IRC logs for #openttd on OFTC at 2012-05-02
⏴ go to previous day
00:09:42 *** drac_boy has joined #openttd
00:15:20 *** theholyduck has joined #openttd
00:24:52 <drac_boy> hm...another of these slow night I guess
01:59:35 *** theholyduck has joined #openttd
02:02:17 *** HerzogDeXtEr has joined #openttd
04:56:20 *** Eddi|zuHause has joined #openttd
04:57:32 *** kkb110_ has joined #openttd
05:22:40 *** telanus has joined #openttd
06:00:56 *** Firartix has joined #openttd
06:42:28 *** sla_ro|master has joined #openttd
06:59:13 *** roboboy has joined #openttd
07:54:38 *** Hazzard has joined #openttd
07:56:34 <Terkhen> hi dihedral and Hazzard
07:56:56 <dihedral> what's new in the world of openttd? :-P
08:04:26 <MNIM> scantily clad women on the sides of busses.
08:05:03 <MNIM> scantily clad women on the side of aircraft?
08:07:24 <Terkhen> there are no people in OpenTTD
08:07:33 <Terkhen> just abstract passengers
08:09:58 <__ln__> then why is there a distinction between passengers and livestock
08:11:41 <MNIM> they're bits! zeroes and ones!
08:12:03 <NGC3982> bits are not zeroes and ones
08:13:09 <NGC3982> it used to, but the national zerg institute of pasadena, LA stated that from 2011-01-01, a bit is defined by the number of slots filled with zerglings, or non-zerglings.
08:13:26 <Terkhen> because abstract passengers take up less space than abstract livestock
08:13:31 <NGC3982> parity is checked with an added hatchling.
08:13:58 <Terkhen> abstract space, of course
08:23:47 <CornishPasty> Isn't mail a passenger too?
08:24:32 <Hazzard> Is there a GRF that makes it so planes can carry only passengers?
08:25:36 *** ZxBiohazardZx has joined #openttd
08:30:24 <Terkhen> Hazzard: to my knowledge, no
08:30:57 <Hazzard> Seems like something that would be pretty easy
08:34:29 <Terkhen> yes, if you plan to change the default aircraft you only need to redefine their cargo properties
08:35:14 *** LordPixaII has joined #openttd
08:36:26 <ZxBiohazardZx> i just realised how stupid forests are in Arctic mode
08:36:33 <ZxBiohazardZx> you can only place them ABOVE the snowline
08:37:01 <ZxBiohazardZx> while in most countries i know, the forests are BELOW or ON the snowline, as above the snowline/treeline there simply wont grow anything
08:37:21 <ZxBiohazardZx> ofc treeline =/ snowline, but in general, if you go too high, shit wont grow...
08:39:32 <Terkhen> I don't remember if OpenGFX+ Industries allowed you to change that
08:41:14 <Hazzard> How do you use 'cargo_allow_refit'? I keep getting an unrecognized identifier error and I can't find anything on the wiki.
08:42:04 <Terkhen> the cargo properties changed some time ago, I have not used them
08:42:11 <Terkhen> since they were changed
08:44:29 *** tokai|mdlx has joined #openttd
08:44:42 <Terkhen> I don't know the new properties, I'm reading
08:45:58 <NGC3982> what was the name of that grf that starts 1830 with that planet engine?
08:48:46 <Hazzard> Maybe I don't have the newest NML
08:48:55 <Hazzard> When I paste that in I get the same error
08:50:09 <Terkhen> that changed recently, yes
08:50:37 <Terkhen> I can give you an example with the old syntax if you want it, my own code in OpenGFX+ Road Vehicles follows it
08:52:23 <Terkhen> but when/if you move to future versions of nml you will need to change it
08:53:03 <Terkhen> removing specific cargos from an allowed cargo class with the old system needs XORing a bit and I don't exactly remember how it was :P
08:53:30 <Terkhen> I don't know how that will affect the dual passenger/mail capacity of aircraft
08:53:53 <Hazzard> Oh, I am actually working on something different
08:54:12 <Hazzard> I want my RV to carry goods, but not all the other express stuff
08:54:26 <Hazzard> (along with some other CC)
08:54:53 <Terkhen> that is the cargo file from ogfx-rv
08:55:14 <Terkhen> you can use it as an example, but note that you would need to remove the #defines and the \ at line endings
08:56:02 <Hazzard> Do you need to specify non refitable cargo classes?
08:56:40 <Terkhen> cargos can have more than one cargo class
08:57:30 <Terkhen> if it has both a refittable and a non refittable cargo class, your vehicle won't carry it
09:02:11 <MNIM> apart from the obvious literal translation, I mean
09:02:13 <Eddi|zuHause> it says "The Spiders" - "The Bulls" - "The Pigs"
09:02:55 <Eddi|zuHause> or if you want to interprete something into it, something along the lines of "the cop-bastards are crazy"
09:03:22 <Eddi|zuHause> which is *completely unintentional* :p
09:04:50 *** TGYoshi has joined #openttd
09:15:16 *** th_gergo has joined #openttd
09:20:19 *** th_gergo1 has joined #openttd
09:28:50 *** th_gergo has joined #openttd
09:42:37 *** Firartix has joined #openttd
10:14:11 *** Hazzard has joined #openttd
10:28:40 *** tokai|noir has joined #openttd
10:28:40 *** ChanServ sets mode: +v tokai|noir
10:47:48 <ZoeB> given the 32bit graphics now available, is it safe to say that OpenTTD is moving forwards with modern technology, not just slavishly imitating the original TTD?
10:49:40 *** goodger has joined #openttd
10:52:27 *** valhallasw has joined #openttd
10:52:56 <ZoeB> and assuming that's the case, are there any plans to support recordings of music (say, Ogg Vorbis files) instead of MIDI files at some point, for more realistic (and consistent sounding, from one machine to the next) music?
10:58:09 <Ammler> @tell Zuu looks like it is always releases, if it's empty, so use it as default?
10:58:09 <DorpsGek> Ammler: Error: I haven't seen Zuu, I'll let you do the telling.
10:58:54 <Ammler> if I would see him, I could indeed do it self :-P
11:00:55 <DorpsGek> Ammler: Zuu was last seen in #openttd 11 hours, 51 minutes, and 19 seconds ago: <Zuu> Night
11:07:46 *** drac_boy has joined #openttd
12:17:15 *** goodger has joined #openttd
12:30:15 *** Hazzard has joined #openttd
12:31:52 *** Arafangion has joined #openttd
12:46:24 <Hazzard> How do I use callbacks so the cargo capacity will be the same for all cargos?
12:47:15 <Hazzard> Or any other way to do it
12:49:41 <TrueBrain> Ammler: its not the bot that is stupid, but your expectations of it ;)
12:50:36 <Ammler> assumed a feature of another bot in a german community
12:50:54 <TrueBrain> I kept the bot very stupid, as people tended to abuse it a lot :'(
12:51:30 <Ammler> I guess, the supybot feature would be called note(s)
12:54:07 <TrueBrain> blender can be weird when importing 3ds models .. wrong center of objects
12:54:36 *** theholyduck has joined #openttd
12:55:01 <TrueBrain> at least another attempt
12:55:16 <TrueBrain> I rarely get anywhere, as there are many minor issues witha lot of stuff
12:56:04 *** th_gergo has joined #openttd
13:07:10 *** Hazzard has joined #openttd
13:08:32 <TrueBrain> it would just be so cool if I was a blender expert :D
13:10:40 <Rubidium> TrueBrain: to me you are! ;)
13:11:57 <TrueBrain> that says more about you than me, sadly :(
13:14:41 <TrueBrain> my biggest issue atm is, that if you link a blend from another blend, you cannot change his parent
13:14:57 <TrueBrain> so, you need to append to object, but then if you modify it in the original blend, it doesnt update
13:15:03 * TrueBrain thinks it makes little sense
13:15:32 <telanus> Afrikaans is now 100% translated
13:19:11 <TrueBrain> ah, it has to be done via groups; why not just make it more complicated #idontlikeovercomplicationofthings
13:23:49 *** roboboy has joined #openttd
14:00:41 <NGC3982> i just noticed that "conditional order jump" isnt translated in the latest stable 1.2.0 (swedish).
14:08:34 <ssdg> I have noticed a little glitch and I was wondering if someone experienced it too.
14:08:34 <ssdg> When playing with friends on a server one of use hosts, when deleting railroads, they disappear from our screens, then a few seconds later we are disconnected. When we come back, the railroads are back from the deads... it's really weird
14:09:04 <planetmaker> you disconnect with a desync I assume. Which version do you play, ssdg?
14:09:22 <planetmaker> I remember that bug from an ancient version and remember it be fixed
14:10:06 <planetmaker> If you don't play OpenTTD 1.2.0 or a (recent) nightly, try that
14:11:43 <ssdg> It's a desync indeed. We play the 1.2... I'll check with the server owner.
14:14:13 * ssdg is still waiting for answers
14:15:26 <ssdg> a 1.2.0 on ubuntu everywhere. (the ubuntu version differs, but the openTTD version is the same)
14:15:32 <planetmaker> you don't need to check with server owner. Unless he uses a self-compiled version
14:15:45 <planetmaker> if it's not the same, you usually can't connect.
14:15:56 <ssdg> A self compiled version it is. (on ARM)
14:16:06 <planetmaker> yours? The server?
