IRC logs for #openttd on OFTC at 2014-06-17
⏴ go to previous day
00:01:10 *** KWKdesign has joined #openttd
01:04:58 *** luaduck_zzz has joined #openttd
01:05:14 *** luaduck_zzz is now known as luaduck
02:45:54 *** Midnightmyth has joined #openttd
04:05:11 *** Hazzard_ has joined #openttd
04:56:18 *** Eddi|zuHause has joined #openttd
05:06:49 *** KWKdesign has joined #openttd
05:44:28 *** DiabloD3 has joined #openttd
05:44:39 *** DiabloD3 is now known as Diablo-D3
05:46:22 *** tokai|mdlx has joined #openttd
06:13:43 <DorpsGek> Commit by fonsinchen :: r26650 trunk/src/saveload/saveload.cpp (2014-06-17 06:13:38 UTC)
06:13:44 <DorpsGek> -Fix [FS#6041]: Support save/load chunk lengths of up to (1 << 32) - 1
06:40:09 *** andythenorth has joined #openttd
06:43:31 <andythenorth> I’ve forgotten everything I know about nfo
06:43:37 <andythenorth> what kind of action is this?
06:43:38 <andythenorth> 51 * 11 00 00 \b1 01 FF \wx0546 1A FF \wx0549
06:46:42 *** Myhorta has joined #openttd
06:52:05 <andythenorth> it looks like buy menu sort order
06:52:10 <andythenorth> nforenum hates it
06:52:46 <planetmaker> I'd recon it's properties 01 and 1A as extended bytes
06:54:33 <Eddi|zuHause> you're saying "change 1 property" and then set two?
06:55:05 <andythenorth> it’s nmlc output
06:55:29 <planetmaker> though I thought it would meant to change one ID
06:55:38 <planetmaker> too long ago that I wrote or read nfo
06:56:42 <andythenorth> it’s definitely generated by sort()
06:56:50 <Eddi|zuHause> ah, so it's not prop 01
06:57:00 <Eddi|zuHause> it's 01 props of ID FF \wxblah
06:57:23 <planetmaker> why then \b1 01 ?
06:57:38 <andythenorth> maybe I have a really old nforenum
06:57:38 <Eddi|zuHause> one ID, and one prop
06:57:39 <andythenorth> /!!Error (141): ID 12 07 out of valid range (00..73).
06:57:49 <andythenorth> since when is 73 the max ID for a train?
06:57:53 <Eddi|zuHause> andythenorth: too old nforenum, i suppose
06:57:55 <planetmaker> you got an ancient nforenum?
06:58:13 <Eddi|zuHause> andythenorth: or ancient data files in some config directory
06:58:16 <andythenorth> NFORenum trunk r958
06:58:20 <Eddi|zuHause> i think i had that problem a long time ago
06:59:14 <andythenorth> this is like archeology
06:59:51 <planetmaker> NFORenum 6.0.4 r980 - Copyright (C) 2004-2013 by Dale McCoy
07:02:33 <planetmaker> public [970:efcb99e7b53c default] 2013-11-08 23:53 +0100
07:02:33 <planetmaker> Fix #5279: allow higher vehicle IDs when a GRF version 8 NFO is given
07:03:19 <andythenorth> so it’s my nforenum
07:14:02 <andythenorth> the realsprites are all garbled
07:14:24 <andythenorth> and the railtypes are borked
07:14:38 <andythenorth> and as expected, many strings are failing
07:15:12 <planetmaker> yes, the pesky details nml guards you against ;)
07:15:47 <andythenorth> I wonder what goes wrong in the nml->nfo step
07:16:16 <andythenorth> I need to compare the linked and non-linked nfo
07:16:22 <andythenorth> but anyway, later
07:16:23 <andythenorth> thanks for the help
07:16:42 <Eddi|zuHause> what's wrong with the sprites?
07:21:39 <andythenorth> crops are wrong I think
07:21:58 <andythenorth> again, that nfo is pure nmlc output
07:22:09 <andythenorth> so something is transposed
07:23:47 <andythenorth> aren’t width and height reversed in nfo?
07:25:35 <Eddi|zuHause> although i've never seen someone use nfo32 and then produce a v7 grf
07:28:37 <andythenorth> some crops are coming out with 1 1
07:52:18 *** jjavaholic has joined #openttd
07:54:54 *** Myhorta has joined #openttd
07:56:25 *** andythenorth has joined #openttd
07:56:44 <andythenorth> why is nmlc outputting train prop 1D 3 times?
