IRC logs for #openttd on OFTC at 2013-09-26
⏴ go to previous day
00:44:27 *** roboboy has joined #openttd
01:01:04 *** Eddi|zuHause2 has joined #openttd
01:02:37 *** fjb is now known as Guest330
01:07:51 *** fjb is now known as Guest332
01:32:21 *** fjb is now known as Guest334
01:52:05 *** Extrems has joined #openttd
02:12:56 *** Eddi|zuHause2 is now known as Eddi|zuHause
03:19:52 *** robotboy has joined #openttd
04:56:17 *** Eddi|zuHause has joined #openttd
05:38:32 *** Pensacola has joined #openttd
05:39:02 *** roadt__ has joined #openttd
05:47:53 *** Pereba_ has joined #openttd
05:53:55 *** Pereba_ is now known as Pereba
06:36:07 *** zeknurn has joined #openttd
07:26:52 *** Devroush has joined #openttd
08:08:45 *** Longtomjr-phone has joined #openttd
08:18:57 *** tdammers has joined #openttd
08:19:10 *** LordAro has joined #openttd
08:21:03 <tdammers> question about FIRS: I have a port, and I deliver cargo to it like crazy, but it doesn't seem to ever produce more than 144 crates of engineering supplies
08:21:36 <tdammers> have I hit a limit there, or is there a way to increase production more?
08:23:11 <tdammers> replaces the industry chain
08:24:22 <matkum> interesting, interesting. will try that. thanks
08:24:54 <tdammers> yeah... vanilla gets old at some point
08:25:06 <tdammers> makes managing trains a lot more complex
08:26:26 <planetmaker> tdammers, it might well be its maximum, yes
08:26:35 <planetmaker> what does its description say?
08:27:23 *** Longtomjr-phone has quit IRC
08:27:34 <planetmaker> mind to check the docs of the FIRS version which you use. It changes....
08:28:04 <tdammers> yeah, I looked up the documentation, but I can't find anything about limits there
08:31:41 <planetmaker> tdammers, do you deliver all three required cargo types within 30 days?
08:31:55 <planetmaker> within the same month?
08:32:08 <tdammers> the port is in Gung Ho mode all the time
08:32:17 <planetmaker> gung ho is max production
08:32:34 <tdammers> so it never increases further?
08:32:44 <planetmaker> but I would believe it adjusts output to input... hm
08:33:12 <V453000> if gung ho, then gung ho
08:33:17 <V453000> input doesnt matter in such case
08:33:23 <tdammers> I just built a second port next to the first one
08:33:24 <V453000> I like it, means you need to split traffic to multiple ports
08:34:14 <tdammers> well, by now I have reached the point where I can just cough up the cash for building a port wherever I please without making a dent in my budget, so heh
08:34:17 <planetmaker> he... I'd solve it such that I make a terminal station. which then distributes by means of very short-route vehicles to different ports :-)
08:35:02 <V453000> sub optimal if ports are all over the map far away :P
08:35:39 <planetmaker> yes. For ports next to eachother like suggested
08:35:59 <tdammers> I built them both in the catchment areas of the relevant stations
08:36:15 <tdammers> that might not have been such a good idea
08:36:26 <tdammers> because now the old port still sucks up most of the cargo
08:36:44 <planetmaker> oh, not a bad actually. transfer delivered cargo to that station. and make short delivery routes with vehicles to the individual ports.
08:36:50 <planetmaker> but pickup can remain centralized thus
08:37:44 <tdammers> yeah, I think I'll do that
08:38:11 <tdammers> I have HEQS loaded, too, so I can actually haul quite a bit with road vehicles
08:38:40 <tdammers> (the later trucks are actually *faster* for livestock than the best NARS trains...)
08:39:13 <planetmaker> for deliveries over 5 tiles or so a train is totally the wrong vehicle category anyway
08:39:37 <planetmaker> just a loop with trucks in drive-through road stops can't be beaten in those cases :-)
08:39:51 <planetmaker> or two rows of opposing dead-end truck stops
08:40:05 <planetmaker> connected by + -type roads among all
08:49:24 <tdammers> I like to make two loops with drive-through truck stops inside them, and then connect the main road such that the trucks use them in both directions
08:49:29 <tdammers> practically doubles their capacity
08:49:44 <tdammers> four truck stop tiles can then accommodate sixteen trucks
08:50:39 <planetmaker> most bad-ass is to make a loop with truck stops which belong to alternating stations ;-)
08:50:49 <planetmaker> one delivery the other pickup
08:50:54 <planetmaker> it's a real money press
08:51:03 <planetmaker> if used with big station spread ;-)
08:51:12 <planetmaker> (yes, it's cheating)
08:52:02 <tdammers> if the station really has to be huge because it handles massive amounts of cargo, then I wouldn't consider large spread "cheating"
08:52:11 <tdammers> but if it's abused for walking, well...