14:17:36 <ssdg> Maybe you put the finger and the reason why our game glitches
14:20:25 <ssdg> Maybe we can try running a pre-packaged version on a i386 computer
14:24:03 <ssdg> Is there a trusted source for ARM packages somewhere?
14:24:16 <TrueBrain> in Debian most likely :)
14:24:26 <TrueBrain> they are ARM freaks, last I checked ;)
14:24:38 <TrueBrain> so the server runs ARM, and you all desync? (all clients)?
14:25:11 <TrueBrain> now that is interesting :)
14:25:17 * ssdg spent too much time playing with spies on RedAlert
14:25:45 <ssdg> But only when one of us is working on railroads. (by removing some of them)
14:25:48 <TrueBrain> optimal we would like to get a desync log, but that might be kinda hard :D
14:25:57 <TrueBrain> does it happen always, or just sometimes?
14:26:43 <ssdg> Everytime we delete a section of railways. Not necessarily immediatly
14:26:59 <TrueBrain> a few minutes go by between desync checks
14:27:01 <michi_cc> Different architectures are always nice, anyone still remembers the bools-are-4-bytes-on-PPC fun? :)
14:27:14 <TrueBrain> I remember much more fun with PPC michi_cc ;)
14:27:48 <ssdg> TrueBrain: about that desync log. Say we had a few desync in the past, that we cant play as of this moment and one of us have an access on the server. What do we do?
14:27:50 <TrueBrain> ssdg: what systems do the clients run? (32bit/64bit, arm/intel/amd/...)
14:28:30 <TrueBrain> hmm, I am kinda rusty in dsync debugging, I dunno if release builds come with the desync log switch, or that it requires a recompile?
14:28:34 <ssdg> Mine is an intel (i7) running ubuntu 32 bits. (I guess one of my friens does run something similar)
14:28:35 <TrueBrain> I guess planetmaker does?
14:29:13 <peter1138> i fucking hate linux
14:29:44 <peter1138> my desktop pcs can't connect to a vpn because the vpn manager doesn't support client side certificates
14:29:47 <ssdg> peter1138: I hate your system too
14:30:01 <peter1138> my android phone, on the other hand, can
14:30:14 <blathijs> ssdg: As an extra datapoint in the testing
14:30:16 <TrueBrain> blathijs: was there ever someone who ran the ARM as server?
14:30:28 <TrueBrain> or are the ARM builds tested at all?
14:30:34 <blathijs> TrueBrain: Not by me :-)
14:30:43 <TrueBrain> I truly wonder if there ever has been a case before :P
14:31:11 <ssdg> We'll try this... when back from work
14:31:24 <peter1138> iirc the debian stuff just tests that it compiles
14:31:28 <planetmaker> well. 1.2.0 has one desync iirc related to house counts. Frosch fixed it recently
14:31:40 <TrueBrain> planetmaker: but that has no relation to removing railroad
14:31:47 <TrueBrain> otherwise we would have had a 1.2.0-r1 already ;)
14:31:53 *** telanus1 has joined #openttd
14:32:18 <TrueBrain> planetmaker: not being able to remove any railroad from any MP game, is kinda .....
14:32:22 <TrueBrain> how to say it .. critical? :)
14:32:22 <planetmaker> we don't have that for the houses either. And it made the game under that circumstances unplayable :-)
14:32:30 <planetmaker> like desync in 5 seconds
14:32:32 <TrueBrain> there is a nuance difference ;)
14:32:51 <planetmaker> not quite. The difference is: you can't do anything then ;-)
14:33:15 <planetmaker> when you immediately desync
14:33:36 <planetmaker> it's like server kicks you a few seconds after connect :-P
14:33:47 <planetmaker> under certain circumstances. Anyway, nvm
14:33:53 <TrueBrain> dude ... a dsync on every removal of every track versus a dsync with certain grfs in certain cases ...
14:33:59 <TrueBrain> there is a nuance difference there
14:34:04 <TrueBrain> first is _very critical_, second is annoying
14:34:23 <planetmaker> anyway, ssdg, what you could do:
14:34:44 <planetmaker> debug_level desync=3 on the server
14:34:58 <planetmaker> save the game. Reload server-side
14:35:37 <ssdg> Can I come back in 4~5 hours?
14:35:40 <planetmaker> note which game you loaded (might be the game you play, we'd need the savegame you start this debugging with)
14:35:50 <planetmaker> not sure you can. But you'll be welcome to :-)
14:36:17 <ssdg> did I make a language error here? (I might, I'm not native)
14:36:29 <planetmaker> no, you didn't. I'm splitting hairs
14:36:30 <TrueBrain> one warning: it makes a lot of data; you kinda want to desync ASAP, then close the server up :D
14:37:26 <planetmaker> Return when convenient and I will try to talk you through, provided I'm here.
14:37:54 <planetmaker> can't promise that atm, though, might be afk. But others could :-)
14:38:18 <planetmaker> desync bugs are annoying and hard to track
14:38:28 <ssdg> Thanks for the warnings. I'll see you once the data is "harvested". And thank you all for your help. (and niceness)
14:38:49 <planetmaker> you want maybe a gigabyte of free space or so. Depending on how long it takes to trigger the desync
14:39:20 <planetmaker> might go much faster and just need a few savegames
14:39:26 <planetmaker> really can't tell :-)
14:40:00 <ssdg> Is it easier with smaller maps? (or other tricks)
14:40:33 <TrueBrain> the faster you desync, the easier for us :)
14:40:48 <planetmaker> It'll make a lot of savegames, so it'd use less space. The map size might matter, but might not
14:41:11 <planetmaker> Just play the normal game you want to play and have the desync debugging switched on on the server
14:41:25 <planetmaker> It won't hurt you playing and will gather data needed to track the bug
14:42:00 <TrueBrain> after a desync, we can "replay" the log, and see where it went wrong (hopefully, in the easiest case :P)
14:42:05 <ssdg> So we can start a new game, build a railroad, delete it and try to force the server to desync. and send you a lot of savegames?
14:42:12 <planetmaker> i.e. I suggest to just continue the game you have problems with
14:42:43 <ssdg> (we restarted the game for other reasons... but we had the problem with every game)
14:42:51 <TrueBrain> how did you manage?
14:43:01 <TrueBrain> if you cannot remove any rail track?
14:44:08 *** frosch123 has joined #openttd
14:44:45 <ssdg> TrueBrain: it's a torture.
14:45:24 <TrueBrain> random question you might be able to answer now: does it happen when you destroy the tile, or if you remove the tile? (bulldozer versus red-square, so to say)
14:45:31 <TrueBrain> you now made me curious :D :)
14:46:44 <TrueBrain> well, desync logs will tell us more, hopefully :)
14:47:24 <TrueBrain> and just out curiousity, why run a server on an ARM? What kind?
14:47:26 <ssdg> But I *think* the desync happends when the tile is modified in an impossible way (say add a signal on what was a crossroads)
14:47:52 <Rubidium> a desync due to endianness seems most 'logical'
14:48:01 <TrueBrain> ARM is also LE, at least, most of them are :)
14:48:04 <ssdg> It's a plug computer my friend runs as a NAS/multipurpose internet server
14:48:11 *** Hazzard has joined #openttd
14:48:17 <Rubidium> though last time I tried BE it worked, although that's quite a while ago and with s390
14:48:38 <TrueBrain> its an interesting piece of hardware to run OpenTTD on :)
14:48:57 * TrueBrain still waits for his RaspberryPI :P
14:50:01 <TrueBrain> Rubidium: well, strictly, ARM is bi-endian, but most people / OSes run it in LE. I havent seen many BE cases (ofc this can be the exception :D)
14:51:11 <__ln__> nslu2's original firmware is BE, afaik.
14:59:26 <planetmaker> indeed you might want to try the debian pre-compiled binary, too
15:01:07 <TrueBrain> is it me, or doesnt he mention which OS he used?
15:01:31 <michi_cc> In theory it could also be a problem with the sizes of uint8/uint16/etc, even if it is unlikely. We do have an assert_compile check, but it only checks sizeof, which is not required to return a results based on 8-bit bytes.
15:01:46 <ssdg> It's a debian sid IIRC. (checking in progress)
15:02:00 <TrueBrain> debian delivers both BE as LE, as far as I know :)
15:02:05 <__ln__> Rubidium: but if it was an endianness issue (and one that would have to be an old one), it would have been noticed on OS X PPC a long time ago.
15:02:13 <ssdg> would a "uname -a" be usefull?
15:02:19 <TrueBrain> where BE was very unstable last I have seen it, but that has been a while ago, so *shrug*
15:02:35 <TrueBrain> __ln__: does it? Who runs OSX PPC still? :P
15:02:47 <TrueBrain> and who runs it as server?
15:03:02 <michi_cc> More likely would also be a code generator fail for 64-bit arithmetics (i.e. the Money type).
15:03:17 <__ln__> TrueBrain: i do, but not as server.
15:03:45 <TrueBrain> michi_cc: that tbh, sounds much more likely yes :)
15:04:00 <__ln__> TrueBrain: in any case, removing tracks is such a common thing to do that it would be noticed pretty much immediately even if there are only a handful of PPC users.
