Log for #openttd on 4th October 2014:
Times are UTC Toggle Colours
00:03:43  <Eddi|zuHause> <Pikka> andythenorth: I'd forgotten what a PITA articulated vehicles with cargo capacity are :) <-- i have no problem with that. just make sure that each part that should have different capacity should use a different vehicle id instead of magic capacity callbacks...
00:04:50  <Pikka> well, yeah
00:05:13  <Pikka> otoh you just set the capacity to total capacity / vehicle parts
00:05:52  <Pikka> but it also means you have to build the articulate in the buy menu to get the correct capacity :)
00:05:56  <Eddi|zuHause> which is weird if they are not integer-dividable
00:06:12  <Pikka> but they are :)
00:06:23  <Eddi|zuHause> sometimes they're not
00:06:33  <Pikka> no, they always are :P
00:06:56  <Eddi|zuHause> like all my articulated vehicles have 3 (or multiples thereof) parts, but next to none of my capacities are dividable by 3
00:07:25  <Eddi|zuHause> well, almost all
00:07:47  <Eddi|zuHause> but the ones that are not are usually steam engines, which have no capacity
00:08:24  *** liq3 [] has joined #openttd
00:10:36  *** Progman [] has quit [Remote host closed the connection]
00:16:53  *** kiwibun [] has quit [Ping timeout: 480 seconds]
00:46:10  *** oskari89 [] has quit []
00:47:57  *** glx [] has quit [Remote host closed the connection]
00:48:18  *** glx [~glx@2a01:e35:2f59:c7c0:288c:966a:109:24db] has joined #openttd
00:48:36  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
00:55:49  *** DabuYu [DabuYu@] has joined #openttd
00:58:49  *** Flygon [] has joined #openttd
01:00:40  *** MTs-iPad [] has quit [Remote host closed the connection]
01:00:55  *** MTs-iPad [] has joined #openttd
01:05:11  *** Flygon_ [] has quit [Ping timeout: 480 seconds]
01:09:49  *** Flygon_ [] has joined #openttd
01:16:11  *** Flygon [] has quit [Ping timeout: 480 seconds]
01:22:15  *** HerzogDeXtEr [~flex@] has quit [Quit: Leaving.]
01:24:16  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
01:28:32  *** luaduck_zzz is now known as luaduck
01:47:50  *** DabuYu [DabuYu@] has joined #openttd
02:15:05  *** theholyduck [~theholydu@] has quit [Ping timeout: 480 seconds]
02:15:32  <peter1138> DIDDLYDUM
02:23:23  *** KWKdesign [] has quit [Ping timeout: 480 seconds]
02:23:24  *** KWKdesign [] has joined #openttd
02:35:08  *** glx [] has quit [Quit: Bye]
02:36:32  *** luaduck is now known as luaduck_zzz
02:40:25  *** Hazzard_ [] has joined #openttd
02:41:08  *** Flygon_ [] has quit [Read error: Connection reset by peer]
02:41:29  *** Flygon_ [] has joined #openttd
02:41:42  *** rambo [] has joined #openttd
02:42:16  *** rambo is now known as Guest434
02:43:47  *** Pokka [] has joined #openttd
02:45:40  *** Wing_ [] has joined #openttd
02:47:22  *** TheMask96- [] has joined #openttd
02:47:38  *** Netsplit <-> quits: Hazzard, jrambo, TheMask96, zeknurn, Pikka, EyeMWing
02:48:18  *** Netsplit over, joins: zeknurn
02:52:32  *** theholyduck [~theholydu@] has joined #openttd
03:05:48  *** Hazzard_ [] has quit [Ping timeout: 480 seconds]
03:21:36  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
03:26:11  *** DabuYu [DabuYu@] has joined #openttd
03:36:51  *** Pereba [~UserNick@] has quit [Quit: AdiIRC >>>>>>> abyss >>>>>>>>>>>>>>>>>>>>> other irc clients. (]
04:03:23  *** Hazzard [] has joined #openttd
04:22:15  *** SHOTbyGUN [] has quit [Read error: Connection reset by peer]
04:25:31  *** Jomann [] has joined #openttd
04:31:44  *** Wing_ [] has quit [Ping timeout: 480 seconds]
04:56:02  *** Eddi|zuHause [] has quit []
04:56:16  *** Eddi|zuHause [] has joined #openttd
04:58:56  *** MTsPony [] has quit [Ping timeout: 480 seconds]
05:07:19  *** theholyduck [~theholydu@] has quit [Ping timeout: 480 seconds]
05:11:32  *** theholyduck [~theholydu@] has joined #openttd
05:23:40  *** theholyduck [~theholydu@] has quit [Ping timeout: 480 seconds]
05:30:42  *** theholyduck [~theholydu@] has joined #openttd
05:33:44  *** Jomann [] has quit [Remote host closed the connection]
05:38:44  *** andythenorth [] has joined #openttd
05:42:00  <andythenorth> what ho
05:44:57  <Supercheese> what hoagie
05:51:53  *** theholyduck [~theholydu@] has quit [Ping timeout: 480 seconds]
06:05:42  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
06:16:38  *** theholyduck [~theholydu@] has joined #openttd
06:20:05  <Pokka> which how?
06:20:12  *** Pokka is now known as Pikka
06:22:57  *** kiwibun [] has joined #openttd
06:24:32  *** DabuYu [DabuYu@] has joined #openttd
06:25:21  <Pikka> andythenorth, whatwhenwhowhyhow?
06:25:21  <Supercheese> the now brown cow, of course
06:25:28  <Pikka> oh
06:25:30  <Pikka> hoo noo
06:25:33  <Pikka> broon coo
06:37:05  <andythenorth> is it done then?
06:37:09  <andythenorth> did you stay up all day?
06:39:06  *** Flygon__ [] has joined #openttd
06:39:30  <andythenorth> Pikka: my advice, don’t draw bulkhead flats with stakes
06:39:52  <Pikka> mm, steaks
06:40:10  <andythenorth> lodes
06:40:19  <andythenorth> for lode states
06:40:22  <andythenorth> are a PITA now
06:40:25  <Pikka> si
06:40:42  <Pikka> I think I've finished the passenger wogans
06:40:46  <Pikka> silly passenger wogans
06:40:56  <Pikka> I want a dome car in 10cc now :)
06:41:04  <Pikka> dome cars are fancy
06:42:13  <planetmaker> moin
06:42:21  <Pikka> moin pm
06:44:11  <andythenorth> fancy dome
06:45:39  <andythenorth> how about a Köln Dom car? o_O
06:45:43  <planetmaker> new set, Pikka ?
06:45:51  *** Flygon_ [] has quit [Ping timeout: 480 seconds]
06:46:00  <Pikka> old set, planetmaker
06:46:02  <Pikka> NARS recode
06:46:11  <planetmaker> hm, recode?
06:46:21  <planetmaker> what's the deal with recode?
06:46:36  <Pikka> starting over is easier than picking out the BAD FEATURES
06:47:17  <Pikka> copy and paste sprite blocks from the old set, write new code around them :)
06:47:35  <planetmaker> :)
06:47:41  <planetmaker> nars2.5?
06:47:50  *** Flygon__ [] has quit [Read error: Operation timed out]
06:47:57  <Pikka> how did you guess? :)
06:48:22  <planetmaker> :)
06:48:41  <andythenorth> no zellepins though :(
06:48:51  <Pikka> in NARS?
06:50:06  *** Flygon [] has joined #openttd
06:51:30  <andythenorth> yes
06:51:36  <andythenorth> zellepin car
06:52:48  *** kiwibun [] has quit [Ping timeout: 480 seconds]
06:56:56  *** Hazzard [] has quit [Ping timeout: 480 seconds]
06:59:30  *** kiwibun [] has joined #openttd
07:06:49  * Pikka bbeventually
07:06:52  <Pikka> latarz
07:06:54  *** Pikka [] has quit [Quit: Leaving]
07:37:23  *** kiwibun [] has quit [Ping timeout: 480 seconds]
07:48:05  *** Yotson [~Yotson@2001:980:6ac8:1:e0ce:47ac:1ba9:af2c] has joined #openttd
07:49:11  *** tokai|noir [] has joined #openttd
07:49:14  *** mode/#openttd [+v tokai|noir] by ChanServ
07:52:26  *** SHOTbyGUN [] has joined #openttd
07:55:06  *** tokai|mdlx [] has quit [Ping timeout: 480 seconds]
08:02:27  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
08:07:26  *** theholyduck [~theholydu@] has quit [Ping timeout: 480 seconds]
08:08:48  *** luaduck_zzz is now known as luaduck
08:11:13  *** Alberth [~hat@2001:981:c6c5:1:be5f:f4ff:feac:e11] has joined #openttd
08:11:16  *** mode/#openttd [+o Alberth] by ChanServ
08:15:33  *** Pensacola [] has joined #openttd
08:18:10  *** Progman [] has joined #openttd
08:20:17  *** Wolf01 [] has joined #openttd
08:20:43  <Wolf01> hi hi
08:22:49  <Alberth> moin
08:26:53  <andythenorth> lo Alberth
08:27:01  <Alberth> hi andy
08:32:36  *** theholyduck [~theholydu@] has joined #openttd
08:37:14  *** oskari89 [] has joined #openttd
08:39:10  * andythenorth makes trains
08:39:16  <andythenorth> these load states you all like so much
08:39:20  <andythenorth> take fricking ages :)
08:40:35  *** DabuYu [DabuYu@] has joined #openttd
08:40:41  <Alberth> but they're beautiful!
08:41:16  <Alberth> no simple way to automate at least most of it?
08:41:46  <planetmaker> do we need, additionally to the 'play me' button also a 'create game' button? :) meta-meta gaming? :)
08:41:48  <argoneus> good morning, train friends
08:41:59  <planetmaker> also hello Alberth et al :)
08:42:23  <argoneus> wow
08:42:24  <Alberth> eg draw background, draw top of cargo, draw front?