08:09:05 <andythenorth> give this realsprite
08:10:34 <andythenorth> 8680 generated/graphics/tank_car_ng_brit_gen_1_0.png 8bpp 60 10 8 25 -3 -21 normal
08:10:52 <andythenorth> I get 200px of white reported
08:11:16 <andythenorth> but the 200px (8 * 25) at x = 60, y = 10 are not white
08:11:53 <andythenorth> it’s an 8bpp paletted png
08:12:45 <andythenorth> top of nfo file contains “// (Info version 32)”
08:14:01 *** AndroUser2 has joined #openttd
08:19:13 *** DiabloD3 has joined #openttd
08:24:36 <andythenorth> well maybe I’ll spot the issue later :P
08:28:20 *** TheMask96 has joined #openttd
08:31:32 *** montalvo has joined #openttd
09:02:20 *** Flygon_ has joined #openttd
09:03:06 *** TheMask96 has joined #openttd
09:16:13 *** montalvo_ has joined #openttd
09:37:21 *** sla_ro|master has joined #openttd
09:42:10 *** Dan9550 has joined #openttd
09:52:20 *** HerzogDeXtEr1 has joined #openttd
11:11:19 *** montalvo has joined #openttd
11:53:52 *** Myhorta has joined #openttd
12:02:37 *** pthagnar has joined #openttd
12:18:35 *** Myhorta has joined #openttd
13:38:17 *** Eddi|zuHause is now known as Eddi|zuHause2
13:54:19 *** Eddi|zuHause2 is now known as Eddi|zuHause
14:09:00 *** Myhorta has joined #openttd
14:52:21 *** andythenorth has joined #openttd
14:59:23 <andythenorth> I’ve done something to make grfcodec barf on my realsprites
14:59:27 <andythenorth> it thinks they’re all white
15:01:14 <planetmaker> maybe they are? As the path to the real sprites is wrong?
15:01:25 <andythenorth> that appears to be the case
15:01:30 <andythenorth> simply not using the right realsprite
15:01:34 <andythenorth> dunno why though
15:03:41 <andythenorth> where’s the action 3 :P
15:04:01 *** Alberth has joined #openttd
15:04:01 *** ChanServ sets mode: +o Alberth
15:05:15 <andythenorth> planetmaker: got 5 mins to verify something for me?
15:07:27 <andythenorth> with Iron Horse, make nfo from iron-horse.nml with nmlc, and encode with grfcodec....
15:07:58 <andythenorth> my nml and iron-horse repos are both in a state of ‘work in progress’ :P
15:09:09 * Alberth got here in time today?
15:09:54 * andythenorth has broken something
15:10:46 * andythenorth wonders why grfcodec wants null_trailing_part.png so many times
15:13:52 <andythenorth> I thought it was valid to do: realsprites, action 2, more realsprites, action 2, varaction 2, action 3
15:16:12 <Eddi|zuHause> andythenorth: some versions of grfcodec had issues with interlaced pngs
15:16:35 <Eddi|zuHause> but that affected only 32bpp for me
15:16:37 <andythenorth> I think this might be EAndythenorth somewhere
15:16:40 <andythenorth> waiting to see what pm finds
15:16:53 <planetmaker> so, andythenorth, I don't get any white pixel warnings with grfcodec
15:17:17 <andythenorth> ok so the problem is newly introduced thanks
15:18:01 <andythenorth> planetmaker: which grfcodec?
15:18:04 <planetmaker> after I installed python3-ply, python3-pillow and found the lang dir and how to teach nmlc about it :P
15:18:24 <planetmaker> it should warn me about them, should it?
15:20:55 *** tokai|noir has joined #openttd
15:20:55 *** ChanServ sets mode: +v tokai|noir
15:20:58 <andythenorth> the pattern of sprites loaded is different for me
15:21:04 <andythenorth> far more null_trailing_part.png
15:21:30 <andythenorth> I’ve broken something :P
15:22:10 <planetmaker> I ran make on a clean iron-horse. nml2nfo'ed the nml. And then grfcodec'ed the resulting nfo.
15:23:44 <andythenorth> it’s good that it works
15:23:54 <andythenorth> fixing my code is easier than fixing grfcodec or nmlc :P
15:24:40 <planetmaker> you did not try with a clean checkout?
15:25:10 <planetmaker> local clones are dirt cheap to get by...
15:25:41 <planetmaker> as you've got an OS with a sensible file system it only costs a few links
15:25:45 <andythenorth> I might try that later, right now I have my own nmlc too, with uncommitted changes
15:25:51 <andythenorth> I could fix it I guess
15:26:05 <planetmaker> well, clone that, too ;)
15:29:35 <andythenorth> are there any ill effects expected from repeating the grfcodec header comments block?
15:31:00 <andythenorth> planetmaker: may I have the nfo you generated? o_O
15:37:24 <planetmaker> ./3443.nfo:1: Warning: Found 9036 fewer sprites than sprite 0 reports.
15:37:24 <planetmaker> Loading generated/graphics/chaplin_0.png(Transparency:100%, Redundancy:100%)
15:37:43 <planetmaker> no white sprites though
15:38:11 <andythenorth> the current expected failure is a spritesheet not being tall enough
15:38:17 * andythenorth is fighting bees
15:38:39 <planetmaker> Loading generated/graphics/chaplin_0.png(Transparency:100%, Redundancy:100%)
15:40:03 <andythenorth> maybe the problem really is the sprites
15:40:17 <andythenorth> nfo always does that
15:40:52 <planetmaker> it does. Thus I dearly avoid it
15:41:12 <planetmaker> reminds me... I wanted to do a station set
15:46:34 <andythenorth> planetmaker: it says nothing about loading null_trailing_part.png?
