IRC logs for #openttd on OFTC at 2008-09-03
⏴ go to previous day
00:01:53 *** Marine_ has joined #openttd
00:19:01 *** welshdragon has joined #openttd
00:33:10 *** Eddi|zuHause has joined #openttd
00:48:16 *** welshdragon is now known as welshdra-gone
01:46:21 <Belugas> ignoring = CargoChangeInfo(index, numinfo, prop, &buf, bufend - buf);
01:46:31 <Belugas> should not it be someting like
01:47:09 <Belugas> ignoring = CargoChangeInfo(index, numinfo, prop, &buf, bufend - buf) != CIR_SUCCESS; ?
01:53:56 *** De_Ghosty has joined #openttd
02:07:18 <glx> Belugas: [17:05:51] <@peter1138> glx, I think more changes are necessary there. That fix the warning, and work the same way it used to, but not really work properly.
02:10:07 <Belugas> not a problem, it's just a warning anyway :)
02:14:01 *** Frostregen_ has joined #openttd
02:19:41 *** Frostregen_ is now known as Frostregen
02:20:30 <Tekky> did the original Transport Tycoon Deluxe feature inflation or was that a feature added later by TTDPatch/OpenTTD?
02:20:50 *** roboboy has joined #openttd
02:47:33 *** Jerimiah40 has joined #openttd
02:51:39 <DaleStan> Tekky: Inflation is a TTD original.
02:59:52 *** Jerimiah40 has joined #openttd
03:07:09 *** elmex_ is now known as elmex
04:07:38 *** De_Ghosty has joined #openttd
05:06:35 <roboboy> When I compile openTTD with vsc05 it never seems to get past determining version number
05:20:04 *** valhalla1w has joined #openttd
05:51:08 *** [alt]buster has joined #openttd
05:51:11 *** [alt]buster is now known as [com]buster
06:31:34 *** Jerimiah40 has joined #openttd
06:32:01 <roboboy> compiling in vsc05 never gets further than determining version and then the status returns to ready from build started
07:14:26 *** Phoenix_the_II has quit IRC
07:16:24 *** Wezz6400 has joined #openttd
07:17:51 *** Doorslammer has joined #openttd
07:23:12 <planetmaker> cargodest test day :)
07:36:54 <Rubidium> should we daemonize the fork?
07:37:24 <TrueBrain> that at least puts him in the background
07:39:43 *** Vikthor has joined #openttd
07:42:06 *** peter1138 has joined #openttd
07:42:06 *** ChanServ sets mode: +o peter1138
07:49:18 *** Yeggzzz is now known as Yeggs-work
07:51:14 *** Celestar has joined #openttd
07:51:20 <CIA-1> OpenTTD: rubidium * r14233 /trunk/src/ (7 files in 2 dirs):
07:51:20 <CIA-1> OpenTTD: -Feature/Fix [FS#2172]: save the palette of the loaded NewGRFs in the savegame,
07:51:20 <CIA-1> OpenTTD: so joining with a server using Windows palette will make a client with the DOS
07:51:20 <CIA-1> OpenTTD: palette do palette conversion and (thus) not cause a desync due to the different
07:51:20 <CIA-1> OpenTTD: palettes disabling different NewGRFs.
07:51:25 <Celestar> Eddi|zuHause: you around?
07:56:42 <Celestar> Eddi|zuHause: you had that problem with two trains: one going A-B-C-D-E and another going F-B-D-G right?
07:57:11 <Eddi|zuHause> yeah. but they were trams ;)
07:57:34 <Celestar> do you have such a savegame (preferably no other vehicles) and/or want to test a patch?
07:59:43 <Eddi|zuHause> hm, i still have that game, but it has lots of other vehicles
08:02:04 <Eddi|zuHause> you can show me the patch, though ;)
08:08:22 <Celestar> doing some preliminary test (=
08:09:09 <Celestar> I've already thought of a first route balancer (=
08:09:16 <Forked> I was late for work again today..
08:09:32 <Forked> expanding the service my company offers to a city called Kristiansand
08:09:40 <Eddi|zuHause> Forked: easy solution: quit work
08:10:00 <Forked> eddi: if only I was rich :p
08:12:32 <Celestar> REAL men make money without work
08:13:09 <Forked> hmm. I think I like Google Chrome.
08:14:33 <Celestar> when is the abort-condition of a for loop tested?
08:15:11 <Rubidium> for (; <this>; ) <- that one?
08:16:08 <Rubidium> for (x; y; z) a == { x; while (y) {a; z;}
08:16:23 <Eddi|zuHause> for (a;b;c) {d;} is an abbreviation for a; while (b) {d; c;}
08:20:56 <peter1138> Forked, it's okay. It's not really anything as special as they're making out ;)
08:21:04 <peter1138> And where's the source if it's open source...
08:21:35 <peter1138> And its 'back' behaviour is a little annoying.
08:22:01 <Celestar> building cargodest with --enable-debug=3 is unfun :P
08:23:08 <Celestar> dunno what's sooo slow about it
08:25:51 *** Sir_Bob has joined #openttd
08:27:45 <peter1138> Hmm, that has to be wrong :o
08:28:17 <Rubidium> shouldn't not optimising be slightly faster?
08:28:33 <Celestar> I mean not the building
08:28:48 <Celestar> fast forward with a SINGLE vehicle hardly gains any speed
08:29:12 * roboboy waits for openttd to compile
08:29:42 <Rubidium> Celestar: absolutely nothing is inlined so almost every "simple" function gets function call overhead/can't get optimised etc.
08:30:23 <Celestar> Rubidium: yeah, but I didn't notice it to that extend without cargodest
08:30:41 <Celestar> but then again, I'm making extensive use of std::map which is about 3 orders of magnitude slower without inlines
08:31:12 *** Jerimiah40 has joined #openttd
08:34:17 <Ammler> nice, joining with dos originals seems dummy safe now. :-)
08:41:57 * Celestar wonders how good g++ is at detecting loop invariants
08:42:20 <Ammler> Rubidium: now, I see only one little problem, which I do not care, if someone likes to load a save local with his DOS newgrfs...
08:42:23 <Celestar> especially in highly nested loop when the invariant is some member function with highly nested templates
08:43:26 <Rubidium> Ammler: what would be the problem?
08:43:38 <Rubidium> he can always switch the palette of the loaded newgrfs
08:44:41 <Ammler> well, OTTD does tell him, that the md5 is different and then he should realize to toggle...
08:44:57 <Ammler> the GRF is orange then?
08:47:33 <Ammler> anyway, many thanks, that will solve many confusion on our server, I hope :-)
08:49:03 <Eddi|zuHause> Celestar: i think templates are resolved first, and then the standard g++ magic starts
08:50:08 <roboboy> I compiled a debug version of the trunk and it tells me that I have invalid language packs even if I coppy it to a working OTTD install
08:50:32 <roboboy> I did not get any language files when I compiled
08:50:33 <Celestar> Eddi|zuHause: yet I'm not sure whether to trust the optimizer fully (=
08:56:27 <Eddi|zuHause> roboboy: MSVC? make sure you compile strgen first
09:01:18 <roboboy> im recompiling but this time cargodest instead of trunk
09:05:47 *** elmex_ is now known as elmex
09:09:25 *** reldred|gone is now known as Reldred
09:09:56 *** Vikthor has joined #openttd
09:33:38 *** DorpsGek has joined #openttd
09:33:38 *** ChanServ sets mode: +o DorpsGek
09:35:48 * Celestar just learnt that valgrind will not find out-of-bound access for arrays
09:36:51 <Eddi|zuHause> i have a little problem properly following my passengers, as trams don't have a detailed view :(
09:36:56 *** TrueBrain has joined #openttd
09:36:59 *** |Bastiaan| has joined #openttd
09:37:15 <Kloopy> Are you using the compile fark for the cargodest test, then?:)
09:37:25 <Celestar> Eddi|zuHause: yeah :(
09:37:31 <Celestar> Eddi|zuHause: use trains? ;)
09:37:47 <Celestar> hm Rubidium where do I find the options that are set when using --enable-debug=?
09:37:48 <Eddi|zuHause> kind of problematic through the city :p
09:44:04 * roboboy leaves compiling cargodest alone as he is getting erors
09:46:21 <roboboy> I shall login here from my server that im compiling on
09:48:14 *** robotboy has joined #openttd
09:50:32 <Eddi|zuHause> hm... is there a way to hide all station signs that start with a certain prefix? (like "WP" for stations that are only used as waypoints)
09:50:43 <robotboy> c:\openttd development\cargodest\src\src\routing_classes.h(32) : fatal error C1083: Cannot open include file: 'boost/graph/adjacency_list.hpp': No such file or directory is one of the erors
09:50:59 <Eddi|zuHause> you need to install boost
09:51:14 <Eddi|zuHause> as described on the wiki
09:51:26 <robotboy> BSCMAKE: error BK1506 : cannot open file '..\objs\win32\release\routing.sbr': No such file or directory being the other
09:51:59 * peter1138 updates a patch from 13388 ... mmm, conflicts ...
09:52:02 *** stillunknown has joined #openttd
09:52:16 <Celestar> robotboy: did you install boost at all?
09:53:08 <robotboy> nope. where does it say to install it? On the cargodest page? if so then I only read it quickly
09:54:29 <Celestar> Cannot open include file: 'boost/graph/adjacency_list.hpp' <= apparently you need to install it somewhere in the default include path
09:56:17 <robotboy> do I just download and install the entire package?
09:56:56 <Celestar> you got space constrains? (=
09:59:46 <robotboy> its great for logging into irc from locations that dnt need/have a client or java installed
10:04:41 *** roboboy has joined #openttd
10:06:08 * roboboy waits for boost to download
10:07:05 <roboboy> its beeing verry slow
10:09:20 <roboboy> I may want to get it elsewhere if its not finished when I get back from dinner and watching TV
10:14:25 <Eddi|zuHause> dbg: [routing] [Passagiere] :INFO: Vehicle arrived at <109> without order <- how can i find out where that is?
10:15:10 <Progman> where? 109 is the station
10:15:32 <Progman> but the question is, which is the vehicle ;)
10:15:37 <Eddi|zuHause> i want to know which station that is
10:15:46 <Eddi|zuHause> as in get the name, or scroll to
10:16:15 <Progman> afaik there is a command to get the name of the ID
10:19:44 <Eddi|zuHause> hm... something is wrong with the start time display
10:21:58 <Celestar> Eddi|zuHause: I did have the station name there, but it was so horribly slow
10:22:34 <Eddi|zuHause> yes, it's not needed there, it's sufficient to run another command to find out the name
10:22:47 <Eddi|zuHause> i just wanted to know which command ;)
10:29:56 *** DorpsGek has joined #openttd
10:29:56 *** ChanServ sets mode: +o DorpsGek
10:30:08 *** TrueBrai- has joined #openttd
10:30:28 <CIA-1> OpenTTD: smatz * r14234 /trunk/src/cargopacket.cpp: -Fix: feeder share was computed wrong when splitting cargo packet
10:30:47 *** TrueBrai- is now known as TrueBrain
10:30:58 <TrueBrain> that should be the last of it :)
10:31:29 <TrueBrain> unexpected reboots :p
10:31:33 <TrueBrain> well .. 'unexpected'
10:33:35 <TrueBrain> now we can finally boot 64bit guests
10:35:16 <Eddi|zuHause> Celestar: unfortunately not
10:36:26 *** lobster_MB has joined #openttd
10:40:52 <DorpsGek> dih: Error: guest64 is not in #openttd.
10:40:54 * roboboy thinks somehing is going on with SF downloads or his connection at the moment
10:41:22 <TrueBrain> roboboy: most likely: SF :p
10:41:31 <TrueBrain> what are you trying to download? :)
10:41:50 <roboboy> boost to compile cargodest
10:42:00 <TrueBrain> dih: that is a big download ;)
10:42:48 <Eddi|zuHause> well, it's only 2 DVDs when you strip the porn :p
10:43:04 <TrueBrain> really, Intel BIOS sucks .. they disable Virtualizing by default .. and for what reason? :(
10:43:05 <peter1138> Grr, Windows just bluescreened :o
10:44:19 <Kloopy> Wasn't TrueBrain going to be really nice and use the compile farm for cargodests this afternoon? :P
10:44:24 <peter1138> SmatZ, that doesn't actually affect payments, does it?