08:42:31  <argoneus> you reduced my glorious name to some latin buzzword
08:42:33  <argoneus> I will remember this
08:43:05  <andythenorth> I have automated bulk cargo colouring
08:43:11  <andythenorth> and I could automate the other loads
08:43:25  <andythenorth> but setting up the automation ~= drawing
08:43:32  <planetmaker> :)
08:43:39  <andythenorth> automation is better if the sprites change though...
08:43:46  <planetmaker> drawing steel coils etc is hard to automate?
08:43:53  <andythenorth> nah
08:43:57  <planetmaker> hm... rendering :D
08:44:28  <Alberth> all these 32bpp lovers around here :p
08:45:09  <argoneus> I just realized RGB covers only 24 bit colors
08:45:13  <argoneus> where did the other 8 bits go
08:45:27  <argoneus> alpha?
08:45:36  <Alberth> they hide in the opacities
08:45:46  <Alberth> at least in some formats :)
08:47:07  <Alberth> or the last byte is ignored, alignment at 32bit makes things way easier and faster
08:47:47  <planetmaker> indeed. RGB is obviously 3 bytes while RGBA is four :)
08:47:57  <argoneus> oh
08:48:00  <argoneus> I was wondering this
08:48:08  <argoneus> when I was trying to learn SDL etc.
08:48:22  <argoneus> I've read people suggest that in a tile-based game, tile size should be a power of two
08:48:31  <argoneus> 16x16, 32x32, 64x64, for 'performance' reasons
08:48:36  <argoneus> is this a myth or does it actually matter?
08:48:58  <argoneus> I never really found any explanation for it
08:50:20  <Alberth> it's mostly founded in the CPU itself, with buses to memory etc
08:50:39  <argoneus> so it's easier for the CPU to deal with calculations in power of two?
08:51:05  <Alberth> not the CPU itself, but eg video memory is not on the CPU chip
08:51:21  <Alberth> it's outside the CPU, so the CPU needs to write data to it
08:51:43  <Alberth> which basically means "write value V to address X"
08:52:15  <Alberth> if your data crosses address-boundaries, you need 2 writes at least
08:52:54  <Alberth> if you also want to keep the parts of the value you don't want to write to, you first have to read the values as well
08:53:23  <argoneus> oh
08:53:41  <Alberth> thus read at x to v1, read at x+1 to v2, fixup v1 and v2 to the new values, write new v1 to x, write new v2 to x+1
08:53:46  <argoneus> so it's not as much that 48x48 is less performant than 32x32 or 64x64, but 48x48 will have the same perf. results as 64x64?
08:54:16  <Alberth> depends on the size of a value
08:54:28  <argoneus> but reading about all this
08:54:35  <peter1138> traditionally, powers of 2 were easier to compute
08:54:40  <argoneus> I feel like for the older systems
08:54:46  <argoneus> the developers did some crazy optimizations
08:54:52  <peter1138> i haven't benchmarked it to see if it's still the case
08:55:29  <peter1138> multiplication to a power of 2 can be reduced to a shift instruction
08:55:50  <argoneus> isn't that multiplication in general?
08:55:54  <Alberth> it still holds for things like multiplication and division, since CPUs compute in powers of 2
08:56:17  <argoneus> 1100 (12) * 2 -> 11000 (24)
08:56:40  <Alberth> argoneus: yes, powers of 2
08:56:49  <Alberth> it fails eg for  * 3
08:56:50  <argoneus> ohhh, I read it wrong
08:56:55  <argoneus> I read "multiplication of a power of 2"
08:57:00  <planetmaker> peter1138, yeah as it can be shifted it should still be quicker and remain so
08:57:26  <peter1138> planetmaker, i've not benchmarked them :)
08:57:45  <argoneus> but would the performance on such a thing matter today?
08:57:55  <planetmaker> I haven't either, but the last time I looked up the performance benchmark it was significant difference there
08:57:56  <argoneus> (most gamedev studios would probably say "nah, just buy a better computer")
08:58:18  <Alberth> argoneus: it depends on how fast you want to go
08:58:32  <planetmaker>
08:58:36  <argoneus> I mean
08:58:45  <argoneus> how did people make 48x48 tiles, then, to keep the performance?
08:58:49  <argoneus> make them 64x64 and downscale?
08:58:58  <argoneus> or is that even worse
08:59:45  <Alberth> probably just do 48x48, it's not exactly as bad as 64x64
08:59:53  <Alberth> at least 16 rows are skipped
09:00:19  <peter1138> I don't think it's something you need to worry about these days.
09:00:42  <Alberth> indeed
09:01:30  *** FLHerne [] has joined #openttd
09:02:07  <planetmaker> we do with our map array :)
09:02:50  <peter1138> Have you ever benchmarked a non-power of 2 map?
09:04:20  <planetmaker> no. Just saying that our code says we do worry :)
09:04:58  <Eddi|zuHause> non-power-of-two multidimensional arrays make calculation of indices complicated
09:05:03  <planetmaker> though map size is also different... it's how our variable is stored
09:05:27  <planetmaker> changing that could again break newgrf :P
09:05:44  <Eddi|zuHause> and non-power-of-two array entries just get pumped up to power-of-two alignment
09:07:51  <Alberth> I doubt 129 bytes become 256 bytes :)
09:08:18  <peter1138> planetmaker, that pdf seems to ignore any core i3-5 & 7 range.
09:08:46  <planetmaker> so it seems. Just the first somewhat suitable hit on google :P
09:09:03  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
09:09:03  <Eddi|zuHause> Alberth: depends on processor architecture :p
09:09:06  <peter1138> Weird for a document dated this year.
09:09:28  <peter1138> It also mentions Sandy Bridge in the change log, but no where else.
09:09:37  <planetmaker> academia... scraped the old machines of the IT department probably
09:09:56  <peter1138> Hmm, actually, Intel SBR... Sandy Bridge?
09:10:27  <peter1138> HWL... Haswell
09:10:34  <peter1138> They've made up their own codenames :(
09:11:45  <argoneus> haswell can compete with nvidia gpus
09:11:53  <argoneus> in terms of how fast they start melting things around them
09:11:57  *** oskari89 [] has quit []
09:12:31  <argoneus> I swear, I wouldn't think my cpu should be able to go to 95C while benchmarking
09:12:36  <argoneus> but it's normal, because haswell
09:22:26  *** TomyLobo [~foo@] has joined #openttd
09:24:08  <Eddi|zuHause> or maybe you should upgrade your cooling?
09:24:14  *** Cybertinus [] has quit [Remote host closed the connection]
09:24:28  *** Cybertinus [] has joined #openttd
09:26:12  * andythenorth should upgrade his coffee
09:26:24  <andythenorth> also, what shall we make today?
09:27:23  <Eddi|zuHause> the same thing we make every day, pinky.
09:27:53  <argoneus> Eddi|zuHause: I have a hyper 212 evo
09:28:04  <argoneus> it doesn't get past 70C while gaming
09:28:07  <argoneus> 80C is safe for haswell
09:28:10  <argoneus> but benchmarks melt it
09:28:28  <argoneus> even video rendering doesn't push 75C
09:28:30  <argoneus> but benchmarks do
09:28:32  *** DabuYu [DabuYu@] has joined #openttd
09:28:56  <andythenorth> Eddi|zuHause: does that make you Brain?
09:29:11  <Eddi|zuHause> somewhat
09:29:14  <andythenorth> also can anyone fix my bugs?
09:29:48  <andythenorth> and shall I ship a new Iron Horse?
09:29:49  <Eddi|zuHause> argoneus: i've really no clue how you're doing that, i have "trouble" getting my computer to 60 on full load, and i already reduced fan speesds
09:29:52  <andythenorth> questions questions
09:29:56  <argoneus> Eddi|zuHause: do you have a haswell cpu?
09:30:10  <Eddi|zuHause> does that matter?
09:30:14  <argoneus> yes it does
09:30:21  <argoneus> haswell is ~10C hotter than the 3rd gen
09:30:59  <argoneus> my old phenom II didn't get hot at all
09:31:04  <argoneus> it was like 55C while rendering
09:31:29  <argoneus> also, Eddi|zuHause, there are different types of full load
09:31:32  <Eddi|zuHause> like i said, all you're saying is that your cooling is insufficient
09:31:34  <argoneus> do you mean full load as in rendering video for example
09:31:38  <argoneus> or actual benchmarks
09:31:38  <andythenorth> could we teach industries to plant their own trees?
09:31:54  <argoneus> I bet if you run linpack for 10 minutes
09:31:58  <argoneus> you will reach inhuman temperatures too
09:32:30  <Eddi|zuHause> andythenorth: same way as when we discussed that same topic 3 years ago :p
09:32:45  <argoneus> I mean
09:32:48  <andythenorth> got a logs link? o_O
09:32:54  <andythenorth> I like the sawmill behaviour
09:32:56  <argoneus> I get 40C idle, 60C gaming, 70C full normal load
09:32:58  <argoneus> that's not really hot
09:33:00  <andythenorth> I don’t like planting trees
09:33:06  <Eddi|zuHause> something about custom fields, i presume
09:33:14  <andythenorth> yeah
09:33:20  <andythenorth> but tmwftlb
09:33:33  <andythenorth> presumably we already have a tree planting loop?
09:33:41  <Eddi|zuHause> andythenorth: but that's why the sawmill is available in tropic only, where trees grow 257 times as fast
09:34:01  <andythenorth> oic
09:34:16  <Eddi|zuHause> normal tree loop plants a tree every 256 tile loops
09:34:33  <Eddi|zuHause> tropic tree loop plants an additional tree every tile loop
09:35:04  <peter1138> Remember when fields used to stay around after the farm disappeared?
09:35:13  <andythenorth> yes
09:35:20  <andythenorth> nice
09:36:06  <Eddi|zuHause> now they complete one growth cycle and then disappear?