15:50:56 <andythenorth> that is the probelm
15:53:25 <andythenorth> compiling 6.0.3 doesn’t resolve it
15:53:38 <andythenorth> lots of compiler warnings too
15:57:11 <andythenorth> planetmaker: can I have your grfcodec 6.0.3 binary?
15:58:16 *** HerzogDeXtEr has joined #openttd
15:59:45 *** Hazzard has joined #openttd
15:59:59 *** Progman has joined #openttd
16:00:21 <planetmaker> andythenorth: how will that help you? It's build non-static for my OS (debian wheezy x64)
16:00:39 <andythenorth> sorry, assumed you had an OS X compiled version :P
16:01:03 <planetmaker> even then, it would be non-static for 10.6
16:01:11 <planetmaker> that wouldn't work on your machine either
16:01:24 <andythenorth> so grfcodec is game-overed currently for me
16:02:38 <planetmaker> you got libpng and zlib available?
16:02:46 <planetmaker> thus yours is built with those?
16:03:38 *** TheMask96 has joined #openttd
16:06:06 <andythenorth> just installed zlib
16:06:13 <andythenorth> fewer compile warnings now (just one)
16:06:23 <andythenorth> but still the spritesheet lines error
16:09:46 <andythenorth> I don’t have this issue with road-hog
16:09:57 <andythenorth> grfcodec works fine there, produces a valid grf
16:21:14 <Alberth> all sprites are the same?
16:23:10 <andythenorth> between the two grfs? Or in Iron Horse?
16:25:33 <Alberth> I thought the error happened when you process 2 sprites at the same time? (didn't carefully read it though, could be wrong)
16:25:50 <Alberth> if all sprites have the same layout, you won't notice it
16:28:59 <andythenorth> this is plausible
16:29:26 <andythenorth> there are multiple spritesheets for each action 3 in Iron Horse
16:29:36 <andythenorth> whereas only one per action 3 in Road Hog
16:29:44 *** Midnightmyth has joined #openttd
16:30:25 <andythenorth> it’s as though it picks up the wrong set of bounding box sizes etc
16:36:03 <andythenorth> if I find-and-replace to just one spritesheet, it compiles
16:37:04 <Alberth> you doubted the existence of the bug?
16:37:37 <planetmaker> seems to work on linux, though, Alberth :)
16:37:56 <andythenorth> I wanted to isolate the cause :P
16:38:21 <andythenorth> so do I finally have to give in, and do newgrf dev in virtualbox?
16:38:26 <Alberth> planetmaker: interesting :)
16:38:52 <planetmaker> but dunno the amount of local modifications andy has in each of the tools .P
16:39:15 <andythenorth> grfcodec, zero :P
16:39:51 <andythenorth> I think we can assume grfcodec is game-over on OS X Mavericks
16:40:06 <planetmaker> I wouldn't understand why
16:40:59 <Alberth> downgrading to 0.6.3 not an option ?
16:41:13 <andythenorth> 6.0.3 shows same issue for me
16:41:26 <andythenorth> I have to compile it myself, and I suspect the issue is the compile
16:41:46 <andythenorth> I assume it’s building with clang or whatever
16:41:58 <andythenorth> I could build gcc, and try forcing it to compile with gcc
16:42:27 <andythenorth> the nfo planetmaker generated also fails
16:42:40 <andythenorth> the only other thing it could be is the pngs, and I see nothing wrong there
16:42:40 <Alberth> hmm, good point, no idea how stable clang is
16:43:02 <Alberth> or pnglib, but that's very unlikely as well
16:43:04 <andythenorth> none of you can compile with Mavericks, and none of us with Mavericks can fix it :)
16:44:25 <Alberth> if it's really the compiler I wouldn't know how to debug that
16:44:39 <Alberth> let alone fixing it :)
16:45:01 <planetmaker> andythenorth, nml has a regression folder. Do those nfo work for you with grfcodec?
16:45:07 <planetmaker> with nml as well?
16:45:44 <andythenorth> hmm issue occurs with pcx too
16:46:02 <Alberth> that rules out the png lib stuff
16:46:17 <andythenorth> currently my nml tests are broken
16:46:17 <planetmaker> andythenorth, can you make a minimal test case?
16:46:21 <andythenorth> but I’ve discounted my nml
16:46:27 <andythenorth> I have planetmaker’s nfo
16:46:38 <planetmaker> why are nml tests borked?
16:46:54 <andythenorth> because I have a patch
16:47:02 <andythenorth> I could fix that, but I think it’s non-relevant
16:47:12 <andythenorth> your nfo fails too
16:47:13 <planetmaker> for grfcodec, sure is
16:47:42 <andythenorth> reminds me to make my patch pass tests though :P
16:53:28 *** frosch123 has joined #openttd
16:56:05 <andythenorth> it’s exactly as snail says - as though grfcodec is garbling which spritesheet to use
16:58:55 * andythenorth considers restructuring the grf to a single spritesheet
16:59:02 <Alberth> since not all compilers do that, it's likely to be either a bug in the OSX compiler, or the source relies on undefined behavior
16:59:28 <Alberth> where gcc then does a happy thing, and apple does the non-happy thing
16:59:54 <Alberth> throw a little pillow into the equation?
