IRC logs for #openttd on OFTC at 2010-09-03
⏴ go to previous day
00:00:51 <planetmaker> and the problem w/o release: the bug doesn't show ;-)
00:01:54 <planetmaker> (according to Y3xo who obviously tried)
00:02:54 <glx> yes because memory access in debug are different
00:03:12 <glx> but I know where to put breakpoints
00:03:24 <Lakie> Debug includes some memory around allocations to detect memory overwrites
00:04:59 <Lakie> And various other checks release does without.
00:06:02 <planetmaker> hm. So a clamp is missing
00:17:56 <glx> 10-(uint)20 4294967286 unsigned int
00:26:35 <glx> so when you divide by 2 it's a valid int
00:49:17 *** ClampyLubsClarey has joined #openttd
01:17:57 *** ChanServ sets mode: +v tokai
01:30:21 *** robotboy has joined #openttd
03:05:43 *** Adambean has joined #openttd
03:27:47 *** roboboy has joined #openttd
03:56:21 *** De_Ghosty has joined #openttd
04:00:18 *** [com]buster has joined #openttd
04:06:18 *** [com]buster is now known as Combuster
04:12:35 *** HerzogDeXtEr has joined #openttd
04:59:31 *** Eddi|zuHause has joined #openttd
06:04:26 *** HerzogDeXtEr has joined #openttd
06:10:32 *** TomyLobo has joined #openttd
06:18:40 *** Kurimus has joined #openttd
06:23:07 *** Cybertinus has joined #openttd
06:26:22 *** |Jeroen| has joined #openttd
06:31:27 *** [com]buster has joined #openttd
06:31:56 *** [com]buster is now known as Combuster
06:37:54 *** ^Spike^ has joined #openttd
06:54:02 <Mek2> hello, is it possible to set up a dedicated server in a way that when noone is playing in 5 years, then a new game is started?
06:57:09 <Yexo> there are no standard configuration options for that
06:57:18 <Yexo> but of course it's possible when using for example ap+
06:58:11 *** thvdburgt has joined #openttd
06:59:51 *** roboboy has joined #openttd
07:02:51 <Eddi|zuHause> but you could pause the game if no players are present... then no 5 years will pass...
07:03:43 <Mek2> yes I know about that option
07:04:03 <Mek2> but people tent not to connect the server when for example, 20 years have passed (even if there are no companies playing)
07:05:52 <Yexo> Mek2: but even then it's not a default option, you'll have to code it yourself
07:46:08 *** TruePikachu has joined #openttd
07:59:20 *** robotboy has joined #openttd
08:31:29 *** Progman has joined #openttd
08:33:49 *** robotboy has joined #openttd
08:41:17 *** JVassie_ has joined #openttd
08:53:30 *** [com]buster has joined #openttd
08:53:33 *** [com]buster is now known as Combuster
09:05:52 *** p01ymer has joined #openttd
09:24:17 *** Adambean has joined #openttd
09:27:44 *** fonsinchen has joined #openttd
09:33:34 *** r0b0tb0y has joined #openttd
10:18:26 *** JVassie has joined #openttd
10:19:10 *** wollollo has joined #openttd
10:41:29 *** rhaeder has joined #openttd
10:51:48 *** r0b0tb0y has joined #openttd
11:00:29 *** Devroush has joined #openttd
11:00:47 *** perk111 has joined #openttd
11:12:18 *** KenjiE20 has joined #openttd
11:33:42 *** Kurimus has joined #openttd
11:33:42 *** resistance.oftc.net sets mode: +ov orudge orudge
11:42:51 *** fonsinchen has joined #openttd
11:51:52 *** robotboy has joined #openttd
11:53:41 <ABCRic> Rubidium: any luck on yesterday's issue with the new sprites?
12:07:55 <ABCRic> hi glx, any luck on yesterday's issue with the new sprites?
12:14:19 *** Combuster has joined #openttd
12:15:10 *** Kurimus has joined #openttd
12:15:10 *** resistance.oftc.net sets mode: +ov orudge orudge
12:22:14 *** fonsinchen has joined #openttd
12:30:32 * TrueBrain slaps orudge for not having a backlog of PMs
12:33:46 *** George|2 has joined #openttd
12:52:00 *** r0b0tb0y has joined #openttd
12:52:31 <CIA-2> OpenTTD: glx * r20722 /trunk/src/statusbar_gui.cpp: -Fix (r20719): signed/unsigned side effect makes a small negative value become a big positive value
12:54:54 <planetmaker> hm... can (int)((uint)a - (unit)b) be negative?
12:55:51 <glx> planetmaker: the problem is ((uint)a -(uint)b)/2
12:56:26 <glx> small negative int -> big uint, /2 -> big int
12:56:53 <planetmaker> so the /2 takes precedence over the cast?
12:57:05 <glx> no the cast takes precedence
12:58:00 <glx> so now it's big uint -> small negative int -> small negative int
12:59:22 <planetmaker> well. my question was what take goes first: ((int)a)/2 or (int)(a/2)
12:59:31 <planetmaker> the first version would still fail
13:04:32 <glx> outdated opengfx was a good thing ;)
13:06:00 <planetmaker> actually an even better thing is the message that people should update ;-)
13:06:17 <planetmaker> but that is / was no bug ;-)
13:06:23 <glx> yes the red box can explain the crash
13:06:31 <glx> even if it should not crash
13:06:58 <glx> but who reads the red boxes ?
13:08:00 <planetmaker> Vitus, you must be the odd exception to prove the general rule
13:08:11 <Ammler> add a textfield to the redbox and let it only close it the typed text in the box is equal to the message
13:08:50 <planetmaker> What should I do, if I accidentially switched to arabic and get that message in a translated form?
13:09:23 <glx> just move the box so you can change the language
13:10:06 <Vitus> Ammler: I still think that most people wouldn't think about what they typed in there :D
13:10:08 <CIA-2> OpenTTD: smatz * r20723 /trunk/config.lib: -Fix: don't define _FORTIFY_SOURCE when not compiling with GCC, other compilers would likely barf on GLIBC code
13:10:35 *** norbert79 has joined #openttd
13:10:43 <norbert79> Good afternoon everyone
13:12:10 <Vitus> Is Rubidium around? I've got one question about FS#4108.