09:36:10  <peter1138> yeah
09:36:18  <peter1138> looks odd
09:36:24  <Eddi|zuHause> would make newgrf fields easier if they were removed immediately on close
09:37:17  <peter1138> someone™ should make them turn to dirt and then slowly to grass and then trees and then the hedges/fences go away slowlyt
09:38:15  <Eddi|zuHause> except for the hedges/fences part that's only one function call :p
09:38:37  <peter1138> true
09:38:38  <andythenorth> someone should just run the explosion sprite on every tile of an industry when it closes
09:39:17  <peter1138> Eddi|zuHause, just keep a reference to the industry around until all the tiles are gone?
09:39:37  <peter1138> the tiles already know which industry they belong to, iirc
09:40:03  <andythenorth> 2 mins to compile Iron Horse :(
09:40:05  <andythenorth> fricking slow
09:40:12  <Eddi|zuHause> peter1138: but what if the tiles have a cycle that doesn't actually get to a "close down" state?
09:40:12  <andythenorth> it’s only got ~50 vehicles in it
09:41:01  <Eddi|zuHause> peter1138: yes, i think the problem was that things like newgrf variables cannot be resolved anymore after the industry closed
09:41:22  <peter1138> only because the industry has gone
09:41:47  <peter1138> so introduce an industry-closing stage
09:42:28  <peter1138> lot of effort for fields
09:42:42  <Eddi|zuHause> yes, and have a time limit how long "fields" can linger, if they don't shut down themselves
09:43:37  <Eddi|zuHause> (where "fields" could also be meadows, trees, power lines, ...)
09:43:52  <peter1138> batteries
09:44:11  <Eddi|zuHause> i've not spent more than 5 minutes in toyland :p
09:44:28  <Eddi|zuHause> over the past 10 years or so :p
09:44:52  <andythenorth> child no. 1 likes toyland
09:44:56  <andythenorth> but prefers ‘normal land'
09:45:12  *** frosch123 [] has joined #openttd
09:45:43  <Eddi|zuHause> andythenorth: opengfx-toyland?
09:46:00  <andythenorth> not so much
09:46:10  <andythenorth> opengfx not so much around here
09:47:44  <peter1138> hmm, weird, houses in the middle of nowhere :S
09:49:19  <andythenorth> o_O
09:50:13  <andythenorth> hmm loads of ottd commits
09:50:17  * andythenorth pulls
09:51:21  <peter1138> Just languages.
09:53:12  <andythenorth> sprite aligner
09:53:13  <andythenorth> stuff
09:53:23  <andythenorth> also I can’t get Iron Horse 1.2.0 to show in my game :(
09:53:25  <andythenorth> bit odd
09:53:27  <andythenorth> maybe I broke it
09:53:54  <peter1138> reddit is scary, i keep agreeing with V453000 on it :S
09:54:43  <andythenorth> scary indeed
09:58:15  * andythenorth broke bananas
09:58:16  <frosch123> do the 8bpp lovers also love 1x zoom?
09:58:37  <frosch123> V453000: why do you provide the 8bpp sprites only in 1x zoom?
09:58:59  <frosch123> is that what people want? or do you confuse zoom levels with colour depths?
09:59:12  <peter1138> Provide them in 4x zoom, and you won't get the blue fringe.
09:59:29  <frosch123> or do the ottd-scaled sprites look bad? :p
10:00:36  <__ln__> fonsinchen: <fonsinchen> maybe. The _LA part doesn't make any sense as there is no such country as 'LA' (or maybe it's Los Angeles).  <---
10:01:10  *** kiwibun [] has joined #openttd
10:01:45  *** HerzogDeXtEr [~flex@] has joined #openttd
10:06:13  *** zeknurn [] has quit [Read error: Connection reset by peer]
10:06:57  *** zeknurn [] has joined #openttd
10:13:03  *** kiwibun [] has quit [Ping timeout: 480 seconds]
10:21:59  *** Progman [] has quit [Remote host closed the connection]
10:23:43  *** KouDy [~koudy@] has quit [Ping timeout: 480 seconds]
10:29:41  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
10:33:56  <frosch123> pff, still no proper testcase for optimising nml
10:34:05  *** KouDy [~koudy@] has joined #openttd
10:34:23  <frosch123> while iron-horse takes 120 seconds to compile, only 6 seconds are spend in nml
10:34:43  <Eddi|zuHause> frosch123: some older CETS?
10:34:51  <frosch123> the rest is some andy magic
10:34:57  <Eddi|zuHause> before i took out all the expressions
10:35:13  <frosch123> Eddi|zuHause: does it have separate nml files?
10:35:30  <Eddi|zuHause> back then it didn't
10:35:31  <frosch123> i want to look into a proper partial compile + linking apprach
10:35:32  *** Brumi [] has joined #openttd
10:35:50  <frosch123> i could still use iron-horse though, it seems to have the right souces
10:36:22  <Eddi|zuHause> you can try current CETS, it depends on eddi-nml though (which adds the "comment" construct)
10:36:38  <Eddi|zuHause> that has loads of nml files
10:37:09  <frosch123> or i work on the yeti memory issue, that at least has a proper reason and fixing plan :)
10:37:20  <andythenorth> frosch123: Iron Horse is the wrong case
10:37:26  <andythenorth> it’s already been ‘optimised’
10:37:28  *** arroyoc [] has joined #openttd
10:37:28  <Eddi|zuHause> which are output to nfo, and then combined to be fed into grfcodec
10:37:48  <Eddi|zuHause> grfcodec on full CETS is also not instant :)
10:38:15  <frosch123> Eddi|zuHause: yeti takes 30 minutes to compile with grfcodec; it takes 8 seconds with nml if the spritecache is ready :p
10:38:21  <andythenorth> frosch123: also most of the 120 seconds are in nml
10:38:24  <frosch123> so, the potential is in nml
10:38:31  <frosch123> you just have to awake it :)
10:38:43  <andythenorth> Iron Horse uses identical method to CETS
10:38:51  <peter1138> o_O
10:39:06  <Eddi|zuHause> frosch123: yes, the difference between CETS and YETI is that in CETS the code is huge, and in YETI the graphics are huge
10:39:07  <frosch123> andythenorth: my measurement says 6 seconds, which is not 'most of 120 seconds
10:39:16  <frosch123> either your or my measurement is wrong then :p
10:39:19  <andythenorth> your measurement is wrong :)
10:39:36  <frosch123> ok, then i will add more details
10:39:45  <andythenorth> try FIRS instead
10:39:48  *** shansen [~shansen@] has joined #openttd
10:39:53  <andythenorth> I gave up on partial compiles with FIRS
10:40:08  <andythenorth> I don’t think it’s actually possible without changing the design
10:40:22  <peter1138> ooh, objects window needs love
10:40:28  <andythenorth> possible / possible without significant risk of compile failing :P
10:41:19  <Eddi|zuHause> frosch123: eddi-nml is now up to date
10:41:35  <V453000> frosch123: well, 8bpp x1 is mandatory, 32bpp x4 is additional ... I dont see much point in adding 8bpp x4 because mostly people either want 8bpp and same amount of pixels in zooms in compare to the rest of the world, or full 32bpp, with extra zoom, regardless how the rest looks
10:42:12  <V453000> ottd-scaled sprites aint that bad
10:42:49  <V453000> peter1138: the blue fringe will be removed with some more fiddling :)
10:43:01  <Eddi|zuHause> CETS seems to compile so far
10:44:23  <peter1138> V453000, i forgot old-1x is mandatory
10:44:43  <V453000> yeah :d
10:44:44  <Eddi|zuHause> one of those legacy thingies
10:44:59  <Eddi|zuHause> i wonder why that was actually done
10:45:35  <V453000> I honestly didnt even think about making 8bpp x4
10:45:40  <V453000> doesnt seem useful to me
10:46:12  <V453000> people only turn on 8bpp blitter because they want 32bpp newGRFs to fit better with old 8bpp newGRFs, both zoom and palette wise
10:46:30  *** ToBeFree [] has joined #openttd
10:47:35  <andythenorth> frosch123: there are a lot of python subprocess calls to nmlc
10:47:38  <andythenorth> in Iron Horse
10:48:07  *** DabuYu [DabuYu@] has joined #openttd
10:48:53  *** Jomann [] has joined #openttd
10:50:58  <peter1138> Eddi|zuHause, might be necessary if you have some zoom levels disabled. Dunno though.
10:51:33  <planetmaker> <peter1138> ooh, objects window needs love <-- I actually think our main toolbar should gain a button which allows building objects
10:51:42  <planetmaker> people constantly ask where they can build that stuff
10:51:43  <V453000> if you disable zoom levels and use YETI in 32bpp, it will just downscale it I believe
10:51:59  <peter1138> I'm only talking 2x GUI :)
10:52:10  <peter1138> I'd completely forgotten about it, as I never use objects.
10:52:42  <andythenorth> what are objects?
10:53:05  <peter1138> Exactly
10:53:14  <Eddi|zuHause> Error on sprite 65540. <-- i wonder why that is still there... that was fixed ages ago
10:53:26  <peter1138> 65540 ...
10:53:55  <Eddi|zuHause> yes. it references the sprite 65536, and nforenum barfs on that
10:54:33  <andythenorth> 105s compile, 101s in nml and grfcodec
10:54:38  <peter1138> Limitation disturbs me very sprites
10:54:40  <andythenorth> Iron Horse is a bad test case though
10:54:52  *** Brumi [] has quit []
10:54:53  <peter1138> Goes it throw out limitation?
10:55:28  *** rambo [] has joined #openttd
10:55:53  *** rambo is now known as Guest478
10:56:07  <Eddi|zuHause> frosch123: if you wanna have a stab at partially-compiling CETS, the autogenerated file currently handles the merging, and filtering out the multiply present headers
10:57:13  *** Guest434 [] has quit [Ping timeout: 480 seconds]
10:57:31  <andythenorth> also is makefile, unlike IH which does same thing in python
10:58:22  <peter1138>
10:58:37  <andythenorth> :D
11:01:42  <andythenorth> nice embiggen also
11:02:29  <andythenorth> who drew those shameful boots?