17:01:07 <andythenorth> I am seriously considering using PIL to concatenate all the spritesheets to one
17:01:22 <andythenorth> it’s probably not The Right Solution
17:02:32 <planetmaker> andythenorth, did you try to build grfcodec with gcc?
17:02:46 <planetmaker> on my machine I have like gcc_select which allows me to choose the version of gcc to use
17:02:50 <andythenorth> I’d need to get gcc I think
17:03:07 <planetmaker> obviously it will need to be different with clang. But getting gcc should not be hard, no?
17:03:25 <planetmaker> might not be default (anymore), but not at all? Honestly dunno, just asking
17:03:51 <planetmaker> appstore might know :)
17:03:53 <Alberth> if it's a nice ./configure tool, CC=gcc ./configure may be enough
17:04:47 <Alberth> although I may be underestimating the niceness of apple configurations :)
17:05:20 <andythenorth> searching implies that reinstalling gcc might be a bad idea
17:07:58 <planetmaker> Alberth, that seemed to have worked for me. A few years ago
17:08:38 <andythenorth> I think I’m going to not install gcc
17:08:47 <andythenorth> I don’t have recent backups, and Bad Things Might Happen
17:09:17 <planetmaker> it's not like an obscure app
17:09:26 <planetmaker> which phones home and sells your children
17:09:58 <andythenorth> might break existing clang setup
17:10:18 <andythenorth> if I had backups I’d just do it
17:10:28 <andythenorth> but I’m away from home, using a phone to get net access
17:20:22 * andythenorth is out of things to google
17:20:28 <andythenorth> therefore accepting defeat
17:23:11 <DorpsGek> andythenorth: snail was last seen in #openttd 6 weeks, 3 days, 0 hours, 21 minutes, and 8 seconds ago: <Snail> tbh I have too much fun to draw manually
17:23:29 <DorpsGek> andythenorth: snail_ was last seen in #openttd 2 years, 2 weeks, 6 days, 22 hours, 41 minutes, and 49 seconds ago: <Snail_> yep, so the question is, would it be possible to implement labels for vehicles that could also be set through a callback
17:24:55 <Rubidium> so the 6884 grfcodec issue is purely OS X?
17:25:09 <andythenorth> probably only Mavericks
17:25:26 <andythenorth> my bet is on clang / LLVM
17:25:31 <andythenorth> I’ll put €20 on it
17:25:46 <andythenorth> and €10 on a dependency, as a hedge bet :P
17:26:51 *** oskari89 has joined #openttd
17:33:17 <frosch123> andythenorth: what does the finished grf look like?
17:33:29 <frosch123> is it correct, can you attach it too?
17:34:21 <frosch123> my bet is on different int promotion stuff
17:35:25 <frosch123> the grfcodec hacking style is prone to fail for such things
17:35:59 <Rubidium> clang 3.3 seems to do the right (tm) thing
17:36:16 <Rubidium> now going to replace 3.3 with 3.5
17:42:29 <Rubidium> unless my clang doesn't actually use llvm
17:43:21 <andythenorth> frosch123: there is no finished grf
17:43:25 <andythenorth> it bails during encode
17:43:59 <Rubidium> hmm... llvm-clang exists as well, but fails horridly upon linking
17:46:08 <planetmaker> hm... does clang compile openttd?
17:46:27 <planetmaker> or doesn't it link it? Likely I'm missing something...
17:47:05 <Rubidium> clang 3.3 on Debian compiles it
17:47:41 <planetmaker> CC=clang CXX=clang ./configure ?
17:50:03 <Rubidium> that's roughly the error I get for grfcodec when using llvm-clang
17:50:21 <Rubidium> though clang, which supposedly uses llvm (or not?!?) works just fine
17:51:38 <Rubidium> it's a typical "didn't link to stdlibc++"
17:55:57 <Rubidium> clang 3.5 compiles OpenTTD on Debian as well
17:57:47 <planetmaker> default on wheezy seems 3.0
17:58:03 <Rubidium> oh, use clang++ instead of clang
17:58:52 <planetmaker> doh. That's looking better :)
18:02:44 *** tycoondemon has joined #openttd
18:10:06 <andythenorth> OpenTTD does compile for me btw, using whatever clang Mavericks has
18:10:45 <planetmaker> with clang++ it does for me, too (clang 3.4)
18:20:25 *** Tramvai has joined #openttd
18:21:10 <Tramvai> Hey guys. I need some help with my trains: the damn trains will go through my unloading station without unloading and to go their home 'depot'
18:21:15 <Tramvai> Why doesn't ANY depot work?
18:21:23 <andythenorth> what is a home depot?
18:21:35 <Tramvai> I don't know why they prefer one depot over the other
18:21:43 <Tramvai> I think it counts as the depot I bought the train in?