13:16:08 <norbert79> Can't get any idle time for him, so maybe he is around
13:16:57 <planetmaker> oh folks. Don't ask whether X or Y is around. Ask the question or keep silent
13:17:11 <norbert79> planetmaker: Sorry pa for waking you ;-)
13:17:15 <Belugas> hello sir planetmaker :)
13:17:32 <planetmaker> I didn't mean you this time, norbert79 :-P
13:17:45 <norbert79> "...this time..." :D
13:19:09 * ABCRic is back from lunch, sees glx fixed the problem, hits UpdateOTTD.bat
13:20:41 <ABCRic> yay! no more crashing! :)
13:21:46 <ABCRic> wow. OpenTTD sure runs fast if full animations and detail are turned off
13:22:09 <Vitus> Full animations are the key thing here
13:22:55 <glx> not noticeable when not in fast forward
13:22:56 <Vitus> (if you are talking about FF, of course)
13:23:51 <CIA-2> OpenTTD: smatz * r20724 /trunk/config.lib: -Fix: debug builds with LTO enabled didn't have debug information and were not optimised at all, causing many compile-time warnings
13:24:40 * ABCRic is trying to betterize his script so it updates OpenGFX as well
13:25:00 <planetmaker> the link to the latest nightly is kinda fixed
13:25:05 <ABCRic> and I will *try* to avoid all the questions this time. :)
13:25:08 <Vitus> Betterize, what a strange word.
13:25:16 *** Kurimus has joined #openttd
13:25:16 *** resistance.oftc.net sets mode: +ov orudge orudge
13:26:32 * r0b0tb0y sugests the word improve to ABCRic
13:26:38 <norbert79> Damn, I just didn't know, that my initiative would be such a failure... No reaction from the hungarian side at all, mainly it seem they don't care at all... :S
13:27:34 <ABCRic> r0b0tb0y: I like to make up new words :D
13:28:09 <planetmaker> norbert79, do yo expect zillion replies in like 24h?
13:28:29 <planetmaker> Many people are not around on a daily basis
13:28:41 <norbert79> No, but at least 1 from those, who are also very active on the forums...
13:28:50 <planetmaker> And it's not like the number of Hungarians is uncountable which roam the forum ;-)
13:29:31 <norbert79> planetmaker: Hey, You were very active :)
13:30:14 <planetmaker> but I've nothing to do with any of the addressed newgrfs
13:30:22 * r0b0tb0y wonders if there are more Japanese or other Asian's on here/TT-F than Australians
13:30:50 <planetmaker> my bet would be that not, r0b0tb0y
13:31:15 <r0b0tb0y> of course I'm only interested in active users
13:35:09 <norbert79> planetmaker: Well, my plan is more like creating GRF-s with all the graphics and ideas what the guys at openttd.info have created so far, or create new ones, but instead of having 50-100 sepereate GRF's they would be only 5 using big categories
13:35:57 <norbert79> Like they have created one GRF for some changes in the roads, one for one type of building, etc... I want to collect all these and releasse it as the TTR GRF
13:36:16 <norbert79> (Total Town Replacement)
13:36:29 <planetmaker> well. you might not want to call it TTR
13:36:38 <norbert79> More like HTTR of course
13:36:46 <planetmaker> but yes. If the licenses of the sprites allow, you can just do that.
13:37:03 <norbert79> thats the plan, thats why the entry, and the open question :)
13:37:10 <planetmaker> And... asking whether X is a good idea usually does not yield useful answers anyway
13:37:17 <planetmaker> Either it gets done. Or it doesn't
13:37:39 <norbert79> True, maybe I am too much used to organized enviroments...
13:37:45 <planetmaker> Asking for licenses is better done by directly writing to the people who did the stuff than an open forum poll / thread
13:39:19 <norbert79> good idea, though like with the Ikarus Set I wanted to build up a communication 'channel' so we will not work in paralell, but more like transferring the results directly to the bigger project
13:39:44 <planetmaker> forums are littered with "I want to do great stuff, please contribute" and "it would be a nice idea, if *someone* would..." ;-)
13:39:58 <planetmaker> The 'big' project will gain momentum if *someone* actually starts it.
13:40:04 <planetmaker> I'm sure of that :-)
13:40:36 <norbert79> This is weird you know, because in multinational companies it's just the opposite
13:40:53 <norbert79> normally they just start something without asking first, they create something and end up wioth a pile of crap
13:41:02 <planetmaker> Everything lives from *someone* doing what s/he things would be nice
13:41:03 <norbert79> and here you have to start with it first :)
13:41:25 <norbert79> In big companies it's just totally the oppoisite :)
13:42:10 <planetmaker> The only thing you need to really take care of: before you publish anything, make sure you may use the sprites you used. Either by the license or by explicit agreement of the artists
13:42:49 <norbert79> I am way informed about that, actually partially I have to deal with these like license on a daily basis
13:42:52 * r0b0tb0y is looking forward to his roadtrip on a bus tomorrow
13:43:30 * ABCRic 's TortoiseHG isn't working so well, needs upgrade to cheetahHG
13:45:34 <planetmaker> CC licenses are also cool. When I use GPL I still prefer v2 as it's compatible with OpenTTD :-)
13:46:01 <planetmaker> ---> either CC-BY or GPL v2 for my. Maybe CDDL.
13:46:09 <norbert79> Hmm, I think GPL 3 is still compatible as it's just a bit more advanced, yet for simplicity I would use GPL 2 then, just not making things more complicated
13:46:34 <planetmaker> IIRC the free software association has some issues with GPL v3
13:46:44 <norbert79> I thought those have been resolved by now
13:48:15 <planetmaker> I don't know. Maybe.