11:16:33  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
11:17:28  *** Phoenix_the_II [] has quit [Read error: Connection reset by peer]
11:21:46  *** Alberth [~hat@2001:981:c6c5:1:be5f:f4ff:feac:e11] has left #openttd []
11:26:44  *** shansen [~shansen@] has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz
11:27:27  *** Brumi [] has joined #openttd
11:29:00  *** Guest478 [] has quit [Read error: Connection reset by peer]
11:29:20  *** arroyoc [] has quit [Quit: arroyoc]
11:29:27  *** johnrambo [] has joined #openttd
11:29:40  *** DabuYu [DabuYu@] has joined #openttd
11:31:25  *** kiwibun [] has joined #openttd
11:32:47  *** Supercheese is now known as Guest480
11:32:48  *** Supercheese [~Superchee@] has joined #openttd
11:37:23  *** Guest480 [~Superchee@] has quit [Ping timeout: 480 seconds]
11:41:11  *** kiwibun [] has quit [Remote host closed the connection]
11:41:20  <frosch123> hmm, now i get 130s compile = 52s andy magic + 23s spawning sub-processes + 54s in nml
11:42:54  *** shansen [~shansen@] has joined #openttd
11:43:00  <planetmaker> probably you need osx ;)
11:44:19  <andythenorth> :P
11:45:01  <andythenorth> I maintain that you should profile one of the grfs that doesn’t have a mental partial compile using grfcodec ;)
11:45:09  <andythenorth> bad science
11:45:28  <frosch123> andythenorth: problem is that all grfs that take long do something weird, in a different way to all others
11:45:44  <planetmaker> :)
11:46:15  *** shansen [~shansen@] has quit []
11:49:12  <DorpsGek> Commit by peter1138 :: r26950 trunk/src/object_gui.cpp (2014-10-04 11:49:05 UTC)
11:49:13  <DorpsGek> -Codechange: GUI-scale dancing for NewGRF objects window.
11:49:37  <peter1138> hmm, what's left
11:49:53  <frosch123> a font gui :p
11:49:57  <peter1138> hah
11:50:10  <peter1138> i'll just double up the sprite font :D
11:50:14  <peter1138> (not really)
11:53:06  <andythenorth> frosch123: try FIRS
11:53:12  <andythenorth> it’s more amenable to profiling
11:53:14  <andythenorth> less weird
11:53:27  <andythenorth> there’s a templating step, but nothing crazy
11:53:29  <planetmaker> or nuts
11:53:37  <planetmaker> or cets
11:53:45  <andythenorth> cets isn’t valid
11:53:54  <planetmaker> and in the end you could always do compilation on pre-processed output
11:54:08  <andythenorth> yes
11:54:14  <planetmaker> but that won't show the weiredness going on in order to arrive at the pre-processed output
11:54:14  <andythenorth> FIRS assembles one .nml file
11:54:45  <andythenorth> the weirdness is non-significant in time for me and Eddi
11:54:49  <frosch123> hmm, ok, found the most time usage in a function that i neglected earlier :p
11:54:58  <andythenorth> it’s provably slow nmlc
11:55:35  <andythenorth> and the time increases out of proportion to input file length
11:55:40  <planetmaker> andythenorth, if half of the time is spent in pre-processing it's irrelevant? <frosch123> hmm, now i get 130s compile = 52s andy magic + 23s spawning sub-processes + 54s in nml
11:56:08  <andythenorth> he’s profiling wrong :)
11:56:34  <andythenorth> actually maybe not in that case
11:56:37  <andythenorth> I should check
11:56:58  * andythenorth checks
11:57:22  <frosch123> ah, ugly code formatting
11:57:36  <frosch123> nml source was hiding a function call in the dark :p
11:58:09  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
11:58:24  <andythenorth> ok so templating is _slow_  for Iron Horse
11:59:19  <andythenorth> definitely
11:59:57  <andythenorth> I think chameleon hates something about the way I’m loading templates
12:00:17  <frosch123> andythenorth: you call nml too often
12:00:24  <frosch123> that's your issue :)
12:00:30  <andythenorth> I do?
12:00:32  <andythenorth> where?
12:01:07  <andythenorth> I only have one call to it
12:01:31  <frosch123> you spend 77 seconds in executing nmlc, 68 seconds of that are in the startup of nmlc, indpednent of input
12:01:58  <andythenorth> like I said, Iron Horse is a bad test case ;)
12:02:03  <andythenorth> it has a partial compile
12:02:07  <frosch123> 23 seconds for spawning the process and loading python, 45 seconds to initialise nml
12:02:32  <andythenorth> try make NO_MP=False COMPILE_FASTER=True
12:02:48  <andythenorth> what’s the commonest compile cases when developing a grf?
12:03:01  <andythenorth> o_O
12:03:21  <frosch123> andythenorth: your solution is to not spawn nml multiple times, but to make nml process multiple files
12:03:32  <andythenorth> avoiding the startup overhead?
12:03:41  <frosch123> exactly
12:03:57  <andythenorth> I couldn’t see how to make nmlc process multiple files in parallel
12:04:41  <andythenorth> this is optimised for the cases of (1) ‘changed graphics, no code changed’ (2) changed single vehicle properties
12:04:50  <frosch123> andythenorth: nml has a main function, which calls a "nml" function with the parameters
12:04:57  <andythenorth> it may well be negative-optimised for the case of whole grf compile
12:05:02  <frosch123> just invent a new command line syntax to pass multiple input and output files
12:05:21  <andythenorth> and move the multiprocessing calls inside nmlc?
12:05:26  <andythenorth> with a -j flag or such?
12:05:28  *** DabuYu [DabuYu@] has joined #openttd
12:06:13  <frosch123> andythenorth: no, instead of spawning nml 70 times for 70 files, spawn in 7 times with 10 files each
12:06:24  <andythenorth> yes that would be smart
12:08:51  <andythenorth> just fill up queues and dispatch them
12:10:17  <andythenorth> whenever I touch multiprocessing, I think of this
12:10:58  <frosch123> hmm, ok, actually this issue appleis to every partial compile strategy
12:11:18  <frosch123> we won't be able to change nml start-up times to less than 0.6 seconds
12:11:53  <frosch123> if we split nml into single files per vehicle, we will get 100 files, and > 60 seconds compile time
12:12:22  <andythenorth> for a full re-compile
12:12:30  <andythenorth> assuming no checking if file is changed
12:13:01  <frosch123> if one wants to optimise that by calling nml for multiple files at once, that totally breaks the Makefile approach
12:13:11  <frosch123> doesn't it?
12:13:36  <andythenorth> I think so
12:13:48  <frosch123> so, how to proceed? make the Makefile more complicated to collect nml calls somehow? or abandon makefile and make nmlc behave like javac with custom dependency resolution?
12:14:37  <andythenorth> which approach is more tractable?
12:14:41  <andythenorth> done > correct
12:15:10  <frosch123> i guess it needs figuring out how likely a full compile actually is
12:15:23  <frosch123> i.e. can we avoid a full compile when changing english.txt :)
12:15:28  <andythenorth> well
12:15:37  <andythenorth> currently not, due to string handling :P
12:15:52  <andythenorth> train_1.png shouldn’t require a recompile though
12:16:15  <andythenorth> that is working brilliantly for Iron Horse, however hokey the implementation :)
12:17:00  <andythenorth> I can change any pngs and come out with a <10s compile, including pixa processing
12:17:46  <andythenorth> hence Iron Horse gets shipped faster :)
12:18:27  <andythenorth> I also have a hack in FIRS that just turns off all industries except one
12:18:41  <andythenorth> which breaks a few things, but is needed for testing industry layouts, sprites etc
12:31:12  <fjb> Moin
12:33:56  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
12:41:04  *** Alberth [] has joined #openttd
12:41:07  *** mode/#openttd [+o Alberth] by ChanServ
12:41:25  <peter1138> Blah at offsets
12:43:30  <peter1138> Hmm, helicopters are fine :S
12:43:31  <andythenorth> +1
12:43:36  * andythenorth fixing buy menu
12:45:49  <andythenorth> should vehicles overlap, or be clipped for height?
12:46:25  <Eddi|zuHause> <frosch123> if we split nml into single files per vehicle, we will get 100 files, and > 60 seconds compile time <-- that's still faster than 5 minutes for reading a 1MB file, and throwing an OOM because nmlc (or ply) need 4GB RAM
12:49:08  <Eddi|zuHause> anyway, gtg...
12:49:31  <Eddi|zuHause> frosch123: if you have a saner method of processing headers without duplicating them in the final grf, i can ditch eddi-nml
12:49:50  <peter1138> Apparently drawing a 4x airport preview sprite is... slow?
12:52:44  *** theholyduck [~theholydu@] has quit [Ping timeout: 480 seconds]
12:59:46  <peter1138> dbg: [sprite] Fixing lru 165223, inuse=29193560
12:59:49  <peter1138> hmm
13:01:48  <peter1138> so the LRU counter goes haywir
13:01:49  <peter1138> +e
13:03:23  *** theholyduck [~theholydu@] has joined #openttd
13:04:46  <Alberth> either that, or you're drawing sprites straight from the hard disk :p
13:05:02  <peter1138> i don't even know what it's for
13:05:46  <peter1138> Well, not precisely.l
13:08:51  <peter1138> Okay...
13:09:01  <peter1138> I took out the DrawSprite call, and it's still laggy :p
13:11:24  *** theholyduck [~theholydu@] has quit [Ping timeout: 480 seconds]
13:11:38  <peter1138> huh... if i move the window mostly offscreen... it stops being laggy
13:11:47  <peter1138> and if i bring the window back... it's still not laggy
13:14:29  <peter1138> hmm, sometimes it is.