18:22:51 <planetmaker> trains don't have that concept
18:22:57 <planetmaker> the whole game doesn't
18:23:10 <Tramvai> I did try making depots easy to access BEFORE the station.
18:23:10 <andythenorth> depot routing issues are usually due to signalling
18:23:30 <andythenorth> I turn off breakdowns + servicing, because depot routing is quite broken
18:23:33 <planetmaker> then give them explicit depot orders in their schedule. Then they'll just go there when you allow them
18:23:43 <planetmaker> otherwise they visit the nearest depot straight when they need
18:24:23 <Tramvai> The train feels like it needs service during its trip to the main station and then it suddenly decides to go to a depot that is AFTER the station and just ends up doing a very long run with a full load...
18:24:25 <Tramvai> It just blows my mind
18:26:47 <planetmaker> when it wants to visit a depot it goes there without stop
18:26:55 <planetmaker> thus, tell it when to visit one
18:27:09 <planetmaker> and make sure there's one where it then can go without long detour
18:28:18 <Tramvai> So, it would help if I just put some depots on the railroad in the middle of nowhere?
18:28:37 <Tramvai> So it can take its piss break there?
18:28:46 <andythenorth> turn off breakdowns :)
18:29:06 <planetmaker> Tramvai, please read again :)
18:29:13 <planetmaker> put the depot visit in the trains' orders
18:29:23 <Tramvai> I don't really feel like defining depots for all the trains I have. :S
18:29:26 <planetmaker> and make sure that then is a depot at the place they're told to visit it
18:30:20 <planetmaker> then make sure that depots are *before* the split of any lines
18:31:12 <andythenorth> so until I can install GCC, my faster-newgrf-compile project is halted :P
18:31:23 <andythenorth> Alberth: how is your lexer coming along? o-O
18:31:48 <Alberth> it fails at regression test 012 currently
18:31:58 <andythenorth> tests are useful :)
18:32:30 <V453000> assuming 90degree turns to be off to make the image work
18:33:26 <V453000> doesnt matter, shorter = better
18:33:31 <V453000> in case you need to autoreplace twice in a row :P
18:34:03 <V453000> if it is the default (150yrs), then you need to wait half a year before you can autoreplace again
18:34:14 <V453000> or just in case trains went through some depot like overflow in the half year
18:34:47 <Alberth> I just the default setting :)
18:34:54 <V453000> it is just a breaking condition which is nice to have so small that it doesnt prevent trains from autoreplacing ever, because you are in control of the depot places
18:35:40 <V453000> but you use breakdowns, which alone is punishable by death :P
18:35:46 <andythenorth> you should turn 90’ turns back on as well
18:35:49 <andythenorth> disabling them is stupid
18:35:51 <Alberth> Tramvai: I make all my junctions such that trains can go in any direction from any direction
18:36:18 <V453000> junctions: yes, terminus stations are fine though :P
18:36:35 <V453000> but you commented on YETI thread Alberth so you can live for now =D
18:37:20 <andythenorth> but V453000 must die for 90º turns being off
18:37:38 <V453000> I dont give a shit I can play with them on, but most people hayt it :P
18:37:48 <Alberth> he must, I also have them turned off, it looks toooooo stupid to allow
18:37:58 <andythenorth> yes but it breaks your ships
18:38:01 <V453000> I would just make my terminus not have that X for 90
18:38:03 <andythenorth> they will get lost on rivers
18:38:13 <V453000> it doesnt break MY ships :D nor MY rivers
18:38:16 <Alberth> nah, just make enough room for them to turn
18:38:18 <V453000> 0 ships, rivers disabled
18:39:39 <V453000> ps I got my BDMT factory sketched up (: just need to model it now :0
18:40:28 <andythenorth> BDMT - is that a sub-dom thing?
18:40:38 <andythenorth> I think you should keep your perversions out of the game
18:42:40 <V453000> building materials, same label as firs
18:42:44 <planetmaker> Alberth, I agree, too stupid. Yet 'on' is default still :(
18:43:06 <V453000> planetmaker: original acceleration is default, PBS is default, and many more are equally intelligent
18:43:20 <V453000> I thought the original acceleration is unquestionable
18:43:28 <planetmaker> pbs as default is quite reasonable. acceleration is not :)
18:43:56 <V453000> TWO way PBS is the most confusing signal in the whole game
18:44:10 <V453000> looks 1-way, behaves 2-way, has no use, only in specific cases
18:44:15 <V453000> -> makes sense to have it default?
18:44:27 <Alberth> obviously it does? :)
18:44:47 <V453000> I can accept that dumb people want to have 1-way PBS as default, obviously simple block would still be better, but 2-way PBS is just nonsense
18:45:03 <fonsinchen> I always use 2-way PBS it for terminus stations.
18:45:22 <V453000> 2-way PBS is only useful for 2-way stations, terminus can do without it
18:45:23 <fonsinchen> A lot of people build their first signals when building stations.
18:45:32 <Alberth> fonsinchen: yes, that's a specific case
18:45:49 <V453000> I am ignoring the penalty usage openttdcoop uses 2way PBS for
18:45:50 <planetmaker> maybe it's a confusing naming rather than a confusing signal?