13:48:43 <planetmaker> I didn't follow that too closely lately as I'm content with my licenses
13:48:54 <norbert79> never mind, I have nothing in my hands yet, so I think I will spend more time on this when it's time :)
13:49:20 *** George|2 has joined #openttd
13:49:20 *** George is now known as Guest525
13:49:20 *** George|2 is now known as George
13:49:21 <planetmaker> the only important thing really is IMHO: allow derivatives and modifications
13:49:49 <norbert79> Sure, that's the minimum
13:49:55 <norbert79> and I will also share all the sources too
13:50:34 <ABCRic> curse the MSVS Express registration system
13:52:08 *** robotboy has joined #openttd
13:53:08 <ABCRic> I like it how they put the tutorials on the home page though. Open MSVB and it'll show some links to tutorials on how to make some programs... It's a good start :)
13:54:30 *** George is now known as Guest526
13:56:03 *** p01ymer has joined #openttd
14:12:30 *** George is now known as Guest527
14:14:18 <CIA-2> OpenTTD: smatz * r20725 /trunk/config.lib: -Codechange: simplify make_compiler_cflags
14:18:51 *** KritiK_ has joined #openttd
14:23:46 *** KritiK_ is now known as KritiK
14:38:52 <norbert79> Good night robotboy
14:40:09 *** p01ymer1 has joined #openttd
14:52:13 *** r0b0tb0y has joined #openttd
14:54:23 *** George is now known as Guest532
14:59:08 *** ABCRic_ has joined #openttd
15:01:11 *** PhoenixII has joined #openttd
15:02:49 *** KritiK_ has joined #openttd
15:02:57 *** George|2 has joined #openttd
15:03:36 *** ABCRic_ is now known as ABCRic
15:04:54 *** George is now known as Guest535
15:04:54 *** George|2 is now known as George
15:07:02 *** zachanima has joined #openttd
15:08:25 *** ClampyLubsClarey has quit IRC
15:08:30 *** Phoenix_the_II has quit IRC
15:08:44 *** KritiK_ is now known as KritiK
15:09:56 *** lobster has joined #openttd
15:20:34 <norbert79> planetmaker: Does NML come as Linux binary only?
15:20:53 <planetmaker> it's a python module
15:20:59 <planetmaker> that's OS independent
15:21:38 <planetmaker> nmlc is a bash script, though, IIRC
15:21:44 <planetmaker> so that might need mingw or msys
15:22:05 <planetmaker> But there might be written possibly a batch file, too
15:22:47 *** lobster has joined #openttd
15:22:55 <norbert79> I am thinking about just installing cygwin then... Does it require a GUI?
15:23:37 <planetmaker> hm... it basically is a wrapper which calls the main nml file
15:23:44 <planetmaker> that could be written as a bat file, too
15:23:51 <planetmaker> no, it's a non-gui thing
15:24:06 <ccfreak2k> The installer uses a GUI.
15:24:17 <planetmaker> might be. But not for me ;-)
15:25:26 <norbert79> I have collected some materials for the GRF, I am quite curious how well I am able to handle this... I am planning going for an easy way and starting with a plain building first
15:26:00 <norbert79> I think I am gonna build a flat/condo first, hungarian style :)
15:28:28 <planetmaker> he, yeah. Fancy things are not yet supported for houses in NML anyway
15:28:54 <planetmaker> It's missing the action2 variables so far
15:29:10 <norbert79> I think I will still go for the hard way, and go from scratch
15:29:36 <norbert79> this is the same way I learned BUILD for Duke 3D too :)
15:29:37 <planetmaker> if you do it in nml, you need to write (at least the code) from scratch anyway
15:30:50 <planetmaker> you shouldn't use NML for houses, if you imediately need variable access. Like "use snowy version, if snow" or "use this version, if date < 1930"
15:31:13 <norbert79> nah I think I will go for basic
15:31:20 <norbert79> NML would be more useful later on
15:31:25 <planetmaker> Just defining houses should work
15:31:47 <Yexo> planetmaker: that's still possible, using var[number] syntax
15:31:47 <norbert79> Understanding the standard, decompiling some stuff
15:32:00 <norbert79> and first creating the graphics :)
15:32:12 <planetmaker> norbert79: that's... not useful to try to understand the de-compiled NFO
15:32:23 <planetmaker> drawing graphics is, though
15:32:48 <planetmaker> because the de-compiled grf is un-commented nfo which has nothing to do with how NML would allow you to work
15:33:01 <planetmaker> Yexo: interesting :-)
15:33:05 <norbert79> Thats why I am not planning using NML from start :)
15:33:17 <norbert79> Need to understand first how a GRF file works
15:33:42 <planetmaker> well... you define the properties. And you associate graphics
15:33:55 <planetmaker> The latter can be done in a nearly arbitrarily conditional way
15:34:05 <norbert79> yeah, same like BUILD works... You attach vectors and define the textures ;-)
15:34:18 <planetmaker> (and properties can be changed depending upon them, too. At least some)
15:40:33 <dihedral> my new phone just arrived :-)
15:41:05 <ccfreak2k> I created a very nice form in Visual Studio.
15:41:17 <ccfreak2k> Then I realized I have NO idea of how to actually use it in C++ code.
15:41:42 <ccfreak2k> I mean, I could probably do it in Qt because it's event-driven.
15:43:21 <ABCRic> you could always try doing it using Visual Basic...
15:44:16 *** lobster has joined #openttd
15:45:56 <ccfreak2k> I could use VB.NET actually yes.
15:47:10 *** George is now known as Guest541
15:49:39 *** b_jonas has joined #openttd
15:51:30 *** frosch123 has joined #openttd
15:53:08 <norbert79> have to rush, bye everyone!
15:53:15 *** norbert79 has left #openttd
15:59:52 *** wollollo has joined #openttd
16:05:02 <glx> ccfreak2k: just do it in C# :)
16:05:22 <ccfreak2k> That would be, like, the same thing.
16:05:55 <ccfreak2k> Qt would be advantageous because it's event-driven AND somewhat portable.
16:06:15 <glx> windows is event driven too
16:06:17 <ccfreak2k> Qexcept QI Qhave Qto Quse QQt's Qclasses.
16:06:29 <Lakie> Well, Qt is more portable than .net but both wpf/winforms and qt are event driven by nature
16:06:38 <ccfreak2k> Also I'm not familiar with win32 -at all-.
16:07:03 <glx> all win32 stuff is hidden with .net :)
16:07:12 <Lakie> Well, the majority of it. :)
16:07:15 <glx> no need to know windows deeply
16:07:26 <ccfreak2k> glx, I'm aware of that, as I've used VB.NET.
16:09:33 <glx> the other option for portable stuff is wxwidgets
16:10:06 <ccfreak2k> I tried that. I really hated it.
16:10:47 <ccfreak2k> Really this project is probably best done in Java, but Java ain't my cup 'o tea.
16:11:28 <Lakie> I assume mono (is that the one) isn't quite upto speed with .net 3.5 yet? :(
16:13:18 <ccfreak2k> Mono is the one, but I'd rather not rely on it.
16:13:51 <Lakie> Well obviously native is better, but it depends on the platforms you need really.
16:15:57 * Lakie kind of only touched J2Me which he wasn't all that fond of.
16:17:37 <Eddi|zuHause> how much could possibly break if the "full animation" setting did actually affect all animation (e.g. newgrf callbacks), not just palette animation?