15:04:13 <TrueBrain> __ln__: but how many of that handful run it as client/server? :P
15:04:24 <TrueBrain> my point is, the fact that you dont hear anything, doesnt mean it works
15:04:36 <TrueBrain> it might have been dismissed as: OSX ..
15:05:11 <TrueBrain> no negative sadly doesn't imply a positive :(
15:05:31 <Rubidium> TrueBrain: if it's dismissed as being OSX, then it would still be in the bug tracker with [OSX] Desync ... or [OSX] Clearing rail costs loads of moneys
15:05:49 <TrueBrain> Rubidium: who says any of those few users went via the bug-tracker?
15:05:56 <TrueBrain> again, no negative doesn't make a positive :)
15:06:02 <ssdg> Our server is a "Linux korriban 2.6.32-5-kirkwood #1 Wed Jan 12 15:27:07 UTC 2011 armv5tel GNU/Linux"
15:06:19 <__ln__> i thought OTTD bug-tracker is the place where cool people hang around regularly. no?
15:06:40 <TrueBrain> el .. was that the LE version? my memory sucks :P
15:07:33 <michi_cc> Should be, ARMv5TE is the device and the L probably means little.
15:07:54 <TrueBrain> Debian has EL and HF atm
15:07:57 <TrueBrain> where HF = Hard Float
15:08:37 <__ln__> armv5teb returns quite a few results on google at least
15:08:49 <TrueBrain> yeah, it is still "work in progress", the BE port of Debian
15:08:52 <TrueBrain> many many many issues :(
15:09:24 <Rubidium> my BE PPC qemu 'works' with r24111
15:09:47 <TrueBrain> so he runs an LE version, dismissing any of the Endian suggestions ... which makes this problem more interesting tbh :D
15:10:29 <michi_cc> So, next question: Compiler version used?
15:10:31 <TrueBrain> it might mis-detected BE/LE?
15:10:41 <TrueBrain> guess we need a config.log too :D
15:10:46 *** Zeknurn has joined #openttd
15:11:01 <TrueBrain> ssdg: do you have the config.log of the server? :)
15:11:32 <TrueBrain> owh, doesnt tell the endian result, nor the gcc version
15:11:36 <__ln__> now is it even possible to accidentally run something as BE on an ARM
15:11:38 <TrueBrain> I did such a bad job there :'(
15:12:19 <TrueBrain> __ln__: no, but for all we know OpenTTD things it is BE (gcc still makes a LE binary). It is just highly unlikely, as many more things would break :P
15:12:48 <TrueBrain> a crash.log tells more than the config.log, omg
15:13:37 <TrueBrain> its strange, as we do know the version in the configure
15:13:42 <__ln__> TrueBrain: but doesn't OpenTTD base its opinion on executing some binary...
15:13:42 <TrueBrain> it is just never written to the config.log :'(
15:14:01 <TrueBrain> __ln__: it runs a small app that checks the order of a short()
15:14:51 <TrueBrain> unsigned char endian_test[2] = { 1, 0 };
15:14:57 <TrueBrain> if (*(short*)endian_test == 1 )
15:16:08 *** th_gergo1 has joined #openttd
15:16:08 <__ln__> looks like i'm missing a lot of fun by not being an early adopter of raspberry pi. :( (yes, debugging cross-platform issues is my idea of fun)
15:17:32 <ssdg> __ln__: maybe you could enjoy qemu ;)
15:17:37 <Rubidium> then OpenTTD's bug tracker has a whole lot of fun for you! ;)
15:17:53 <TrueBrain> Rubidium: isnt it a pre to have those devices to start with? :D
15:19:27 <__ln__> i have a real ARM already though, a sheevaplug. but it obviously doesn't have any video output. (yes, X over SSH would be a possibility)
15:19:58 <TrueBrain> so get moving you lazy bumb :D
15:21:15 <__ln__> i gotta go have a beer with some german people first, but tonight i'll see what the sheevaplug can do. :)
15:22:06 <TrueBrain> I doubt you will be in any condition to do that then ;)
15:29:14 *** sla_ro|master has joined #openttd
15:29:41 *** andythenorth has joined #openttd
15:30:26 <TrueBrain> now I need to figure out how I can read Groups in an external Blend file ... lets seeeeeeeee
15:40:07 *** Prof_Frink has joined #openttd
15:48:17 *** Progman has joined #openttd
15:52:09 <TrueBrain> yippie! My blender script works :D Automation ftw :D
15:53:13 *** ElusiveParticle has joined #openttd
15:53:38 <ElusiveParticle> http://www.elusiveparticle.hostoi.com/ Excuse me, could one of you possibly do me a favor and visit my website? it's full of science stuff, I just want you to check it out It's interesting and full of video links, if you won't or you think it's some virus, don't worry about it it's not, just worry about your education, this is science please!
15:53:48 <frosch123> @kban ElusiveParticle
15:53:48 *** DorpsGek sets mode: +b *!b89e1566@ircip1.mibbit.com
15:53:49 *** ElusiveParticle was kicked by DorpsGek (frosch123)
15:53:49 <TrueBrain> @kban ElusiveParticle goodbye
15:53:50 <DorpsGek> TrueBrain: Error: ElusiveParticle is not in #openttd.
15:54:39 <TrueBrain> hmm .. how do I make 8bpp graphics out of my 32bpp?
15:54:45 <TrueBrain> any automated process you can let lose on it?
15:54:55 <frosch123> none which is worth the result
15:55:22 <TrueBrain> honestly, I dont really care how good it looks :P
15:55:51 <andythenorth> don't bother, 8bpp is dead :)
15:55:58 <TrueBrain> I still need to feed it in grfs ;)
15:56:09 <frosch123> TrueBrain: then just provide a black image
15:56:27 <TrueBrain> initial results via GIMP werent that bad
15:56:54 <andythenorth> draw a pixel art gravestone
15:57:10 <TrueBrain> I would get emails asking wtf it represents :P
15:58:10 <andythenorth> put in pixel font "for the avoidance if doubt, this is a gravestone"
15:58:20 <andythenorth> "representing the death of 8bpp"
15:59:15 <TrueBrain> anyway, not really helping :P
15:59:26 <TrueBrain> gimp has a command line ... now I need to know the command ...
16:05:59 *** mahmoud has joined #openttd
16:09:24 <TrueBrain> worlds worst command line :P
16:11:10 <TrueBrain> or so, yes; also, totally undocumented, it still boots the GUI stuff up, on error it just tells you there was an error (NO DEBUGGING WHAT SO EVER), ......
16:11:23 <frosch123> you cannot call just call a function, you first have to open several parantheses
16:11:49 <TinoDidriksen> So script ImageMagick instead?
16:12:00 <TrueBrain> TinoDidriksen: just looking at that yes
16:13:21 * andythenorth should write a pixa routine for upscaling pixels
16:16:21 <TrueBrain> imagemagck does something, at least I get a grey image :D :P
16:20:21 <TrueBrain> lol, greyscaled image
16:24:13 *** th_gergo has joined #openttd
16:25:00 *** Devedse has joined #openttd
16:41:18 *** |Jeroen| has joined #openttd
16:42:11 <NataS> i don't understand why people want an extra zoom level in
16:42:22 <NataS> and not extra zoom levels OUT
16:43:25 <TrueBrain> what has one to do with the other? (except them containing the word "zoom")
16:43:31 <TrueBrain> kinda a different audiance for both
16:51:09 *** telanus has joined #openttd
17:08:12 <TrueBrain> owh, some tools ... a png with 256 colours, greyscale, 8bit
17:08:57 <Eddi|zuHause> you can have python in gimp
17:09:46 <TrueBrain> ah, the alpha channel is giving troubles
17:10:49 *** Zeknurn has joined #openttd
17:11:54 *** Rhamphoryncus has joined #openttd
17:21:44 <TrueBrain> argh, why cant I just convert a RGBA to an indexed image .......
17:21:50 <TrueBrain> its so easy i nGIMP, via the freaking menu :(
17:27:32 <frosch123> hmm, once i connected two computers with the serial port and simulated a mouse with one of them, to automate a task on the other one
17:28:15 <Zuu> frosch123: sounds interesting and far more advanced than using the windows script host :-)
17:28:37 <frosch123> though i had to try quite long until the mousemovement worked out, because i failed to notice that it would be a good idea to disable mouse acceleraion :)
17:28:53 <Zuu> Especially as you only probably can do relative mouse movement. :-)
17:29:15 <Zuu> Although I guess one can reset to a corner easily.
17:29:25 <andythenorth> TrueBrain: load it in PIL, load a palette from another file (e.g. I have one you could use), apply the palette, save
17:29:30 <frosch123> Zuu: after each iteration of the task i moved the mouse hard top-left to recalibrate
17:29:32 <andythenorth> it's about 9 lines of python
17:29:39 <andythenorth> if you want to resize at that point you can
17:29:42 <TrueBrain> andythenorth: can that be automated?
17:29:45 <andythenorth> alpha channel might dick you around a bit
17:29:58 <andythenorth> TrueBrain python imaging library
17:30:07 <andythenorth> considered to be inferior to image magick, but meh
17:30:15 <andythenorth> IM looks complicated, PIL is simple
17:30:25 <TrueBrain> got an example script?