10:44:29 <Kloopy> Meaning you'd not have to download boost?
10:44:50 <Celestar> peter1138: only feeder shares
10:45:33 <SmatZ> peter1138: as Celestar said :) the only problem is that the vehicle that loads these packets may show invalid negative income
10:45:45 <Ammler> Kloopy: he is just waiting for us to ask him to :-)
10:46:03 <Ammler> (he has nothing else to do) ;-)
10:46:21 <SmatZ> but the real money are still computed from distance(orig. src -> dst)
10:46:29 <Celestar> Kloopy: only if i've 'fixed' the source for the test
10:46:35 <Celestar> and I want some changes in first
10:46:39 <Kloopy> What's broken with it, Celestar? :)
10:46:45 <Kloopy> You're the man! You can fix it! :D
10:47:08 <roboboy> but I want to compile myself anyway
10:47:25 <Kloopy> Fair enough, roboboy. :)
10:48:27 <Eddi|zuHause> OVERDRIVE: 2 <- i think that's too little information to really know if it's doing what i want it to do
10:48:34 <Celestar> Kloopy: payment stuff is one thing (in trunk now)
10:49:00 <Celestar> Eddi|zuHause: enable debug routing level 1, then you get a BIT more information. You'll get even more in about 30 minutes, k?
10:49:15 <Ammler> how do I pull a hg repo in a svn repo?
10:49:21 <Eddi|zuHause> i enabled that, but it doesn't say anything
10:49:30 <Ammler> (if it is possible...)
10:49:54 <Eddi|zuHause> except the other message about the wrong station
10:50:11 <Celestar> Eddi|zuHause: DONT use intermediate stations without orders, just don't use them
10:50:23 <Eddi|zuHause> yes, it was a mistake
10:52:00 <peter1138> Celestar, 'ghost' orders are for after merge?
10:52:50 <Eddi|zuHause> merge should have priority now, i think
10:54:44 <Celestar> peter1138: I think so
10:54:55 <Celestar> peter1138: I think the GUI stuff should have priority now
10:55:38 <Kloopy> Celestar: In Multiplayer, are cargodest patch settings for EVERYONE or can you have some companies with cargodests enabled and some without?
10:55:45 <peter1138> They're for everyone.
10:57:59 <Celestar> they must be basically
10:58:52 *** Dred_furst has joined #openttd
11:02:09 * roboboy wonders where else to get boost or the bit he needs
11:02:15 *** Jezral is now known as TinoDidriksen
11:02:29 <Eddi|zuHause> when newgrf ports are extended to tram stations, there should be the ability to place certain road bits over them
11:02:47 <Celestar> roboboy: www.boost.org ?!
11:03:58 <roboboy> I said else as there downloads are from sourceforge and it is taking forever to get
11:04:17 <roboboy> unless I leave it overnight
11:05:13 <roboboy> that was quick compared to before
11:05:51 <Celestar> er use another mirror?
11:08:43 <Eddi|zuHause> apparently i have an "OVERDRIVE: 7" route ;)
11:09:34 <roboboy> do I just need graph?
11:10:06 <roboboy> or should I just unzip the entire boost folder?
11:10:06 <Celestar> roboboy: it it depends heavily on other stuff
11:10:43 <Rubidium> Celestar: what do you want to know exactly about the debuglevels?
11:11:50 *** Sir_Bob has joined #openttd
11:12:00 <Eddi|zuHause> dbg: [routing] [Passagiere] :Overdrive used at <63>
11:13:51 <Eddi|zuHause> is it possible that the function gets called a lot of times?
11:14:15 <Eddi|zuHause> i got kinda spammed with output, even though only 1 passenger is in the tram
11:14:16 <Celestar> Eddi|zuHause: every time yes.
11:14:31 <Celestar> Rubidium: er ... er ... what was my question again? :P
11:15:04 <Rubidium> 11:37 < Celestar> hm Rubidium where do I find the options that are set when using --enable-debug=?
11:15:10 <Eddi|zuHause> hm, no, it's the wrong tram
11:15:10 <Rubidium> 11:37 < Celestar> hm Rubidium where do I find the options that are set when using --enable-debug=?
11:15:12 <Celestar> Rubidium: no problem I found it (=
11:15:24 <Rubidium> argh... stupid putty :(
11:16:51 <Eddi|zuHause> the tram was going the other direction, coming from the main station, so it was probably rather full ;)
11:17:01 <Eddi|zuHause> so it appears to be working alright
11:18:22 <Forked> reminds me, I saw some strange behaviour earlier, but will look at it when I get home.
11:18:41 <Celestar> Program received signal SIGFPE, Arithmetic exception.
11:19:03 <Forked> A train going A-C-B never loaded a single passenger.. while trains going A-B-C got them all - even those to C
11:19:15 <Rubidium> sounds like a generic division by 0
11:19:18 <Forked> but that might not have been in the latest version :)
11:20:14 <roboboy> so I ned to include the boost files?
11:20:40 <roboboy> ie add them to the include folders list in vs05
11:22:24 <Eddi|zuHause> i wondered yesterday... why am i the only person who finds those electrified/unelectrified bugs?
11:23:33 <TrueBrain> Eddi|zuHause: because you are a geek? :p
11:23:55 <fjb> Eddi|zuHause: because everybody uses maglev. :-P
11:25:38 <Celestar> peter1138: I'd need some help with the GUI stuff
11:25:52 <Celestar> I'm waaaaay to stupid to do it correctly today
11:26:19 <TrueBrain> General Anouncement: Windows Sucks! (takes 422 seconds to compile a Windows binary, where all other targets take around 180 seconds)...
11:26:48 <Celestar> General Anouncement: Zero is needed Zero
11:27:30 <SmatZ> TrueBrain: remove -O3 :) (if it is still there ;)
11:31:42 * roboboy leaves cargodest to compile
11:33:15 <SmatZ> TrueBrain: can you set lower optimisation anyway?
11:33:58 <TrueBrain> dunno, I don't work with MSVC :p
11:34:07 <SmatZ> iirc, win binaries are much bigger than linux ones, seems it does much more inlining
11:34:30 <SmatZ> I fail to find anything at the nightly page now :-)
11:37:18 *** Brianetta has joined #openttd
11:39:22 *** Vikthor has joined #openttd
11:40:57 <robotboy> BSCMAKE: error BK1506 : cannot open file '..\objs\win32\release\routing.sbr': No such file or directory I still get this eror
11:42:31 <Celestar> roboboy: is that the FIRST error you get?
11:42:44 <robotboy> does boost need to be in the include variable or the lib variable?
11:43:00 <robotboy> no I still get the boost error
11:43:12 <Celestar> then you have boost in the wrong place
11:43:29 <Celestar> you need a file that is "boost/graph/adjacency_list.hpp"
11:43:38 <Celestar> the PATH is important!
11:43:41 <robotboy> does it go as an include or lib
11:45:10 <robotboy> so C:\Boost\Include would include all the boost files in the include variable
11:45:31 <Celestar> then you need C:\Boost\Include\boost\graph\adjacency_list.hpp
11:45:39 *** Jerimiah40 has joined #openttd
11:46:08 <Brianetta> Celestar: When I said, "The passengers get off the train at every single stop. There, they decide which vehicle to use next," it was really to drive home the point that every station is somewhere that changes happen. That the passenger has *no inclination whatsoever* to stay on the train if there's a "better" one. That they don't actually get off until the decision is made is only semantically different.
11:46:54 <Eddi|zuHause> i beg to differ :p
11:47:10 <Brianetta> Eddi: Only loading times make it different
11:47:16 <peter1138> But isn't there a penalty for transfering?
11:47:31 <robotboy> lets see if rebuilding fixes it otherwise I will be off to bed
11:47:34 <peter1138> Hmm, I thought there was :o
11:47:43 <Celestar> peter1138: there is a penatly for each stopover, no matter whether transferring or transitting
11:47:49 <Eddi|zuHause> peter1138: no, because it does not know in advance if the intermediate stop is really a transferring stop
11:47:58 <Celestar> peter1138: I'm working on coughing something up about this
11:47:58 <Brianetta> There probably should be. Passengers like to keep their seat, and it's time consuming and expensive to unload trains of cargo.
11:48:11 <Celestar> Brianetta: I'm already working on it
11:48:17 <Celestar> but it's something for version 2
11:48:22 *** Doorslammer has joined #openttd
11:48:27 <peter1138> Eddi|zuHause, right... and if it doesn't know about it all the way you could end up with bouncing cargo.
11:48:30 <Brianetta> One day we'll have cargo packets bound to cargo wagons
11:48:41 <Brianetta> and alter the consist in shunt yards
11:48:50 <Brianetta> so a given wagon will get where it's going
11:48:57 <Brianetta> but the "train" as an entity will be fluid
11:49:11 <Celestar> Brianetta: cargopackets ARE bound to cargo wagons
11:49:21 <Brianetta> I mean bound in the literal sense
11:49:49 <Brianetta> Until the mid-20th century, many cargo wagons were owned by the customer
11:49:58 <Brianetta> They just wanted their wagon to get somewhere
11:50:11 <Brianetta> It's spend several hours per night in marshalling yards
11:50:23 <Brianetta> but the cargo wouldn't be unloaded except by the customer
11:50:49 *** Reldred is now known as reldred
11:50:54 <peter1138> In that case you're transporting the wagons. The contents are irrelevant.
11:51:08 <peter1138> Sounds like quite a major gameplay change.
11:51:11 <Brianetta> I'm not saying our wagons shouldn't be our wagons; they should be "leased" to the customer at pick-up time
11:51:26 <Brianetta> So, we go fetch coal, as before
11:51:54 <Brianetta> but not all the wagons might be going the same way
11:52:24 <Brianetta> Trains could be split or merged
11:52:41 <Ammler> isn't it still that way with containers?
11:53:03 <Brianetta> but the trains consist isn't altered for those
11:53:07 <TrueBrain> Brianetta: a better addition to the game might be that you can 'buy' coal at a mine for a given price, and sell it to a powerstation at a given price, where those prices change depending on demand (so no longer it matters how long it travels)
11:53:24 <Brianetta> TrueBrain: That's how people perceive it at the moment
11:53:31 <Celestar> first thing we need a realistic braking
11:53:36 <Celestar> then we need realistic reversing
11:54:01 <Ammler> TrueBrain: then you would need storage stations :-)
11:54:03 <Brianetta> [12:48] <Brianetta> One day we'll have.....
11:54:23 <Celestar> peter1138: have you finished them? :P
11:54:34 <Eddi|zuHause> argh... i wondered why my trams are suddenly going late at the same time
11:54:38 <TrueBrain> but I guess the game goes away from TT then a lot ;)
11:54:45 <Brianetta> I think buying/selling cargo departs radically from the game's premise
11:54:47 <Eddi|zuHause> i forgot to place a signal before the level crossing...
11:54:56 <Brianetta> People will want private industries, too
11:55:18 <Brianetta> Besides, the cargo might be unaffordable
11:55:36 <Brianetta> You're paid to carry them, but could you, as a train company, afford to pay outright for them?
11:56:15 <TrueBrain> Brianetta: in that case you would need to get a contract, to just move them from A to B
11:56:58 <Brianetta> TrueBrain: I thought the contract was just hidden from play, exactly like the passengers' tickets
11:57:19 <Brianetta> I never made the assumption that any of the cargo on board my trains belonged to me
11:57:39 <TrueBrain> it now for sure is ;)
11:57:50 <Brianetta> Cargo destinations removes the final stumbling block, wherein I get to choose where other peoples' cargo goes
11:57:55 <Brianetta> (thankgs Celestar, peter1138)
11:58:10 <TrueBrain> well, for sure it can;t be the 'final' one
11:58:23 <Brianetta> In this field of "it's not my cargo" it is
11:58:32 <TrueBrain> but for sure it is a good move forward ;)
11:58:38 <Brianetta> There are other stumbling blocks, but they're unrelated (:
11:59:18 <TrueBrain> but I still wish OpenTTD would have different types of gameplay
11:59:20 <TrueBrain> that would just be fun :)
11:59:32 <TrueBrain> and why does my Debian install hang at 'Validating libtext-iconv-perl' :(
12:00:03 <Eddi|zuHause> TrueBrain: how much micromanagement do you expect people to put in selling cargo at the right time?