16:17:43 *** ABCRic_ has joined #openttd
16:18:04 *** ABCRic is now known as Guest544
16:18:04 *** ABCRic_ is now known as ABCRic
16:19:56 <Lakie> Ah, that answers a small question I had, I was wondering if my object animations should be subject to that switch the other day...
16:21:00 <frosch123> Eddi|zuHause: it would only desync, nothing to bother :p
16:21:55 <Eddi|zuHause> frosch123: i just remember back when i last tested ECS, it was extremely slow...
16:22:13 <frosch123> that is due to the animation frame length callback
16:22:48 <frosch123> if you want to fix that, add a property to define a granularity for it, so it does not need calling every tick
16:23:36 <Eddi|zuHause> i don't think i'll use ECS again any time soon...
16:23:51 <frosch123> i.e. currently if that callback is enabled every animated tile calls a callback every tick, whether the frame is finished. change that to (tick % some_granularity == 0)
16:26:19 <Lakie> Sounds not so great for performance...
16:26:33 <Lakie> Does it not say in most specs use it very sparingly?
16:26:39 <frosch123> yup, the frame-length callbacks are the most troublesome ones :)
16:28:32 <frosch123> @calc 41097010 / 31 / 74
16:28:47 <frosch123> 18000 animated industry tiles
16:29:01 <frosch123> though recent ottd place less industries on map creation than back in 2008
16:30:43 <Lakie> Depends on how many of those use the callback though
16:30:58 <frosch123> well, 18000 using the callback :p
16:31:12 <frosch123> no idea how many other animated tiles there are :)
16:31:45 <Lakie> How long does the average callback take to process?
16:32:27 <Eddi|zuHause> hm... was r20506 considered for backporting?
16:36:45 <Eddi|zuHause> does that need a savegame bump?
16:37:38 *** NukeBuster has joined #openttd
16:38:04 <Eddi|zuHause> i think it would be worthwile to try to backport it...
16:38:29 <Yexo> it doesn't need a savegame bump
16:38:39 <Yexo> but it doesn't fix a bug either, it's more a small feature
16:38:49 <Yexo> which could be backported, but imo it's not that important
16:40:25 <Eddi|zuHause> imho it is a bug...
16:44:26 *** zachanima has joined #openttd
17:22:51 *** zachanima has joined #openttd
17:31:25 <DorpsGek> SmatZ: zachanima was last seen in #openttd 6 weeks, 3 days, 18 hours, 6 minutes, and 55 seconds ago: <zachanima> I travel a lot more than I can afford ^^;;
17:34:20 <zachanima> I just don't talk a lot
17:34:50 * OwenS thinks DorpsGek needs to learn about who's online
17:35:27 <avdg> maybe his name tells a lot about him :p
17:36:42 <avdg> dorp -> town gek -> crazy
17:37:38 <avdg> yeah, idiot is a better translation
17:39:25 <zachanima> also anyone reading 'town crazy' would think 'village idiot' anyway
17:39:39 <zachanima> except crazy not in any way implying idiocy
17:42:43 *** BlackTow3x has joined #openttd
17:45:55 <CIA-2> OpenTTD: translators * r20726 /trunk/src/lang/ (9 files): (log message trimmed)
17:45:55 <CIA-2> OpenTTD: -Update from WebTranslator v3.0:
17:45:55 <CIA-2> OpenTTD: belarusian - 5 changes by KorneySan
17:45:55 <CIA-2> OpenTTD: finnish - 2 changes by jpx_
17:45:55 <CIA-2> OpenTTD: french - 4 changes by glx
17:45:55 <CIA-2> OpenTTD: italian - 4 changes by lorenzodv
17:45:55 <CIA-2> OpenTTD: korean - 2 changes by junho2813
17:46:43 <Eddi|zuHause> hm... i have no idea why, but 2D games in wine seem to be extremely slow...
17:47:41 <Eddi|zuHause> "wine itself" isn't doing much, it just provides some general API mapping
17:47:52 <Eddi|zuHause> it's not an emulator, after all
17:47:58 <avdg> I know, but it replaces libraries
17:48:00 <Eddi|zuHause> the game code does run native...
17:49:26 *** BlackTow3x has left #openttd
17:49:57 <Eddi|zuHause> and when 15 year old games run slower than 3 year old ones, something seems wrong
17:50:48 *** Alberth has joined #openttd
17:51:59 *** Brianetta has joined #openttd
17:52:24 <fjb> The games are old and can not go faster any more. Give them a wheel chair.
17:55:42 <ABCRic> Or sell all the trains and buy some MagLevs :P
17:57:36 <avdg> hmm, doesn't maglevs boost the production a lot? :p
18:01:24 <Lakie> They just increase the servicing ratings which can lead to better production
18:06:44 <Eddi|zuHause> wow... the sky just looks so surreal...
18:06:58 <Eddi|zuHause> the clouds are red and black, it looks like lava
18:07:25 *** ajmiles has joined #openttd
18:07:59 * Prof_Frink moves the sky sideways
18:08:02 <avdg> I know that red and orange color at sunset means the sky is 'cold'
18:08:16 <avdg> just heared from somewhere :p
18:08:52 <Prof_Frink> Red sky at night, shepherd's delight.
18:10:25 <Eddi|zuHause> the sky is blue, the clouds are red
18:10:53 <Eddi|zuHause> the clouds that are not near the sun anymore are black
18:11:17 <Eddi|zuHause> and there was a cloud that was on the edge, that was red and black mixed...
18:19:07 <ccfreak2k> I really hate Qt Creator's insistance on snapping widgets to the grid.
18:19:25 <ccfreak2k> It's in stark contrast to how Visual Studio will snap to various cardinal points of other widgets and the window.
18:21:30 <ABCRic> could anyone tell how the heck can I compile OpenGFX?
18:21:46 <ABCRic> It seems the guide doesn't exactly specify *how* to do it.
18:22:02 <Eddi|zuHause> ABCRic: you mean type "make?
18:22:05 <Alberth> or is there not a Makefile?
18:22:26 <ABCRic> but if I type make, it just returns an error.
18:22:54 <Alberth> something along the lines of 'you don't have grfcodec'?
18:23:01 <Eddi|zuHause> "an error" is a really helpful description
18:23:24 <ABCRic> it syas: make: *** No rule to make target 'opengfx.obg', needed by 'all'. Stop.