18:46:05 <V453000> names dont help when people look at it
18:46:14 <V453000> still it is a signal without normal use
18:46:25 <Alberth> planetmaker: the signal is confusing, it's too easy to make a complete mess of your network
18:46:32 <V453000> making new player use it is really the worst option, perhaps aside combo signal :D
18:46:43 <fonsinchen> It's probably a good idea to have the signal be default that newbies use first
18:46:43 <planetmaker> well, terminus stations give them a normal use. But using them everywhere indeed is very stupid
18:46:47 <planetmaker> and will mess up things
18:46:58 <V453000> terminus stations _dont_ need 2way PBS either
18:47:02 <planetmaker> maybe the default pbs signal should be the 1-way pbs one
18:47:10 <planetmaker> they don't. Indeed
18:47:15 <Alberth> terminus can be done without 2-way pbs, for newbies
18:47:28 <fonsinchen> You can do without pbs there, but it's more complicated.
18:47:30 <V453000> for everyone, the 2way pbs is just wasted space there Alberth
18:47:36 <fonsinchen> you need pre/post signals then
18:47:49 <planetmaker> fonsinchen, no, you don't. You can do w/o any signals at terminus
18:48:01 <planetmaker> only entry 1-way pbs suffices
18:48:04 <Alberth> V453000: indeed, I never place them, except in pass-through stations used from both sides
18:48:14 <V453000> I think 1-way PBS would make a ton more sense. I still think it is vital for people to understand the system of block before diving into paths though.
18:48:22 <V453000> yes, that is what I call 2-way station Alberth
18:48:55 <fonsinchen> interesting ... I never though of building a terminus without signals
18:49:02 * fonsinchen goes and plays a bit
18:49:26 <Alberth> it saves an entire tile :)
18:49:47 <Alberth> although I still build a straight track first from a platfom, as it looks better
18:50:21 <fonsinchen> Also that exit is not exactly a safe waiting position. But I get the logic
18:50:51 <V453000> proper networks can assume exits to be always free :P
18:51:08 <andythenorth> I’d never thought about that
18:51:31 <fonsinchen> V453000: then you can place the exit signal(s) even closer to the station.
18:51:32 <Tramvai> V453000: Have they changed the minimum service time to 30 days now?
18:51:36 <andythenorth> also stations without signals?
18:51:52 <Alberth> fonsinchen: that exit looks like a block signal to me
18:52:04 <V453000> Tramvai: there might be some way to set it by clicking or editing config, but 30 is fine too
18:52:32 <Alberth> Tramvai: double click the line for a popup?
18:52:36 <Tramvai> Editing the cfg seems to be the only way... In-game it won't go lower than 30. Tried both using the arrows and editing through double click.
18:52:41 <V453000> fonsinchen: the station is Very far from perfect or even good, but yeah you can put it closer
18:52:59 <V453000> Tramvai: I did it somehow by some fiddling, but cant tell anymore :d 30 is fine tho
18:53:01 <Tramvai> Yeah, opened up the popup, tried 5, didn't change a thing. 30 is the lowest it accepts.
18:53:17 <V453000> could have been some combination of switching to %, setting it to 5 and then back
18:53:27 <V453000> doesnt really matter in the end, 1 month is very shoret
18:53:49 <Alberth> less than a minute real-time :p
18:55:39 <Alberth> Tramvai: you know V plays without breakdowns and without servicing, right?
18:56:16 <Alberth> ie his trains never go to a depot, normally
18:56:25 <Tramvai> I just don't understand why the breakdown and servicing system is that bad right now...
18:56:52 <Tramvai> Seems to be a fundamental part of the game
18:57:08 <Tramvai> Yet it's seriously broken... apparently
18:57:34 <Tramvai> Or are my train drivers retarded?