12:05:23 <Eddi|zuHause> man... i definitely need orders that are both "no loading" and "no unloading"
12:05:42 <Eddi|zuHause> "go via" does not let me timetable
12:06:04 <Eddi|zuHause> i want the freight trains to go to a siding and wait there for the express train to pass
12:06:37 <Brianetta> You could, for now, put the sidings in a location that neither accepts nor provides cargo
12:07:27 <Eddi|zuHause> the sidings are station tracks near a town
12:07:44 <Eddi|zuHause> stations are not in the middle of the pampa...
12:10:27 <Celestar> pampa ... south american plains
12:11:06 <Eddi|zuHause> it's a common phrase in german for a place in the middle of nowhere ;)
12:13:33 <Kloopy> Bah, whoever built that station failed. Have they not read the wiki about catchment areas?
12:13:52 <peter1138> They were obviously playing a cargodest build :)
12:13:54 *** |Bastiaan| has joined #openttd
12:15:04 <Eddi|zuHause> yeah, that's roughly my layout, but it does not guarantee in anyway that there is no acceptance
12:17:11 <Eddi|zuHause> it's funny how the surrounding photos are grey, and the area around this station is green ;)
12:18:29 <hylje> that's how catchment looks like
12:20:57 *** [alt]buster has joined #openttd
12:25:12 <peter1138> maps.live.com is better around this area.
12:25:59 <Eddi|zuHause> that sounds like microsof
12:26:32 <peter1138> Hmm, there are 5 lines through that tiny countryside station ;)
12:27:08 *** [alt]buster is now known as [com]buster
12:28:08 <peter1138> And a Pendolino further up.
12:29:28 <Kloopy> They're a nice ride, those.
12:35:03 <Eddi|zuHause> dbg: [routing] [Passagiere] :Overdrive used at <64> <- i don't understand that one, 64 is the main station, no vehicles go through there, all end there... how can overdrive occur there?
12:35:25 <Celestar> Eddi|zuHause: not sure (yet) :P
12:36:29 <peter1138> Hmm, someone just did a transaction for €10,000
12:36:31 <Eddi|zuHause> instead of taking the train to the other end of the city, which they planned at first
12:37:09 <Rubidium> peter1138: you have to notify the CIA of such large transactions?
12:41:50 <Celestar> TrueBrain: can we somehow start the compile farm on the current cargodest code base?
12:42:40 <glx> Celestar: msvc fails due to latest merge ;)
12:43:52 <glx> r14234 cargopacket.cpp:275
12:44:14 <glx> Erreur,...2,...error C2666: 'OverflowSafeInt<T,T_MAX,T_MIN>::operator /' : les 4 surcharges ont des conversions similaires,...d:\developpement\ottd\trunk3\src\cargopacket.cpp,...222
12:44:28 <glx> overloading problem, MSVC can't determine what to use
12:45:00 <Celestar> glx: then you should have that error in trunk as well
12:45:20 <Celestar> glx: any ideas how to get around that?
12:45:33 *** Jerimiah40 has joined #openttd
12:45:48 * glx needs to find how we solved it
12:54:16 <Eddi|zuHause> hm... missing a route...
12:56:01 <Eddi|zuHause> peter1138: that live.com site fails horribly in konqueror...
12:59:13 *** Jerimiah40 has joined #openttd
13:07:32 <Tekky> hi, did I understand correctly that the trunk and cargodest repositories were broken 10 minutes ago? Do they compile now or should I wait before getting the latest SVN release?
13:07:59 <glx> you can wait if you are a MSVC user :)
13:08:41 <Tekky> yes, I am a user of MS Visual C++ 2008 :(
13:13:59 <dih> i would say lol, but i thought i'd be more supportive after you expanded MSVC :-P
13:15:19 <peter1138> glx: ((uint)cp->count) ?
13:15:29 <peter1138> glx, why the extra braces?
13:16:40 *** Vikthor has joined #openttd
13:17:13 <Celestar> er .. static_cast<uint>cp->count ?
13:18:03 <TrueBrain> Celestar: is cargodest already compilable?
13:18:16 <Celestar> TrueBrain: no, I'm waiting for the final fix of glx
13:18:23 <Celestar> I dunno why MSVC fucks that up
13:18:55 <glx> TrueBrain knows how it nicely fails
13:19:01 <Celestar> cp>-count IS a unsigned short?!
13:19:36 <glx> and uint16 is not one of the 4 "supported" types in templates
13:19:52 <Celestar> I didn't now MSVC sucked THAT badly
13:20:28 <Eddi|zuHause> my routing cache is wrong :(
13:20:30 <Celestar> are we we committing this cast?
13:20:35 <Eddi|zuHause> or my route network
13:20:46 <Celestar> Eddi|zuHause: how did you manage to mess it up?
13:21:05 <Eddi|zuHause> i think it was through adding intermediate orders
13:21:10 <Eddi|zuHause> but i have no real idea
13:21:24 <Celestar> Eddi|zuHause: reset it, see if it works
13:21:42 <Celestar> if it does, bad luck. if it doesn't I might have a chance debugging it?
13:21:44 <Belugas> are we this committing cast ?
13:21:47 <Eddi|zuHause> yeah, that helps
13:22:00 <glx> I'm writing the commit message
13:22:17 * Celestar would prefer static_cast over C-Style casts btw (=
13:22:30 <Belugas> [09:20] <Celestar> are we we committing this cast? <--- twisting the words in a different meaning ^_^
13:22:33 <glx> indeed it looks "cleaner"
13:22:46 <Celestar> glx: yes and you spot it better
13:22:55 <Celestar> going "BEEP BEEP BEEP BEEP" in your brain
13:23:25 <glx> but count or cp->count are unsigned
13:23:45 <peter1138> I meant C-style as opposed to C++-Style...
13:26:36 <CIA-1> OpenTTD: glx * r14235 /trunk/src/cargopacket.cpp: -Fix (r14234): compilation with MSVC was broken
13:27:05 <Tekky> when working with classes and polymorphism, it is important to distinguish between static_cast and dynamic_cast. But I don't see any necessity to use this distinction with signed and unsigned integers, so I would personally prefer C-style casts in this situation.
13:28:06 <Noldo> Tekky: there are other types of casts too
13:28:46 <Tekky> Celestar: cargodest is also broken, isn't it? Isn't a trunk merge with cargodest necessary now?
13:29:15 *** Jerimiah40 has joined #openttd
13:29:24 <Noldo> problem with c-style cast is that it's sometimes reintrpret and sometimes static cast
13:29:45 <Celestar> Tekky: it's already done
13:31:14 *** reldred is now known as reldred|gone
13:32:48 <Celestar> oh it's still pushing?!
13:33:26 * eekee is compiling r14234. jas just missed getting cargo dest?
13:33:32 <Tekky> Celestar: what does "pushing" mean? Is that the process of merging with trunk?
13:33:47 <Celestar> Tekky: pushing is publishing my local changes to a server, in this case, hg.openttd.org
13:34:49 <Tekky> Celestar: Ah, so when you merge with trunk, you first do so locally and then update the server with the cargodest repository?
13:35:06 <Celestar> Tekky: yes. In hg, all commits are local only
13:35:14 <Celestar> push then moves it onto the server
13:35:17 <glx> there's not real hg repository :)
13:35:45 <glx> you could pull changes from Celestar's local copy directly
13:35:52 <TrueBrain> decentral, is the word :)
13:35:53 <Celestar> IF I run hg serve (=
13:36:13 <Celestar> TrueBrain: can we like run the compile farm on cargodest at 1600CEST?
13:36:14 <TrueBrain> Celestar: and, is it ready now? :p
13:36:15 <eekee> oh hg is like git in that way?
13:36:22 <Celestar> TrueBrain: ready yes. tested no
13:36:32 <Celestar> TrueBrain: you can also just run it and see what fails to compile (=
13:36:36 <TrueBrain> Celestar: well, I can't schedule it, but I can trigger it now, or in 24 minutes .. what ever you want
13:37:10 <Celestar> TrueBrain: in a few minutes
13:37:31 <glx> there's still a warning (but it's in trunk too)
13:37:31 <Celestar> TrueBrain: You shall be informed
13:38:37 <glx> InfoChangeResult to bool conversion
13:39:45 *** KillaloT has joined #openttd
13:39:58 <glx> but I won't commit it as peter1138 said he'll fix it in a better way
13:44:32 <Celestar> TrueBrain: you have a go for compile (=
13:44:58 <TrueBrain> but now I don't want to :p
13:45:40 *** Doorslammer is now known as Slammer
13:46:02 * Celestar hands TrueBrain a beer
13:47:42 *** Slammer is now known as roDoslammer
13:47:54 <TrueBrain> [2008-09-03 13:46:46] Created job 0000007 (hg://hg.openttd.org/developers/celestar/cargodest.hg/)
13:47:54 <TrueBrain> [2008-09-03 13:46:46] Created proc 0000048 (0000007, win64)
13:47:54 <TrueBrain> [2008-09-03 13:46:46][Thread 1] Starting proc 0000048
13:47:57 <TrueBrain> (and many more of those lines
13:49:16 <peter1138> So do we the pleasure of being able to do these things ourselves, or is our dedicated server TrueBrain/Rubidium only territory?
13:49:37 <TrueBrain> peter1138: the latter, of course
13:49:41 <TrueBrain> that is logic, not?
13:50:18 <TrueBrain> peter1138: as we can't do magic, it takes time to create such tools which all the developers understand, and can use .. for now we can't even produce nightlies on a nightly base
13:50:59 <TrueBrain> 230 seconds to produce a debian binary :)
13:51:31 <peter1138> 230 seconds... this server must suck.
13:51:49 <TrueBrain> you ever did: make mrproper; ./configure; make
13:52:00 <TrueBrain> see how long it takes on your computer ;) (make -j1 btw)
13:52:59 <peter1138> It's off at the moment so I can't double check :p
13:53:17 <TrueBrain> so it sounds about right :)
13:53:26 <peter1138> 230 seconds is nearly 4 minutes.
13:53:27 <Tekky> are you talking about the new LeaseWeb server?
13:54:10 <TrueBrain> peter1138: yup :) About 20 seconds is wasted in unpacking a tar with all the required files, and an other 20 seconds to pick the binary and required files
13:54:18 <TrueBrain> an other 20 to boot up the system and shut it down
13:54:23 <TrueBrain> so all by all, it is not that slow ;)
13:54:50 <eekee> what do you call engines when a single purchase gets you an engine at both ends of the train?
13:55:42 <TrueBrain> Windows 64bit: 519 seconds
13:55:46 <TrueBrain> now that is a lot ... stupid MSVC :(
13:55:57 <glx> windows startup is slow too
13:56:07 <peter1138> Feh, takes me 18 seconds for a full (debug) build ;)
13:56:12 <TrueBrain> glx: it beats Debian to it
13:56:27 <TrueBrain> takes my own system here 2 minutes 42 seconds to do a full recompile
13:56:38 <TrueBrain> glx: that is called ripping away all unneeded shit from Windows :p
13:56:40 <glx> MSVC release builds are slow
13:56:45 <TrueBrain> it doesn't even boot the disk manager ...
13:57:34 <peter1138> Hmm, you run the compile as root? :o
13:57:38 <glx> I can configure and make win9x after starting win32 release build, and win9x is finished first
13:58:07 <TrueBrain> glx: sounds about right ;)
13:58:12 <TrueBrain> peter1138: how do you mean?