18:24:43 <Eddi|zuHause> that doesn't sound right...
18:24:55 <ABCRic> other than that, everything on Makefile.local is commented.
18:24:58 <Eddi|zuHause> you sure you extracted it correctly?
18:25:10 <ABCRic> the guide just said to uncomment INSTALLDIR =
18:26:10 <ABCRic> extracted? I didn't extract anything. I'm trying to set it up so it compiles and installs using a bat file, like I did with OpenTTD itself.
18:26:15 <Alberth> Makefile.local is for customizing for your local system
18:26:18 *** Sacro_ is now known as Sacro
18:28:53 <Eddi|zuHause> ABCRic: replace "extract" with "download" or "checkout" ad libitum
18:30:10 <ABCRic> Eddi|zuHause: *where*?
18:30:15 <Alberth> Eddi|zuHause: 'clone' with hg :)
18:31:36 <Eddi|zuHause> hm... furry thing that makes *miau* needs attention
18:32:08 <Belugas> kitty kitty kitty kitty!
18:32:39 <ABCRic> I already cloned the repository, installed/extracted MinGW, MSYS, TortoiseHg, grfcodec, NFORenum, etc. etc.
18:33:42 <ABCRic> I couldn't find any ".renum" directory
18:34:18 <Eddi|zuHause> directories starting with . are usually hidden
18:34:39 <ABCRic> Show hidden files and directories is enabled.
18:35:17 <Eddi|zuHause> then i can't help you anymore
18:36:27 <Eddi|zuHause> i have neither windows nor opengfx...
18:36:46 <Lakie> You might find .renum under the user folder
18:37:33 <Lakie> It depends on various factors of working directory though
18:37:41 <ABCRic> searched for it, manually and through search.
18:37:52 <Eddi|zuHause> hm... furry thing that doesn't make *miau* anymore fell asleep on my arm...
18:38:00 <Eddi|zuHause> how do i reach the mouse now?
18:38:21 <ABCRic> don't mention the mouse!
18:38:28 <ABCRic> or furry thing will no longer be asleep
18:38:58 <Eddi|zuHause> it's a mouse of little interest to furry thing
18:39:12 <Eddi|zuHause> it's too big and doesn't have a tail
18:39:18 <Alberth> we used to have such a furry thing that only cared very much about 'fish'
18:39:18 <Lakie> When you run Renum it doesn't say anything like "Created .nforenum directory in C:\Users\Lakie."?
18:40:32 <ABCRic> says NFORenum 4.0.0 - Copyright 2004-2010 Dale McCoy.
18:40:41 <ABCRic> Processing from standard input.
18:41:06 <Eddi|zuHause> well, occasionally we have dead things laid out on the floor...
18:42:32 <Lakie> Have you tried .nforenum rather than .renum in your searches, ABCRic?
18:44:41 <APTX> I want to play around with OpenTTDs pathfinding algorithms, for a school project. Does OpenTTD have some debug code for paths/measuring performance? Any general hints about working on this part of the game?
18:51:08 <Eddi|zuHause> APTX: the pathfinder internals was coded by a template-guru who came here a few years ago and nobody really understands that code ;)
18:51:37 <APTX> I don't really mind templates
18:51:52 <Eddi|zuHause> no, he _REALLY_ was a guru ;)
18:52:39 <APTX> I want to play around with the algorithms for ships, as rail/road has all the complex segmentation of tracks
18:52:42 <Eddi|zuHause> APTX: look in the sources under src/yapf/*
18:52:53 <Eddi|zuHause> and don't blame me if you end up in a mental hospital ;)
18:53:34 <Eddi|zuHause> the original ship pathfinder might be easier to grasp
18:53:37 <APTX> so... is there any debugging code for that you (or anyone else) knows of?
18:54:17 <ABCRic> Lakie: no dice. Additionally, Error 666 occurred while attempting to search the entire C: drive.
18:54:24 <Eddi|zuHause> i'm sure it has debug output, check "debug_level" in the ingame console
18:54:38 <ABCRic> being error code 666 a full system freeze...
18:54:59 <APTX> Eddi|zuHause: thanks, I'll be looking into it
18:55:02 <ABCRic> except for the mouse cursor, of course.
18:55:17 <ABCRic> I hate when everything freezes, *except* the mouse cursor.
18:55:39 <Eddi|zuHause> ABCRic: might be a faulty hard drive
18:55:46 <yorick> ABCRic: or might be windows
18:55:59 <ABCRic> my bet is on Windows...
18:56:39 <SteelSide_> btw, why is the original pathfinder set as default for ships? I haven't changed it but I was curious..
18:56:41 <ABCRic> or a virus, but I doubt that.
18:56:55 <Eddi|zuHause> SteelSide_: performance issues
18:56:57 <SteelSide_> cause my ships behave odd unless I plant buoys
18:56:58 <ABCRic> SteelSide_: it's apparently better than the other ones
18:57:00 <yorick> SteelSide_: because the other pathfinders suck on ships
18:57:23 <ABCRic> well, actually, I usually change the ship pathfinder to NPF.
18:57:29 <yorick> someone made a region-yapf...but that was discontinued
18:57:40 <ABCRic> 'cos I only use a few ships :D
18:57:50 <yorick> ABCRic: yapf is probably in all ways better than npf
18:57:54 <yorick> including ship-performance
18:57:58 <APTX> Isn't it because all others are slow for ships?
18:57:58 <Eddi|zuHause> SteelSide_: the YAPF was optimised for rails, where large stretches without any branches appear, but on water practically every tile is a branch, and thus the search-space explodes
18:58:17 <Eddi|zuHause> yorick: there was no NPF for ships
18:58:41 <APTX> <Eddi|zuHause> SteelSide_: the YAPF was optimised for rails, where large stretches without any branches appear, but on water practically every tile is a branch, and thus the search-space explodes <- exactly what I want to play with really
18:58:55 <SteelSide_> by the way is there any idea of implementing signals for roads?
18:58:56 <yorick> APTX: it's at src/yapf/*
18:59:13 <yorick> SteelSide_: there is a traffic lights patch
18:59:20 <ABCRic> actually, yorick, for ships, the original pathfinder is recommended, NPF says nothing and YAPF is not recommended
18:59:25 <Eddi|zuHause> SteelSide_: there are plenty of ideas, but not all of them make sense
18:59:30 <yorick> ABCRic: who do you think recommended it
19:00:04 <ABCRic> so NPF is obviously better than YAPF for ships.
