05:09<@Alberth>as well as the original 8bpp art
05:09<roidal>ah, so the baseset-format was change too
05:09<roidal>to provide 32bpp?
05:09<@Alberth>at least for the sets that have such high resolution / full colour grapihcs, that is
05:10<@Alberth>zBase is a baseset with 32bpp, and high resolution
05:10<@Alberth>I don't like it though, too clean to my taste
05:11<@Alberth>afaik there is a 32bpp wiki page
05:11<@Alberth>listing stuff
05:11<@Alberth>no idea how up to date it is
05:16-!-HerzogDeXtEr [] has joined #openttd
05:24-!-oskari89 [] has joined #openttd
05:24-!-andythenorth [] has joined #openttd
05:31-!-Wolf01 [] has joined #openttd
05:51*andythenorth refactors compile :|
05:51*andythenorth probably needs to teach make about deps
05:54<Wolf01>lol... I ordered a display dock from microsoft back in december, shipping in 4-6 weeks, put 1 week for the shipping itself... where's my product? still not shipped, contacted the customer support "we are sorry, we'll ship it in 8-12 weeks", I hope it will come before I'll change the phone when it will become old :(
05:55<Wolf01>back to train driving session
05:57<andythenorth>nmlc takes 26s, but the compile takes 40s :(
05:57*andythenorth must find the slow
05:58<Wolf01>buy a faster pc?
06:00<Wolf01>meh, no steam loco tutorial, had to figure out alone how to not blow up the loco just after the first acceleration
06:05*andythenorth tries all the pythons
06:09<andythenorth>yeah, new computer is probably the solution
06:14-!-Progman [] has quit [Remote host closed the connection]
06:17<argoneus>good morning train friends
06:20-!-Antheus [] has quit [Quit: Bow down to your Lord Antheus]
06:20-!-Antheus [] has joined #openttd
06:35-!-andythenorth [] has quit [Quit: andythenorth]
07:22<_johannes>the real problem about the templates is imo not the runtime
07:22<_johannes>it's that you can't read the code as a human
07:23<_johannes>for example, rail nodes have an attribute named m_key , but where is that m_key actually defined
07:23<_johannes>it's in some base class, which depends on many templates...
07:24-!-^0 [] has joined #openttd
07:25<@Alberth>ag m_key
07:25<_johannes>ag ?
07:26<@Alberth>grep, but tuned for searching source code
07:26<_johannes>for example, yapf_costrail.hpp, line 329... where is that m_key defined? I have no clue...
07:28-!-frosch123 [] has joined #openttd
07:29<@Alberth> is the ag output
07:30<@Alberth>so yapf_node.hpp or yapf_node_rail.hpp
07:30<frosch123>hmm, "ag" is shorter than "~/my<tab>"
07:30-!-Supercheese [] has quit [Read error: Connection reset by peer]
07:31<_johannes>I just installed "ack"
07:31<_johannes>ag is not available for gentoo...
07:31<@Alberth>$ rpm -qf `which ag`
07:31<frosch123>_johannes: i suspect the "a" stans for "alberth"
07:31<@Alberth>nope, it doesn't :)
07:32-!-Supercheese [] has joined #openttd
07:32<frosch123>hmm there is a silversearcher-ag package
07:33<_johannes>for which distro frosch123 ?
07:33<_dp_>"The command name is 33% shorter than ack, and all keys are on the home row!"
07:33<_dp_>who cares about home row probably doesn't use querty...
07:34<Wolf01>fuck, steam locomotives are really a mess, I can't even figure out how I made it moving, I think if I chose the one with full simulation it would blow up in some seconds
07:34<frosch123>_dp_: there are people using german keyboard layout, who claim that vim is any good
07:34<@Alberth>I used ack for a while, but there was an edge case where it didn't do what I wanted
07:35<_johannes>anyways, Alberth , I think it's from yapf_node.hpp, but the type is Tkey_
07:35<@Alberth>template type?
07:35<frosch123>but well, there are also people who do not use a 40 year old arcane keyboard layout, who think that vim using ESC as escape key is any good
07:35<_johannes>yes, and I don'T know where this is from
07:36<@Alberth>yep, reading that code is a challenge :)
07:36<frosch123>oh, it uses colours, how unusual
07:37-!-^0 [] has quit [Quit: Page closed]
07:38<_dp_>frosch123, I know next to nothing about both vim and german layout so no idea what're you trying to say %)
07:39<frosch123>vim uses "/" for searching
07:39<@Alberth>_johannes: bottom of yapf_common.hpp lists the overall template
07:39<frosch123>which is a convenient key on an us layout
07:39<frosch123>but is shift+7 on a german layout, which is the worst for a hotkey
07:39<@Alberth>_johannes: ag CYapfT
07:40<_johannes>Alberth: just saw where it is defined: struct CYapfRail1 : CYapfT<CYapfRail_TypesT<CYapfRail1 , CFollowTrackRail , CRailNodeListTrackDir, CYapfDestinationTileOrStationRailT, CYapfFollowRailT> > {};
07:40<@Alberth>no german keyboard remapping available?