13:15:23  <peter1138> hmmmmmm
13:15:38  <peter1138> and if my game window is larger... no problem
13:24:20  *** shansen [~shansen@] has joined #openttd
13:26:28  *** MTs-iPad [] has quit [Remote host closed the connection]
13:26:43  *** MTs-iPad [] has joined #openttd
13:27:06  *** theholyduck [~theholydu@] has joined #openttd
13:32:31  *** Bobjr [] has joined #openttd
13:32:48  <Bobjr> Hey guys! I have a weird problem and I was hoping you could help me
13:33:02  <Alberth> o/
13:33:20  <Bobjr> So I am running the opengfx graphics package (Just so we are all on the same page) and my water is blinking.
13:33:32  <Bobjr> Like if I zoom out all the water is constantly blinking/refreshing
13:33:43  <Bobjr> Like the movement of the water is causing a color change?
13:33:52  <Bobjr> Hello Alberth!
13:33:55  <frosch123> that's how it is supposed to work
13:34:04  <frosch123> disable "full animation" in settings, if it annoys you
13:34:36  <Bobjr> Oh cool. Yeah it was annoying me quite a bit :P
13:35:28  <Bobjr> Thank you very much
13:36:49  <peter1138> You'll miss out on the other animations though.
13:37:38  <Bobjr> Hm. Its weird I never noticed that the ocean moved like that
13:37:41  <Bobjr> Till today
13:38:15  <planetmaker> :)
13:38:44  <frosch123> gah, dynamically typed languages :/
13:42:58  *** Myhorta [] has joined #openttd
13:48:10  *** MTsPony [] has joined #openttd
14:03:33  *** oskari89 [] has joined #openttd
14:11:35  *** EyeMWing [] has joined #openttd
14:14:04  <andythenorth> @seen danmack
14:14:04  <DorpsGek> andythenorth: danmack was last seen in #openttd 3 days, 12 hours, 54 minutes, and 18 seconds ago: <DanMacK> Hey all
14:14:23  <SpComb> yay look it's in production now!
14:14:52  <andythenorth> ugly boats you have there
14:17:20  <peter1138> Wait a minute, that game looks familiar...
14:21:17  <__ln__> yeah, it resembles TTD
14:21:23  <frosch123> it has unusual zoom levels though
14:22:33  <frosch123> SpComb: that's a screenshot right, no running game?
14:22:48  <frosch123> ah yeah, the url says so
14:23:29  *** sla_ro|master [slamaster@] has joined #openttd
14:26:08  <andythenorth> hmm
14:26:49  <peter1138> pikka was green, spcomb red, me peach, andy blue
14:26:56  <peter1138> skiddles brown
14:27:00  <peter1138> and someone grey
14:27:41  <andythenorth> templating on Road Hog is insanely faster than templating on Iron Horse
14:27:46  <peter1138> scrolls a bit quicker in ottd :D
14:27:47  <andythenorth> same templater, same approach
14:37:11  *** DabuYu [DabuYu@] has joined #openttd
14:41:06  <MTsPony> hi all. where is it suitable to point to the author of
14:41:13  <MTsPony> and license file fr
14:41:17  <MTsPony> errr
14:41:36  <MTsPony> when using a heightmap or parts of it as a server maintainer?
14:42:47  <frosch123> if you load a savegame the viewport always points to the same location you saved the game with
14:42:53  <frosch123> that also applies to all joining clients
14:43:04  <frosch123> you can use that to put signs with text there
14:43:11  <andythenorth> hmm just found 5s of slow in a 21s templating step
14:43:15  <frosch123> or you can run a bot to respond on commands
14:43:43  <MTsPony> can i put it in the message when someone joins the server? would that be appropriate?
14:44:07  * andythenorth wonders if python is passing around copies of the vehicle object
14:44:18  <andythenorth> with setup overhead every time
14:44:32  <frosch123> MTsPony: sure, why not?
14:44:53  <MTsPony> kthx :)
14:50:17  <Bobjr> Can a cargo ship pick up gold?
14:53:02  <Bobjr> Because my cargo ships just kinda parks at the port that is right next to the gold mine and then sails away
14:53:07  <Bobjr> No cargo onboard
14:53:12  <Bobjr> And the dock says it supplies gold
14:53:22  <peter1138> is the ship refitted to gold?
14:53:35  <Bobjr> You have to refit a cargo ship?
14:53:57  <Bobjr> Wow alright, I see now
14:54:08  <Bobjr> Sorry i'm new, didn't know that it worked that way, haha
14:54:23  <peter1138> no worries :)
15:00:34  *** TheMask96- [] has quit [Ping timeout: 480 seconds]
15:05:51  *** TheMask96 [] has joined #openttd
15:13:49  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
15:18:50  <andythenorth> frosch123: maybe not now, but Iron Horse templating step might be a little faster now
15:19:16  <andythenorth> at least for me, 38s vs 54s or so (using single process)
15:19:43  <andythenorth> might be faster on second run, due to compilation of templates to python byte code
15:21:47  <andythenorth> would be interested to know if you see any difference at all (but don’t worry if you’re doing other stuff)
15:25:59  * peter1138 boggles
15:31:32  *** fjb [] has quit [Remote host closed the connection]
15:32:53  <andythenorth> hmm
15:32:53  *** shansen [~shansen@] has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz
15:33:04  *** shansen [~shansen@] has joined #openttd
15:33:28  *** Netsplit <-> quits: tneo, andythenorth, @planetmaker, ST2, Pinkbeast, blathijs, NGC3982, Prof_Frink, Markk, guru3,  (+10 more, use /NETSPLIT to show all of them)
15:34:10  *** Netsplit over, joins: JGR_, Prof_Frink, XeryusTC, KenjiE20, LordAro, TheMask96, oskari89, andythenorth, NGC3982, __ln__ (+10 more)
15:34:11  *** KenjiE20 [] has quit [Read error: Connection reset by peer]
15:36:40  *** DabuYu [DabuYu@] has joined #openttd
15:37:24  *** fjb [] has joined #openttd
15:38:59  *** MTs-iPad [] has quit [Ping timeout: 480 seconds]
15:40:23  *** shansen [~shansen@] has quit [Quit: My MacBook Pro has gone to sleep. ZZZzzz
15:42:05  *** shansen [~shansen@] has joined #openttd
15:46:44  *** shansen [~shansen@] has quit []
15:48:57  <andythenorth> hmm
15:49:23  <andythenorth> just found a single function that accounts for 34s of a 38s processing time
15:50:55  <andythenorth> specifically this is super slow: if year in range(variant.intro_date, variant.end_date):
15:50:55  <andythenorth> ah
15:50:56  *** guru3 [] has quit [Read error: Connection reset by peer]
15:51:00  <andythenorth> intro_date min value is 0
15:51:16  <andythenorth> end_date max value is 5000001
15:51:24  <andythenorth> that’s
15:53:29  <andythenorth> :P
15:53:29  <andythenorth>             if year >= variant.intro_date and year < variant.end_date:
15:53:42  *** tneo [] has quit [Ping timeout: 480 seconds]
15:53:45  *** funnel_ [~funnel@] has joined #openttd
15:54:10  *** tneo [] has joined #openttd
15:54:25  <Alberth> should be a tad quicker :p
15:54:55  *** ProfFrink [] has joined #openttd
15:54:58  *** Prof_Frink [] has quit [Read error: Connection reset by peer]
15:55:08  *** ProfFrink is now known as Prof_Frink
15:56:47  *** JGR_ [] has quit [Ping timeout: 480 seconds]
15:57:22  <andythenorth> ridiculously faster
15:57:37  *** planetmaker [] has quit [Ping timeout: 480 seconds]
15:57:42  *** JGR [] has joined #openttd
15:57:44  <peter1138> Ah, it's bjarni-code from r8770
15:57:51  <peter1138> 7 years :S
15:58:06  *** SHOTbyGUN [] has quit [Read error: Connection reset by peer]
15:58:09  *** planetmaker [] has joined #openttd
15:58:12  *** mode/#openttd [+o planetmaker] by ChanServ
15:58:37  <andythenorth> 38s -> 5s for single process compile
15:58:46  <andythenorth> 21s -> 2.5s for multiprocess compile
15:59:14  <Alberth> find a few more such things for speedup :)
15:59:23  <peter1138> 16:50 < andythenorth> specifically this is super slow: if year in range(variant.intro_date, variant.end_date):
15:59:24  <andythenorth> I’ve been commenting stuff in and out
15:59:32  <peter1138> does it put a switch entry in for each year or something?
15:59:51  <andythenorth> it’s building a random switch
15:59:54  <andythenorth> in that case
16:00:06  *** guru3 [] has joined #openttd
16:00:38  <peter1138> -Fix r8661: corrected some aircraft sprite placement issues in the depot window
16:00:38  <andythenorth> the resulting switch only has 2 or 3 actual switch options
16:00:44  <Alberth> 'range' builds a list with every value in the range. Then you linearly search for the yeat
16:00:46  <peter1138> +							  y + max(spr->height + spr->y_offs - 14, 0), // tall sprites needs an y offset
16:00:51  <peter1138> i wonder which "tall" sprites this affectrs
16:00:56  *** Hazzard [] has joined #openttd
16:01:05  <peter1138> seems to me it will only affect talls sprites which have an incorrect offset...