18:58:17 <Alberth> probably your expectations and how it really works are not well aligned :)
18:58:34 <Tramvai> I haven't really paid attention to it much before, but I've just recently noticed what my trains are doing... And it really brought my piss to a boil
18:58:58 <Tramvai> They seem to prefer some depots over the other... that's what's wrong
18:59:21 *** zeknurn has joined #openttd
19:00:57 <Alberth> if you tell them to get serviced every 30 days, they will try to do so every 30 days, to a depot that is near
19:01:09 <andythenorth> or a depot that is ‘near’ according to the pathfinder
19:01:13 <andythenorth> not physically near
19:01:19 <andythenorth> even if it involves going down a sidetrack
19:01:32 <andythenorth> or in the case of RVs, 20 tiles in the wrong direction
19:01:51 <DorpsGek> Commit by rubidium :: r26651 /trunk/src (3 files in 2 dirs) (2014-06-17 19:01:45 UTC)
19:01:52 <DorpsGek> -Fix: make sure an abs is used that supports int64 when using abs on those variables
19:01:54 <andythenorth> or for trains, into a station that they are too long for
19:02:08 <andythenorth> thereby blocking the network
19:02:34 <planetmaker> hehe, yeah. One of the reasons to prefer one-tainlength networks
19:02:49 <Alberth> my trains are mostly all about the same length :)
19:02:49 <andythenorth> “Breakdowns: off"
19:02:53 <planetmaker> trains on wrong paths then won't cause havoc
19:03:25 <andythenorth> rather than clang / llvm
19:05:07 <andythenorth> I wonder if libpng is the issue
19:05:37 <planetmaker> libpng warning: iCCP: Not recognizing known sRGB profile that has been edited <-- I get that, too
19:06:18 <planetmaker> on this system (fedora20). Not on my debian
19:06:31 <andythenorth> whines about ‘convert'
19:06:37 <andythenorth> command not found
19:07:03 * andythenorth gets imagemagick
19:07:15 <andythenorth> this is going to hurt my phone bill
19:08:17 <DorpsGek> Commit by frosch :: r26652 /branches/1.4 (6 files in 4 dirs) (2014-06-17 19:08:07 UTC)
19:08:18 <DorpsGek> [1.4] -Backport from trunk:
19:08:19 <DorpsGek> - Fix: Segmentation fault when encountering a .obg/.obs/.obm with empty string/zero length MD5 checksums [FS#6038] (r26637)
19:08:20 <DorpsGek> - Fix: The 'Load' button was not properly enabled/disabled for old savegames without NewGRF information (r26634)
19:08:21 <DorpsGek> - Fix: If the video driver fails to supply a list of resolutions, display an error message [FS#6012] (r26629)
19:08:33 <Alberth> Tramvai: easy, specify a depot order in the orders of the train
19:08:47 <Tramvai> I got a lot of trains, it's not very efficient :S
19:09:02 <Alberth> then it will only visit that depot at the moment the order says so
19:09:04 <planetmaker> you don't use any shared orders?
19:10:03 <planetmaker> yup. They're tasty. Nut flavour with a trace of cat. Or so
19:10:05 <Tramvai> I just specify orders and clone the train...
19:12:48 <Alberth> basically, you share the same set of orders between trains. Change them once to change them for every train using those shared orders
19:13:16 <Tramvai> Alright, converting the trains right now. That's great, thanks.
19:13:26 <planetmaker> ctrl+click on other train
19:13:38 <planetmaker> then the train will get shared orders with the train you clicked
19:13:47 <Tramvai> Yeah, the wiki page explained it fairly well.
19:13:51 <planetmaker> or ctrl+clone to clone an existing. k :)
19:14:17 <frosch123> basically, get some glue and make your ctrl key perma-pressed
19:14:42 * andythenorth is now proud owner of imagemagick
19:14:58 *** pthagnar has joined #openttd
19:15:06 <DorpsGek> Commit by frosch :: r26653 /branches/1.4 (8 files in 3 dirs) (2014-06-17 19:14:59 UTC)
19:15:07 <DorpsGek> [1.4] -Backport from trunk:
19:15:08 <DorpsGek> - Fix: Incorrect usage of string commands in the base language [FS#6037] (r26642, r26640, r26639, r26632)
19:15:23 <andythenorth> frosch123: a new exciting error, after running that command
19:15:24 <andythenorth> sprites/graphics/chaplin_0.png: PNG file is not a 256 colour file!
19:15:34 <andythenorth> so it’s no longer paletted
19:15:38 <andythenorth> so I need to change the nfo?
19:16:39 <andythenorth> currently they have 8bpp specified
19:17:31 <frosch123> nah, then the imagemagick command is wrong
19:17:47 <frosch123> actually, interesting... color profiles on 8bpp data
19:17:47 *** montalvo_ has joined #openttd
19:17:55 <frosch123> that sounds like silly crap :p
19:18:47 <planetmaker> frosch123, no, not really. They define how actually each colour looks like
19:20:39 <V453000> HOW IN THE HELL CAN AN ERROR _EXCITE_ YOU ANDY
19:20:44 <V453000> of all things possible
19:24:03 <andythenorth> V453000: because it’s different to the old error
19:24:08 <andythenorth> new always = exciting, no?
19:24:30 *** montalvo has joined #openttd
19:24:53 <V453000> when I get a new error, the 4 letter words usually echo through the room
19:25:02 <V453000> I guess you can describe that as exciting
19:25:17 <V453000> though not the most fit word probably
19:26:40 *** Supercheese has joined #openttd
19:27:31 <andythenorth> wondering if I can downgrade libpng to the version snail reports as working with 6.0.3
19:28:33 <planetmaker> you use macports (or brew)?
19:28:47 <andythenorth> I (stupidly) have both brew and macports versions right now
19:28:52 <andythenorth> I am going to bin the brew version
19:28:55 <fonsinchen> 4-letter words ... beer?
19:29:44 <fonsinchen> so many 4-letter words
19:31:22 *** FLHerne has joined #openttd
19:32:17 <andythenorth> how do I get macports to give me older libpng?
19:32:35 <frosch123> yeti, nuts, firs, fish, most words are 4 letter words
19:34:04 <V453000> OIL_ is a nice 4 letter word
19:34:47 <planetmaker> do you still have that maybe, andythenorth ?