19:00:07 <SteelSide_> I was just thinking that rails got this bunch of signals (thou no non-hack way of creating a priority mainline.. hmmm!) and roads haven one
19:00:13 <yorick> ABCRic: apparently I changed my mind on that
19:02:22 *** ChanServ sets mode: +v glx_
19:05:20 *** perk111 has joined #openttd
19:08:44 <Alberth> SteelSide_: trucks don't even collide
19:09:15 *** guru3_ is now known as guru3
19:09:16 <ABCRic> meh. I still think we need waypoints for roads.
19:09:31 <Yexo> ABCRic: if you're interested in pathfinding you could also take a look at some ai code, finding a new rail/road route has the same exploding search tree as the ship pathfinder
19:09:59 <Yexo> you can build a drive-through road stop and use non-stop orders
19:10:32 <ABCRic> Yexo: I would look at the code, but I probably wouldn't understand it :/
19:10:51 <Yexo> I'm quite certain it's easier to understand than yapf :p
19:11:13 <Yexo> oh, sorry, I mixed up you and APTX
19:11:15 <ABCRic> I know I can build a station, but that's like building a rail station instead of a rail waypoint.
19:11:55 <Yexo> ABCRic: well, some people (*couch*mb*couch*) keep saying that a rail waypoint is a rail station
19:12:08 <Yexo> APTX: about the pathfinding, read a bit up
19:12:25 <Yexo> <APTX> <Eddi|zuHause> SteelSide_: the YAPF was optimised for rails, where large stretches without any branches appear, but on water practically every tile is a branch, and thus the search-space explodes <- exactly what I want to play with really
19:12:34 <Yexo> <Yexo> if you're interested in pathfinding you could also take a look at some ai code, finding a new rail/road route has the same exploding search tree as the ship pathfinder
19:13:18 <ABCRic> a waypoint... is a station? that makes no sense!
19:13:33 <Yexo> in a newgrf a rail waypoint is coded the same as a station (it has the special station class "WAYP")
19:14:04 <Yexo> and I think (not really sure) that in ttdpatch there is no difference when building them either
19:14:42 <ABCRic> but afaik, trains cannot stop at waypoints
19:14:49 <Alberth> I once heard that you can reduce the search space explosion by sorting yet-to-do tiles on nearest to the target
19:15:04 <ABCRic> as such, they cannot un/load cargo there
19:15:08 <Alberth> I did not have the chance to test that idea
19:15:26 <Yexo> Alberth: welcome to A* :)
19:15:52 <Yexo> the NoAI pathfinder libraries, NPF and YAPF already do that
19:16:05 <Alberth> A* fails on large spaces where all distances are equal
19:16:15 <SmatZ> "nearest to the target" = guessed nearest to target
19:16:31 <Yexo> of course, otherwise you'd already know a path :)
19:16:43 <SmatZ> that would be too easy ;0
19:17:15 <Yexo> Alberth: I don't get that, it doesn't matter if all distances are equal
19:17:30 <Alberth> no, that is the primary ordering. Within the smallest distance of those, the nearest to the target, without taking obstacles into account
19:17:44 <Yexo> A* takes a long time if the best path is far from the guessed best path (ie you can't go in a straight line, but have to take a long detour)
19:19:10 <Yexo> I'm still not sure what you mean with "Within the smallest distance of those, the nearest to the target, without taking obstacles into account"
19:19:55 <Alberth> you have a set of positions that are most promising (in the sum of traveled + estimation to target)
19:20:07 <Yexo> if you sort by guessed distance from target rather than by (distance so far + guessed distance from target) you'll indeed find a path sooner
19:20:19 <Yexo> but then the path you find is not always optimal
19:20:54 <Alberth> within that set, you take a position closest to the target without taking obstacles into account
19:23:14 <Yexo> that does require quite some extra bookkeeping
19:23:44 <Yexo> depending on how you store the open list
19:28:23 <APTX> <Alberth> you have a set of positions that are most promising (in the sum of traveled + estimation to target) <- actually it does, depending on the estimation to target
19:29:06 <SmatZ> Alberth: feel free to code YAPF2 :)
19:29:30 <APTX> there are rules the heuristic needs to comply for A* to return an optimal path
19:29:44 <Alberth> First I need to discover why my save/load change causes a crash :(
19:31:01 <Alberth> APTX: I know, I have read and programmed the algorithm too (but not in openttd) :)
19:31:44 <Alberth> unfortunately, before I knew about min heap :)
19:32:24 <ABCRic> SmatZ: you gotta give a better name. something like APFFL
19:32:41 <ABCRic> Another PathFinder For the List
19:32:55 <SmatZ> ABCRic: ah... I thought A is for Alberth's
19:33:11 <ABCRic> and it's pronounced apfle, which means something in german...
19:33:27 <ABCRic> pie, or something like that :P
19:34:40 <Alberth> I was more thinking about ONNANPF
19:34:55 <Alberth> Oh No, Not Another New Path Finder
19:34:55 <ABCRic> ah, yes, I remember now. I saw it on a restaurant's dessert menu: apfle shtrudel mit ice und schokoladen sauce <-- completely sloppy german
19:34:57 <SmatZ> that's sounds... umm.... :)
19:35:39 <ABCRic> basically, it was apple pie with ice cream and chocolate sauce.
19:35:45 <Rubidium> nah, NoPF would be more appropriate
19:36:01 <Alberth> only if written in Squirrel
19:36:04 <ABCRic> oooh, that sounds like a great idea!
19:36:26 <ABCRic> vehicles moving around completely ignoring their orders :D
19:36:41 <APTX> This code doesn't look so bad
19:37:10 <Rubidium> Alberth: NoAI is technically the API only :)
19:44:10 <Eddi|zuHause> you might be able to cut down pathfinder time if you go simultaneously from the target backwards, as from the source forward
19:44:26 <Eddi|zuHause> but un planar graphs, this effect might be lower
19:45:44 <Eddi|zuHause> on "random" graphs this might cut down time by the order of sqrt(n), on planar graphs it might be n/2
19:47:26 <Rubidium> but... it's basically a random graph :)
19:48:29 <Eddi|zuHause> the distribution of bridges and tunnels might not be random enough for this effect ;)
19:48:53 *** StefanR has joined #openttd
19:49:33 <APTX> in practice, it's less efficient than it sounds as you're not sure which node is the one that would connect the paths
19:50:00 <Eddi|zuHause> one approach to ship pathfinding might be to partition the water into convex clusters
19:50:45 * avdg just thinks about fitting rectangles
19:50:49 <Eddi|zuHause> within one convex cluster, pathfinding is a straight line
19:50:50 <SmatZ> Eddi|zuHause: yeah, I was thinking about that idea
19:50:54 <SmatZ> but never implemented it
19:51:27 <Eddi|zuHause> a few years back i have seen a screenshot by someone who tried to distribute buoys over the map like that
19:51:46 <StefanR> Hey guys i get an error while i compile openntd in 64bit with MSVC++ 2010. The Error: error C2061: Syntaxfehler: Bezeichner '__RPC__out_xcount_part'. Could someone help me?