16:01:08  <peter1138> -s
16:01:33  *** MTs-iPad [] has joined #openttd
16:03:22  *** funnel_ [~funnel@] has quit [Quit: leaving]
16:04:29  <andythenorth> just break it and see who complains
16:04:46  <andythenorth> wonder how I could teach my compile a -j option
16:04:55  <andythenorth> dunno, sounds like shlex
16:05:06  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
16:09:55  <andythenorth> hmm
16:09:56  <andythenorth> odd result
16:10:19  <andythenorth> using python multiprocessing pool results in compile in ~30s instead of ~60s
16:10:29  <andythenorth> even if the pool contains only 1 worker
16:13:36  <peter1138> Well the height scaling thing doesn't work right anyway
16:14:18  <peter1138> (And never did)
16:14:38  <andythenorth> stupid editor, it defaults to show .orig files after a revert
16:14:38  <peter1138> So I'll take that out and if it's needed, it can be done properly :p
16:15:17  *** funnel_ [~funnel@] has joined #openttd
16:15:52  *** DabuYu [DabuYu@] has joined #openttd
16:17:56  *** liq3 [] has quit []
16:18:14  * andythenorth wonders if pool.join() causes block until all workers complete
16:18:30  <andythenorth> also give me my cpus back please
16:19:57  <peter1138> They're overrated.
16:34:19  <LordAro> workers, or cpus?
16:34:52  *** shansen [~shansen@] has joined #openttd
16:37:19  *** KenjiE20 [kenjie20@] has joined #openttd
16:38:39  *** funnel_ [~funnel@] has quit [Quit: leaving]
16:38:58  <andythenorth> that fix cuts Road Hog compile time in half too
16:39:12  *** funnel_ [~funnel@] has joined #openttd
16:39:17  *** argoneus [] has quit [Remote host closed the connection]
16:39:19  *** argoneus [] has joined #openttd
16:39:54  *** KWKdesign [] has quit [Ping timeout: 480 seconds]
16:40:03  *** funnel_ [~funnel@] has quit []
16:40:28  *** KWKdesign [] has joined #openttd
16:40:30  <DorpsGek> Commit by peter1138 :: r26951 /trunk/src (10 files) (2014-10-04 16:40:23 UTC)
16:40:31  <DorpsGek> -Codechange: Do the GUI-scale dance for vehicle, depot, and purchase lists.
16:41:57  <andythenorth> \o/
16:43:48  <andythenorth> frosch123: so due to your profiling, I just made 3 grfs compile faster :D
16:43:50  <andythenorth> thanks
16:44:23  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
16:44:27  <frosch123> yay, i shoveled work on someone :)
16:45:00  <Alberth> and he thanks you for it :p
16:45:14  <Rubidium> so how many compiles do you need to do to break even?
16:45:27  <frosch123> i think i am 50% done with yeti memory fix
16:45:35  <Rubidium> ... queue obligatory xkcd link
16:46:43  <andythenorth> it’s funny
16:46:56  <andythenorth> even though I have spent a few hours poking at compiles
16:46:57  <andythenorth> it was fun
16:46:59  *** shansen [~shansen@] has quit [Ping timeout: 480 seconds]
16:47:01  <andythenorth> and I learnt some things
16:47:07  <Rubidium> true
16:47:08  <andythenorth> so doing cost-benefit on the time
16:47:14  <andythenorth> I come out ahead :P
16:47:21  <Rubidium> lots of things we do here won't have much benefit w.r.t. that xkcd link
16:47:33  <andythenorth> also, when I want to see if a change is working in game, faster compile is worth £$€
16:47:44  <andythenorth> especially if I want to get it done and go to sleep :P
16:47:58  <peter1138> so what's the total cost?
16:48:02  <andythenorth> 20p
16:48:10  <andythenorth> value of compiles increases exponentially after 11pm
16:48:11  <andythenorth> also
16:48:17  *** luaduck is now known as luaduck_zzz
16:48:26  <Rubidium> $ 7,814,042 ?
16:48:52  <andythenorth> canadian, or US?
16:49:10  <Rubidium> probably US
16:49:42  <Rubidium> although... apparantly the master server + content server are worth more than half a million
16:50:11  <Rubidium> (more than grfcodec + nforenum combined)
16:50:39  <Rubidium> hmm... never mind
16:50:44  <peter1138> andythenorth, *sigh* i mean total compile time before and after your optimizations.
16:51:02  <Rubidium> apparantly coverity left some stuff, so the master + content server are only worth a quarter of a million ;(
16:53:24  *** zeknurn [] has quit [Quit: .]
16:53:31  *** DabuYu [DabuYu@] has joined #openttd
16:53:36  <Alberth> (17:58:37) andythenorth: 38s -> 5s for single process compile
16:53:36  <Alberth> (17:58:46) andythenorth: 21s -> 2.5s for multiprocess compile
16:53:42  *** zeknurn [] has joined #openttd
16:54:27  <peter1138> 3Yeah that's slightly faster.
16:54:38  <peter1138> Didn't I see 130s earlier?
16:55:11  *** Bobjr [] has quit [Quit:  HydraIRC -> <- In tests, 0x09 out of 0x0A l33t h4x0rz prefer it :)]
16:58:30  *** gelignite [] has joined #openttd
17:00:40  *** funnel_ [~funnel@] has joined #openttd
17:02:43  *** glx [] has joined #openttd
17:02:46  *** mode/#openttd [+v glx] by ChanServ
17:06:01  *** luaduck_zzz is now known as luaduck
17:09:10  <andythenorth> 114s from a cold cache
17:09:28  <peter1138> But that's rare I guess
17:09:41  <andythenorth> yeah
17:09:52  <andythenorth> testing with warm cache
17:10:44  <peter1138> So what's the point of marico?
17:12:39  *** DDR [] has joined #openttd
17:12:55  <andythenorth> because people love it
17:13:24  <andythenorth> peter1138: you never had this?
17:13:27  <andythenorth> my uncle had some
17:13:36  <peter1138> Nope.
17:14:05  <peter1138> In summary, newobjects are about as useful as newstations...
17:14:55  <Rubidium> but newobjects don't fill the "small" static station table of TTDP
17:14:57  <frosch123> they are there, so you can reduce station spread to prevent coop style cheating
17:15:24  <andythenorth> hmm, that compile was 155s
17:15:26  <peter1138> heh
17:16:47  * peter1138 fiddles with AIs and finds another bit that needs to be fixed for GUI-scale :p
17:18:22  <andythenorth> new compile is 124s cold cache, 72s warm cache
17:18:33  <peter1138> And 2.5s when?
17:19:28  * andythenorth tests more
17:19:36  <andythenorth> 41s with 16 worker processes
17:20:02  <andythenorth> 6.5s if only graphics are changed
17:20:11  <andythenorth> 2.5s was the fastest time for just the templating step
17:20:24  <andythenorth> but there’s make, grfcodec, docs, etc
17:21:43  <andythenorth> vanilla nmlc has zero capability to check deps
17:21:57  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
17:22:02  <andythenorth> so for a single changed pixel, it will re-compile everything
17:22:06  <peter1138> hmm, installing civilai doesn't install pathfinder.road
17:22:17  <andythenorth> nmlc does have a sprite cache though o_O
17:22:44  <andythenorth> anyway, grfcodec ftw
17:22:45  <peter1138> ah
17:23:14  <andythenorth> if we want to get away from crazy andythenorth partial compiles, we could start by teaching nmlc to dep check
17:24:08  <Rubidium> just implementa sprite cache for grfcodec ;)
17:24:23  <andythenorth> o_O
17:25:11  <peter1138> Hmm, wonder what town cars are needed...
17:25:58  * andythenorth wonders what the nmlc sprite cache caches
17:26:11  <Rubidium> the compressed sprite AFAIR
17:26:22  <andythenorth> it’s all just squiggles to me :D
17:29:50  <andythenorth> the makefile seems to have about 3s startup time
17:29:54  <andythenorth> interesting
17:30:53  *** DabuYu [DabuYu@] has joined #openttd
17:32:40  *** fjb [] has quit [Remote host closed the connection]
17:33:32  <andythenorth> if I circumvent make, a sprites only compile is 0.8s
17:33:44  <andythenorth> maybe I write a local script to do that
17:34:33  *** fjb [] has joined #openttd
17:38:53  *** theholyduck [~theholydu@] has quit [Ping timeout: 480 seconds]
17:39:08  *** funnel_ [~funnel@] has quit [Quit: leaving]
17:41:09  *** funnel_ [~funnel@] has joined #openttd
17:44:03  *** funnel_ [~funnel@] has quit []
17:45:23  <DorpsGek> Commit by translators :: r26952 trunk/src/lang/japanese.txt (2014-10-04 17:45:15 UTC)
17:45:24  <DorpsGek> -Update from WebTranslator v3.0:
17:45:25  <DorpsGek> japanese - 54 changes by guppy
17:51:12  <peter1138> Hmm, what do I need to do to get the unread news icon on the status bar?
17:51:58  <frosch123> set news to "off"
17:52:51  <frosch123> iirc it flashes when a news item spawns, which is filtered, and only goes to message history
17:53:02  <peter1138> errr
17:53:23  <peter1138> newspaper: off < still getting the newspaper popup
17:53:56  <peter1138> nice, found a bug those :D
17:53:58  <peter1138> ...
17:54:02  <peter1138> nice, found a bug though :D
17:55:41  *** funnel [] has quit [Quit: leaving]
17:57:38  *** funnel [] has joined #openttd
18:00:37  *** Pensacola [] has quit [Remote host closed the connection]
18:13:17  <DorpsGek> Commit by peter1138 :: r26953 trunk/src/news_gui.cpp (2014-10-04 18:13:10 UTC)
18:13:18  <DorpsGek> -Codechange: GUI-scale for company manager face on news popup.
18:19:28  <DorpsGek> Commit by peter1138 :: r26954 /trunk/src (ai/ai_gui.cpp newgrf_gui.cpp) (2014-10-04 18:19:22 UTC)
18:19:29  <DorpsGek> -Codechange: GUI-scale for AI/GS settings and NewGRF settings windows.