17:30:25 <frosch123> hmm, actually ... i think i used two serial ports to simulate mouse and keyboard, and i connected the mouse only because the windows serial keyboard only allowed ascii, so i could not simulate alt+hotkeys
17:31:02 <andythenorth> TrueBrain: I could write one, but no time right now, bathing a child :P
17:32:20 <TrueBrain> hmm, GIMP does not what he claims
17:32:26 <TrueBrain> he removes colour entries from the colour map
17:32:31 <TrueBrain> while I do uncheck it
17:32:38 <TrueBrain> so .. my colourmap image is broken to start with :'(
17:32:44 <TrueBrain> I need a png which contains all the OpenTTD colours :P
17:33:17 <andythenorth> TrueBrain: all 32bpp?
17:33:26 <TrueBrain> win I guess, I dunno :P
17:33:35 <andythenorth> DOS is de-rigeur
17:33:46 <TrueBrain> ANY will be fine for me :P
17:34:22 <TrueBrain> do you have any idea how long I search for that?
17:34:39 <frosch123> there are gimp palette files there
17:34:51 <TrueBrain> but I need an ImageMagick
17:35:27 <andythenorth> requires a paletted image called 'foo.png' :P
17:35:41 <andythenorth> will produce whatever 256 colours are in the palette, with index numbers
17:35:47 <andythenorth> it's a nice PIL example
17:36:04 <TrueBrain> awh, it is not part of Blender
17:37:31 <Eddi|zuHause> the easiest way to generate a palaetted image is "grfcodec -d -p{1,2} <whatever.grf>"
17:37:43 <TrueBrain> well, with that png file convert does what it should do, only it doesnt keep the fucking colormap :(
17:37:46 <TrueBrain> Eddi|zuHause: easiest? I disagree :P
17:40:04 <__ln__> so the steps to reproduce the RPI problem would be to: 1) build a track, 2) destroy a track, 3) anti-profit?
17:41:27 <__ln__> the forum post doesn't mention network nor multiplayer though
17:42:18 <__ln__> yeah, i'll begin investigating a bit
17:45:41 <frosch123> __ln__: i agree with albert, first check the sizes of uint32 and co
17:45:57 <frosch123> hmm, though maybe we have an assert_compile for those
17:46:24 <CIA-1> OpenTTD: translators * r24196 /trunk/src/lang/ (14 files): (log message trimmed)
17:46:24 <CIA-1> OpenTTD: -Update from WebTranslator v3.0:
17:46:24 <CIA-1> OpenTTD: afrikaans - 33 changes by telanus
17:46:24 <CIA-1> OpenTTD: belarusian - 6 changes by KorneySan, Wowanxm
17:46:24 <CIA-1> OpenTTD: croatian - 4 changes by VoyagerOne
17:46:25 <CIA-1> OpenTTD: czech - 4 changes by Paragulis
17:46:25 <CIA-1> OpenTTD: dutch - 8 changes by habell
17:52:51 *** FLHerne has joined #openttd
18:02:45 <TrueBrain> how nice .. I now have it to make me a 256 palette png
18:02:50 <TrueBrain> it first does the colours it used
18:02:53 <TrueBrain> then fills it with the rest
18:02:57 <TrueBrain> +++++ software I am using here!
18:06:59 <TrueBrain> lol: last updated May 2005
18:16:53 *** Chris_Booth has joined #openttd
18:19:43 <TrueBrain> known issue? am I doing something wrong? is there a better source?
18:22:02 <frosch123> it's a broken win palette
18:22:13 <TrueBrain> owh, it is from the original grf
18:23:14 <TrueBrain> explains a few things :D
18:23:29 <TrueBrain> (why my earlier attempts failed, but with the png from andythenorth it worked)
18:23:51 *** ZxBiohazardZx has joined #openttd
18:24:48 <ZxBiohazardZx> is there a way to change industry parameters in opengfx? --> forests only above snowline is stupid if you ask me, i prefer forests on hillslopes, or in the valley's rather then on top of my mountain.....
18:25:01 <ZxBiohazardZx> and ofc sprites are drawn with snow
18:25:24 <ZxBiohazardZx> and/or make them snow-aware ?
18:26:09 <ZxBiohazardZx> also NataS here?
18:30:56 <ZxBiohazardZx> was wondering about the cuba game :P
18:31:02 <ZxBiohazardZx> want me to start a new one
18:31:12 <ZxBiohazardZx> on Cargodist or the Chrill pack?
18:31:15 <NataS> yes, and Actualy, can we do a diffrent map?
18:32:27 <NataS> this is my favorite map, however it does have a rare newgrf, I have removed it and played a full game without any problems though.
18:32:49 <NataS> also, what newgrfs do you like?
18:32:51 <ZxBiohazardZx> i think roma had some similar maps
18:32:53 <ZxBiohazardZx> but whatever you want
18:33:01 <ZxBiohazardZx> as long as the grfs are on banana's im cool with anything
18:33:27 <NataS> I like TRS, ISR, Japan stations, FISH AV8, HEQS and EGRVTS
18:33:32 <NataS> that's what I had for the last game lol
18:33:44 <andythenorth> have CHIPS with your FISH
18:34:15 <NataS> i like where chips is going, but it needs more things before it can replace IRS
18:34:25 <andythenorth> it's not trying to replace ISR
18:34:39 <__ln__> trunk compilation time on 1.33GHz PPC: 27m30s, 1.2GHz ARM: 43m9s
18:34:39 <ZxBiohazardZx> FISH, Dutchsets (dunno why i just like it), ISR (cause its epic) CHIPS (has some nice tiles, but lacks some others)
18:34:53 <ZxBiohazardZx> Andy you created CHIPS right?
18:35:07 <NataS> and having one station set that does everything well and is consistant with itself would be ideal
18:35:12 <ZxBiohazardZx> darn i forgot what tile i lacked on it
18:35:13 <NataS> right now I load ISR and japan stations
18:35:26 <ZxBiohazardZx> i had a nice station in my mind but i lacked one sort of CHIPS tile to make it complete :P
18:35:36 <NataS> but that leads to clutter, inconsistancy, and unused station types.
18:35:39 <andythenorth> if you remember let me
18:35:51 <andythenorth> CHIPS has room for another 5 or 6 tiles I reckon
18:35:54 <andythenorth> then it's done, done done
18:36:15 <ZxBiohazardZx> do you have a list of current tiles
18:36:28 <ZxBiohazardZx> <-- to lazy to open ingame just to check the current tiles :P
18:36:32 <NataS> will CHIPS have more buildings and more complex stations?
18:37:20 <ZxBiohazardZx> lemmy just go ingame and check andy
18:37:22 <andythenorth> most things it lacks are available elsewhere
18:37:32 <ZxBiohazardZx> unless you have a full spritesheet laying around
18:37:40 <ZxBiohazardZx> yeah but i found something others also lacked
18:37:47 <ZxBiohazardZx> or at least didnt have a matching look for :)
18:37:50 <NataS> in one screenshot I saw what looked like an ISR building on a CHIPS platform, what is that?
18:37:56 <NataS> some sort of custom grf?
18:38:34 <ZxBiohazardZx> 0.7.0 is latest? (i doubt so)
18:39:42 <andythenorth> it will get to 1.0 quite fast
18:39:43 <ZxBiohazardZx> so you have hte 4 base-baretiles and the 12 additional tiles :P
18:39:51 <ZxBiohazardZx> i think DWE had some nice tiles
18:39:57 <ZxBiohazardZx> but they didnt quite match the styles
18:40:13 <andythenorth> the new objects thing by Quast65 looks good
18:42:13 <ZxBiohazardZx> you mean the additional ISR tiles from him? (dock ones?)
18:42:42 <TrueBrain> okay, giving up trying to convert a 32bpp to a 8bpp palette ... omfg ... who would have thought ..
18:46:07 <ZxBiohazardZx> andythenorth i miss some FIRS styled total indoor unloaders (like a huge shed, or some tiles like the FIRS aluminiumplant/cementplant) and some raised tiles with some random FIRS stuff on it might be nice as well
18:46:22 <ZxBiohazardZx> aka muddy tracks for the sand/forests?
18:47:09 <ZxBiohazardZx> a bit less muddy then the platforms you have now, a tad more greenish on the sides, if that made/makes any sense
18:47:19 <__ln__> the bad news: it's fucking slow over wi-fi and X11; the good news, i cannot destroy a piece of track because it costs £241.591.910.343 and i don't have that much.
18:47:50 <ZxBiohazardZx> aka mud tiles should be more like the Quarry platforms, fading towards the "normal" green grass (or arctic brown grass, whatever is used)
18:48:45 <TrueBrain> __ln__: enjoy figuring out why :D
18:49:06 <TrueBrain> make sure to check the Money class, check uints and ints etc, make sure BE/LE is all in order, etc etc :)
18:49:34 <TrueBrain> __ln__: and if you cant find it, please do feed us with information like gcc version, architecture, and basic other information; I am sure you know which we like :D
18:49:40 <ZxBiohazardZx> <__ln__> / TrueBrain, the costs are insane due to inflation that is huge?
18:51:09 <ssdg> TrueBrain: still on the ARM arhitecture case study?
18:51:42 <__ln__> 21:47 < __ln__> the bad news: it's fucking slow over wi-fi and X11; the good news, i cannot destroy a piece of track because it costs £241.591.910.343 and i don't have that much.