19:51:49 <Eddi|zuHause> not sure how far that was from a proper implementation, or who it was
19:52:20 <Rubidium> StefanR: what version of OpenTTD?
19:52:42 <StefanR> the code of the stable version
19:53:07 <Rubidium> that doesn't have a MSVC 2010 project file and its conversion doesn't quite work
19:53:38 <Rubidium> having said that, __RPC__out_xcount_part isn't anywhere in OpenTTD's source code
19:54:31 <StefanR> maybe it's better with the full error :P. ---> >C:\Program Files (x86)\Microsoft SDKs\Windows\v7.0A\Include\objidl.h(11280): error C2061: Syntaxfehler: Bezeichner '__RPC__out_xcount_part'
19:55:10 <Rubidium> so there's something fishy with Windows' SDK
19:55:17 <Rubidium> (no clue how to solve that though)
19:55:37 <APTX> I don't see the benefit from most of the yapf templates, but it looks pretty neat
19:56:29 <michi_cc> APTX: Without the templates you'd need virtual methods which can't be inlined.
19:57:43 <Eddi|zuHause> APTX: let's say it's about a factor of 10-20 faster than NPF, while the algorithm is basically the same, NPF is using objects/pointers, while YAPF is using templates
19:57:50 <SmatZ> the more work for compiler, the faster run :)
19:58:26 <SmatZ> however, recent gcc is able to de-virtualize methods while compiling in whole-program mode (lto)
19:58:39 <SmatZ> I haven't had a look at it yet though...
19:58:56 <ABCRic> and the faster OpenTTD runs, the happier people will be :D
19:59:24 <SmatZ> indeed - reminds me of one patch at tt-forums thread :)
19:59:34 <APTX> it's not like I'm questioning it doesn't help... I'm wondering how I could fit an algorithm in that doens't use any of the yapf structures
20:01:55 <Rubidium> hook it in the vehicle code, i.e. there were the "split" between yapf, npf and old is done
20:02:54 <APTX> Thanks, that's very useful
20:08:14 *** NukeBuster has left #openttd
20:11:53 <Vitus> Rubidium: Thanks for clearing up FS#4108. I've got one last question, though. Lowering the terrain actually causes some graphical glitches (http://totalniparba.wz.cz/dam_tf.png), is this mistake in NewGRF, too?
20:14:21 <Lakie> I suspect thats more of again an autoslope not being disabled problem
20:15:18 <Lakie> As I think it uses do-not-draw-foundations
20:15:44 <Rubidium> yup, that's basically the NewGRF's fault
20:15:47 <Lakie> Which means it draws at the lowest tile's corner height, hense the issue
20:16:06 <Vitus> Alright, thank you very much Lake and Rubidium :)
20:16:10 <Rubidium> though... that brings me to the question: what does TTDP do when the disable autoslope callback is enabled but missing?
20:16:32 <Rubidium> in OpenTTD it will allow autoslope as it's implemented that way for industries as well
20:16:48 <Lakie> My from understanding it'll just do whatever houses / industries do
20:17:22 <Lakie> As the callback handling is the same wrapper used by both
20:19:31 <Rubidium> lovely... in-sync behaviour!
20:21:13 * Lakie is half expecting someone to ask why rv's don't drive on the road the object uses...
20:21:39 *** r0b0tb0y has joined #openttd
20:22:26 <Rubidium> "because wally forgot to draw the army people blocking it because there might be a terrorist threat and as such this high vulnerability target must be secured"
20:32:49 <CIA-2> OpenTTD: yexo * r20727 /trunk/src/settings_gui.cpp: -Fix [FS#4106]: compilation with --disable-ai was broken (cirdan)
20:46:09 <Eddi|zuHause> * Lakie is half expecting someone to ask why rv's don't drive on the road the object uses... <-- in theory, if the airport-statemachines get capable of handling articulated engines, one could get arbitrary road paths in objects
20:46:57 * ABCRic needs a translation on that
20:47:18 <SmatZ> Eddi|zuHause: iirc, such a feature request was rejected for possibly high time requirements while pathfinding
20:47:33 <SmatZ> eg. doing CB everytime PF gets at that tile is very costy
20:47:44 <SmatZ> much more costy than simple map array access
20:48:08 <SmatZ> also, the object shouldn't change roadbits while a vehicle is driving on it...
20:49:36 <SmatZ> well, CB wouldn't be a good solution
20:50:03 *** r0b0tb0y has joined #openttd
20:50:30 <Eddi|zuHause> SmatZ: well, pathfinder shouldn't call the statemachine-callback. it should have a different method to supply a "follow track" result
20:50:37 *** r0b0tb0y is now known as roboboy
20:51:20 <Eddi|zuHause> it might be a static mapping of enter-trackdir and exit-trackdir, for all possible combinations
20:54:37 <Eddi|zuHause> i mean. it's a valid argument, but not necessarily a reason to drop the idea
20:59:59 <avdg> just a question: is it really bad to assign 2 extra bytes each tile on an ingame map?
21:00:29 <Terkhen> @calc 2048 * 2048 * 2
21:01:32 <Eddi|zuHause> that's *only* 6MB
21:01:59 <DorpsGek> Eddi|zuHause: 1.22222222222
21:02:16 <avdg> just thinking about an algoritme for the ship pathfinder *but can't code c++ -_-*
21:02:40 <Eddi|zuHause> avdg: water tiles have plenty of free bits
21:03:34 <avdg> hmm.. so I don't have to worry at all
21:12:43 <StefanR> hey i'm back :) i fix the problem with windows 7 sdk but now i get this error: warning C4005: 'EWOULDBLOCK': Makro-Neudefinition
21:12:43 <StefanR> 4> C:\Program Files (x86)\Microsoft Visual Studio 10.0\VC\include\errno.h(132): Siehe vorherige Definition von 'EWOULDBLOCK'
21:13:00 <StefanR> 4>..\src\genworld_gui.cpp(336): error C2664: 'ShowErrorMessage': Konvertierung des Parameters 3 von 'int' in 'WarningLevel' nicht möglich
21:13:00 <StefanR> 4> Konvertierung in einen Enumerationstypen erfordert explizite Typumwandlung (static_cast-Operator oder eine Typumwandlung im C- oder Funktionsformat)
21:13:48 <StefanR> i'm sorry that this is in german...