18:32:03  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
18:38:37  *** theholyduck [~theholydu@] has joined #openttd
18:50:46  *** theholyduck [~theholydu@] has quit [Ping timeout: 480 seconds]
18:56:05  *** theholyduck [~theholydu@] has joined #openttd
19:01:09  *** Supercheese is now known as Guest516
19:01:10  *** Supercheese [~Superchee@] has joined #openttd
19:05:24  *** Guest516 [~Superchee@] has quit [Ping timeout: 480 seconds]
19:06:36  *** Alberth [] has left #openttd []
19:23:50  <DorpsGek> Commit by rubidium :: r26955 /trunk (3 files in 2 dirs) (2014-10-04 19:23:43 UTC)
19:23:50  <DorpsGek> -Fix [FS#6126]: owner of road depot road types were not properly changed upon bankruptcy causing crashes when trying to remove the depot in certain situations
19:28:40  <Supercheese> What on Earth is going on with the train depot tooltip, "Trains - drag vehicle with left-click to add/remove from train, right-click for information. Hold Ctrl to make both functions apply to the following chain"
19:28:50  <Supercheese> What is "both functions" referring to?
19:29:10  <Supercheese> I get that Ctrl allows you to remove the remaining cars from a trian
19:29:17  <Supercheese> after the one dragged
19:29:30  <Supercheese> but what is the other "function"?
19:29:57  <Supercheese> train*
19:30:03  *** DabuYu [DabuYu@] has joined #openttd
19:30:44  <frosch123> Supercheese: both: left- and right-click
19:30:58  <frosch123> left-click: drag single vehicle
19:31:08  <frosch123> ctrl-left-click: drag vehicle chain
19:31:14  <frosch123> right-click: show vehicle info
19:31:23  <frosch123> ctrl-right-click: show vehicle chain info
19:31:24  <Supercheese> oh hot dang
19:31:33  <Supercheese> I have never done ctrl+right click before
19:31:37  <Supercheese> neat
19:31:50  <frosch123> yeah, it's the only place with that key combination :p
19:32:15  <Supercheese> Well, I've got my Something New Learned Today done early it seems
19:33:16  <peter1138> overloaded tooltips!
19:33:20  <peter1138> I didn't know about that either.
19:33:31  <Supercheese> Heh
19:33:32  <frosch123> we need html formatting in tool tips :)
19:33:46  <frosch123> <ul> and <b> and such
19:34:00  <Supercheese> you have color codes already
19:34:10  <Supercheese> {RED} THIS IS IMPORTANT
19:34:11  <frosch123> but we have no <ul><li>
19:34:20  <Supercheese> :P
19:42:37  <peter1138> Ok... 2x is fine, but with 4x, the settings circle icon is not visible, or flickers :S
19:44:54  <frosch123> the expand/collapse circle in adv settings?
19:46:11  <peter1138> Ah, it's just an overflow :)
19:46:47  <frosch123> where is the int8? :p
19:47:15  <peter1138> line isn't high enough, so it wraps, but it's unsigned, so... boom.
19:52:01  <Supercheese> Huh, another typo...
19:52:30  <DorpsGek> Commit by peter1138 :: r26956 trunk/src/cheat_gui.cpp (2014-10-04 19:52:23 UTC)
19:52:30  <DorpsGek> -Codechange: GUI-scale for cheat window.
19:57:54  <Supercheese>   easy fix
20:11:23  <peter1138> Supercheese, probably should the "shares"
20:11:32  <peter1138> Like shift-click is "shows"
20:12:22  <Supercheese> Well, the three other strings immediately following all say, "Ctrl+Click will share the orders."
20:12:30  <Supercheese> the TRAIN_DEPOT is the odd man out
20:12:34  <peter1138> ok
20:12:58  <Supercheese> One of these strings is not like the other... ♪♫
20:18:32  <DorpsGek> Commit by peter1138 :: r26957 trunk/src/station_gui.cpp (2014-10-04 20:18:25 UTC)
20:18:33  <DorpsGek> -Codechange: Spread the station cargo icons out by GUI scale.
20:26:16  *** EyeMWing [] has quit []
20:34:01  <andythenorth> peter1138: so is there a bit I can flip to turn on GUI scale?
20:34:51  <DorpsGek> Commit by rubidium :: r26958 /branches/1.4 (30 files in 12 dirs) (2014-10-04 20:34:41 UTC)
20:34:52  <DorpsGek> [1.4] -Backport from trunk:
20:34:53  <DorpsGek> - Cleanup [Squirrel]: remove many unused bits and pieces
20:34:56  <DorpsGek> Commit by peter1138 :: r26959 /trunk/src (settings_gui.cpp settings_gui.h) (2014-10-04 20:34:43 UTC)
20:34:57  <DorpsGek> -Codechange: GUI-scale for settings window.
20:35:04  <peter1138> yikes
20:35:16  <andythenorth> timing
20:35:17  <Supercheese> concurrent commits
20:35:26  <peter1138> Oh, 1.4 backport, that won't be in trunk :)
20:35:48  <peter1138> andythenorth, only if you compile it
20:36:19  *** heffer_ [~quassel@2001:41d0:2:b5ce:de1::1] has joined #openttd
20:36:19  <andythenorth> I can compile
20:37:53  * andythenorth looks
20:38:13  *** Yotson [~Yotson@2001:980:6ac8:1:e0ce:47ac:1ba9:af2c] has quit [Quit: .]
20:40:19  *** Jomann [] has quit [Remote host closed the connection]
20:41:53  <andythenorth> so I need to set ZOOM_LVL_GUI somewhere?
20:41:58  <peter1138> zoom_type.h
20:43:16  *** heffer_ [~quassel@2001:41d0:2:b5ce:de1::1] has quit [Remote host closed the connection]
20:43:17  *** heffer_ [~quassel@2001:41d0:2:b5ce:de1::1] has joined #openttd
20:43:44  <andythenorth> got it
20:44:30  <andythenorth> slow compie is slow :(
20:45:03  <andythenorth> that is *fricking awesome*
20:45:21  <Supercheese> If I had a nickel for every time andy complained about slow compile times... well, I wouldn't have too much, but maybe enough to buy a candy bar
20:45:33  <Supercheese> :P
20:46:40  <andythenorth> at the price of candy in the UK
20:46:45  <andythenorth> you’d have half a mars bar
20:47:18  * andythenorth tries 2X
20:47:28  <andythenorth> NORMAL is a bit intense on a 13” screen
20:47:30  <Rubidium> are they that cheap in the UK?
20:47:35  <peter1138> haha
20:47:50  <andythenorth> best thing in the game since the last best thing
20:48:03  <andythenorth> the last best thing was window size pinning
20:48:03  <peter1138> Changing that define did mostly work anyway
20:48:12  <peter1138> Well, until you actually got to playing the game.
20:48:41  <peter1138> Need to sort out the close box :p
20:49:48  <andythenorth> oh it’s beautiful
20:49:53  <andythenorth> 2x fits perfectly on my screen
20:50:03  <andythenorth> it’s like 1994 again
20:50:09  * andythenorth is stuck in the past :(
20:50:19  <Supercheese> Party like it's 1995
20:50:56  <andythenorth> oh my eyes
20:51:10  <peter1138> Of course, it someone™ provides 2x scaled sprites, that'll not be so blocky.
20:51:13  <andythenorth> peter1138: you should see this on a retina-ish screen
20:51:13  <peter1138> *if
20:51:22  <andythenorth> the blocky is nice
20:51:27  <andythenorth> it’s so crisp and bright
20:51:31  * andythenorth can’t contain self
20:51:35  <andythenorth> excuse me for a minute
20:51:39  <peter1138> I only have ~85 dpi 22" screens.
20:53:24  <andythenorth> yeah everything should be redrawn to 2s
20:53:26  <andythenorth> 2x
20:53:29  <andythenorth> let’s just overlook that
20:53:47  <peter1138> biggui has some stuff already
20:54:05  <peter1138> i guess someone™ will merge that into opengfx or something
20:54:07  <andythenorth> I mean everything :P
20:54:10  <andythenorth> FIRS and all
20:54:28  <peter1138> :S
20:57:54  <Supercheese> voxels
20:58:01  <peter1138> cubicles
20:58:01  <Supercheese> solve all scaling problems :P
20:58:12  <frosch123> damn, peter was faster
20:58:15  <Supercheese> (and introduce many other problems)
20:58:17  <Rubidium> vauxhauls?
20:58:32  <Supercheese> foxholes
21:02:12  *** oskari892 [] has joined #openttd
21:03:24  <Supercheese> Hmmm, genders not available in conditional order strings
21:05:54  *** Pikka [] has joined #openttd
21:07:50  <Supercheese> ah, isn't a problem
21:08:03  <Pikka> yim bom bala boo
21:08:44  <frosch123> Pikka: we need svg graphics for all gui icons!
21:09:19  <Pikka> we do?
21:09:22  *** oskari89 [] has quit [Ping timeout: 480 seconds]
21:09:30  <Supercheese> or other vector format
21:09:32  <frosch123> maybe? :)
21:10:31  <frosch123> hmm, wait... i think that android guy even started drawing some
21:11:31  <frosch123> <- indeed
21:11:47  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
21:13:45  *** heffer [felix@2a01:4f8:160:9ffe:5054:ff:fe23:5f04] has quit [Remote host closed the connection]
21:15:43  *** heffer_ [~quassel@2001:41d0:2:b5ce:de1::1] has quit [Remote host closed the connection]
21:15:45  *** heffer [] has joined #openttd
21:25:18  *** heffer [] has quit [Ping timeout: 480 seconds]
21:29:11  <peter1138> Hmm, so just the X close button...
21:29:21  <peter1138> And a couple of little bits.
21:30:01  <peter1138> Hmm, I found the news icon, but it flashes up very briefly...
21:30:03  <Supercheese>
21:30:07  <Supercheese> Impressive
21:30:14  <Supercheese> (as frosch mentioned)
21:30:49  <peter1138> nice
21:31:01  <Supercheese> there's your vector format
21:31:05  <frosch123> peter1138: it's probably a silly icon :p
21:34:55  *** sla_ro|master [slamaster@] has quit []
21:36:59  <peter1138> There's a bug there. It's sending SBI_DELETED all the time
21:37:00  *** heffer [] has joined #openttd
21:37:30  <peter1138> SBI_NEWS_DELETED i mean
21:37:46  <frosch123> you still mean the flashing icon?