08:52:34 <tdammers> then you can just cover an entire city with just four road station tiles :x
09:06:16 <tdammers> hah, both port are in gung ho mode now
09:09:54 *** oskari89 has joined #openttd
09:14:12 *** montalvo has joined #openttd
09:14:20 *** TheMask96 has joined #openttd
09:20:03 *** TomyLobo has joined #openttd
09:20:37 *** tokai|noir has joined #openttd
09:20:37 *** ChanServ sets mode: +v tokai|noir
10:11:06 *** sla_ro|master has joined #openttd
10:28:06 *** TheMask96 has joined #openttd
11:25:25 *** TheMask96 has joined #openttd
11:27:17 <oskari89> When did the OpenTTD developement start?
11:32:01 *** Supercheese has joined #openttd
11:37:59 <Xaroth|Work> in the summer of '69
11:54:15 <blathijs> oskari89: The first version in the changelog (0.1.1) is from 2004, so before that at least
11:54:48 <LordAro> no one knows when ludde actually started, do they?
12:01:32 *** Stimrol has joined #openttd
12:01:35 <AndreasB> oskari89: Why do finland put letters on the end of names to make them "finnish" ?
12:01:50 <AndreasB> I would assume "Oskari" is the same name as Oskar in norway?
12:02:21 <AndreasB> Finland has got very weird names also, like Aama and Aapo
12:13:10 <i> damn that kind of fucks up highlights XD
13:17:32 *** TheMask96 has joined #openttd
13:21:46 *** DanMacK has joined #openttd
13:33:52 *** oskari892 has joined #openttd
14:13:19 <LordAro> perhaps a dev can answer this: now that limits is part of the spec/stdlib, is the massive amount of defines in stdafx.h (l29-onwards) really needed? even for those 'strange' OSs/compilers? or is OTTD not properly c++11 yet?
14:14:38 <blathijs> LordAro: I don't think any of the strange OSs and compilers support C++11 yet
14:14:42 <TinoDidriksen> <stdint.h> you mean? That will work in all compilers from 2010 onwards, including VC++ 2010. It's not in 2008, which I think OTTD still supports.
14:14:49 <blathijs> I don't even think GCC supports it completely yet
14:15:13 <TinoDidriksen> GCC 4.8 has full C++11 language support. 4.9 will have full library.
14:15:20 <TinoDidriksen> Clang has full both.
14:15:37 <blathijs> TinoDidriksen: Cool.
14:15:38 <TinoDidriksen> But just stdint.h you can use everywhere.
14:15:57 <planetmaker> removing support for old(er) gccs is not exactly desirable unless it makes things way more complicated
14:16:00 <TinoDidriksen> (note stdint.h, not cstdint)
14:17:19 <LordAro> even debian stable has gcc 4.7
14:17:32 <LordAro> (well, 4.6 for powerpc and similar)
14:17:38 <Japa> the worst thing is doing own implementation of stdint
14:19:06 <TinoDidriksen> Since stdint.h is C99, it's in all GCC versions. stdafx.h claims MorphOS doesn't have stdint.h, which sounds bizarre...isn't that GCC?
14:19:52 <blathijs> Wasn't morphos that OS that stuck to GCC 2.95?
14:20:22 <TinoDidriksen> Breaking support for that should be considered an act of mercy.
14:20:38 <blathijs> There was one a few years back that only did 2.95, but that might have been long fixed
14:21:17 <TinoDidriksen> It also says SunOS/Solaris doesn't have stdint.h, but Solaris 10 and onwards does.
14:22:53 <TinoDidriksen> Seems the stdafx.h header is still mitigating for problems from 2006 that just aren't here today, with exception of VC++ 2008 - so, is 2008 support important?