07:40<frosch123>also vim uses ESC as a common key, which was in the position of capslock on the machine vi was designed for; but which is a silly to reach key nowadays
07:40<_dp_>shift+7 doesn't sound too horrible, and I bet you can rebind it if you want
07:41<@Alberth>/me finds esc position very convenient (but I am left-handed :p )
07:41<frosch123>Alberth: i only use vim when there is no other editor available, as such i always use the default configuration
07:41<@Alberth>makes sense
07:42<frosch123>_dp_: i have trouble to press shift+7 with one hand; it's possible, but usuallyi use two
07:42<@Alberth>use ? :p
07:42<frosch123>is that equivalent to / ?
07:42<@Alberth>yes, but up instead of down
07:43<@Alberth>if you use N instead of n, you get reverse match
07:43<@Alberth>and you get mostly the normal search
07:44<frosch123>_dp_: anyway, did you consider alternatives to ScrollCompanyClientsTo? like extending ScriptViewport::ScrollTo to work in ScriptCompanyMode ?
07:45<_dp_>what is ScriptCompanyMode? you mean issuing commands on behalf of a company?
07:45<_dp_>I think it's more clear for it to be issued as deity
07:46<frosch123>the difference is that using the CompanyMode would also allow human clients to issue a command like that
07:46<_dp_>hm, yeah, but do they need to?
07:47<frosch123>well, forcing a scrolling on all clients is already on the level of "does it need to do that" :p
07:47<frosch123>usually there is a goal list or story book, where you can click on something which then triggers the scrolling
07:48<_dp_>also there is other problem actually, it would be better to have command that scrolls clients individually, not whole company
07:48<_dp_>but gs doesn't know a thing about clients
07:48<frosch123>well, i would be annoyed by any automatic scrolling, if i am currently building something
07:49<frosch123>so, i would always prefer some entry in the goal or story guy
07:49<frosch123>why would you want to scroll single clients?
07:49<frosch123>the client who issued something is likely already in the position
07:49<frosch123>you would want to tell the other clients in the company
07:49<_dp_>I'm only going to use it when clients join/start company
07:49<_dp_>to focus on claimed town in cb
07:50<frosch123>ah, on joining
07:50<frosch123>that's like a completely different thing :p
07:51<_dp_>I'm also thinking of automatically claiming some random town for new company, and scrolling is a must for that
07:52<frosch123>ok, i will think a bit more about that "join company" context
07:52<_dp_>yeah, mb it wold make more sense to extend joining itself, but there is no place for gs
07:52<frosch123>it would also make sense for other scripts to display the goal/status to joining clients
07:53<frosch123>like in silicon valley i only display the goal on game start or when a company is funded
07:53<frosch123>not when joining
07:53<_dp_>also I can imagine some gs that may need scrolling in the middle of the game if some major event happens
07:53<frosch123>that i would file under annoying :p
07:54<_dp_>not if it happens once in a game
07:56<Wolf01>yes, I agree with frosch123, gnomoria has it and it's really annoying when you are building something, last time I mined an entire layer because a lizard was sighted
07:58<_dp_>even I argee on that, but we're just talking about api here, it's up to gs to use it wisely
07:58<Wolf01>at least in ottd if the construction fails, it doesn't construct even the pieces which don't fail
07:58<frosch123>yes, it all depends on context :)
07:59<Wolf01>it could be good for tutorials, but you have to disable the user interaction before doing that
07:59<_dp_>Wolf01, that one is not true, many things do construct partially
07:59<Wolf01>you mean with autorail/road?
08:00<_dp_>autorail, area clear
08:00<Wolf01>oh, the area clear with drag&drop has been finally implemented?
08:01<Wolf01>wait, maybe I'm confusing with the diagonal clearing
08:01-!-andythenorth [] has joined #openttd
08:04<_dp_>btw, for tutorial(s) there is ScrollTo function. this one is more for tutorial-ish multiplayer things)
08:07<frosch123>Wolf01: don't imitate eddi :p
08:08-!-gelignite [] has joined #openttd
08:09<Wolf01>I'm doing my best efforts
08:14<_dp_>speaking of drag&drop construction... it probably shouldn't work across multiple viewports)
08:14<frosch123>you probably mean news and vehicle viewports
08:15<frosch123>because the extra viewports were specifically added for that
08:15<frosch123>i think it's even disabled in the news viewport
08:15<_dp_>yeah, leads to building weird things when news suddenly pops up
08:16<_dp_>is it? never had this issue with news myself but I recall someone complaining
08:17<_dp_>only happened to me with extra viewport that I had pinned to watch some town
08:18<Wolf01>I usually do that with long straight lines or tunnels
08:18<frosch123>hmm, maybe i only dreamed that :p
08:19<_dp_>you build tunnel only on one end so that's fine anyway
08:19<Wolf01>not if you don't know where it ends ;)
08:19<_dp_>and for long lines it sounds overcomplicated, why not just zoom out?
08:19<Wolf01>because touch
08:20<_dp_>I mean you can use extra viewport to see other end, it's not cross-viewport constuction
08:20<Wolf01>yes, I meant that
08:20<Wolf01>for tunnels
08:24-!-Mazur [] has quit [Remote host closed the connection]
08:24<Wolf01>with touch screen and 2x zoom in (because in normal zoom I miss the tile) is difficult to scroll
08:24<Wolf01>so for > 30 tiles tracks it is really useful
08:26<_dp_>hm, I bet polyrail is even more useful then xD
08:27<_dp_>also, just checked, constructing works on news, no restrictions at all
08:27<Wolf01>even a double track autorail with signal would be handy in that case
08:32-!-srhnsn [] has joined #openttd
08:40<_dp_>I'm considering doing a patch that adds a third type of economy, say, "stable", in which production doesn't change at all.