18:52:53 <TGYoshi> [random breakthrough] Any of you have any idea how to calculate the distance between a point and a straight line? [/random breakthrough]
18:53:18 <andythenorth> ZxBiohazardZx: tile mixing with default ground from the tile are desired, but are pretty much impossible due to station spec
18:53:35 <andythenorth> at least to do well
18:57:41 <__ln__> btw, i also confirmed that destroying tracks is not that bloody expensive on PPC.
18:58:23 <ZxBiohazardZx> andythenorth yeah true, i just dislike teh whole brown thing as mud, i would prefer it to blend in a bit with the tile next to it :P
18:58:44 <ZxBiohazardZx> as i thought stations where an overlay so you could just make a transparant tile and have it solved
18:59:35 <ZxBiohazardZx> its plain math :P
18:59:45 <ZxBiohazardZx> you have point (x,y) or (x,y,z)
19:00:06 * Rubidium is stupid... trying to reproduce the arm issue on ppc
19:00:06 <ZxBiohazardZx> R2 = <x,y> R3 = <x,y,z>
19:00:30 <ZxBiohazardZx> ok so you need point -> line R2
19:00:33 <__ln__> R2 is also a tiny robot in star wars.
19:00:45 <ZxBiohazardZx> yeah yeah TGYoshi just needs some math done :P
19:00:48 <ZxBiohazardZx> 1 sec phone though
19:00:53 <ZxBiohazardZx> ill explain after phone :P
19:01:01 <andythenorth> ZxBiohazardZx: blend might be possible, we tried lots of things
19:01:05 <TGYoshi> Dont forget vertical lines :3
19:01:08 <andythenorth> can't remember which ones work
19:01:15 <andythenorth> I think we always get green grass
19:01:30 <ZxBiohazardZx> im even to lazy to explain wolfram is your friend for it
19:01:39 <TGYoshi> Thats what my problem is :P
19:01:48 <ZxBiohazardZx> andy if its 100% transparent, then why is it green
19:02:15 <TGYoshi> My lines arent ax+by+c, they are actually line [parts] defined by two points
19:02:17 <andythenorth> ZxBiohazardZx: base tile
19:02:28 <TGYoshi> (so they can also be vertical etc..)
19:02:31 <ZxBiohazardZx> GTYoshi any line can be represented by a vector
19:02:34 <ZxBiohazardZx> its vector math
19:03:29 <ZxBiohazardZx> the beauty of math is that eventually they always find a formula that is true in ALL X/Y combinations
19:03:38 <ZxBiohazardZx> that is vertical, horizontal and even R^3
19:04:17 <ZxBiohazardZx> a horizontal line is 0 1 a vertical line is 1 0
19:04:23 <ZxBiohazardZx> the vector equation still holds
19:04:35 <ZxBiohazardZx> regardless of values
19:04:41 <ZxBiohazardZx> goddamn Mila Kunis is hot
19:05:37 <andythenorth> ZxBiohazardZx: if you have pictures of structures you think should be considered...post links?
19:05:54 <ZxBiohazardZx> will do andythenorth, im just trying to think of a structure
19:06:05 <ZxBiohazardZx> i just lacked stuff to stick to the cement / aluminium plant
19:06:27 <ZxBiohazardZx> as they have a more modern architecture and there isnt a stationset that truely fits with that
19:06:47 <ZxBiohazardZx> so maybe some grey-toned warehouses like on FIRS (also your creation afaik so )
19:07:36 <ZxBiohazardZx> TGYoshi 12 down is more relevant for your problem btw, you can do it with 2 points in (any) space
19:08:02 <andythenorth> ZxBiohazardZx: have a look at what pikka did for finescale track grf depots
19:08:02 <ZxBiohazardZx> equation 12 down :P
19:08:26 <ZxBiohazardZx> got a hotlink for me ? XD
19:08:43 <ZxBiohazardZx> TGYoshi on the wolfram site i linked, there are (12) on the right side to indicate equation stuff
19:09:00 <ZxBiohazardZx> from 12 down its about any line between 2 points and any point in the space around it
19:09:20 <ZxBiohazardZx> yeah i liked the grey bedding on the side, and also the depot tiles are very nice
19:09:32 <andythenorth> I can include any of those
19:09:39 <andythenorth> pikka has done both angles ;)
19:09:45 <ZxBiohazardZx> i seriously think we lack some good depot tiles for normal scale tracks as well, i always used the bigger depots so i could make them linking up
19:09:59 <ZxBiohazardZx> well the middle one looks nice for usage in a station as well if you ask me
19:10:06 <ZxBiohazardZx> and bottom right one
19:10:12 <TGYoshi> ZxBiohazardZx ahhhh I see, thanks - random v variabele
19:10:17 <ZxBiohazardZx> the other tiles are nice but less likely to be usable for stations i think
19:10:30 <ZxBiohazardZx> less realistic layouts, in those cases the building would be on 1 side
19:10:47 <andythenorth> I might just use all of them :P
19:11:02 <ZxBiohazardZx> haha not bad, pikka always has some nice graphics
19:11:48 <ZxBiohazardZx> i also like that trackbed on the right there :P
19:12:06 <ZxBiohazardZx> anyway i still hope you can get muddy to be less, eeehm brown flat surface
19:12:16 <ZxBiohazardZx> maybe make it green/brownish as in the forest tiles?
19:12:27 <andythenorth> [shrug] it matches the mud tiles in FIRS
19:12:35 <andythenorth> not the opengfx one though
19:12:51 <ZxBiohazardZx> im comparing to FIRS
19:12:54 <andythenorth> forest style tiles is a valid idea
19:13:10 <ZxBiohazardZx> hence i suggested for aluminium plant & cement plant
19:13:18 <ZxBiohazardZx> you lack tiles for some nice industry combinations
19:13:29 <andythenorth> every time the ground tile is discussed, I have to remind myself why it's not possible :P
19:13:31 <ZxBiohazardZx> eg the mud ones for me would be nice for forests, and also the wooden hut with it
19:13:45 <ZxBiohazardZx> but it turns out to be a bit to muddy for me
19:13:52 <ZxBiohazardZx> and just turns into 1 huge brown square
19:13:59 <ZxBiohazardZx> rather then being "realistically muddy"
19:14:22 *** FLHerne has joined #openttd
19:14:29 <andythenorth> ok, the issue is that the tile underneath is not terrain aware
19:14:42 <andythenorth> so you get temperate grass, arctic grass, or tropic grass
19:14:43 <TGYoshi> ZxBiohazardZx thank you, figured it out
19:14:47 <ZxBiohazardZx> because spec demands so?
19:15:00 <ZxBiohazardZx> im starting to hate the spec
19:15:02 <andythenorth> so no snow, no partial snow, no desert, no partial desert
19:15:07 <ZxBiohazardZx> same for stupid forests above snowline
19:15:27 <ZxBiohazardZx> andy is that because the spec expects the station tile to cover the entire square?
19:15:52 <andythenorth> it's also because the station spec is fricking insane
19:15:52 <ZxBiohazardZx> eg original gfx covered full ground-square, so all newgrfs should also do so and dont have option on having transparency?
19:16:02 <ZxBiohazardZx> so no chance on any fix/changes there
19:16:11 <andythenorth> it's also because rail types can't really handle anything else properly
19:16:20 <andythenorth> it's also because of historical reasons
19:16:28 <ZxBiohazardZx> well the rails are tile/terrain aware
19:16:32 <ZxBiohazardZx> just stations are not
19:16:47 <andythenorth> and it's also because it was thought useful to provide a 'magic sprite' that provided the rails to newgrfs, but the magic sprite is not terrain aware
19:17:01 <andythenorth> it's resolvable, but someone would have to resolve it
19:17:07 * __ln__ is recompiling with debug symbols, d'ogh
19:17:09 <ZxBiohazardZx> yeah as with most suggestions
19:17:17 <ZxBiohazardZx> suggestion is good, but implementation is a biatch
19:17:17 <andythenorth> and because the station spec is insane...it's preferred not to pile more on it
19:17:36 <ZxBiohazardZx> i know all about that, im king of coming up with suggestions that require spec rewrites and are close to impossible
19:17:40 <ZxBiohazardZx> but they would be nice :P
19:17:54 <ZxBiohazardZx> im gonna go download that movie
19:17:58 <ZxBiohazardZx> i think mila is hot in it
19:18:09 <ZxBiohazardZx> and she being lesbo with zoe saldana is not a bad movie
19:20:30 <Eddi|zuHause> yeah... thunderstorm...
19:21:11 <Eddi|zuHause> wheather... it's what's "outside"... where the pizza comes from... you know?
19:21:32 <ZxBiohazardZx> weather (no h?)
19:22:28 <Eddi|zuHause> you gonna ride around on any spelling mistake of a non-native speaker now?
19:23:16 <Rubidium> Eddi|zuHause: with monday and such? ;)
19:24:28 <ZxBiohazardZx> Eddi yes specially since i have seen you typing without that many of them
19:31:48 *** Biolunar has joined #openttd
19:34:00 <__ln__> something called "Open Transport Tycoon" has made it on the raspberrypi.org frontpage.
19:34:27 <andythenorth> make sprite 1012 etc terrain aware?
19:34:38 <andythenorth> or rather, make whatever returns sprite 1012 to stations terrain aware?