14:27:36 <Japa> MSVC 2008 doesn't have stdin.h, as far as I know, but 2010 does.
14:27:48 <TinoDidriksen> 's what I said...
14:28:23 <LordAro> which is part of c++98, as far as i can tell...
14:28:41 <LordAro> don't even need [c]stdint[.h]
14:28:57 <TinoDidriksen> LordAro, <limits> does not define the stdint.h types or #defines, which is what stdafx.h emulates.
14:29:37 <LordAro> but why are the defines needed? what's wrong with just "static const UINT32_MAX = std::numeric_limits<uint32>::max();" ?
14:29:42 <LordAro> ottd is c++, after all
14:29:44 <TinoDidriksen> Neither does <climits>. Standard C++98/03 has no idea what uint64_t or UINT64_MAX is.
14:29:53 *** montalvo has joined #openttd
14:30:10 <LordAro> it knows what a long int is
14:30:22 <TinoDidriksen> But not long long
14:30:54 <TinoDidriksen> long may be 32 or 64 bits, so you can't just ask numeric_limits<long>...
14:31:17 <TinoDidriksen> You need stdint.h to get it right, or replicate everything stdint.h is doing for you.
14:31:31 <LordAro> fine, i'll concede that point
14:31:44 <LordAro> but you might as well do it the 'proper' way for the others
14:32:06 <TinoDidriksen> Which is what? int may be 16 or 32 or 64 bits.
14:32:52 <LordAro> how often is it not 32 bits though? is it really worth supporting those?
14:34:50 <TinoDidriksen> I would also remove those in favour of stdint.h giving you concrete fixed widths, with no guessing or assertions needed.
14:36:53 <LordAro> maybe, but i've never seen a bug report or anyone complaining about their compiler/os giving them a compile error at that point
14:37:25 <TinoDidriksen> Agreed, current platforms stick to those, even if that violates the Standard.
14:38:06 <TinoDidriksen> Strictly speaking, int must be the native CPU word size, which on 64bit CPUs clearly can't be 32bit - yet implementations keep it 32bit.
14:38:24 <planetmaker> LordAro, debian stable is also a newly released distro
14:38:44 <planetmaker> you can't generally assume that everyone has that. At least old stable should be supported, usually :-)
14:38:51 <LordAro> i thought things seemed relatively up-to-date :)
14:39:47 <LordAro> ok, gcc 4.4 in oldstable
14:40:36 *** sla_ro|master has joined #openttd
14:41:19 <TinoDidriksen> The only real issue against stdint.h is VC++ 2008 - but is that a concern, when 2010, 2012, and 2013 Express are free?
14:41:49 <Japa> TinoDidriksen, winXP support?
14:42:09 <planetmaker> how does the icc treat those things? what about mingw compiler in its packages? and yes, win 9x/XP support :-)
14:42:10 <LordAro> 2010 supports xp, iirc
14:42:12 <TinoDidriksen> 2010 deploys to XP.
14:42:24 <TinoDidriksen> ICC and MinGW has stdint.h
14:42:32 <TinoDidriksen> It's C99, so they have it.
14:42:38 <planetmaker> and... osx compiler uses 4.0 or 4.2
14:42:39 <Japa> There's really no reason to support it then
14:42:56 <planetmaker> TinoDidriksen, how do you know which one our CF uses? :-)
14:43:56 <TinoDidriksen> Apple has been on GCC 4.2 for many years, since it was the last GPLv2 version. I can't imagine you'd still be on 4.0, or is that a 10.4 machine?
14:44:28 <TinoDidriksen> Either way, they both have C99 headers...
14:45:50 <planetmaker> err... no... even 10.6 ships with 4.0 and 4.2
14:46:18 <planetmaker> and the CF might as well be a cross-compiler based on 4.0
14:46:28 <planetmaker> (yes, we cross-compile the osx on a linux machine :D )
14:47:16 <TinoDidriksen> I've done a fair bit of homework on stdint.h support, as might be evident. I've enforced its use for some scientific software - old deployed OSs on various outdated hardware platforms? Universities got lots of those. stdint.h works everywhere.
14:49:46 <planetmaker> yeah, CF uses likely a custom-built gcc 4.0 compiler
14:49:57 <planetmaker> it's darwin-9. Which usually is 4.0
14:51:51 <LordAro> you do realise 4.0 is almost as old as ottd? :L
14:54:01 <LordAro> and don't most osx developers use clang now, anyway?