14:00:11 <TrueBrain> glx: in this case, 2 debian targets and source + docs compiled, in the time it took win64 to compile
14:01:09 <glx> but MSVC release builds are highly optimised
14:02:26 <TrueBrain> gcc -O3 is known to over optimize too
14:02:38 <TrueBrain> when you try that for OSX target, you have to start running ;)
14:02:44 <Celestar> not as bad as it used to
14:02:52 <Celestar> er .. doesn't gcc on OSX go up to -O5 ?
14:03:04 <TrueBrain> Celestar: possible, I know -O4 was instant crash
14:03:13 <TrueBrain> so never checked any higher values :p
14:03:32 <Celestar> -O5 takes the whole OS with it
14:03:43 <Celestar> -O6 then causes a quantum singularity
14:03:44 <TrueBrain> Celestar: at least makes it fast :p
14:04:04 <TrueBrain> 455 seconds for win32 ..a good 60 seconds faster than win64 ..
14:04:08 <glx> at least it showed that int x = a() << 16 | b(); doesn't always call a before b
14:04:39 <Celestar> x = a() + b() <= There is no definition which function is called first.
14:04:40 <glx> yes and msvc optimisation showed it nicely
14:04:44 <TrueBrain> glx: you should never trust the order of such statements :)
14:05:00 <Celestar> don't make use of UBs (=
14:05:09 <glx> there was some stuff like that with Random() calls
14:05:25 <TrueBrain> Celestar: well, isn't NULL also undefined for C89 or so?
14:05:33 <Celestar> TrueBrain: undefined.
14:05:38 <Celestar> TrueBrain: not undefined behaviour
14:05:45 <Celestar> read: implementation dependent
14:06:41 <TrueBrain> btw, it amazed me that MS is really so stupid to make ftell() ... smart? Well, that aint the word for it ..
14:07:13 <glx> fopen(file, "rb"); <-- the "b" flags has been added for MS ;)
14:10:21 <TrueBrain> ah, cargodest finished
14:10:28 <TrueBrain> (Well, the compile-farm part)
14:10:40 <eekee> British craftsmen have a derogatory phrase that could be applied to quite a lot of things in computing: "Trying to be clever."
14:10:45 <TrueBrain> that should be it ;)
14:11:12 <Celestar> cool. Thanks TrueBrain
14:11:13 <TrueBrain> look, it even has a changelog! :p
14:12:32 <TrueBrain> peter1138: if you want to know, the server itself can compile OpenTTD in 35 seconds. But that is utilizing all 4 cores :)
14:13:20 <TrueBrain> (which isn't the smartest idea, knowing the website and stuff do want CPU time too ;))
14:13:47 <glx> and some servers want memory too ;)
14:14:00 <TrueBrain> memory there is plenty :p
14:14:36 <TrueBrain> (lighttpd is restricted to avoid it consuming ALL the memory)
14:14:48 <Gekz> make it use 3 cores then
14:15:02 <glx> it just dies when it reaches the limit ?
14:15:29 <TrueBrain> the compile-farm does .. 3 targets compile at the same time, each at their own core :)
14:16:01 <TrueBrain> glx: yes, OOM-killer kicks in
14:16:07 <TrueBrain> (not much you can do about that in Linux land)
14:16:12 <eekee> wouldn't it get better cache usage by doing 1 task on 3 cores?
14:16:24 <Gekz> Linux seriously fails for memory handling
14:16:41 <glx> Gekz: I think windows is worse
14:16:48 <TrueBrain> eekee: VirtualBox is used, so no ;)
14:16:53 <Gekz> you cant do anything on Windows to prove that it fails
14:16:58 <Gekz> it doesnt let you get to that point
14:17:25 <eekee> mem handling is generally impossible to do well in a multi-tasking environment
14:40:41 *** Frostregen has joined #openttd
14:46:15 *** ChanServ sets mode: +v tokai
14:48:17 <TrueBrain> Celestar: OSX sucks
14:48:21 <TrueBrain> we tried like 10 ways already
14:48:26 <TrueBrain> and .. OSX 10.5 sucks
14:53:02 <eekee> I 'spect OS X requires hardware opengl
14:53:30 <Rubidium> Gekz: ofcourse you can cross-compile... but those binaries won't boot on Intel OSX 10.5
14:53:41 <TrueBrain> OSX is nice .. just that they dropped their GCC patches is less nice
14:54:25 * eekee smells big trouble brewing
15:16:10 *** Swallow has joined #openttd
15:17:54 *** Jerimiah40 has joined #openttd
15:21:55 <eekee> hmm, I could use a way to make a train go to depot only after it's finished loading/unloading
15:22:17 <CIA-1> OpenTTD: glx * r14236 /branches/noai/ (94 files in 16 dirs): [NoAI] -Sync: with trunk r14194:14235
15:22:44 <eekee> or to go to depot (and stop) only when it comes to the service at .. order
15:29:03 * yorick wonders if it's possible to switch password encription to sha
15:29:55 <yorick> someone told me he used a "publicly available openttd password cracker" to steal company passwords
15:30:06 <yorick> and he actually got the company passwords from somewhere
15:31:04 *** Jerimiah40 has joined #openttd
15:31:26 *** mortal` has joined #openttd
15:31:55 <blathijs> yorick: You mean a server admin that recovered passwords of clients?
15:32:33 <yorick> some absolute stranger that said he had used a publicly available openttd password cracker to steal the passwords that were used on the server...
15:32:41 <yorick> and yes, he got them correctly
15:33:18 *** roDoslammer has left #openttd
15:34:35 <eekee> yorick: you mean he picked some random server and was able to get in?
15:34:58 <yorick> he got the company passwords and the server password
15:36:25 <TrueBrain> dih: stop toying with yorick, he doesn't like it :p
15:36:58 <yorick> well, I know the tools to bruteforce openttd exist
15:37:04 <eekee> glibc SHA1* functions don't look too different to the MD* functions
15:37:08 <TrueBrain> dih: I was kidding :)
15:37:36 <TinoDidriksen> If there's access to the hashes, it's easy to do lookups in rainbow tables...but why would he have access to those?
15:38:00 <yorick> I dont think he had access to the hashes
15:38:04 <dih> yorick: use ap+, remove the rcon password and use !rcon ;-)
15:38:12 <TrueBrain> what is sent over the network nowedays? md5?
15:38:18 <peter1138> Switching hash does not help against bruteforce at all.
15:38:29 <TinoDidriksen> Salted, you mean.
15:38:40 <TrueBrain> what is the plain text for: 60b725f10c9c85c70d97880dfe8191b3
15:38:41 <yorick> yes, he didn't try the rcon password
15:39:11 <TrueBrain> (hint, use www.google.com)
15:39:38 <DaleStan> Are there not an infinite number of plaintexts that produce that hash?
15:39:58 <TrueBrain> DaleStan: depending on your string length, but there is one simple plaintext solution for this string
15:39:59 <peter1138> Without a limit on input length, yes.
15:40:03 <TinoDidriksen> Yes, but you know passwords are usually not very long, so easy to make tables.
15:40:15 <TrueBrain> (and in fact, the md5sums of all dictonary words are findable via google)
15:40:30 <peter1138> To get the hash you need to either read OpenTTD's memory, or snoop the network.
15:40:35 <TrueBrain> just to show you how uncool md5 is, when you use plaintext words which can be found in a dictonary
15:40:35 <peter1138> Easier to bruteforce.
15:40:38 <TinoDidriksen> Hence why salting with longer unique strings is very important these days.
15:40:51 <peter1138> I do not use secure passwords for multiplayer games...
15:41:09 <dih> yorick: your's and rortom's bot can be used to bruteforce easily
15:41:10 <TrueBrain> peter1138: I sure hope nobody is using his uber-password for MP games :s
15:41:19 <peter1138> TrueBrain, you'd be surprised...
15:41:22 <dih> i am so not happy with that bot
15:41:28 <TrueBrain> peter1138: I always am .. makes me very sad :(
15:41:40 <yorick> dih: if someone writes a thing to hash the passwords, yes
15:41:51 <dih> yorick: rcon is not hashed
15:42:12 <yorick> how would you check rcon password correctness
15:42:18 <yorick> it has no error if it's incorrect
15:42:21 <dih> besids the hashing is not _that_ difficult to rewrite into a python bot
15:43:00 <TinoDidriksen> CHAP with SHA1 is a good start.
15:43:04 <yorick> that'd require a timeout every time, dih
15:43:05 <dih> yorick: just read the openttd code, & translate it to python
15:43:20 <yorick> python strings are immutable
15:43:23 <dih> a timeout is not an issue
15:43:47 <peter1138> Doesn't CHAP require the password to be stored in plaintext?
15:43:48 <yorick> it takes 1 second for each failed try
15:44:04 <yorick> python has no string bitwise stuff
15:45:41 <SmatZ> does client ever receive any password, even in hashed form?
15:45:53 <yorick> only way is to bruteforce
15:46:00 <dih> all it gets is the servers unique_id
15:46:00 <SmatZ> no ... so how could he "dehash" it?
15:46:26 <dih> and that is _so_ hard to get
15:46:27 <TinoDidriksen> Point of CHAP is that what is sent over the wire is not possible to reuse. If you add salting with a known salt on both sides then you don't need to store in plain text either.
15:46:36 <dih> if the client can hash it's own stuff
15:46:41 <dih> why can not a python client
15:46:57 <TinoDidriksen> Python can. There's plenty of modules for that.
15:47:06 <Belugas> one of them is called pepper
15:47:19 <TrueBrain> yorick: the passwords used on the server he supposenly broke into, were they dictonary words?
15:47:51 <yorick> TB: I dunno, I think they were partly dutch, partly english with some numbers mixed in
15:48:11 *** frosch123 has joined #openttd
15:48:51 <TrueBrain> yorick: then that would be a no ;)
15:49:20 * dih starts looking in his dict to find a half dutch half english word with digits in the middle
15:49:30 <eekee> wouldn't better encryption slow the bruteforce?
15:49:50 <SmatZ> if it was his server, he could set company passwords to whatever he wanted ;)
15:50:13 <dih> anyway - the rcon is not encrypted
15:50:13 <yorick> SmatZ: well, but someone breaking in to a normal server is a disaster
15:50:23 <yorick> dih: he didn't crack the rcon
15:50:29 <dih> so there is no need to discuss the encryption on someone hacking the rcon password
15:50:38 <dih> yorick: as far as you know, he did not
15:50:47 <Ammler> it might be better to store company passwords plain and tell that
15:50:58 <TrueBrain> yorick: well, if he is for any real, he comes in here and tell us how he did it
15:51:04 <Ammler> so they could be saved and nobody has to worry about.
15:51:46 <yorick> dih: cracking the rcon password would take approximatly 2.77777778 × 10^58 hours
15:51:58 <yorick> assuming a timeout of 1 second
15:52:03 <dih> yorick: that's what you guess... yes
15:52:22 <TrueBrain> yorick: why a timeout of 1 second? You can do much much more requests per second
15:52:37 <yorick> TrueBrain: only if the server responds, it is correct
15:52:38 <TrueBrain> introduction to bruteforce: you send many many many tries per second, with the command: terminate server
15:52:42 <TrueBrain> when you notice the server is gone
15:52:45 <TrueBrain> go back 5 minutes in your tries
15:52:57 <TrueBrain> that wy, you never have to wait for a server reply
15:53:31 * dih pats yorick on the head ;-)
15:53:32 <TrueBrain> what dih says works too, if you can monitor the result
15:53:45 <TrueBrain> (we did my trick over SSH :p)
15:53:46 <dih> you will sooner or later get a result
15:54:20 <Ammler> dih: but the problem is not yoricks bot, it would be openttd :-)
15:54:22 <yorick> the server console would get flooded first
15:54:48 <yorick> and possibly an auto-ban after 10 failed tries
15:55:04 <dih> Ammler: the bot makes it a lot easier to start brute forcing
15:55:33 <yorick> dih: *someone* created remote admin tool that uses rcon, too
15:56:13 <yorick> lets not put in the fact I helped him out
15:56:38 <dih> well - those things were bound to show up sooner or later
15:56:53 <yorick> first, he sent rcon commands over UDP and expected a result :-P
15:57:14 <yorick> then, he sent the rcon commands ofer TCP and instantly closing the connection, still expecting a result :P
15:57:26 <mortal`> lol, that's horrible yorick
15:58:07 <dih> it sounds like he tried to follow an rcon tool for like quake3
15:58:16 <dih> and tried turning that into tcp
15:58:19 <yorick> those server error codes should get a meaning in console ;)
15:59:01 <yorick> error 10054, client closed connection, how are the normal openttd hackers supposed to know what that means
15:59:02 <dih> there should be a way to tell if something is a bot or a real client
15:59:19 <dih> yorick: grepping the code?