08:40-!-efess [] has joined #openttd
08:40<_dp_>frosch123, what do you think of it?
08:41-!-Quatroking [] has quit [Ping timeout: 480 seconds]
08:41<_dp_>also in that case is it ok to just switch setting type from bool to uint8 (or whatever matches the size) or should I add new one?
08:42<Wolf01>but industries should start with middle-high production, because if all the industries start with 32t of production it will piss of everyone
08:42<@Alberth>_dp_: that's something you can handle in newgrf
08:43<_dp_>Alberth, I know there is grf already but it doesn't work with custom industry sets
08:44<@Alberth>it won't ever work with industry sets, as the newgrf decides
08:45<_dp_>yet current smooth-switch does work
08:47<@Alberth>not really, newgrf doesn't understand smooth economy
08:48<_dp_>as I see it works unless grf has a custom callback
08:49<_dp_>though all popular ind sets probably have it...
08:49<frosch123>i considered adding an "economy speed" parameter
08:50<frosch123>which reduces calls to "random production changes"
08:50<frosch123>slows down smooth economy, and is readable by newgrf
08:50<_dp_>sounds fine for me too
08:51<frosch123>about bool->uint8, we did that multiple times already, it just needs the right savegame conversion code
08:52<frosch123>also smooth economy is a BAD FEATURE
08:52<frosch123>it quite high on the realism bullshit ladder
08:55<_dp_>idk about realism, but drying out industries on long-running mp server is bad for sure :p
08:55<_dp_>and any production change isn't good for goal competitiveness
08:55<frosch123>from single player gameplay point of view: the original double/half production changes were rare, and they called for player action
08:56<frosch123>the smooth economy changes happen every month to all industries, there is no immediate implication on gameplay
08:56<frosch123>it just obfuscates what is happening, snice the news messages are completely useless
08:56<_dp_>for sp and non-goal mp any economy is fine, it's not a major factor
08:57<frosch123>also when rolling a dice every month, you could as well skip rolling a dice
08:57<andythenorth>economy is meh
08:57<frosch123>randomness only makes sense when it is random
08:58<_dp_>yeah, skipping a roll exactly is what I was thinking with that "stable" economy)
08:59<frosch123>anyway, bye till evening
08:59<_dp_>though I'm a bit confused with newgrfs now, do random changes happen in newgrfs internally or there are separate callbacks for smth like "base" production and random fluctuatitons?
09:00<_dp_>because I don't want to stop ecs production increase or firs multipliers with this, just eliminate randomness
09:01<_johannes>can a Tile be multiple squares? like, e.g., a whole 2x4 station is one Tile, consisting of 2x4 squares?
09:01<_johannes>probably not because there is a TileArea type?
09:03<_dp_>you mean TileIndex? It's just an integer so no
09:04<_johannes>_dp_: ok, so a TileIndex represents exactly one square? (with square, I mean the small squares you see in a game)
09:05<_dp_>_johannes, kinda, it's basically just a x and y coordinates merged together
09:05<_johannes>_dp_: thanks
09:06-!-Mazur [] has joined #openttd
09:11<_dp_>_johannes, btw, those very squares are called tiles :) and there is Tile struct that stores info about each, I forgot about it at first
09:12<_johannes>ah I saw that one...
09:13-!-AdmiralKewl [~pcc31@] has quit [Read error: Connection reset by peer]
09:13-!-AdmiralKewl [~pcc31@] has joined #openttd
09:13<_dp_>_johannes, so, TileIndex is exactly what it's called, an index in array of tiles (aka map)
09:14-!-sim-al2 [] has joined #openttd
09:32-!-Clockworker [] has joined #openttd
09:34-!-andythenorth [] has quit [Quit: andythenorth]
09:37-!-andythenorth [] has joined #openttd
09:38-!-Clockworker_ [] has joined #openttd
09:40<andythenorth>frosch123: how does smooth economy work? :P
09:40<andythenorth>and why doesn’t andythenorth just know this
09:44*andythenorth ponders a GS
09:45-!-DDR [] has joined #openttd
09:45<andythenorth>industries will close one-by-one unless arbitrary goals are met for each one
09:45<andythenorth>winner is whoever has most industries left after n years
09:45*andythenorth must to chores
09:45-!-andythenorth [] has quit [Quit: andythenorth]
09:45-!-Clockworker [] has quit [Ping timeout: 480 seconds]
09:46-!-drac_boy [] has joined #openttd
09:46<drac_boy>btw anyone here from uk really thought the waterpans was only for nonstop express trains? well, heres something to the contrast
09:47<drac_boy>I'm not sure but that sure looks like one of the very common freight 0-6-0's as well
10:03-!-AdmiralKew_ [~pcc31@] has joined #openttd
10:09-!-AdmiralKewl [~pcc31@] has quit [Ping timeout: 480 seconds]
10:11<drac_boy>weird thing is even when I was using ukrs a bit before I don't think I ever bothered buying the 0-6-0 at all
10:11<drac_boy>ah well its just a grf after all :)
10:14<drac_boy>either way do have one silly uic-related question tho... do you just write "two C'C' units" or is there a more informal way to describe these particular locomotives that was made up of at least two or more separate bodies permanently kept together?