14:55:50 <planetmaker> Will you build us a toolchain which cross-compiles with, say, SDK 10.7 or 10.8?
14:56:18 <planetmaker> As that's the real difficulty in this endeavour
14:56:44 <LordAro> i'm sure you could find someone willing to compile natively
14:56:54 <LordAro> (yes, i realise that's not a brilliant answer)
14:56:55 <TinoDidriksen> No, but I got a Mac mini standing around doing my building...got an easy way to nightly build?
14:59:03 <planetmaker> LordAro, no, it need be accessible by the CF to run builds on demand, triggered by the CF
14:59:56 <TinoDidriksen> How is that triggered? SSH?
15:07:46 <planetmaker> but yes, we would need ssh to that server in any case
15:13:54 <LordAro> anyway, i guess 'sorry' for my lack of thinking towards backward compatibility - i guess a combination of arch linux and watching http://channel9.msdn.com/Events/GoingNative/ over the last few days influenced me a bit and filled my head with lots of exciting things :)
15:14:17 *** TomyLobo has joined #openttd
15:15:28 <TinoDidriksen> I'll be so happy in mid-2015 when I can start using C++11 in production code...
15:15:31 <planetmaker> thinking compatibility sometimes sucks :-) But... it ensures also a happy userbase :-)
15:16:04 <planetmaker> what will have changed then, TinoDidriksen ?
15:16:23 <TinoDidriksen> Ubuntu 10.04 LTS will drop out of support.
15:16:58 <planetmaker> if you use your own toolchain with possibly updated compiles...
15:18:01 <TinoDidriksen> Unfortunately, I can't push that on the aforementioned universities with ancient setups.
15:18:53 <LordAro> make something awesome which only works on newer setups? :P
15:19:53 <planetmaker> if gcc < 4.8; then print "update, you dork"; else print "have fun!"; fi
15:21:15 <LordAro> me and Alberth have not yet had to worry about backward compatibility with freerct - we both have gcc 4.8.1 and our userbase consists of about 100 people :L
15:22:07 <LordAro> + freerct is only really just out of the 'experimental' phase
15:25:57 <planetmaker> LordAro, that's something different really. New projects have no need for backward compatibility
15:27:02 <LordAro> i know, we're just now probably going to go for full c++11 usage
15:27:13 <LordAro> although you 2 have rather shaken my opinion on that :L
15:27:55 <LordAro> (we already use static_assert, but are about to try [new to both of us :L] using shared_ptr)
15:28:43 <TinoDidriksen> As a new project, I'd say go for it. Platforms will catch up as you code along. VS 2013 has very usable C++11 support.
15:29:01 <TinoDidriksen> By the time you're ready for a wider audience, it shouldn't be a problem.
15:29:13 <LordAro> ha, windows. currently, i don't think even mingw can compile it :p
15:29:23 <LordAro> you're welcome to provide patches though :D
15:29:43 <LordAro> (include/library issues, iirc)
15:30:14 <LordAro> i do keep meaning to re-write the compile system
15:30:32 <TinoDidriksen> ...the compile system? You're not using something like CMake?
15:31:08 <LordAro> no, no, but the configure script is fairly new (i committed 2 weeks ago :L) and the paths are all still hardcoded
15:31:27 <oskari892> Can GameScripts used on dedicated servers?
15:31:38 <TinoDidriksen> I suggest looking at CMake then. It works really well cross-platform.
15:31:39 <oskari892> Or multiplay at all?
15:31:58 <planetmaker> of course, oskari892
15:32:16 <LordAro> TinoDidriksen: careful mentioning CMake around here, it's not well liked ;)
15:32:35 <LordAro> + writing your own is much more 'fun' :3
15:32:41 <planetmaker> creates the same amount of issues as it solves ;-)
15:33:13 <TinoDidriksen> I use CMake for all my projects. Works wonderfully on all platforms and toolchains.