19:34:55 <ZxBiohazardZx> andy who your asking/talkin to lol
19:36:13 *** telanus1 has joined #openttd
19:37:17 *** telanus2 has joined #openttd
19:38:27 <ZxBiohazardZx> ugh hate that google maps wont give me 3d for the building i want to see :P
19:38:43 <ZxBiohazardZx> looking @steel mills near IJmuiden Netherlands, Corus plants
19:38:52 <ZxBiohazardZx> to give you an idea what i mean for CHIPS
19:40:00 <ZxBiohazardZx> enclosed sheds with shunters pulling teh ores from the harbor towards teh actual steel mills
19:40:21 <ZxBiohazardZx> and then similar shunters pulling the sheeted steel-rolls to gathering tracks again
19:40:25 *** Eddi|zuHause2 has joined #openttd
19:41:34 <ZxBiohazardZx> i know its not ideal to see it topside only
19:41:46 <ZxBiohazardZx> but the stuff you see there is steelmills in the netherlands
19:41:49 <Wolf01> [21:38:47] <ZxBiohazardZx> looking @steel mills near IJmuiden Netherlands, Corus plants <- why did I read Coruscant?
19:41:58 <ZxBiohazardZx> haha Wolf01 cause you can
19:42:30 <ZxBiohazardZx> anything down to the harbor from the "waterweg" is part of the mill & terrains
19:43:52 <ZxBiohazardZx> anyway mainly trying to show andythenorth that CHIPS might be able to offer some nice enclosed tiles (not to many) that fit well with both the aluminium factory from FIRS and cementplant from FIRS
19:45:26 <ZxBiohazardZx> yeah mainly its sheds
19:45:40 <ZxBiohazardZx> although in there the actual steel mills (hot, cold, etc etc) are inside the sheds :P
19:45:48 <ZxBiohazardZx> so matching with the FIRS factories would be nice
19:46:02 <ZxBiohazardZx> so forest tiles & modern-FIRS-industrial tiles :P
19:46:05 <ZxBiohazardZx> 2 suggestions :P
19:46:45 <ZxBiohazardZx> imma go watch tv, brb if at all tonight :P
19:47:58 *** Eddi|zuHause2 is now known as Eddi|zuHause
19:53:55 <TrueBrain> and __ln__, how is it going? :)
19:54:48 <__ln__> i figured i had compiled without debug symbols, and recompiling is soon finished
19:55:24 <TrueBrain> ssdg: and yeah, somewhat 'good' news is that __ln__ can reproduce it locally too, so an ARM issue; now the question would be: what? :P
19:55:53 <TrueBrain> ugh, I do dislike people calling this game Open Transport Tycoon. It is called OpenTTD ffs!
19:57:52 <__ln__> indeed, and the big reason is that "Transport Tycoon" is (or at least was) a registered trademark with a known owner.
19:57:55 <Zuu> TrueBrain: Indeed, I had to hold back myself from posting a silly comment on the 3.14 website about the naming.
20:00:22 <TrueBrain> __ln__: and we are called OpenTTD :P
20:00:28 *** Doorslammer has joined #openttd
20:01:28 <__ln__> TrueBrain: that's right, but the title screen used to say "Open Transport Tycoon" in the early days.
20:01:40 <TrueBrain> long long long long long ago :)
20:01:43 <Zuu> There is also one comment that imply that OpenTTD trunk is unstable. Something I don't think is very true. Sure a few times a year it might have a fatal problem, but its not really that fragile as they make it sound.
20:01:46 <TrueBrain> then again, the title screen used to be sea :P
20:02:02 <TrueBrain> Zuu: indeed, "unstable" is a big word
20:02:56 <Rubidium> "debian unstable" unstable ;)
20:09:55 <Eddi|zuHause> was somewhere between r22434 and now a fix for bounding box drawing? my old revision hangs when i press ^B, in newer ones it doesn't...
20:16:06 * Rubidium wonders what Greyhound is up to
20:16:15 * NGC3982 is in love with his current sky.
20:16:20 <Rubidium> cause their pricing makes no sense
20:16:26 <NGC3982> venus, aldebaran and a lot of colors.
20:16:39 <DorpsGek> Eddi|zuHause: 14.6666666667
20:16:44 <DorpsGek> Eddi|zuHause: 117.333333333
20:18:26 <Rubidium> Toronto -> Montreal costs CAD 20 (via Ottawa). Toronto -> Ottawa costs CAD 39 and Ottawa -> Montreal costs CAD 21
20:19:30 <Rubidium> and you'd be in the exact same busses
20:19:38 <Wolf01> that's right, the more you travel, the less you pay :D
20:19:48 <Rubidium> the train costs CAD 78 though :(
20:20:35 <Wolf01> bah, trains cost all the same in the world, and tt doesn't make an exception
20:21:21 <Belugas> trains are more comfy than busses?
20:22:22 <Rubidium> definitely not all trains are
20:23:01 * ssdg is sorry, there is the french presidential debate on TV and it's quite hard to follow
20:23:05 <Belugas> since /me only travles by car...
20:23:30 <Wolf01> if I didn't have to wait 2 hours for check-in, I'll travel by plane from venice to milan, it costs less and take less time than train (which costs about 50€ and takes 3.5-4 hours), too bad the nearest airport is at 80km from my house :(
20:23:55 <Belugas> commuters bus are a bit different then inter cities ones...
20:24:11 <Eddi|zuHause> Wolf01: some people complain that the airport is closer to their house :p
20:24:17 <Belugas> i mean.. traveling, like "relatively" long distance
20:25:02 <ssdg> Congratulations TrueBrain on reproducing my bug by the way
20:25:10 <Rubidium> Wolf01: don't take checked luggage, then it's only 30-45 minutes if you check in via internet
20:25:47 <Wolf01> but I have 1 hour of travel to reach the airport
20:26:19 * Rubidium knows it is at least 15 minutes from the train station to the (almost) furthest gate at AMS
20:26:32 <Rubidium> for inner Schengen travel
20:38:34 <ssdg> TrueBrain: We reproduced it here too. with the debug settings. Interested?
20:38:56 <TrueBrain> __ln__: and, how is it going on your end?
20:39:11 <TrueBrain> ssdg: well, it turns out to be an ARM issue, seemly with Money (a special 64bit integer)
20:39:18 <TrueBrain> so a desync log etc are not super useful
20:39:26 *** ZxBiohazardZx has joined #openttd
20:39:30 <TrueBrain> ssdg: did you try the debian package?
20:40:16 <ssdg> Not Yet. (or it was during my commute and I didn't know about it). (Yes, I take real non-virtual trains too :) )
20:41:27 * ssdg is quite confuse. how Money causes a bug in tracs?
20:41:54 <TrueBrain> we dont know yet :D
20:41:58 <ssdg> Or did you call money what might just be a bigbigint ?
20:42:00 <TrueBrain> it might be something completely different
20:42:02 <Rubidium> imagine the amount you have to pay on your computer is 1$ and 100.000.000.000$ on the server
20:42:07 <TrueBrain> might be a gcc that fucks up .. who knows :)
20:42:23 <TrueBrain> Money is a type we defined which is practical a 64bit integer
20:42:29 <TrueBrain> it has some .. additional code attached to it
20:42:39 <TrueBrain> INT64_MAX + 1 = INT64_MAX
20:42:43 <Rubidium> then on the server you don't have enough money to remove it, whereas on the client you have; thus it's being removed for you but not on the server
20:42:59 <__ln__> TrueBrain: there would seem to be some crazy value in Money's m_value right after it's been initialized to zero, or something. but i don't know why yet.
20:43:18 <TrueBrain> __ln__: destroying airports, town houses etc, does work?
20:43:26 <TrueBrain> its only railways that give this issue?
20:43:43 <Rubidium> try roads as well, as you get money from they too
20:44:03 <ssdg> In fact, I have to translate a bit of "how-to" give me one sec
20:44:07 <TrueBrain> building and removing "land owner" signs?
20:44:44 <__ln__> it would seem that destroying anything that costs money to destroy does work at least
20:45:18 <__ln__> selling an engine works too
20:46:14 <ssdg> 1) Choose a tile going down, say top on S-E, bottom on NW
20:46:14 <ssdg> 2) build a track on it, said track going from the hill to the valley.
20:46:14 <ssdg> 4) Remove said track and builld a perpendicular one (creating a cliff)
20:48:15 <__ln__> seems like i'm doing a full rebuild again
20:49:11 <TrueBrain> __ln__: in the meantime, do you have gcc version for us
20:49:19 <TrueBrain> can you tell if the BE/LE detection went okay?
20:55:21 * ssdg is open to critics if some info is missing
20:57:02 <__ln__> TrueBrain: g++ (Debian 4.4.5-8) 4.4.5
20:57:54 * NGC3982 is so gonna make that soylent green scenario grf.
20:58:05 <__ln__> endian_check reports TTD_LITTLE_ENDIAN
20:58:28 <TrueBrain> so that leaves a compiler issue I guess ...
21:01:05 <Rubidium> given 4.4.5 is broken, it must be broken for a long time
21:01:45 <Rubidium> which would mean lots of broken versions might be floating around
21:02:44 <TrueBrain> Rubidium: sadly, you talk in crypt to me, and I have no clue if you are sarcastic (which you tend to be if you disgree on a statement, no offense, just trying to figure out what you try to say): do you mean that GCC 4.4.5 is known to be broken, or are you trolling?