10:16*drac_boy might be missing something but isn't sure heh
10:19<Eddi|zuHause>yes. C'C'+C'C'
10:22<drac_boy>ah so thats where the plus came from...guess I'll try remember that for everything else
10:23-!-_johannes [] has quit [Quit: Lost terminal]
10:24<@Alberth>it's fine if you write it down :)
10:25<drac_boy>as long as you're not writing every single little things daily, this is what it could look like?
10:25<drac_boy>mind you I'm just joking of course :p
10:26<@Alberth>just scan everything :p
10:26<drac_boy>alberth yeah as I recall he wished all these filing cabinets away into the form of a blue imac :)
10:26<drac_boy>had watched a bit of that movie some time ago
10:27<@Alberth> how people spend their day :)
10:31-!-srhnsn [] has quit [Quit: srhnsn]
10:31*drac_boy gives alberth ten thousand yellow post it's just cause
10:35-!-Clockworker [] has joined #openttd
10:36-!-Quatroking_ [] has joined #openttd
10:41-!-Clockworker_ [] has quit [Ping timeout: 480 seconds]
10:43-!-liq3 [] has quit []
10:50-!-_johannes [] has joined #openttd
10:55-!-andythenorth [] has joined #openttd
11:05*andythenorth remembers some FIRS thing in the makefile was slow
11:05<andythenorth>sed over the hg log or something
11:06<andythenorth>Iron Horse makefile is unexpectedly slow
11:09<andythenorth>can’t remember what we changed
11:09<drac_boy>umm, changelog? ;)
11:11<andythenorth>maybe this
11:21-!-glx [] has joined #openttd
11:21-!-mode/#openttd [+v glx] by ChanServ
11:23<andythenorth>changed that, saved about 5s
11:24<drac_boy>going to eat soon so bye anyhow mr.coder :) heh
11:24-!-drac_boy [] has left #openttd []
11:25<@Alberth>too much $(shell ...) there
11:25*andythenorth wonders about the makefiles
11:25<andythenorth>I have 6 or so of them in active projects
11:25<andythenorth>all different
11:26<@Alberth>version stuff (around line 150) is now in iirc
11:27<@Alberth>it seems to start a few dozen shells for major stuff like replacing a small piece of text
11:27-!-ckraniak [~ckraniak@2600:100c:b204:dae5:e67d:79ba:e8da:32b7] has quit [Ping timeout: 480 seconds]
11:27<@Alberth>you can probably even code that in the makefile itself :p
11:28<@Alberth>imho, code it in python, and throw out all the crap
11:28<andythenorth>I don’t know what it does
11:29<andythenorth>which makes me scared of it :)
11:29<andythenorth>it’s eating 6 seconds or so on every compile
11:29<andythenorth>seems to be overkill for the result
11:29<@Alberth>maybe replace it with
11:30*andythenorth looks
11:30<andythenorth>(the 6s is the total extra time to run the makefile, on top of the components that build the grf)
11:30-!-Ketsuban [~ketsuban@2a02:c7d:a33e:4100:55e3:3d6e:77f3:3f3d] has joined #openttd
11:31<andythenorth>what is
11:31*andythenorth has missed a memo here :)
11:31<@Alberth>isn't one in bb?
11:31<@Alberth>may be GC specific thouigh
11:33<@Alberth> doesn't seem to have one
11:33<andythenorth>there is one in BB
11:34*andythenorth wonders how to implement that
11:35<@Alberth>opengfx-mars repos also have one
11:35<andythenorth>it returns some vars I think
11:35<andythenorth>and I need to parse those
11:35*andythenorth guessing
11:36-!-gelignite [] has quit [Quit:]
11:38<@Alberth>it returns revision, branch, modified flag and some more data
11:38<@Alberth>BB makefile parses it (not very good, I see though)
11:39<@Alberth>VERSION_INFO := "$(shell ./"
11:39<@Alberth>REPO_VERSION := $(shell echo ${VERSION_INFO} | cut -f2)
11:39<@Alberth>REPO_TAG := $(shell echo ${VERSION_INFO} | cut -f5)
11:39<@Alberth>REPO_DATE := $(shell echo ${VERSION_INFO} | cut -f7)
11:39<Eddi|zuHause>i think make-nml uses the current date or something
11:47<@Alberth>UNIX2DOS=$(shell which unix2dos 2>/dev/null) <-- make that UNIX2DOS=unix2dos
11:48<@Alberth>no point in inserting the absolute path in the variable
11:48<frosch123>that will likely break stuff
11:48<frosch123>"which" is used in those cases to result in the empty string, when it does not exist
11:48<frosch123>later there are ifs in the makefile, which check for empty string
11:49<frosch123>however, you can make it a := instead of a =
11:49-!-Clockworker_ [] has joined #openttd
11:49<@Alberth>for optional stuff, ok
11:49<@Alberth>for required stuff, like nml, you can skip that
11:50<@Alberth>or hg
11:56-!-Clockworker [] has quit [Ping timeout: 480 seconds]
12:04<andythenorth>how do I merge default into a working branch with hg?
12:04-!-ConductorCat [] has quit [Ping timeout: 480 seconds]
12:04<andythenorth>I have just read the manual
12:04<andythenorth>but not clear
12:04<andythenorth>child #1 has been adding trains to iron-horse in a branch
12:05-!-Clockworker__ [] has joined #openttd
12:05<andythenorth>and I want to pull recent changes from default into that branch
12:05<andythenorth>is it safest to save a patch from default and apply it?