21:37:50  <peter1138> yes
21:37:55  <frosch123> what else should it do but flash?
21:38:01  <frosch123> stay forever?
21:38:10  <peter1138> It's meant to stay there for a second, maybe 2.
21:38:18  <frosch123> oh
21:38:22  <peter1138> It's literally flashing for a frame.
21:39:31  *** DabuYu [DabuYu@] has joined #openttd
21:40:48  <frosch123> it worked around 1.2 :p
21:41:07  <frosch123> i think we replaced some generic circle with that special sprite back then
21:41:14  <peter1138> yeah
21:41:17  * Supercheese recalls the circle
21:42:20  <peter1138> yeah...
21:42:25  <peter1138> it hmm
21:43:39  <andythenorth> is pikka chops
21:43:51  <andythenorth> what ho
21:43:51  <Pikka> frequently
21:43:57  <andythenorth> NARS 3 yet?
21:44:01  <Pikka> 2.5
21:44:08  <Pikka> wogans progress
21:44:23  <andythenorth> what are you waiting for?
21:44:28  <andythenorth> I have drawn a whole wagon today
21:44:36  <Pikka> I'm not drawing any wagons
21:44:38  <andythenorth> and spent 8 hours making my compile 5 seconds faster
21:44:38  <Pikka> only coding
21:44:43  <Pikka> huzzah
21:44:51  <andythenorth> coding schmoding
21:45:10  <Pikka> you only have to compile it six thousand times and it will totally have been worth it
21:45:41  <Pikka> well I can't release the set without schmoding
21:46:11  <Pikka> I know there's a school of thought which says the way to make progress is to post sprites in the forum, but...
21:47:07  <frosch123> do they still exist?
21:47:20  <andythenorth> can’t you schmode a schmode generator?
21:47:28  <andythenorth> frosch123: sprites exist
21:47:30  <Pikka> people who just post sprites in the forums, frosch123?
21:47:46  <frosch123> i do not follow forums much :p
21:47:54  <andythenorth> any reason? o_O
21:48:10  <Pikka> too full of bullies, andythenorth
21:48:16  <Supercheese> and drama
21:48:17  <frosch123> i thought everyone ieither posts them to devzone hg, or keeps them private
21:48:33  <frosch123> andythenorth: it's just not interesting
21:49:17  <Wolf01> 'night
21:49:26  *** Wolf01 [] has quit [Quit: Once again the world is quick to bury me.]
21:52:02  <andythenorth> frosch123: could I make it interesting somehow? o_O
21:52:06  <andythenorth> or is it the wrong question?
21:52:45  <frosch123> everything of relevance is said or linked here
21:53:25  <andythenorth> order is very important in that sentence
21:53:33  <andythenorth> try rearranging it, it’s definitely not true
21:53:50  <andythenorth> (it becomes definitely not true...)
21:53:58  <frosch123> you mean "linked from" ?
21:54:31  <andythenorth> here everything said is of relevance or linked
21:54:32  *** Progman [] has joined #openttd
21:54:35  <frosch123> anyway, if there is something interesting on the forums, it's usually carried here
21:55:20  <andythenorth> peter1138: that newspaper at the end is giant!
21:55:32  <frosch123> graphics dev forum was never particulary interesting though
21:55:45  <andythenorth> shocking assertion
21:55:59  <andythenorth> where would andythenorth be without it?
21:56:41  <frosch123> i look into your topics every month, i find nothing that you didn't say here as well :)
21:57:19  <andythenorth> well I can’t fill up the forums with 5-line teddy-bear coding thoughts
21:57:27  <andythenorth> also, in the forums, I might get answers
21:57:31  <andythenorth> which is not the point
21:57:41  <frosch123> wasn't there a forum game subforum?
22:02:18  *** oskari89 [] has joined #openttd
22:02:28  <peter1138> andythenorth, isn't it just
22:02:36  *** oskari89 [] has quit []
22:08:18  <andythenorth> frosch123: there are dark bits of the forum I never see
22:08:28  <andythenorth> I believe there’s all kinds of things
22:08:39  <andythenorth> I found a thing called ‘locomotion’ in there the other day
22:08:43  <andythenorth> looked a bit odd
22:08:57  <andythenorth> also I think there’s a racism and sexism sub-forum
22:09:01  *** oskari892 [] has quit [Ping timeout: 480 seconds]
22:09:04  <andythenorth> or is that #tycoon?
22:10:08  <SpComb> the dark side of the forums
22:12:10  <andythenorth> huzzah
22:12:13  <andythenorth> giant cursor is giant
22:12:21  <Pikka> at least off-topic, forum games and locomotion keeps the kids out of the proper forums :) most of the time...
22:12:43  <Pikka> how giant?
22:13:35  <andythenorth> as big as my little finger
22:13:40  <andythenorth> well, the tip of it
22:14:02  <frosch123> night
22:14:05  *** frosch123 [] has quit [Quit: be yourself, except: if you have the opportunity to be a unicorn, then be a unicorn]
22:14:22  <andythenorth> need giant industry chain view
22:14:26  <andythenorth> actually maybe not
22:14:47  <andythenorth> giant small map is giant too \o/
22:17:15  * andythenorth ponders the UI chrome icons
22:17:24  <andythenorth> maybe they just need drawing better
22:21:08  * andythenorth bulldozes some rivers
22:21:09  <andythenorth> in my way
22:21:22  <andythenorth> 2x zoom dynamite!
22:21:43  <andythenorth> with animation!
22:22:13  <andythenorth> feature request: demolishing a station shows explosion on all tiles
22:22:21  <andythenorth> a bit like ‘realistic’ vehicle smoke :P
22:22:31  <andythenorth> increase the explosion effect sprite limit
22:23:22  <andythenorth> also bed time
22:23:23  <andythenorth> :P
22:23:35  *** andythenorth [] has quit [Quit: andythenorth]
22:33:28  *** HerzogDeXtEr [~flex@] has quit [Quit: Leaving.]
22:43:49  *** matma6 [] has joined #openttd
22:43:57  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
22:43:59  <matma6> hello
22:44:47  <matma6> i'm curious how openttd detects where mouse points
22:45:26  <matma6> i tried to read source code, but I didn't get anything
23:02:09  <FLHerne> matma6: Looks like it's in window.cpp
23:02:39  <matma6> does version matter?
23:04:00  <FLHerne> Hmm, that only uses the values. Probably comes out of SDL somewhere
23:04:38  *** Progman [] has quit [Remote host closed the connection]
23:08:41  <matma6> But, do you know the concept standing behind this method
23:09:10  <matma6> also, I can find OnHover in window.cpp - is it part of SDL?
23:09:21  <matma6> *can't
23:10:58  <matma6> Please look at line 779
23:11:05  <FLHerne> Looks like the mouse position is found separately by the platform-specific VideoDriver code
23:11:06  <matma6> 	w->OnHover(pt, wid->index);
23:13:21  <FLHerne> e.g win32_v.cpp has 'case WM_MOUSEMOVE:' and then sets the cursor values
23:13:58  *** DabuYu [DabuYu@] has joined #openttd
23:14:03  <glx> then main window pass it to the top children
23:15:02  <FLHerne> matma6: It's a function of struct Window, not sure what that does
23:15:04  *** fjb is now known as Guest535
23:15:05  *** fjb [] has joined #openttd
23:15:08  <matma6> i think video level is to low
23:15:46  <matma6> this counts some deltas and then calls HandleMouseEvent()
23:16:31  <glx> HandleMouseEvent() redirects the event where it should
23:16:43  <matma6> ok
23:16:47  *** blathijs [] has quit [Ping timeout: 480 seconds]
23:16:48  <matma6> Where is it
23:17:17  <matma6> ok
23:17:21  <matma6> i got it
23:17:57  <glx> and events are propagated through windows
23:18:14  <glx> until one element handles it
23:19:36  <matma6> I think that sprite picker does it, am I right?
23:21:21  <matma6> oh, am i right?
23:21:24  <matma6> i click
23:21:43  <matma6> some kind of listener is set
23:21:57  <matma6> whole screen is dirty and needs redraw
23:22:07  *** Guest535 [] has quit [Ping timeout: 480 seconds]
23:22:14  <matma6> listener waits, when coords are good
23:22:38  <matma6> and then remembers actually drawn sprite -> actually pointed
23:23:15  <glx> not really a listener, but a special status is defined and if left click happens it stores the clicked pixel
23:24:26  <glx> sprite picker stuff is always handled before GUI
23:25:53  <matma6> so when pixel is drawn, then sprite is captured?
23:28:09  <glx> yes
23:28:48  <matma6> ok
23:29:09  <matma6> so mouse position is done
23:29:24  <glx> every sprite drawn on the stored pixel are added to the list
23:29:26  <matma6> *so mouse position is computed by the way while drawing
23:29:56  *** Brumi [] has quit []
23:31:37  <matma6> right?
23:32:45  <glx> mouse position is determined by video driver
23:37:48  <matma6> so thx for help
23:38:30  *** gelignite [] has quit [Quit:]
23:42:03  *** Flygon_ [] has joined #openttd
23:42:28  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
23:46:41  *** FLHerne [] has quit [Quit: There's a real world out here!]
23:48:26  *** Flygon [] has quit [Ping timeout: 480 seconds]
23:48:51  *** Flygon [] has joined #openttd
23:50:13  *** DabuYu [DabuYu@] has joined #openttd
23:50:31  *** Flygon_ [] has quit [Ping timeout: 480 seconds]
23:51:03  *** Myhorta [] has quit [Remote host closed the connection]
23:54:14  *** matma6 [] has quit [Quit: Page closed]

Powered by YARRSTE version: svn-trunk