19:34:51 <planetmaker> %% port installed inactive
19:34:59 <andythenorth> having to do manual install
19:35:58 <planetmaker> and you didn't use time machine on that either, no?
19:37:10 <andythenorth> certainly I don’t have a backup nearby to check
19:37:35 <planetmaker> hm, I hoped you had the old version still. I did that on my machine a few times. But I don't clean those
19:42:55 <andythenorth> and built grfcodec 6.0.3
19:43:05 <andythenorth> but same spritesheet bounds error
19:43:33 <andythenorth> so snail has something different to me
19:43:49 <andythenorth> the deps I have are now ~same rev as snails
19:44:31 <Rubidium> does it actually pick up the old png, or does it use the new one?
19:44:47 <andythenorth> I’ve activated it with macports
19:45:08 <planetmaker> if you've mixed brew and macports you can't be sure which is used
19:45:16 <andythenorth> I removed the brew version
19:46:03 <planetmaker> then it should be unique :)
19:56:41 <Rubidium> it's not libpng 1.6.10; it works for me with 1.6.10
19:56:59 <NGC982> R-r-r-r-r-return what you have stolen!
19:57:01 <Rubidium> except the warning about iCCP / the sRGB profile
20:02:05 *** Diablo-D3 has joined #openttd
20:03:15 *** andythenorth has joined #openttd
20:07:18 *** DiabloD3 has joined #openttd
20:09:06 <DorpsGek> Commit by michi_cc :: r26654 /trunk (5 files in 2 dirs) (2014-06-17 20:09:00 UTC)
20:09:07 <DorpsGek> -Fix (r26482): Header file missing from sources.list.
20:10:03 *** michi_cc_ is now known as michi_cc
20:23:21 *** Midnightmyth has joined #openttd
20:32:47 *** Diablo_D3 has joined #openttd
20:33:17 *** Tramvai has joined #openttd
20:33:45 <Tramvai> Can I destroy a train?
20:34:00 <planetmaker> you can crash it or sell it
20:34:02 <Rubidium> just crash another train into it
20:34:25 <Tramvai> I have a train that moves half a block
20:34:34 <Tramvai> All my train traffic is halted
20:34:39 <Rubidium> build a depot really really close
20:34:48 <Rubidium> and force the train to go there
20:35:04 <Tramvai> But still takes like 5 minutes
20:35:20 <Tramvai> And it doesn't want to go there...
20:35:29 <FLHerne> Or turn off breakdowns, because they're a horrible BAD FEATURE that doesn't work properly for gameplay :P
20:37:29 <Tramvai> Had my friend who hosts it turn it off
20:37:48 <Tramvai> They were ridiculous indeed :s
20:38:16 *** funnel_ has joined #openttd
20:45:55 *** funnel_ is now known as funnel
21:28:21 <planetmaker> hm... was Message: Assertion failed at line 1509 of /home/openttd/svn-publicserver/src/saveload/saveload.cpp: IsVariableSizeRight(sld)
21:28:37 <planetmaker> I don't see that in our logs but seem remember... wrongly?
21:28:42 <frosch123> it can happen ever again
21:28:53 <planetmaker> just happend on our PS
21:29:45 <frosch123> but no, no issue of that type was fixed since r26645
21:30:07 <frosch123> do you have a core file or other type of backtrace?
21:30:18 <planetmaker> just the crash.log
21:30:20 <frosch123> we need to know on what chunk it happened
21:32:01 <frosch123> hmm, crash.sav... how does it create a crash.sav when saving causes the crash? :p
21:33:36 <frosch123> did you just start playing, or did it not happen all day?
21:35:01 <planetmaker> we started building on a virgin map.
21:36:51 <planetmaker> some 'planning' stuff might have been build before. Dunno when. So some tracks existed. And also a vehicle. But not sure how long before. V453000 do you know?
21:37:44 <V453000> the vehicle has been built today morning, few game years
21:37:52 <V453000> just check the vehicle age :)
21:38:07 <V453000> the first tracks existed since same time
21:38:14 <V453000> autoclean was on previously which cleaned our company
21:38:16 <frosch123> well, enable core files, and check whether you can trigger it again :)
21:38:33 <planetmaker> how do I enable core files?
21:38:56 <frosch123> ulimit -c unlimited
21:39:05 <frosch123> just put it into .bashrc
21:39:39 <frosch123> or add it to your ottd server start script, or whatever
21:40:00 <frosch123> there is no harm, it only has benefits :)
21:40:52 <Xaroth|Work> there are no limits to removing limits.
21:42:57 <planetmaker> it's started by means of some python scriped and then forked into background
21:49:12 <planetmaker> well, let's see what happens
21:49:36 <planetmaker> I'll also have a look at my bed now. :) Good night
21:50:54 <frosch123> planetmaker: oh, i also assume you compiled with debug symbols
21:52:09 <planetmaker> I actually don't know :)
21:53:12 <planetmaker> config.log says no
21:53:25 <frosch123> well, then also enable those :)
22:12:06 *** luaduck has joined #openttd
continue to next day ⏵