12:06<@Alberth>iirc, you check out the destination, and then merge with the other one
12:07<Eddi|zuHause>hg checkout default, hg merge branch
12:07<@Alberth>hg has rollback, so you can do it, check how it looks, and then decide whether it's good
12:07<andythenorth>ach, I can always delete the repo and clone again
12:07<Eddi|zuHause>or the other way around
12:07<@Alberth>Eddi|zuHause: that's what I do to merge into default :)
12:08<@Alberth>but yeah, it's always tricky, one to look up
12:08<andythenorth>is ‘hg checkout' synomymous with ‘hg up’ in this context?
12:08<@Alberth>it is
12:08<andythenorth>so the commands are identical to git, but the results are different
12:08<andythenorth>this page is totally impenetrable to me, but eh
12:09<@Alberth>that's moving stuff between repos
12:09<andythenorth>works, thanks
12:10<andythenorth>ah, that’s why it’s not answering my question then :)
12:10<andythenorth>people merge between different repos? :o
12:10<andythenorth>instead of just branching?
12:10<andythenorth>or do they mean instances of the repo
12:12<@Alberth>yesm hg recommends making a clone for a new branch, iirc
12:12<@Alberth>at least they did
12:12-!-Clockworker_ [] has quit [Ping timeout: 480 seconds]
12:14*andythenorth back to fixing animated pixels
12:22<_johannes>is there a debug mode in openttd which shows me the tile coords under the mouse pointer?
12:24<_dp_>land info does I think
12:24<frosch123>yes landinfo
12:25<frosch123>when debug is enabled via console, it also prints the map array for that tile to console
12:25<frosch123>also there is the console command "scrollto" which does the reverse
12:25<frosch123>scroll to tileindex
12:26<_johannes>frosch123: just typing "landinfo" in console says "command not found"
12:27<frosch123>it's in the info menu on the very right
12:28<_johannes>ah that question mark thing... how could I forget that
12:31<@Alberth>play the game more often :p
12:31<_johannes>haha I play it for five years, multiple hours each week :D
12:32<_johannes>and still have not discovered half of the game...
12:32*andythenorth too :P
12:32<andythenorth>diagonal land-lower/raise :P
12:32<andythenorth>recently found
12:32<andythenorth>ctrl-drag vehicle groups
12:32<andythenorth>recently found :P
12:33<frosch123> <- always fun to read :)
12:33<_johannes>I heard you can even build diagonal tunnels?
12:33<_dp_>haha, when I learned about ctrl magic I tried ctrl-ing everything xD
12:34<andythenorth>so many things I didn’t know :o
12:36<_johannes>openttd is the game I've most played in my life
12:36<_johannes>by far...
12:37<_johannes>it's ~ 0.5% of my whole lifetime
12:38<frosch123>i am also worried that the most written english i have read, was chat by andy
12:39<andythenorth>can’t be good
12:40<frosch123>well, if i am lucky, i have read more pratchett
12:43-!-Snail [] has joined #openttd
12:48<_dp_>how do newgrfs like firs randomize production?
12:49<_dp_>something tells me that "random production change" callbacks aren't very random)
12:50<frosch123>firs has no randomness whatsoever
12:50<_dp_>I just checked with firs 1.smth and there were some production changes
12:51<_dp_>on monthly callback it seems
12:53<andythenorth>FIRS randomises primary production at game start
12:53<andythenorth>there is no other randomness
12:53<andythenorth>random production change is handled to prevent OpenTTD default behaviour ;)
12:55<frosch123>_dp_: you cannot control industry grfs, which do monthly production changes
12:55-!-Quatroking_ is now known as Quatroking
12:55<frosch123>you cannot slow the months, or you will break other stuff
12:55<frosch123>you can only slow the random production change callback
12:56<frosch123>and offer a new variable to grfs, so they can respond to the setting themself
12:58<_dp_>I'm trying to find out where random fluctuations happen and where it's base production changes
12:59<_dp_>callback is called "Monthly random production change (35) " in wiki
12:59<frosch123>yes, but it is essentially a "monthly" callback
12:59<frosch123>all grfs with supplies or similar use it as "monthly" and not for production change
13:01<_dp_>yeah, that's what I trying to find out
13:02<_dp_>interesting, newgrf can fall back to default changes but that will not take smooth_economy setting into account.
13:04<frosch123>no, smooth_economy is disabled when newgrf does anything with production
13:04<_dp_>how to expose a variable to newgrf?
13:06<frosch123>GetGlobalVariable in newgrf.cpp
13:07-!-__builtin [] has quit [Ping timeout: 480 seconds]
13:07<frosch123>or GetPatchVariable for stuff that is constant during a game
13:08<frosch123>or reasonably constant :p
13:09<_dp_>well, do settings count as reasonable constant?)
13:09<_dp_>looks like...