15:59:31 <yorick> more like grepping head
16:00:47 <yorick> try not to kill too many deers on the way home
16:04:32 <TinoDidriksen> There is by definition no way to tell who's a bot and who's a real client, especially since bad guys won't follow any such rules, or will swich as they see fit.
16:06:03 <yorick> Tino: unless desync check moves to server
16:06:39 <TinoDidriksen> They'll just keep synced.
16:07:14 <TinoDidriksen> Every client is by definition hostile, until proven otherwise by authentication.
16:08:04 <TrueBrain> TinoDidriksen: authentication alone is never enough
16:08:40 <yorick> TinoDidriksen: how would they keep synced without being a client?
16:08:44 <TinoDidriksen> Well true there should be "idiot safeguards", but if someone auths as super-admin, they can't really be denied such access.
16:08:55 <yorick> I think a client would be something that keeps sync
16:08:58 <TrueBrain> TinoDidriksen: I agree :)
16:09:00 <TinoDidriksen> Build a bot on top of the client...
16:09:12 <yorick> then it is still a client
16:09:43 <murray> riding the client like a wild horse
16:09:56 <yorick> or the other way around, building a client on top of a bot
16:10:04 <yorick> which is possible, tho
16:10:18 <yorick> if something can act as a client, it can also act as a server
16:10:35 <TinoDidriksen> Point is that you cannot tell which is which. You can request clients/bots identify themselves, but can't really trust that.
16:12:44 <yorick> no, why would you need that
16:13:00 <yorick> "never trust a client" is the fundamental rule of the openttd protocol
16:16:41 *** [com]buster has joined #openttd
16:17:00 <TrueBrain> yorick: only for OpenTTD?
16:17:31 <yorick> for every other open-source-app protocol
16:17:37 <Belugas> "never trust a client" is the key rule to almost every business
16:17:52 <yorick> closed source and real life is also recommended
16:17:54 <Belugas> add "never trust a clerk" too...
16:18:34 * yorick remembers good ol' multiplayer flash games
16:18:53 <yorick> where people could not die, and fly with the speed of light
16:19:56 <TrueBrain> for fun, talk to bzFlags developers
16:20:07 <TrueBrain> the amount of 'cheats' for that game, are amazing ..
16:20:10 <TrueBrain> and it is so easy ...
16:20:16 <yorick> then shot you down through a wall with a mega gun that sounds like an exploding elephant
16:24:09 <Ammler> are those debian packs at the binary repo up2date?
16:24:25 <yorick> if they're at 0.6.2, then yes
16:24:56 <peter1138> Never trust a yorick.
16:25:11 <yorick> seems like the stables are
16:25:59 <Ammler> ah, they install openttd
16:26:00 <blathijs> Ammler: I guess so, I think that autobuilding of .debs worked
16:26:09 <blathijs> Ammler: But the package is still called "openttd"
16:26:23 <blathijs> I'm working on changing that to "openttd-branchname" as we speak :-)
16:26:45 <Ammler> well, just wondeing, if we could confuse a debian user
16:27:14 <Ammler> because never saw those packs before in nightly archive
16:27:28 <yorick> Ammler: y0U 3a8 confuse 3872590n3
16:27:37 <Ammler> and personally, I would no like to have them installed, but debian user might like that.
16:27:57 <blathijs> Ammler: No, building nightly .debs is a feature of the new compile farm
16:28:15 <Ammler> can you have both at same time?
16:28:19 <TrueBrain> SmatZ: stuff like: going through walls
16:28:23 <Ammler> stable and nightly builds?
16:28:40 <SmatZ> TrueBrain: flying, too?
16:28:41 <yorick> TrueBrain: with guns that sound like exploding elephants?
16:28:43 <blathijs> Ammler: I'm hoping to just get a repository, so you can add a line to sources.list and then just apt-get install openttd-cargodest
16:29:03 <blathijs> Ammler: Not currently, but I have been thinking about that. Should be possible, I guess.
16:29:11 <TrueBrain> Ammler: the debs are new, yes
16:29:14 <eekee> cargodest is still a branch? What was that about merging earlier?
16:29:39 <Ammler> well, as debian users are Freaks, they should know what they do ;-)
16:29:59 <TrueBrain> Ammler: Gentoo users are Freaks
16:30:02 <TrueBrain> Debian users are stupid
16:30:13 <yorick> Ammler: you can even confuse cats
16:30:14 <eekee> ubuntu users are stupid :þ
16:30:29 <Eddi|zuHause> eekee: it doesn't just happen from one second to the next just because someone said "we should merge that soon"
16:30:48 <eekee> Eddi|zuHause: no someone actually said they were merging
16:30:48 <yorick> Ammler: cats don't use debian
16:31:12 <eekee> cats use source mage if they have the patience, gentoo otherwise
16:31:56 <Osai> compilation fails on mac os x 10.5.4
16:32:05 <Eddi|zuHause> eekee: that was a merge the other way round
16:32:19 <Eddi|zuHause> merging recent trunk changes into cargodest
16:33:22 <eekee> Eddi|zuHause: Oh I see ^^'
16:37:18 <Eddi|zuHause> gnah... multi tile "waypoints" don't work properly...
16:37:31 <Eddi|zuHause> all trains want to go through the same track
16:37:36 <Osai> it was something weird, I don't want to talk about it ^^
16:37:36 <Eddi|zuHause> and ignore the other tracks
16:38:05 <Tekky> To whom it may concern: You may now enter "cargodest" in the OpenTTD wiki to reach the appropriate page. I just created a redirect page.
16:38:44 <TrueBrain> YES! My life is now complete :) Tnx Tekky! :)
16:39:13 <Tekky> well, it is a lot easier than typing "passenger and cargo destinations" :)
16:41:47 <peter1138> I just go to it via the 'Recent Changes' page ;)
16:42:28 <TrueBrain> I tried Random Pages
16:42:38 <TrueBrain> did found a lot of crap
16:48:07 <CIA-1> OpenTTD: matthijs * r14237 /trunk/os/debian/ (control control.in rules):
16:48:07 <CIA-1> OpenTTD: -Feature [Debian]: Allow the Debian packaging to change the package name of the resulting package.
16:48:07 <CIA-1> OpenTTD: - The name can be changed at build time, by changing the package name in debian/changelog.
16:48:07 <CIA-1> OpenTTD: - This will allow packages like openttd-svn or openttd-cargodest packages to be built.
16:57:15 <fjb> Eddi|zuHause: You are expecting too much.
16:57:49 <TrueBrain> he said: you are expecting too much
16:58:05 <Eddi|zuHause> i don't even have an idea what i am expecting...
16:58:22 <fjb> You expected some people to find a link a posting above.
16:58:29 <TrueBrain> haha, sounds bad ;)
16:58:54 *** |Jeroen| has joined #openttd
17:03:01 *** Jerimiah40 has joined #openttd
17:04:34 <fjb> Hm, I guess I need to find a good free email account.
17:04:35 *** DaleStan has joined #openttd
17:05:58 <TrueBrain> I have one :) But you can't have it ;)
17:09:10 <TrueBrain> why? If you have my account, I can't mail :(
17:09:42 <fjb> But the account of a true brain would be perfect.
17:13:54 *** DaleStan is now known as Guest5096
17:13:55 *** DaleStan has joined #openttd
17:16:58 <TrueBrain> OdwallaBongwater: he was talking about a GOOD free email account
17:17:06 <TrueBrain> I don't think gmail falls under that category
17:17:18 <TrueBrain> (well, or you think your Privacy isn't important, then you might classify it under good)
17:17:37 * fjb doesn't think either. But could be ok for just a fun account.
17:17:41 <OdwallaBongwater> never had a problem with gmail, dont quite see what you're talking about
17:18:10 <fjb> Big Google is watching you.
17:18:15 <CIA-1> OpenTTD: matthijs * r14238 /trunk/os/debian/ (21 files in 3 dirs): (log message trimmed)
17:18:16 <CIA-1> OpenTTD: -Change [Debian]: Update Debian packaging files to the latest official Debian version.
17:18:16 <CIA-1> OpenTTD: - Packaging files are now identical to those of the official 0.6.2-1 (or,
17:18:16 <CIA-1> OpenTTD: r11138 in the collab-maint subversion repository), with the following
17:18:16 <CIA-1> OpenTTD: - Desktop files are removed, since openttd trunk installs them already.
17:18:18 <CIA-1> OpenTTD: - Changes regarding package names from r14237 are preserved.
17:18:22 <TrueBrain> OdwallaBongwater: you do know the word 'privacy' I hope
17:18:23 *** Jerimiah40 has joined #openttd
17:18:29 <TrueBrain> I know it is a bit snowed under with Google
17:18:31 <TrueBrain> but it really exists
17:18:54 <OdwallaBongwater> I dont see quite what you're getting at though
17:19:23 <blathijs> Rubidium: TrueBrain: You could try to add the following just before the debian/rules binary call in the build-debs process of the compile farm, to change the package name:
17:19:44 <blathijs> dch --create --package openttd-svn -v 0.7.0~svn-1 " * Nightly build"
17:19:45 <TrueBrain> well, you might want to read that thing you signed when you signed up
17:20:01 <OdwallaBongwater> I guess I might
17:20:20 *** OdwallaBongwater is now known as nckomodo
17:20:51 <fjb> So what counts as a good free mail account today? Hard to find any.
17:21:06 <blathijs> Rubidium: TrueBrain: Or perhaps even "-v 0.7.0~svn.r`svnversion`" to get the svn revision in the version number
17:24:54 <nckomodo> TrueBrain I'm not seeing anything that should be cause for alarm, mind pointing it out to me
17:26:10 <TrueBrain> the fact that they index the content of all your mails
17:26:16 <TrueBrain> in order 'to supply you with more direct adverts'?
17:26:31 *** Dred_furst has joined #openttd
17:26:40 <nckomodo> oh god its Dred_furst
17:27:29 <Dred_furst> haha how random is that
17:27:34 <Dred_furst> ANOTHER GOON APPEARS
17:44:15 *** LilDood has joined #openttd
17:52:44 *** Yeggs-work is now known as Yeggstry
18:04:19 <Kloopy> peter1138: I just got a crash on the cargodests binary build. I was looking at a treeview window and just as I went to click, the last passenger was loaded on to a train. When I clicked my mouse, it was on the space where the + was just showing before. It crashed as soon as I clicked.
18:05:13 <CIA-1> OpenTTD: frosch * r14239 /trunk/src/build_vehicle_gui.cpp: -Fix [FS#2267]: The engine-purchase-list-sorter doubled running-cost and halfed capacity of double-headed engines.
18:05:15 <Belugas> i think this crash would be better answered by Celestar, Kloopy
18:05:32 <Kloopy> I agree, but he's not here and I wanted to put it down in writing before I forgot.
18:05:39 <Kloopy> When I see him next, I'll give the same text to him. :D
18:17:43 *** [alt]buster has joined #openttd
18:23:43 *** [alt]buster is now known as [com]buster
18:25:59 <Eddi|zuHause> openttd: /home/johannes/spiele/OpenTTDx/paxdest/src/autoreplace_cmd.cpp:376: CommandCost ReplaceChain(Vehicle**, uint32, bool, bool*): Assertion `GetNextUnit(new_head) == __null' failed.
18:27:56 <Eddi|zuHause> AAAAAH... some idiot switched autosave off!!