21:03:49 <Rubidium> earlier someone talked about GCC 4.6.3 being used to create a broken OpenTTD, now GCC 4.4.5. That means the issue (if in GCC) is already pretty old
21:04:03 <TrueBrain> just realise we talk about an ARM-branch of GCC here
21:04:08 <TrueBrain> it doesnt have to be in the mainstream GCC
21:04:18 <TrueBrain> it can be just some optimization for the ARM platform that is not doing what we want from it
21:06:22 <Zuu> Wolf01: You make a good point why not use "km" as length unit for traveling. Going 8 "mil" sounds much shorter than 80 km, while still being the same distance :-)
21:07:50 <TrueBrain> so a test-case for __ln__ would be: CFLAGS="-fno-cse-follow-jumps" ./configure && make ;)
21:08:02 <Wolf01> Zuu: never heard of that
21:08:51 <Zuu> Wolf01: Its a Swedish thingy to confuse people comming here and thinking that a "mil" is 1.6 km, while it is actually 10. :-)
21:09:08 <ssdg> Zuu: Not quite sure about this
21:09:26 <ZxBiohazardZx> miles are 1.6 kms
21:09:42 <Yexo> Zuu: google "8 mil to km", it says "8 mil = 2.03200 × 10-7 kilometers"
21:09:50 <Rubidium> GCC 4.4.5 does not have that particular fix
21:10:14 <ZxBiohazardZx> thats cause it calculates mil as million haha
21:10:33 <ZxBiohazardZx> 60 miles/hour = 97 km/hour
21:10:43 *** Zeknurn has joined #openttd
21:10:43 <Yexo> Mil, also known as a thou is equal to one thousandth of an inch (0.001 inch).
21:10:50 <Wolf01> still better than furlongs
21:11:08 <ZxBiohazardZx> a strange factor of it
21:11:33 *** andythenorth has left #openttd
21:12:18 <Zuu> English wikipedia: In Norway and Sweden, a mil is a unit of length equal to 10 kilometres and commonly used in everyday language. However in more formal situations, such as on road signs and when there is risk of confusion with English miles, kilometres are used instead.
21:13:00 <Rubidium> "[...] Scandinavian mil ranges from 7.5km to 12km depending on nation or region" ;)
21:13:31 <Zuu> According to wikipedia that was standardized in 1649.
21:13:53 <Yexo> When the Metric system was introduced in Norway and Sweden in 1889 (the actual law having been passed in 1875), the mil was redefined to be exactly 10 km. <- it says 1889 actually
21:13:53 <Zuu> to 10.687 km, and then in 1889 to exactly 10 kilometers.
21:14:32 <ZxBiohazardZx> ugh i hate the new blender versions :(
21:14:40 <ZxBiohazardZx> not joking, i loved the old interface
21:14:49 <ZxBiohazardZx> but they fucked up alot of the stuff i liked about it
21:15:11 <Yexo> TrueBrain: now import nml and call run nmlc directly from blender too :p
21:15:23 <TrueBrain> I can now make a nice Lego model in LeoCAD, export it to 3DS, put it in a blender file, hit Run Script, and enjoy it ingame :D
21:15:39 <Zuu> TrueBrain: You make it sound easy :-)
21:15:51 <TrueBrain> Zuu: after a days work, it is
21:16:15 <ZxBiohazardZx> TrueBrain why not create similar script for LegoCAD?
21:16:35 <Rubidium> ZxBiohazardZx: that'd be cheating ;)
21:17:35 <Yexo> it won't work a second time unless it's completely unloaded/reloaded though
21:17:43 <Yexo> too many global variables
21:18:20 <TrueBrain> it runs nmlc.exe --grf blabla
21:18:26 <TrueBrain> so that would work as many times as needed ;)
21:18:42 <TrueBrain> just atm blender fails to do a second render
21:19:07 <Yexo> since nml is written in python, my initial though was about "import nml; nml.main('', '--grf', 'filename'...);"
21:19:16 <TrueBrain> no, that is too scary for me :D
21:19:19 <ZxBiohazardZx> blender and nml are both python?
21:19:24 <Zuu> As a fan of toyland I look forward to whatever comes out of the lego project. :-)
21:19:29 <ZxBiohazardZx> then yeah it should be directly linkable
21:19:35 <TrueBrain> Zuu: it already looks better, so it cant go worse :P
21:19:49 <ZxBiohazardZx> i think the latest blender actually offered more shit to be compatible for games and creating them (eg sprite creation etc)
21:19:49 <Eddi|zuHause> <Yexo> since nml is written in python, my initial though was about "import nml; nml.main('', '--grf', 'filename'...);" <-- i did that once
21:20:47 <Eddi|zuHause> mainly to avoid the nml exception handling though
21:21:00 <Eddi|zuHause> to attach a proper post-mortem debugger
21:21:19 <TrueBrain> the only "requirement" for my method, is that you put objects you want rendered in a group, where the name is: <whatever>,spriteid,spriteid,spriteid,spriteid
21:21:29 <TrueBrain> where the amount of spriteids give the amount of rotations :P
21:22:04 <TrueBrain> only y_offs is a bitch
21:22:08 <TrueBrain> as it is not the center of the image
21:22:24 <TrueBrain> (where x_offs is), for terrain at least
21:24:27 <Zuu> TrueBrain: Where is the lego/brickland thread? Or is it just locally at your computer? :-)
21:25:08 <Eddi|zuHause> TrueBrain: x_offs/y_offs define the north corner of the tile
21:25:40 <Zuu> TrueBrain: Oh, sorry, I just realized there is a search function at the forum :-)
21:26:20 <Eddi|zuHause> TrueBrain: or the north (hidden) corner of the bounding box for vehicles
21:27:57 <__ln__> what does --enable-debug=3 do compared to --enable-debug?
21:28:40 <Eddi|zuHause> __ln__: afair it's the same thing
21:28:54 <Zuu> well, take your time and have fun with the awsomeness :-)
21:28:55 <TrueBrain> there are several levels of debugging, in most parts of the code
21:28:56 <Eddi|zuHause> --enable-debug defaults to 3
21:29:10 <TrueBrain> Zuu: yeah; if I have the landscape how I like it I will slap it in a thread
21:29:20 <TrueBrain> I am sure some 32bpp gurus will flame me, as I made my own blender and everything
21:29:37 <TrueBrain> the sun is how I think it should be etc
21:29:47 <__ln__> Eddi|zuHause: can't be
21:29:59 <TrueBrain> the default blender has a glitch in camera position, giving slightly below optimal images back, etc :P
21:30:16 <Eddi|zuHause> __ln__: then it was either changed or mis-communicated
21:30:58 <Zuu> The best I ever made in blender before giving up on it was an office tape stand. :-)
21:31:09 <TrueBrain> I never made anything in blender :P
21:31:16 <TrueBrain> I make it in LeoCAD, and import it :D
21:31:21 <Rubidium> Eddi|zuHause: enable-debug defaults to 1
21:31:30 <__ln__> with --enable-debug=3 it takes several minutes to link the openttd binary, and the ARM bug is not present.
21:31:53 <Eddi|zuHause> __ln__: usually the higher debug levels make linking _Faster_
21:31:58 <TrueBrain> owh, lolz, that --enable-debug :D
21:32:05 <TrueBrain> I was tihnking -debug when launching .. oops :D
21:32:13 <TrueBrain> __ln__: =3 is without any optimizations
21:32:25 <Rubidium> uhm.. 2 = -g -fno-inline
21:32:32 <Rubidium> 3 = -g -O0 -fno-inline
21:32:40 <TrueBrain> __ln__: so that confirms compiler issue
21:32:59 <TrueBrain> or, well, it can still be a few other things, but it is very likely :D
21:33:09 <ssdg> Seems to work with the .deb
21:33:47 <TrueBrain> ssdg: well, that at least means you can play the game for now :)
21:34:19 <ssdg> I hope our package will help you find why it didn't work at first
21:34:41 <TrueBrain> well, we are closing down on it, so that is great news :)
21:34:45 <TrueBrain> tnx for all the info etc :)
21:35:33 <TrueBrain> Rubidium: does 4.4.6 contain the fix for that bug report you linked?
21:35:43 *** ZxBiohazardZx has left #openttd
21:35:53 <Eddi|zuHause> ssdg: so, who is winning the debate? :)
21:36:21 <ssdg> Depends of who's looking. That debate almost never changes the trends
21:36:59 <ssdg> It did just once... since the beginging of the 5th republic. (it's roughtly 50 years old)
21:37:32 <Eddi|zuHause> ssdg: has any president ever won in the first round?
21:37:48 <ssdg> I guess Charkes the gaule.
21:38:06 <Eddi|zuHause> he must get 50% of the votes for that?
21:39:30 <ssdg> and actually, I was wrong. Nobody won in the first round
21:40:53 <ssdg> First round: select two persons out of everybody who got 50 mayors/other people with similar jobs. Second round whoever has more than 50% wins
21:40:59 <Rubidium> TrueBrain: I'm getting an internal server error when I want to get the changelog of 4.6.3, so I have no clue :(
21:41:05 <ssdg> First round: select two persons out of everybody who got 50 mayors/other people with similar jobs signatures. Second round whoever has more than 50% wins
21:41:52 <Rubidium> ssdg: and what if it's 48% 49% and 3% blank?