13:09<frosch123>you cannot change the values from GetPatchVariable in multiplayer for example
13:09<frosch123>since you need to reload newgrf
13:10<frosch123>basically, either the newgrf reads the var once on start, and then sets stuff accordingly
13:10<frosch123>or it reads in whenever used
13:12<_dp_>I guess it's better be in globals then
13:12<frosch123>if you put it in globals you cannot make availability of industries and vehicles depend on it :p
13:14<_dp_>so, basically it's a choice between being able to set availability or change it during the game
13:15<_dp_>at least I can imagine how latter would be useful)
13:15<frosch123>there are always corner cases, but some stuff can only be set on start
13:15<frosch123>either a vehicle exists, or it does not
13:15<frosch123>you cannot change a setting, and suddenly all vehicles on tracks are invalid :p
13:16<frosch123>anyway, if you do economy speed, it likely belongs into Global
13:16-!-__builtin [] has joined #openttd
13:17<_dp_>yeah, don't think "speed" is a good word for it though
13:17<_dp_>more like volatility or smth
13:17<frosch123>possibly :)
13:17<andythenorth>meh, can’t make Iron Horse compile faster
13:18*andythenorth experiments
13:18-!-cursarion [] has quit [Quit: Lost terminal]
13:19<_dp_>it may even be more that one setting
13:20<_dp_>smooth economy influences frequency of changes, severity of changes and station rating influence over them
13:20-!-andythenorth [] has quit [Quit: andythenorth]
13:21-!-andythenorth [] has joined #openttd
13:21<frosch123>that would be very smooth economy specific
13:21<frosch123>i doubt any grf would care about that
13:21-!-Snail [] has quit [Quit: Snail]
13:21<frosch123>grfs have their own parameters
13:21<frosch123>like effect of supplies and stuff
13:21<frosch123>those settings make no sense for that
13:22-!-andythenorth [] has quit []
13:22<_dp_>yeah, but now I wonder how do I do all if I'm going to replace smooth economy with some other setting %)
13:22<frosch123>trash smooth economy :p
13:22<frosch123>it's crap
13:24-!-andythenorth [] has joined #openttd
13:26<frosch123>i considered removing it completely and supplying a substitute newgrf
13:26<frosch123>but since it is so crappy, making the grf would be too boring
13:28<_dp_>dunno about that, I think I like it more than original one
13:29<andythenorth>rendering switches is slow :P
13:30*andythenorth just took about 8 seconds off Iron Horse compile
13:30<andythenorth>8 / 38
13:30<andythenorth>20% or so faster :P
13:37<_dp_>original economy only uses 60% station rating threshold
13:38<_dp_>and only x2 /2 changes that are way too big for any king of competitive play
13:38*andythenorth would fix it in newgrf :)
13:44-!-andythenorth [] has quit [Quit: andythenorth]
13:46-!-HerzogDeXtEr1 [] has joined #openttd
13:47<_johannes>is YAPF able to go into a depot if it's the only way to reach a station ?
13:47<frosch123>it's called "overflow depot"
13:47<frosch123>it's a known track layout
13:49<_dp_>I'd really like to have a configurable economy without newgrfs
13:49<_dp_>like wtf, you can confugure the hell out of pathfinder but only have two pitiful bools for economy xD
13:50<frosch123>i'd really remove all the disputable stuff from the game, and make everything via addons :p
13:50<frosch123>+like to
13:50<_dp_>frosch123, that would be an empty program :p
13:50<frosch123>sounds awesome
13:50<frosch123>no bugs
13:52-!-HerzogDeXtEr [] has quit [Ping timeout: 480 seconds]
13:57<_dp_>I'm kinda tempted to return to my original idea and just add "stable" economy
13:57<_dp_>coz it's more or less clear how to stop random changes but not how to scale them
14:01-!-Cursarion [] has joined #openttd
14:01-!-frosch123 [] has quit [Quit: be yourself, except: if you have the opportunity to be a unicorn, then be a unicorn]
14:02-!-frosch123 [] has joined #openttd
14:04-!-Crisco [~oftc-webi@] has joined #openttd
14:05<Crisco>someone recommend a good heightmap to create a save on?
14:08-!-Cursarion^ [] has joined #openttd
14:08-!-Cursarion [] has quit [Read error: No route to host]
14:13<_johannes>frosch123: if a train is at A and I tell YAPF to go to A, then YAPF should return immediatelly, right?
14:14<_johannes>I mean, it's not driving a complicated cycle in order to get to A again?
14:18<_dp_>I wish there was an easier way to check, but looks like all goal servers use smooth economy
14:18<_dp_>also coop and reddit
14:18<_dp_>so if any needs trashing it's original one :p
14:20-!-Crisco [~oftc-webi@] has quit [Quit: Page closed]
14:21-!-andythenorth [] has joined #openttd
14:29<andythenorth>are rosters in newgrfs stupid?
14:33-!-Cursarion^ is now known as Cursarion
14:56-!-Wormnest [] has joined #openttd
14:57<AdmiralKew_>i was wondering where and how you got roosters into the game
15:01<andythenorth>roosters would be better
15:29-!-gelignite [] has joined #openttd
15:32<andythenorth>Iron Rooster?
15:43-!-Progman [] has joined #openttd
15:50<supermop>rooster roster
15:52-!-AdmiralKew_ [~pcc31@] has quit [Quit: Leaving]
15:53-!-AdmiralKew [~pcc31@] has joined #openttd
15:57*andythenorth is baffled by names
15:57<andythenorth>I need to name Iron Horse rosters
15:57<andythenorth>they were going to be shakespeare quotes about horses
15:58<andythenorth>“Forgiveness, horse! why do I rail on thee”
15:58<andythenorth>is not easily understood as ‘African trains'
16:02-!-orudge [] has quit [Remote host closed the connection]
16:03<frosch123>andythenorth: easy, the african roaster of iron horse should obviously be named iron zebra
16:03*andythenorth considered that :)
16:03<andythenorth>but wondered if it was silly
16:04<Ketsuban>The Swahili for "horse" is "farasi".