18:29:06 *** stillunknown has joined #openttd
18:29:31 <Eddi|zuHause> my last savegame is from 12:00 :(
18:29:40 <dih> Rubidium, with all these bots hanging around, which can join a game as an apparent client
18:29:48 <Eddi|zuHause> that's like a year ago (x8)
18:29:58 <dih> would it not be an idea to add a flag to each client
18:30:22 <dih> and set it to false when a client sends the sync packet?
18:30:46 <dih> or does the client decide weather to 'desync'
18:31:53 <peter1138> Or actually just not allow them.
18:32:17 <frosch123> Eddi|zuHause: how the hell did you manage to trigger that assertion
18:32:20 <dih> peter1138, for that one has to know if a 'client' is a client or a bot
18:32:41 <Eddi|zuHause> frosch123: i can reproduce it in my build when i replace wagons that have no engine
18:32:56 <Eddi|zuHause> trying to find a clean trunk build ;)
18:33:22 <frosch123> wagons without engine, ok...
18:34:05 <frosch123> is that a feature of your custom build, or did I missed a trunk feature
18:34:33 <Eddi|zuHause> i don't think any of my patches changed anything regarding autoreplace
18:35:00 <dih> is there any info only a valid client sends to the server?
18:35:00 <peter1138> Mass-update button in the depot window?
18:35:03 *** GoneWacko has joined #openttd
18:35:09 <peter1138> That'll upgrade wagons without an engine, won't it?
18:35:42 <dih> or rather, only a valid client can 'know'
18:35:48 <Eddi|zuHause> yeah, crashes in clean trunk r14128, too
18:36:15 <frosch123> I thought it is not possible to upgrade wagons only
18:36:27 <peter1138> frosch123, so a new rewrite coming up? ;)
18:36:54 <Eddi|zuHause> i always use that button
18:36:57 <peter1138> dih, actually I thought the network connection chain of events was forced. Apparently not.
18:37:09 <peter1138> i.e. connect, download map, etc...
18:37:29 <dih> and then desync is still determined by the client
18:37:35 <dih> which the bots simply skip
18:38:13 <dih> so i am hoping to find something even a valid spectator sends to the server, so that the server can determin if something is a bot or not
18:38:22 <dih> and that should then be something only a real ottd client can 'know'
18:38:51 <Eddi|zuHause> frosch123: well, i could live with those not being upgraded at all
18:38:56 <SmatZ> when you modify OTTD to work as bot
18:39:03 <SmatZ> you will send correct sync values
18:39:20 <dih> but not if it's some python bot :-P
18:39:24 <peter1138> Any solution put in place can be worked around.
18:39:27 <SmatZ> and you can place SendChatMessage("SPAM SPAM SPAM") to the main loop anyway
18:40:25 <dih> how about adding thresholds?
18:40:33 <peter1138> Can bots bypass a server password?
18:40:44 <dih> i.e. kick / ban client after x something in one month
18:40:58 <peter1138> Doesn't a desync happen way before then?
18:41:05 <peter1138> Well, unless it's a modified client...
18:41:21 <dih> peter1138, problem is, when the server asks/waits for the password, the bot can still try to bruteforce the rcon password
18:42:06 <dih> and the connection is kept open
18:42:29 <dih> that is how those dodas did the spaming without being connected
18:43:12 <dih> so something extra would be pretty good
18:43:51 <dih> a threshold of 3 invalid rcon passwords from same ip = kick, 5 = ban
18:44:22 <dih> same with server password
18:44:40 <dih> and not to accept any packet from a client if the server is waiting for a server / company password
18:44:57 <dih> (*any packet = rcon / chat / etc)
18:45:26 <SmatZ> if (cs->status != STATUS_ACTIVE) {
18:45:41 <yorick> < dih> a threshold of 3 invalid rcon passwords from same ip = kick, 5 = ban <-- how are you planning to reach 5 if it kicks on 3?
18:46:22 <dih> i.e. as long as the server is running, memorize each invalid rcon attemt
18:46:37 <yorick> not the per ip solution
18:46:42 <yorick> that changes far too often
18:46:56 <yorick> (proxies should be possible too, only they need to be made)
18:47:26 <dih> besides, ip's dont change as often as you can try to send passwords in a single second
18:47:46 * SmatZ thinks about botnets...
18:47:51 <yorick> SmatZ: doesn't matter, I'll just connect and then bruteforce
18:48:18 <SmatZ> yorick: yeah, but now you can bruteforce passworded server
18:48:33 <yorick> I'll bruteforce the password first :-P
18:48:40 <dih> bruteforce will take years if you 1. get kicked after 3 tries, 2. have to change your ip after 5
18:49:13 * SmatZ remembers applying wrong password at coop several times
18:49:21 <yorick> dih: yea, lets introduce an ip table, maybe just instant ban instead of kick first
18:49:26 <dih> yes - just not _that_ often SmatZ :-P
18:49:27 <SmatZ> maybe 100 would suffice against bruteforcing
18:49:32 <dih> at least not in a row or?
18:49:50 <dih> yorick, hence configureable
18:50:02 <dih> so invalid_password_kick = 0
18:51:00 <dih> reset the counter on success
18:51:06 <yorick> maybe the banning system is flawed too
18:51:13 <dih> and remove the ip from being memorized after ban
18:51:23 <yorick> some people just get a proxy...
18:52:52 <dih> yes - just not as fast as one can send packets
18:53:16 <dih> it's not perfect, it's a start
18:55:23 <frosch123> he, you cannot open the autoreplace gui without a single engine
18:56:15 <yorick> how about engines being autoreplaced also get autoreplaced when bought?
18:57:16 <Eddi|zuHause> frosch123: yeah, i noticed :p
18:57:17 <peter1138> Write the patch for it.
18:58:12 <Eddi|zuHause> i'm sad now, i lost half a day worth of playing... :(
18:58:30 <Eddi|zuHause> i don't understand why autosave was off
18:58:37 <Eddi|zuHause> it was on weekly before...
18:58:50 <yorick> half a day isn't that muhc
18:58:55 <yorick> only a few real seconds ;)
18:59:21 <Eddi|zuHause> how can i kickban yorick?
19:00:07 <Prof_Frink> Eddi|zuHause: You can kick him by tricking him into saying
19:00:16 *** Prof_Frink was kicked by DorpsGek (Wrong channel. Retry in #openttdcoop.)
19:00:17 *** Prof_Frink has joined #openttd
19:00:55 <hylje> that's just kick however
19:01:05 *** peter1138 sets mode: +b *!*Yorick@*.adsl.wanadoo.nl
19:01:05 *** yorick was kicked by peter1138 (Like this?)
19:01:19 *** peter1138 sets mode: -b *!*Yorick@*.adsl.wanadoo.nl
19:01:28 <Eddi|zuHause> yeah, that'd work
19:03:38 <dih> would SEND_COMMAND(PACKET_CLIENT_ACK)(); not be a good start?
19:04:33 <dih> it's not 'fore' anything :-P
19:04:49 <nappe1> peter1138: this is quite hard... :D (Playing Nappe1-2008/05 build right now...) I had to in final testing adjust loan interval to 2500... with 10000 it was just suicide to company to be founded. :D
19:04:55 <dih> it's _for_ making sure one could find out if a client is a bot or not
19:05:37 <Prof_Frink> dih: openttd capcha?
19:06:10 <Prof_Frink> Finding out if someone's a bot.
19:06:27 <peter1138> I have no idea what you're talking about.
19:07:58 <dih> Prof_Frink, yes, i am hoping to be able to distinguish
19:08:03 <dih> server side of course ;-)
19:09:14 <peter1138> Born_Acorn! NewKittenAuth!
19:09:45 <yorick> you'd need something that requires human intervention and is not solvable by noai ais
19:10:13 <yorick> as soon as you know, tell me, maybe I shall accept the challenge
19:10:34 *** larsemil has joined #openttd
19:10:49 <dih> yes, lay rails spelling out the password :-D
19:11:03 <larsemil> i get this error when trying to connect to multiplayer: dbg: [net] Sync error detected!
19:11:19 <yorick> larsemil: what version, what server?
19:11:32 <larsemil> yorick: 0.6.2 my brother on a 100/100 connection
19:11:34 <dih> issue is only, that you have to already distinguish between spectator and bot
19:11:53 <yorick> dih: and how would the server see if a password is correct?
19:11:56 <larsemil> yorick: then i will not be able to play with them.
19:11:57 <Eddi|zuHause> there was some site that had maths formulas as captcha
19:12:08 <yorick> make them upgrade aswell
19:12:08 <dih> yorick, I DONT KNOW YET ;-)
19:12:11 <Eddi|zuHause> and i was like: "well, those are actually EASIER to solve for a bot"
19:12:25 <yorick> Eddi: not if they are IN a captcha
19:12:48 <yorick> only the rapidshare cats n dogs captcha is genious
19:12:57 <yorick> but I cant solve it myself :-P
19:13:00 <larsemil> yorick: well we are in the middle of a game
19:13:05 <Eddi|zuHause> yorick: well, it was simple latex output, so you could easily have OCR'ed the formula
19:13:17 <Eddi|zuHause> and then pipe it into a CAS
19:13:29 <larsemil> well it is the latest stable version, shouldnt that do?
19:14:05 <yorick> nightlies without newgrfs are the only thing more stable
19:14:08 <nckomodo> the thing with math problems is captchas is more of an obscurity thing
19:14:52 <Eddi|zuHause> larsemil: some desync issues can be solved by saving, restarting the server, and loading the savegame again
19:15:20 <yorick> dih: just don't come up with ideas if you don't have any :-P
19:15:39 <larsemil> Eddi|zuHause: well now i got the connection dropped error.
19:16:08 <yorick> dih: moving desync-test to server would be a solution, then you'd have to expand openttd
19:16:08 <Eddi|zuHause> that sounds like a firewall issue
19:16:21 <yorick> firewalls don't cause desyncs...
19:17:12 <larsemil> this is how it is, first i connect to the game, it loads the whole map, it freezes, its moving and recieving info for 1 sec and then it bugs out
19:17:14 <FauxFaux> I bet you could write a firewall that dropped specific packets. :)
19:17:37 <yorick> larsemil: try to enable pause_on_join
19:17:46 *** welshdragon has joined #openttd
19:17:51 <larsemil> everyones game pauses when someone joins
19:18:23 *** welshdragon is now known as tireddragon
19:18:44 *** tireddragon is now known as welshdragon
19:19:31 *** larsemil has joined #openttd
19:19:38 <larsemil> anyone said anything?
19:20:20 <larsemil> so there is no solution except getting a nightly?
19:20:49 <hylje> does it happen in small maps too
19:20:51 <glx> autoreplace/autorenew is active ?
19:21:35 <larsemil> well i have no idea, i just wanted to play a game with some friends. i changed nothing from the 0.6.2 package i downloaded
19:22:16 <glx> there are some known desync causes for 0.6.2 and autoreplace/autorenew is one of them
19:23:43 <larsemil> i got another bug today, when building signals back in 1965 i got the newer ones. :D
19:25:35 <larsemil> well i give up, no luck connection
19:25:55 <larsemil> had some problems in the beginning but then i came in and it was stable for 2 hours, now its frakkin with me again
19:26:15 <Eddi|zuHause> i set the time the new signals appear to 1948
19:26:38 <Eddi|zuHause> which is rather *insert dreaded r word here*
19:32:00 *** yorick was kicked by DorpsGek (ok)
19:32:18 <Belugas> I awas bout to do it!
19:34:34 * murray snaps the cookie in the passing
19:36:56 <Forked> hows the cargodest thingy test going?
19:37:52 <dih> yorick: i am thinking about a solution, not telling you what solution i have
19:38:20 <Forked> sooo.. whats the kick count on yorick?
19:38:42 <dih> this channel only or all others on this network?
19:40:09 <Forked> I think this will do..
19:41:23 <yorick> does that include the "Did you bring snow yet" kicks?
19:41:33 <SmatZ> grep -c "yorick has been kicked from the channel by" *#openttd.log
19:42:04 <Ammler> Forked: join #openttdcoop.dev
19:42:16 <SmatZ> glx: only 1 occurence :-/
19:42:26 <SmatZ> maybe my logs are too little
19:42:33 <yorick> noai, there are the tbs
19:46:53 <TrueBrain> I have a total of 32
19:47:35 <dih> include openttdcoop openttdcoop.dev openttdFairPlay (what else have i forgotten?)