21:42:28 <ssdg> Blank is not taken into account. (and it's wrong)
21:42:33 <Eddi|zuHause> i was under the impression in the second round a simple majority suffices
21:43:08 <Eddi|zuHause> or are really only 2 people allowed in the second round?
21:43:12 <Rubidium> though I wonder what happens when there isn't a majority (equal votes)
21:43:25 <glx> Eddi|zuHause: could be 3, but usually it's only 2
21:43:30 <ssdg> Rubidium: Seems statisticaly improbable to me
21:45:03 *** FLHerne has joined #openttd
21:45:45 <Rubidium> ssdg: less probable than having the one with the fewest actual votes being declared the winner, but that happened as well. Which, for me, would imply needing some margin by which you should win to prevent a few miscounts from selecting the wrong, which would make it more likely that the votes are "equal"
21:47:24 <ssdg> Rubidium: Indeed, errors are possible.
21:48:02 <__ln__> the case of a broken GCC in something as popular as Debian stable leaves open the question: what else is broken, both in OTTD and the whole Debian on ARM...
21:50:18 <Eddi|zuHause> makes you wonder where miscounts have larger effect: in a majority vote, or in a proportional vote
21:50:35 <TrueBrain> __ln__: did you try compiling with that special CFLAG?
21:51:08 <TrueBrain> __ln__: and if you read the bug report Rubidium posted earlier, it is a pretty severe issue
21:51:31 <TrueBrain> Debian on ARM is stabalizing, but is far from stable (same for GCC :))
21:51:57 <__ln__> TrueBrain: i'm in the process of trying that, but the compilation is only at the letter 'n' at the moment
21:52:20 <Rubidium> Eddi|zuHause: the former, though in the latter you'll notice it more often
21:53:24 <Eddi|zuHause> Rubidium: in proportional vote, the effect of a miscount is almost always visible, while in a majority vote, the effect is only visible if it is a close outcome
21:53:55 <__ln__> any serious development work on ARM would require a distcc cluster of sheevaplugs i suppose
21:54:23 <Eddi|zuHause> but if it is visible in the majority vote case, then it has the bigger impact on the overall picture
21:54:24 <Rubidium> yes, but *if* you have an effect in the majority vote it's 100% from the actual truth
21:55:10 <Eddi|zuHause> but you can't have proportional vote in a presidal election, where only one seat is to be distributed :p
21:56:59 <Eddi|zuHause> it gets really silly in hybrid systems like the german parliamentary elections. like the 2009 elections of the Schleswig-Holstein parliament, where the parties with the lower "popular vote" got more seats.
21:57:23 <Eddi|zuHause> the vote was later declared illegal, now they re-vote
21:57:39 <Rubidium> I'd say the American system is the silliest
21:58:20 <Rubidium> (of the ones I know)
21:58:35 <Eddi|zuHause> the british used to have a silly system
21:58:52 <Eddi|zuHause> where the electorate constituencies were defined by area instead of population
21:59:01 <Eddi|zuHause> so there was one constituency with a single voter
21:59:45 <Eddi|zuHause> and the prussians used to have a silly system
22:00:03 <Eddi|zuHause> where 1 noble counted the same as 100 citizens, or 10000 peasants
22:00:13 <Rubidium> in the US you can, in theory, win the presidential elections with only few percent of the votes being cast for you
22:00:36 <__ln__> ah well, it's still compiling and i'm falling asleep, so i guess i'll report the results of the CFLAGS in the morning.
22:01:01 <GT> Hi guys, being forced to newgrf coding by recent development in the 32bpp world I wondered: is one free to use a grfid, or is there some kind of registration needed?
22:01:33 <Eddi|zuHause> Rubidium: in the European parliament elections, one maltesan or luxembourgian vote counts much more than a french or german vote
22:02:26 <Eddi|zuHause> because the number of seats for each country is not decided linearly based on population
22:02:52 <TinoDidriksen> And us smaller countries are happy with that.
22:03:13 <Rubidium> in the US you could be president with 12 votes
22:03:17 <michi_cc> GT: There is no global registry, but bananas will refuse duplicate IDs.
22:03:21 <Eddi|zuHause> so if we ever got a directly elected european president, or even a democratically elected EU government, then similar oddities may apply
22:03:48 <GT> Is there a list to prevent that?
22:03:59 <Rubidium> Eddi|zuHause: that's because they have to have at least one seat?
22:04:10 <TinoDidriksen> The US EC don't have to follow the popular vote; they're free to just vote however they like. Hasn't been many cases of them doing so, though.
22:06:33 <michi_cc> There are 2^32 possible IDs, so just pick something, it will likely be free :)
22:06:41 *** TGYoshi_ has joined #openttd
22:07:30 <Eddi|zuHause> Rubidium: Germany: 96 Seats (one seat per 859.000 inhabitants). Malta: 6 Seats (one seat per 67.000 inhabitants)
22:07:36 <GT> @michi: well, something starting with 32.. might get popular.
22:08:15 <Eddi|zuHause> GT: the usually promoted way is to set the first two characters as your initials. that is reasonably unique
22:08:40 <GT> Ok, Eddi, didn't know that.
22:08:43 <Eddi|zuHause> but that is no universal rule
22:09:43 <GT> for 32bpp-extra I used something like 32E. , I may try to continue that for other grfs
22:11:56 <TinoDidriksen> Use a 32bit rand() result.
22:11:56 <GT> @Michi, BTW, do you agree with my last response on the litght setup issue, that the sun is effectively after 3 o'clock?
22:13:34 *** TrueBrain has joined #openttd
22:14:19 <GT> Hi, TB. Great job on NoGo.
22:14:45 <TrueBrain> lol; I did that months ago! :P
22:14:57 <GT> Not implemented anything yet, but ideas are spinning in my head.
22:15:22 <GT> So what, even months ago, it stays a great job
22:15:27 <TrueBrain> and: BRCK is mine, GRFID wise
22:15:34 <TrueBrain> I dont care what the rest does, but thatone is mine :P
22:15:50 <GT> Huh, et tu Brute, using newgrf?
22:16:18 <TrueBrain> one of the prices to pay for 1 unified system :)
22:17:15 <Eddi|zuHause> this is also funny: "The council of europe (not to confuse with the European council, nor with the EU council)"
22:18:16 <TrueBrain> (and the square block exception is there, as you get that for free when launching Blender :P)
22:18:46 <TrueBrain> I dont understand how Blender works, it is like this magic machine or what-ever :P
22:18:53 <TrueBrain> but, its lighting system is fucking amasing
22:23:12 <TrueBrain> well, enough fiddling for one day; good night all
22:23:17 <TrueBrain> __ln__: please do let me know the results of your tests :)
22:23:22 <TrueBrain> and tnx for the effort etc :)
22:28:52 <GT> Anyway, I kind of get fed up with all the discussions on 32bpp ground tiles, what would be the best way to finally get some newgrfs?
22:36:56 <jnx> Is the cargo distance the distance between stations, or between the industries adjoining the stations?
22:38:23 <GT> I made a python script to convert old tars to nml, I could just start converting, and publish.
22:39:29 <GT> Not usable for sprites with masks yet, but there are hundreds of sprites available without.
22:40:10 <Zuu> jnx: the distance between the station signs
22:40:26 <Zuu> also, it's the manhattan distance
22:40:57 <GT> i.e. the tiles in x + the tiles in y
22:41:48 <GT> Zuu, what about writing junctioneer in HTML5?
22:42:23 <GT> using a canvas could be pretty platform independent
22:44:34 <Zuu> Apart from speed issues in simulation mode, I would not like to spend another half year with more porting after having made a port from SDL to wxWidgets. :-)
22:45:52 <Zuu> The wxWidgets version actually was easier that expected to get to run in Linux. The OpenGL initialization had to be done a bit different but thats pretty much it.
22:46:27 <GT> Can imagine that. I did download the pre-compiled version. Pretty nice, but I did miss road crossings on different levels.
22:47:00 <Zuu> You can do that. Edit the save XML file in a text editor and change the Z values to create level crossings.
22:47:22 <Zuu> Make sure that roads that should go over/under are at least 1,0 apart in z-axis.
22:47:41 <Zuu> There is just no GUI yet to support it :-)
22:47:45 <GT> Really? great. I will try that.
22:47:53 <GT> Real men use the command line
22:48:55 <GT> What about it? You edit it with a text editor
22:49:23 <Zuu> I don't really have anything against it. I have't learned yet what it can do.
22:51:19 <GT> moving square blocks on a canvas would be pretty simple. I just started some experiment, and had some nice graphs pretty soon. (Take into account that I do know a couple of programming languages, for the nono-Windows user it may be a bit harder, but I know you are not one of those)
22:55:15 <Zuu> Looks nice, although one might be less happy when you got 10k+ lines of javascript. But then you'll probably try to only have the render-engine in javascript and keep the simulation engine in C++. Not impossible to do, but sounds like quite some work :-)
22:59:41 *** Hazzard has joined #openttd
23:03:24 <Zuu> GT: By editing the xml file you can also go out of the grid with eg x=5.7 and y=20.52 if you wish to do so. :-) In 0.1.5 there was some interesting crashes you could trigger if you edit the file yourself and make mistakes, but 0.2 should be more resistent against that.
23:04:09 <GT> Right, time for me to sleep too, CU
continue to next day ⏵