16:04<andythenorth>do I use Wart Hog for the equivalent RVs?
16:05<andythenorth>hmm. Igbo for ‘horse’ is ‘inyinya’
16:07<frosch123>no, you would keep the "road", and replace the "hog" with something african
16:07<frosch123>hippos are said to be realted to hogs
16:07<frosch123>but hippo could also work for horse
16:07<frosch123>iron hippo?
16:08<frosch123>sounds more like heqs though
16:08*andythenorth followed a wikipedia link
16:09<andythenorth>frosch123: does this scheme scale? o_O
16:09<andythenorth>Norwegian Trains…Iron ???
16:10<frosch123>there are various horse like things in various areas of the world
16:10<andythenorth>mostly called ‘horse’ I think :D
16:10<frosch123>iron mule
16:10<andythenorth>I should just call the rosters ‘African trains’, ‘UK and Ireland trains’ etc :P
16:11<andythenorth>but then I get complaints they aren’t realistic eh?
16:11<frosch123>iron elephant
16:11<frosch123>iron bear
16:11<andythenorth>so just pick a distinctive animal
16:11<frosch123>iron moose
16:11<andythenorth>iron moose
16:11<andythenorth>oh you beat me :P
16:11<@Alberth>elephant is more appropriate for africa, imho
16:12<frosch123>i picked it for india :)
16:12<andythenorth>I like the random animal idea
16:12<andythenorth>I will find animals that aren’t easily conflated :P
16:12<@Alberth>iron duck
16:12<andythenorth>iron tuna
16:12<frosch123>elephants seem to be more domestic in india than africa
16:12<frosch123>i have never heard about work elephants in africa
16:13<frosch123>anyway, i guess you can find a specific animal for every major geological region
16:14<frosch123>likely you could as well for smaller regions, but then noone would recognise the animal :p
16:15<andythenorth>road moose
16:17<andythenorth>Iron Bushbaby?
16:17<andythenorth>it’s mostly narrow gauge trains
16:18<andythenorth>Iron Lion :D
16:19<frosch123>iron rhino
16:20<frosch123>lion and rhino are both similar to iron, both hard to pronounce together
16:20<andythenorth>interesting :)
16:21<andythenorth>in my accent, Iron Lion flows as two similar sounds
16:21<andythenorth>Iron Giraffe?
16:22<frosch123>i like silly names, so i still like zebra :)
16:23<andythenorth>Iron Okapi?
16:24<frosch123>also nice
16:24<frosch123>would give iron alpace for eastern stuff
16:25<frosch123>hmm, no, south america
16:25<andythenorth>silly old zebras don’t like in west africa, where my trains are
16:25*andythenorth might ignore that
16:26<andythenorth>Iron Hartebeest is no good?
16:26<andythenorth>Iron Camel
16:27*andythenorth thinks Zebra
16:27<andythenorth>not planning any East African trains, and all trains are the same in Africa, maybe?
16:28<frosch123>hmm, for west africa it says mostly gazelle
16:28<andythenorth>Iron Gazelle
16:28<andythenorth>I’ll have to add some description text anyway :)
16:28<andythenorth>if you want Zebra, I’m happy with that
16:29<frosch123>gazelle would work with moose
16:29<frosch123>and stag
16:29<andythenorth>and reindeer
16:29<andythenorth>and bison
16:29<frosch123>that's too christmassy
16:29<andythenorth>norway :)
16:29<@Alberth>roster is only active in december :p
16:29<andythenorth>norway is always christmas
16:30<frosch123>iron bison is like lion again :p
16:30<andythenorth>iron elk
16:30<frosch123>i wondered about iron buffalo for an us set
16:30<andythenorth>buffalo makes me think of mozarella
16:30<andythenorth>“domestic Italian Water Buffalo"
16:31<frosch123>i remember that dioxine stuff
16:35*andythenorth considers making a name generator
16:36<andythenorth>Iron Sandworm?
16:38<frosch123> :)
16:38<frosch123> <- more crowded
16:39<frosch123>"animal world map" has more results like that
16:39-!-FLHerne [] has joined #openttd
16:40<frosch123>hmm i wonder why "iron python" is so funny to me
16:43<andythenorth>frosch123: animal world map is winning :)
16:43<andythenorth>I am not going to comment on iron python
16:43<frosch123>most maps you cannot read though
16:43<andythenorth>iron polar bear
16:44<frosch123>not many trains there :p
16:44<frosch123>iron penguin
16:44<Wolf01>iron cat?
16:45<Wolf01>or more likely, iron zoo
16:45<andythenorth>iron penguin
16:47<_dp_>Iron Tsetse
16:47*andythenorth now inventing rosters just for the name
16:47<andythenorth>iron wallaby
16:47<andythenorth>iron tiger
16:47<andythenorth>frosch123: pick Zebra or Antelope or Lion, then I can make a commit :P
16:48-!-ckraniak [~ckraniak@2600:100c:b202:23a4:24eb:b10d:798d:2df5] has joined #openttd
16:48<andythenorth>Iron Hyena?