19:48:02 <Rubidium> dih: #openttdstupidpythonbot
19:48:05 <TrueBrain> 32 are the channels Dorpsgek is in ;)
19:55:11 <yorick> I haven't been kicked from #openttdstupidpythonbot
19:55:13 <Ammler> kick is quite stupid as most have autojoin...
19:55:42 <glx> Sacro doesn't have autojoin ;)
19:56:34 <Rubidium> I think mode Q is better
19:56:49 *** Progman has joined #openttd
19:56:53 <glx> registered and identified only ?
19:57:30 <Rubidium> no, "global" /ignore
19:58:07 <dih> and that works very well actually
19:59:03 *** Rubidium sets mode: +q *!*@s55924da0.adsl.wanadoo.nl
19:59:37 <dih> uh uh - what you say? i cannot hear you :-P
19:59:38 *** Rexxars has joined #openttd
20:00:13 *** Rubidium sets mode: -q *!*@s55924da0.adsl.wanadoo.nl
20:00:56 <yorick> I didn't even notice ^^
20:01:05 * SmatZ wonders what would happen if yorick removed nm from all other users :)
20:01:40 <yorick> you mean something like @seen *
20:02:34 <SmatZ> ok maybe my client interpretes wrong what Rubidium just did
20:03:00 <SmatZ> hmm is it correct to say "wrong" or "wrongly"/
20:03:34 <glx> I'd say "wrongly" in this case
20:03:46 <SmatZ> [21:59:03] *** Rubidium gives channel owner privileges to *!*@s55924da0.adsl.wanadoo.nl.
20:03:58 <yorick> your client wrongly interprets
20:04:19 <Ammler> [21:59] *** Rubidium gibt *!*@s55924da0.adsl.wanadoo.nl Besitzerstatus.
20:04:33 <dih> SmatZ is sitting on our bouncer?
20:04:38 <Belugas> isn't it "wrongfully"?
20:04:46 <glx> hmm +q is indeed for owner
20:05:31 <SmatZ> dih: I don't filter ctcp version requests
20:05:52 <dih> the bouncer did not reply to my ctcp version request :-P
20:06:00 <glx> but on oftc +o is the max
20:06:05 <SmatZ> [22:05:00] [CTCP] Received CTCP-VERSION reply from SmatZ: Konversation 1.1 (C) 2002-2008 by the Konversation team.
20:06:18 <yorick> === CTCP version reply “Konversation 1.1 (C) 2002-2008 by the Konversation team” from SmatZ
20:07:10 <dih> i configured the bouncer to filter them for me :-)
20:09:42 <dih> now - back to detecting bots
20:09:52 <dih> Rubidium, what would your gut-feeling be to this?
20:09:54 * Prof_Frink points at DorpsGek
20:11:05 <dih> yes - well done Prof_Frink, but i mean inside a game
20:11:59 <Rubidium> dih: well, ban them?
20:12:45 <Prof_Frink> LIE_MODE=0 Are you a bot?
20:13:09 <yorick> dih: how can you detect IRC bots?
20:13:12 <dih> moving desync detection to the server side
20:13:28 <yorick> people will port openttd to python :-p
20:13:59 <yorick> SpComb: I already did the networking for you :)
20:14:09 <SpComb> yorick: let me see your code
20:14:29 <dih> you did the networking for yourself and now want SpComb to use your work
20:14:52 <yorick> dih: that's basically the same, but seen in a more true way
20:15:59 <yorick> and I did not really do the openttd interface...I'm mainly working on the bot itself :-p
20:19:32 <dih> Rubidium, what do you think of configurable thresholds to stop bruteforcing
20:19:41 <yorick> I think it's a nice idea
20:20:11 <yorick> but I'm not rub...and I shall not say anything
20:20:33 <Rubidium> dih: you mean: failed password == kick, no rcon without being fully joined to the map?
20:20:57 <yorick> I would like the failed rcon 3x = kick
20:21:07 <yorick> no rcon without being fully joined is silly
20:21:42 <dih> you wanna be able to rcon while the server is awaiting the server password ?
20:21:53 <yorick> I sometimes use rcon to control the server while waiting for the map
20:22:15 <yorick> and I wanna be able to use rcon before sending the PACKET_CLIENT_JOIN packet
20:22:45 <dih> yorick: just exactly that is the problem
20:22:51 <dih> people like using exactly that stage
20:22:59 <dih> or at least _can_ use exactly that
20:23:14 <yorick> then add the limit there
20:23:15 <SpComb> the only of openttd that I care about is the networking
20:23:24 <SpComb> so I'd jus rewrite it anyways :P
20:24:16 <Belugas> yeah! let's do that!! with XML
20:24:25 <yorick> Belugas: let's do what?
20:24:54 <Belugas> i guess yorick has SpComb on /ignore :)
20:24:56 <Rubidium> yeah, and SOAP and RPC
20:25:14 <dih> why not include an ftp server to upload savegames
20:25:34 <dih> and while we are at it....
20:25:45 <yorick> dih: is on the goals list...
20:25:59 <dih> you are on the goals list
20:26:17 * Prof_Frink was implemented in 0.5
20:26:22 <yorick> a yorick-less openttd python framework
20:26:48 <yorick> Prof_Frink caused too much trouble and was removed from any future versions
20:31:18 * eekee converts an old game from standard trains to ukrs to aleviate boredom
20:34:18 <Belugas> eekee, want something that will REALLY aleviate boredom?
20:35:28 * TrueBrain makes a static link to this, and sees it doesn't resolve ..
20:39:43 *** [com]buster has joined #openttd
20:44:32 <eekee> when I look at the source I get a bit intimidated, not knowing where to start, and I've never really had anything to do with c++ either
20:45:51 *** NukeBuster has joined #openttd
20:45:53 <TrueBrain> I am proud to let you guys know that 54% of the OpenTTD webpage vistors use FireFox :)
20:46:13 <TrueBrain> eekee: if you want to touch (any) source, you need to have a goal :)
20:46:18 <TrueBrain> then the rest comes when you go
20:46:24 <nckomodo> I keep wanting to modify OpenTTD and add planes that can go to waypoints and bomb things
20:46:32 <nckomodo> but I wouldnt know where to start
20:46:43 <TrueBrain> nckomodo: and what do you want to bomb? :)
20:46:49 <eekee> TrueBrain: this is true, and my goals tend to be large
20:47:32 <TrueBrain> eekee: well, I started out doing FaceSelector for OpenTTD (was never added, as it sucked :p)
20:47:34 <TrueBrain> next thing was bigmaps :p
20:47:39 <Eddi|zuHause> i don't use firefox either...
20:47:41 <TrueBrain> so that doesn't matter ;) Just takes longer :)
20:47:44 <nckomodo> actually I think that'd be a fun modification to OpenTTD, a wartime sort of thing
20:47:50 <nckomodo> trucks with AA guns, etc etc
20:48:17 <TrueBrain> nckomodo: you might want to load up an other game
20:48:31 <eekee> TrueBrain: I've been wishing for elevated light rail / monorail, lately
20:48:55 <TrueBrain> eekee: haha, that might be too hard to start with yes :p
20:49:30 <nckomodo> TrueBrain maybe, but I still think it could be fun if you do it right
20:50:46 <eekee> hmm for wartime you'd need ability for vehicles (inc waggons) to shoot at each other, and several new flavours of crash and burn. doesn't sound too awful
20:50:50 <TrueBrain> nckomodo: I doubt there is a 'right' in this case :)
20:50:57 <TrueBrain> hmm .. mapgen .. that is a project I should revive ..
20:51:56 <eekee> oh my nephews love playing dirty. like, building a short bit of rail across a road where their oponent has trucks, and running an engine across when they see a truck coming
20:52:11 <CIA-1> OpenTTD: rubidium * r14240 /trunk/docs/obg_format.txt: -Doc: add a few pointers/"howto"-ish lines to the obg format documentation about what it does and does not do and how you should fill the data.
20:56:38 *** OdwallaBongwater has joined #openttd
20:58:02 *** Brianetta has joined #openttd
20:59:39 <Eddi|zuHause> oh i remember doing that too... back when i was 12...
21:00:42 <eekee> they do it to each other & like it. *shrug* :D
21:04:00 <SmatZ> they enjoy the game, that's important
21:04:16 <eekee> TrueBrain: they're 20 & 24 ;)
21:04:44 <eekee> honestly I don't really 'get' the finer points of playing fair. maybe they're the same
21:05:00 <SmatZ> it is too easy to play unfair in OTTD
21:05:29 <SmatZ> furthermore when you can modify sources...
21:05:30 <eekee> well, it's a business sim...
21:05:49 <eekee> are you suggesting something SmatZ? ^^;
21:06:09 <SmatZ> just you can automate some destroying actions
21:06:19 <SmatZ> though maybe it is more fun to flatten the land manually
21:06:22 <Prof_Frink> peter1138! NewOFT!
21:08:02 *** AlexFili has joined #openttd
21:09:00 <AlexFili> I'm having some major problems with online multiplayer
21:09:15 <AlexFili> after about 10 minutes I get disconnected and my router stops accepting connections for some reason
21:09:30 <AlexFili> im using a wireless router
21:09:57 <AlexFili> with a wireless network card in the pc
21:10:29 <AlexFili> also when im running openttd, my DS and PSP cant connect to the router
21:10:29 *** Wezz6400 is now known as Guest5127
21:10:32 *** Wezz6400 has joined #openttd
21:10:32 <glx> if that works the problem is wifi and not the router
21:10:34 <AlexFili> sounds like openTTD is overloading my router
21:10:55 <TrueBrain> yeah, it tends to use a lot of bandwidth (lol, that was sarcastic :p)
21:11:07 <TrueBrain> OpenTTD is most likely the most network friendly application you have on your computer :)
21:11:21 <TrueBrain> any Torrent application requires MUCH more ;)
21:11:41 <AlexFili> well either way, something with openttd makes my router crash, and i wish i knew what
21:11:41 <Prof_Frink> TrueBrain: I dunno. ls?
21:12:04 <AlexFili> does it matter if the router is right next to the pc?
21:12:26 <TrueBrain> reminds me, does OpenTTD already support upnp? :)
21:12:27 <Prof_Frink> TrueBrain: ls is probably more network-friendly than openttd
21:12:52 <glx> TrueBrain: no upnp support
21:13:17 <TrueBrain> Prof_Frink: and is that a network application?
21:13:33 <Prof_Frink> You never said that.
21:13:36 <fjb> Prof_Frink: Also when you are using nfs?
21:13:45 <TrueBrain> [23:11] <TrueBrain> OpenTTD is most likely the most **network** friendly application you have on your computer :)
21:13:50 <TrueBrain> you should learn to read my friend :)
21:13:52 <glx> btw wireless needs some distance between source and receiver
21:13:54 <eekee> Prof_Frink: try ls -R on a large repo served over 9P
21:14:22 <Prof_Frink> network friendly **application**
21:14:27 <AlexFili> this problem never happened with my old router
21:14:32 <Prof_Frink> Not netork friendly network application
21:14:33 <AlexFili> only since i got a new BT homehub
21:14:55 <TrueBrain> AlexFili: I doubt very much OpenTTD is the reason for your problem :)
21:15:15 <AlexFili> TrueBrain this doesnt happen with any other application
21:15:29 <AlexFili> Counter Strike, torrents, you name it
21:15:34 <AlexFili> it only happens when i play openTTD
21:15:38 <Prof_Frink> AlexFili: Is this the BT homehub that would crash if someone uses BBC iplayer?
21:15:47 <Rubidium> AlexFili: that's just luck I reckon
21:16:10 <AlexFili> i can play iplayer fine
21:16:51 <TrueBrain> it is like blaiming the Minin on the highway for the traffic-jam where it is standing in .. 10 km from the start of the jam :)
21:17:12 <AlexFili> :( the problem only happens when I play openTTD
21:17:22 <Prof_Frink> Except it *is* the micra's fault.