21:15:12 <yorick> avdg: that doesn't mean you can't design something :)
21:16:05 <avdg> just checking the bytes allocation
21:18:37 <Rubidium> StefanR: stable doesn't have warninglevels for that function, actually that function doesn't seem to be called at all in genworld_gui.cpp (of stable)
21:19:00 <Rubidium> or in "trunk" for that matter
21:19:13 <Rubidium> so... it smells like you have patched OpenTTD with something
21:23:39 *** Progman has joined #openttd
21:24:13 <avdg> hmm.. just 1 thing… a pf that uses the index :p
21:34:43 *** Chillosophy has joined #openttd
21:41:23 <CIA-2> OpenTTD: yexo * r20728 /trunk/src/road_cmd.cpp: -Codechange: shuffle some code in CMD_BUILD_ROAD so it doesn't call CMD_LANDSCAPE_CLEAR in test-mode from exec-mode
21:45:29 <ABCRic> mere curiosity, any list of commands available for DorpsGek?
21:47:00 <Ammler> ABCRic: /msg DorpsGek list|help|...
21:51:24 <CIA-2> OpenTTD: yexo * r20729 /trunk/src/ (5 files in 2 dirs): -Fix [FS#4107]: assert when overbuilding object
22:04:14 <Syzgyn> So I've got this recurring issue that I can't seem to find any info on
22:05:12 <Syzgyn> I'll have trains on a one way loop working fine. Then I'll change some of the track around, keeping it totally valid, but the trains won't use the new track segment
22:05:33 <Syzgyn> Instead they'll go up to the new bit, stop, reverse, then hit the nearest one way signal and reverse back, and be stuck
22:06:20 <Syzgyn> I can upload a save file demonstrating it if you'd like
22:07:00 <Eddi|zuHause> Syzgyn: you built non-electrified track
22:07:19 <Syzgyn> I'll feel really dumb if thats the case
22:07:21 <Eddi|zuHause> i'm fairly sure you did ;)
22:07:31 <Eddi|zuHause> happens to all of us ;)
22:07:47 <Syzgyn> well thanks for solving the problem
22:08:20 <Syzgyn> That had been bugging me for a while
22:09:12 <Noldo> Eddi|zuHause: look how I managed to give neutral response instead of the instictive "don't ask to ask!!1"
22:09:17 <ABCRic> That's why I always use MagLevs! xD
22:09:42 <Eddi|zuHause> Noldo: why tell that to me?
22:10:00 <ABCRic> well, technically, Noldo, Syzgyn didn't ask anything...
22:10:09 <Syzgyn> Noldo: I didn't mean to ask to ask, I was typing it out as you responded.
22:10:34 <Syzgyn> I just wanted to preface it so you wouldn't assume it was some stupid question that's on the wiki
22:11:04 <Eddi|zuHause> Noldo: i'm not usually saying "don't ask to ask"...
22:11:08 <Noldo> Syzgyn: don't worry about it, you did good
22:13:16 <Eddi|zuHause> Noldo: i usually either ignore it, or come up with a witty reply like "hey, that's a great problem description"
22:17:28 <avdg> eddi: I have to say it more then I want
22:18:37 <ABCRic> Eddi|zuHause: "hey, that's a great problem description" is often accurate, but some times people are still writing when you respond.
22:18:41 <Zuu> lol, the most serious human game I can find on my computer is a replication of a time controled traffic light with a bit of added dynamics. :-)
22:19:04 <Zuu> The network is more or less otherwise a cross to feed the intersection :-)
22:22:17 <Zuu> It's not very advanced really. The minor approach detectors need to be activated in order to the traffic light to switch from main to minor.
22:25:26 <ABCRic> Any way we can obtain said game?
22:26:59 <Zuu> Sure, I can upload it if you want to take a look
22:28:29 <CIA-2> OpenTTD: yexo * r20730 /trunk/src/object_cmd.cpp: -Fix: crash when removing an object while the newgrf debug window was open for that object
22:30:27 <Zuu> You need last nightly for it
22:31:10 <Zuu> (as I have oversaved the old version that I found)
22:32:57 * avdg thinks he has to look how yapf works
22:34:51 <ABCRic> Zuu: seems to be running fine on 20709
22:35:14 <ABCRic> I have to go now, I'll examine it in detail tomorrow.
22:35:16 <Zuu> Then the save game version did not change between those versions.
22:53:34 *** zachanima has joined #openttd
23:04:25 <CIA-2> OpenTTD: yexo * r20731 /trunk/src/ (core/smallvec_type.hpp landscape.cpp): -Fix (r20739): SmallVector did not have an assignment operator, causing invalid memory reads / double free
23:07:09 <Eddi|zuHause> fixing a revision in the future? cool :p
23:07:28 <SmatZ> Eddi|zuHause: Yexo knew he would break that in r20739 ;)
23:08:13 <Yexo> r20739: -Cleanup: remove operator= from SmallVector
23:09:52 *** zachanim1 has joined #openttd
23:23:16 <Eddi|zuHause> @convert 10 36 10000
23:23:16 <DorpsGek> Eddi|zuHause: Error: 36 is not a valid unit.
23:23:26 <Eddi|zuHause> @convert 10 35 10000
23:23:26 <DorpsGek> Eddi|zuHause: Error: 35 is not a valid unit.
23:23:43 <Eddi|zuHause> @convert 10 16 10000
23:23:43 <DorpsGek> Eddi|zuHause: Error: 16 is not a valid unit.
23:32:59 *** wollollo has joined #openttd
23:41:50 *** zachanima has joined #openttd
23:44:04 *** zachanima has joined #openttd
23:48:27 <CIA-2> OpenTTD: yexo * r20732 /trunk/src/tunnelbridge_cmd.cpp: -Fix: when trying to build a bridge over an object, try to autoremove the object if it's too high
continue to next day ⏵