16:49<frosch123>i don't like lion or hyena
16:49<Wolf01>oh, iron wolf and shut up ;)
16:49<frosch123>i wonder about the difference between antelope and gazelle
16:50<Wolf01>"all gazelles are antelopes, but not all antelopes are gazelles" nice one
16:51<frosch123>well, iron antelope then
16:53<andythenorth>Iron Pika
16:54-!-Alberth [~alberth@2001:981:c6c5:1:be5f:f4ff:feac:e11] has left #openttd []
16:54<frosch123>iron meerkat
16:56<andythenorth>fits the range
16:56<andythenorth>well, only 1000 miles wrong :)
16:56<frosch123>it's the standard cute animal :)
16:57<frosch123>anyway, western africa: go for antilopes
16:57<frosch123>they work well with moose, and buffalo or even mule
16:58<frosch123>i guess don't pick too small animals :p
16:58<frosch123>nor the really big ones
16:58<andythenorth>Iron Bongo
16:58<Wolf01>it's for a passenger service set?
16:58<andythenorth>mostly freight
16:58<Wolf01>because antelope doesn't fit well for a long heavy goods train
16:59<andythenorth>Iron Gorilla?
17:01<frosch123>gorillas don't run
17:02<frosch123>i don't think it needs to be a strong animal
17:03<frosch123>engines are named crocodile, piggy or duck
17:03<andythenorth>Iron Bushbuck
17:03<andythenorth>kind of antelope
17:03<andythenorth>sounds suitable for tropic
17:04<andythenorth>not too big or too small
17:04<andythenorth>map range is super-realistic w.r.t to the trains used
17:07<andythenorth>Iron Bikeshedding? :D
17:11-!-roidal [] has quit [Quit: WeeChat 1.4]
17:16<_dp_>Iron Badger
17:19<andythenorth>I can have Iron Badger for UK?
17:25<Eddi|zuHause>can they have a running sound of "badger badger badger badger mushroom mushroom"?
17:25<andythenorth>ha ha
17:25*andythenorth must to bed
17:25<andythenorth>bye and thanks
17:25-!-andythenorth [] has left #openttd []
17:36-!-Clockworker_ [] has joined #openttd
17:36-!-smoke_fumus [~smoke_fum@] has quit [Quit: KVIrc 4.2.0 Equilibrium]
17:38-!-_johannes [] has quit [Quit: Lost terminal]
17:43-!-Clockworker__ [] has quit [Ping timeout: 480 seconds]
18:04-!-frosch123 [] has quit [Quit: be yourself, except: if you have the opportunity to be a unicorn, then be a unicorn]
18:22-!-M-E [] has quit [Ping timeout: 480 seconds]
18:22-!-gelignite [] has quit [Quit:]
18:22-!-M-E [] has joined #openttd
18:31-!-sla_ro|master [slamaster@] has quit []
18:43-!-Wormnest [] has quit [Quit: Leaving]
18:43-!-Biolunar [] has joined #openttd
18:51-!-drac_boy [] has joined #openttd
18:55-!-Xal [] has joined #openttd
19:06-!-Wolf01 [] has quit [Quit: Once again the world is quick to bury me.]
19:08-!-Clockworker [] has joined #openttd
19:12-!-Snail [] has joined #openttd
19:15-!-Clockworker_ [] has quit [Ping timeout: 480 seconds]
19:19-!-drac_boy [] has quit [Ping timeout: 480 seconds]
19:25-!-smoke_fumus [~smoke_fum@] has joined #openttd
19:34-!-drac_boy [] has joined #openttd
19:34<drac_boy>not sure why that stopped responding...
19:34-!-Progman [] has quit [Remote host closed the connection]
19:34<drac_boy>but anyhow as I was wondering, anyone want a real locomotive that needs a lot more than just a rag and some water? :p
20:03-!-liq3 [] has joined #openttd
20:06-!-FLHerne [] has quit [Ping timeout: 480 seconds]
20:16-!-smoke_fumus [~smoke_fum@] has quit [Quit: KVIrc 4.2.0 Equilibrium]
20:33-!-tokai [] has joined #openttd
20:33-!-mode/#openttd [+v tokai] by ChanServ
20:40-!-tokai|noir [] has quit [Ping timeout: 480 seconds]
20:49-!-Quatroking [] has quit [Read error: Connection reset by peer]
21:05-!-oskari89 [] has quit []
21:08<drac_boy>going sleep anyhow..just can't seem to do much for grfs :-s
21:08-!-drac_boy [] has left #openttd []
22:08-!-Biolunar_ [] has joined #openttd
22:15-!-Biolunar [] has quit [Ping timeout: 480 seconds]
22:16-!-Snail [] has quit [Quit: Snail]
22:20-!-Snail [] has joined #openttd
22:42-!-glx [] has quit [Quit: Bye]
22:55-!-Snail [] has quit [Quit: Snail]
22:55-!-Ketsuban [~ketsuban@2a02:c7d:a33e:4100:55e3:3d6e:77f3:3f3d] has quit [Quit: Leaving]
22:58-!-tokai|noir [] has joined #openttd
22:58-!-mode/#openttd [+v tokai|noir] by ChanServ
23:05-!-tokai [] has quit [Ping timeout: 480 seconds]
23:21-!-ConductorCat [] has joined #openttd
23:29-!-ConductorCat [] has quit [Ping timeout: 480 seconds]
23:30-!-supermop_ [] has joined #openttd
23:36-!-supermop [] has quit [Ping timeout: 480 seconds]
---Logclosed Sun Feb 21 00:00:09 2016