21:17:35 <Rubidium> AlexFili: and it started happening since you got a new router, right?
21:17:53 <TrueBrain> Prof_Frink: there is always a chance it caused a traffic jam, drove back to the end of the line, and sat in the jam it created itself
21:17:59 <TrueBrain> possibilities there are always :)
21:18:09 <Rubidium> AlexFili: ANY hardware that OpenTTD crashes means that the controlling software of that hardware is flawed.
21:18:44 <AlexFili> well for whatever reason this only happens with openTTD
21:18:56 <SmatZ> AlexFili: my old router had bug that it locked up when I had too many open connections...
21:19:18 <peter1138> Hmm, my plan for generic action 0 property skipping fails.
21:19:21 <TrueBrain> SmatZ: which you would notice with any torrent application .. but with OpenTTD? :p
21:19:24 <SmatZ> or when I opened too many TCP connections in too short time
21:19:43 <SmatZ> TrueBrain: I don't know if torrent is UDP or TCP based...
21:19:55 <SmatZ> though... how many connections do you have open in OTTD...
21:19:56 <Rubidium> so the router can't handle 100+ UDP packets in transit?
21:20:01 *** Guest5127 is now known as Wezz6400
21:20:13 <AlexFili> actually that reminds me
21:20:17 <AlexFili> i havent used config on this new router
21:20:24 <AlexFili> I guess i should see if 3979 port is blocked lol
21:20:32 *** Wezz6400 is now known as Guest5129
21:20:35 *** Wezz6400 has joined #openttd
21:20:35 <TrueBrain> SmatZ: OpenTTD? 1 UDP socket, 1 TCP socket, for a server, and each client connected of course :)
21:21:06 <glx> for a client 1 TCP once joined
21:21:07 <SmatZ> TrueBrain: when you open the server list and request newgrf info to be sent, this all is done via UDP, right?
21:21:51 <TrueBrain> SmatZ: and 1 UDP socket, to be exact .. but UDP doesn't bind, so that doesn't really count ;)
21:22:02 <TrueBrain> (UDP isn't really a connection ;))
21:22:15 <TrueBrain> a stateless mean of communication .. :p
21:22:34 <SmatZ> yeah :) that's why am asking if it is done via TCP :)
21:22:38 <glx> it's like using a megaphone?
21:22:40 <SmatZ> I hope I never said that
21:22:52 <TrueBrain> glx: lol, nice comparison :)
21:22:58 <SmatZ> :-D megaphone = broadcast ;)
21:23:41 <eekee> broadcast udp, because some of the recipients may not get the message properly
21:23:57 <TrueBrain> but for OpenTTD to kill any sane TCP stack .. well .. there needs to be something very wrnog in that case :p
21:24:16 <AlexFili> yeah i didnt understand it either
21:24:39 <Rubidium> just complain to BT that your router crashes reliably
21:25:15 *** Guest5129 is now known as Wezz
21:26:03 <AlexFili> I dont suppose anyone here has a BT Homehub?
21:26:20 <glx> I have a freebox v4 and it works very well
21:26:23 * eekee won't ever have one, now :)
21:26:43 <AlexFili> im just wondering how i get past this user/password query
21:26:47 <AlexFili> admin/admin doesnt work
21:27:06 <eekee> oh it might be "root" for both user and pass
21:27:56 <AlexFili> "Your BT Home Hub is preset with Broadband user name and password for automatic connection to the BT Broadband network.
21:27:56 <AlexFili> You can use this page to change the default settings if requested by your Broadband service provider. Please check with your service provider if you need to enter alternative settings and what these settings are.
21:27:56 <AlexFili> Important: If your Hub is already connected to the Internet, you will first need to disconnect before being able to change the username and password."
21:28:08 <AlexFili> damn, guess i need to d/c
21:28:14 <AlexFili> well thanks for all the help
21:28:18 <AlexFili> hopefully i'll be back
21:31:55 <fjb> If we will ever see him again?
21:33:12 *** Wezz6400 has joined #openttd
21:34:03 <TrueBrain> fjb: I am bombing his router, I doubt that :p
21:35:38 <TrueBrain> I have enough bandwidth to do so ... :p
21:35:47 <TrueBrain> I remember the early days of Hotmail
21:35:55 <TrueBrain> some friend was bragging that Hotmail had the best filters
21:35:59 <TrueBrain> it was impossible to mailbomb him
21:36:04 <TrueBrain> the next day he had 100,000 emails in his box
21:36:15 <TrueBrain> (in those days, there were no such things as spam-filters, greylisting, and other 'protections')
21:36:22 <fjb> Too bad all that old tricks are not working anymore. Was fun. Ping of death or later ping with payload "+++ath0". :-)
21:36:24 <TrueBrain> and I happened to have access to a 100 mbit connection ...
21:36:34 <Forked> heh us robotics modems..
21:36:41 <TrueBrain> somehow we never became friends .. makes you wonder ..
21:36:42 <Rubidium> the good old times of jolt :)
21:36:58 <TrueBrain> fjb: you would be supprised how easy it is for me to ping you to death
21:37:08 <TrueBrain> if ICMP ping fails, I always have the ways of TCP pings
21:37:15 <fjb> Not US Robotics. Rockwell chipset in modems.
21:37:46 <TrueBrain> (just TCP pings requires more bandwidth from me, as I don't kill your upload, but your download, which takes slightly longer)
21:38:33 <fjb> TrueBrain: Mircrosoft fixed the ipstack of Windows, so no ping of death anymore... :-(
21:39:13 <TrueBrain> you want me to try it at yours? :p
21:39:30 <peter1138> That payload of +++ath0 always surprised me.
21:39:48 <peter1138> The fact it worked...
21:40:05 <fjb> Try it. But what you are talking about is nor classical ping of death where you needed only one packet to shoot Windows down.
21:40:57 <peter1138> No, I mean surprised that ppp allowed it to go straight through for ICMP, but obviously didn't for TCP...
21:41:14 <TrueBrain> fjb: ah, that ping of death
21:41:18 <TrueBrain> you are right about that :)
21:41:22 <fjb> peter1138: Yes, the payload was echoed back. And the stupid Rockwel chips didn't wait for the pause between "+++" and "ath0".
21:42:20 <Rubidium> fjb: jolt is a much nicer way of teasing Windows users
21:42:28 <fjb> TrueBrain: And you are free to flood my firewall.
21:43:01 <fjb> What is jolt? I'm not in the Windows shooting business anymore.
21:43:35 <Brianetta> A player on my server just accidentally auto-replaced his Deltics with 0-4-0 saddle tanks
21:43:37 <Rubidium> fjb: google for jolt.c
21:43:44 <Brianetta> He doesn't have funds to undo this
21:43:48 <Brianetta> we're all giggling rather
21:45:07 <fjb> Rubidium: Yoah, jolt is that ping of death that I was talking about. Just didn't know it under that name.
21:46:05 *** Osai is now known as Osai^zZz
21:46:44 <Rubidium> fjb: but it isn't deadly; okay the mouse pointer can't be moved while under the influence of jolt, but when stopped it'll return to "normal" operations within a minute
21:47:07 <Rubidium> though usually the people getting jolted reset their computer before you stop jolt
21:48:03 <fjb> You could make Windows 3.11 blusscreen with oversized packtes.
21:48:28 <fjb> TrueBrain: When do you start to ping me?
21:49:04 * Rubidium wonders what kind of connection fjb has
21:52:50 <Brianetta> Rubidium: I can make most desyncs go away by disabling all newgrfs.
21:53:45 <Brianetta> Unfortunately, two non-overlapping sub-sets of newgrfs cause problems.
21:53:55 <Brianetta> so I can't point my finger at any one.
21:56:19 <fjb> Point your fingers at both.
21:57:18 <Eddi|zuHause> don't point naked fingers at dressed newgrfs
21:58:16 <Rubidium> Brianetta: is the waypoint issue caused by any of those subsets, or also without newgrfs?
21:58:35 *** Jerimiah40 has joined #openttd
21:58:35 <Rubidium> and what are the subsets?
21:59:03 <Brianetta> All but stations, certain trains but no stations, no trains but plenty of stations
22:00:18 <Brianetta> Rubidium: Currently, with no station newgrfs, we've had no waypoint sesyncs to my knowledge
22:00:23 <Brianetta> but plenty of others
22:01:02 <Brianetta> First big desync happened as I ordered road reconstruction on a town full of trams
22:01:10 <Brianetta> One of the trams got stuck, too
22:01:33 <fjb> peter1138: :-) TTDP unable to load certain old saves?
22:02:52 <Rubidium> Brianetta: did the tram thingy desync everybody or only newly joined people?
22:03:07 <Brianetta> It desynced three of five players
22:03:17 <Brianetta> we've had sporadic desyncs ever since
22:03:19 <eekee> I've had ottd mess up save games saved with a particular nightly before. ^^; It's fixed now, but I'm currently playing another save that has ufo shadows all over the place from another nightly
22:03:40 <Brianetta> I have not desynced yet, and I've been on and off
22:04:01 <Rubidium> anything different between the three desyncees and the two who didn't desync?
22:04:18 <Rubidium> like different OS/compiler etc
22:05:06 <peter1138> Not the old out of bounds array access..
22:05:09 <Brianetta> Nothing I'm aware of. I think all the desyned players are XP users.
22:05:47 <Brianetta> thinglie's a Fedora user
22:06:06 <Rubidium> and the non-desyncing users?
22:07:01 <Brianetta> Mixture. I'm Ubuntu Linux, Michael's Vista
22:07:29 <Rubidium> did that Michael compile OpenTTD himself?
22:07:36 <Brianetta> I'm running a self-compiled SVN checkout, as is the server
22:08:19 <Brianetta> Michael is using the official download
22:10:06 <Brianetta> I'm going to run without trams or newstatsw in the next game.
22:10:23 <Brianetta> This means I'll have no newgrf waypoints
22:10:29 <Brianetta> and no trams, clearly
22:28:09 *** Jerimiah40 has joined #openttd
22:31:10 <Rubidium> newstatsw is the only newgrf with waypoints?
22:31:31 <Rubidium> can't quickly find anything suspicious in it
22:33:44 <Rubidium> that isn't used on Brianetta's server
22:34:35 <Brianetta> We'll see. If the waypoints thing crops up without newstatsw we'll know it's not because the waypoints were graphically overridden.
22:35:30 <Brianetta> davidpk212 has quoted me in his sig
22:35:36 <Brianetta> [All] Brianetta: Nobody shifts gems and christmas cards like me.
22:35:52 <Brianetta> That was me, with my passenger-free Deltic-based valuables and mail line
22:36:13 <Brianetta> Pikka: Does your suburban renewal touch hotels?
22:36:40 <Pikka> if you mean the big two-tile ones with the pool on the roof, they're gone, yes.
22:36:58 <Brianetta> allow me to upload some shots
22:42:15 <Brianetta> Can you spot it? (:
22:43:14 <Brianetta> I created that company after Pikka called me a back seat engineer
22:44:30 <eekee> hmm no tree grfs on grf crawler
22:45:59 *** Rexxars has joined #openttd
22:48:20 <eekee> there is at least 1 tree replacement grf 'out there' isn't there?
22:50:04 *** Wezz6400 has joined #openttd
22:50:37 <eekee> (I'm not surprised if they're 'stolen' from some other game, there's a heck of a lot of tree sprites iirc)
22:58:17 *** Wezz6400 has joined #openttd
23:18:49 *** Yeggstry is now known as Yeggzzz
23:30:03 *** Wezz6400 has joined #openttd
23:57:25 <Tekky> in the OpenTTD source code, there is the following define in stdafx.h:
23:57:27 <Tekky> #define cpp_offsetof(s, m) (((size_t)&reinterpret_cast<const volatile char&>((((s*)(char*)8)->m))) - 8)
23:57:28 <Tekky> Why not simply use the ANSI offsetof operator?
23:58:55 <Tekky> I don't understand why this ugly hack is required.
continue to next day ⏵