15:34:08 <LordAro> the only experience i have with cmake is in corsix-th, where it broke with my lua5.2 (wanted 5.1)
15:40:22 <Eddi|zuHause> i don't understand cmake, i seem to be missing some basic philosophy behind it
15:41:16 *** DanMacK has joined #openttd
15:44:06 *** abchirk_ has joined #openttd
15:51:53 *** Progman has joined #openttd
15:56:56 *** Djohaal has joined #openttd
16:00:10 *** Prof_Frink has joined #openttd
16:06:07 *** TheMask96 has joined #openttd
16:22:09 *** ntoskrnl has joined #openttd
16:45:27 *** Extrems has joined #openttd
16:45:27 *** mindlesstux has joined #openttd
16:45:27 *** planetmaker has joined #openttd
16:45:27 *** Xaroth|Work has joined #openttd
16:45:27 *** tparker has joined #openttd
16:45:27 *** Twofish has joined #openttd
16:45:27 *** Defaultti has joined #openttd
16:45:27 *** Nothing4You has joined #openttd
16:45:27 *** weber.oftc.net sets mode: +o planetmaker
16:45:42 *** ChanServ sets mode: +v planetmaker
17:16:05 *** DarkAceZ has joined #openttd
17:45:22 <DorpsGek> Commit by translators :: r25799 trunk/src/lang/ukrainian.txt (2013-09-26 17:45:15 UTC)
17:45:23 <DorpsGek> -Update from WebTranslator v3.0:
17:45:24 <DorpsGek> ukrainian - 59 changes by odisseus
17:54:29 *** frosch123 has joined #openttd
17:56:43 *** valhallasw has joined #openttd
18:17:26 *** oskari892 is now known as oskari89
18:29:14 *** andythenorth has joined #openttd
18:36:30 <DorpsGek> andythenorth: danmack was last seen in #openttd 2 hours, 55 minutes, and 3 seconds ago: <DanMacK> Hey all
18:47:39 *** Progman has joined #openttd
19:08:53 *** Speedy` has joined #openttd
19:09:09 *** Speedy` is now known as Speedy
19:24:38 *** jjavaholic has joined #openttd
19:31:37 *** DanMacK has joined #openttd
19:51:20 <LordAro> hmm. who here understands the rvalues and lvalues of c++11?
19:52:19 <LordAro> something to do with "move semantics" whatever they are..
19:54:39 *** Longtomjr-phone has joined #openttd
19:58:15 *** Longtomjr-phone has quit IRC
20:03:17 <frosch123> rvalues and lvalues are no c++11 thing :p they are normal stuff
20:03:49 <frosch123> an Lvalue is something which can be on the Left side of an assignment operator, and Rvalue can be on the Right side
20:04:17 <frosch123> Lvalue is more, because you can assign it. Rvalues can only be read
20:04:36 <Prof_Frink> No, R-values are a measure of thermal insulation.
20:05:01 <TinoDidriksen> LordAro, Freenode's ##C++ (strict) and ##C++-general (friendly) are good places to ask about all of that. Yes, two #s
20:05:04 <frosch123> lvalues are non-const references
20:05:16 <frosch123> rvalues are const values and temporary variables
20:07:10 <LordAro> TinoDidriksen: freenode is scary :)
20:07:43 <TinoDidriksen> Some channels are, hence why I listed ##C++-general which is the friendly version of ##C++
20:08:21 <LordAro> i usually get good advice between here and stackoverflow though
20:08:43 <TinoDidriksen> But really, ##C++ knows a lot...I've learnt so much from that channel in the past ~2 years.
20:08:54 <LordAro> (despite the people here being a bit C-orientated :p)
20:27:39 <frosch123> i thought people here are python oriented
20:30:07 <Xaroth|Work> fewer than you thought, apparently
20:31:58 <Taede> theres probably both, and some others to boot
20:34:57 <jjavaholic> I thought there would be more users here
21:28:55 *** Devroush367 has joined #openttd
21:29:42 *** andythenorth has left #openttd
21:34:46 *** Devroush has joined #openttd
21:35:34 <Eddi|zuHause> jjavaholic: we more often get "boah i didn't think there would be so many" than "boah i thought there were more" though :p
21:45:36 *** jjavaholic has joined #openttd
22:15:09 *** Devroush367 has joined #openttd
22:29:22 *** Devroush has joined #openttd
23:02:06 *** jjavaholic has joined #openttd
23:03:19 *** TomyLobo2 has joined #openttd
23:06:14 *** kais58___ has joined #openttd
23:09:01 *** TomyLobo2 is now known as TomyLobo
continue to next day ⏵