IRC logs for #openttd on OFTC at 2008-05-12
⏴ go to previous day
00:26:56 <CIA-3> OpenTTD: glx * r13058 /trunk/src/smallmap_gui.cpp: -Codechange: make a class of the ExtraViewportWindow.
00:34:17 *** Eddi|zuHause2 has joined #openttd
01:47:52 *** mynetdude|Away is now known as mynetdude
01:53:02 *** mynetdude is now known as mynetdude|Away
02:08:26 *** De_Ghosty has joined #openttd
02:17:42 *** De_Ghost has joined #openttd
02:53:35 *** Eddi|zuHause has joined #openttd
02:58:55 <Eddi|zuHause> spontaneous self reboot... that is weird...
03:13:06 *** Frostregen_ has joined #openttd
03:19:16 *** Frostregen_ is now known as Frostregen
04:33:41 *** roboman has joined #openttd
04:49:39 *** DaleStan has joined #openttd
05:23:04 <CIA-3> OpenTTD: rubidium * r13059 /3rdparty/squirrel/sqstdlib/sqstdstring.cpp: [Squirrel] -Codechange: silence warnings about not checked argument types for a few sprintf calls.
05:52:11 *** Alberth has joined #openttd
06:12:19 *** einKarl has joined #openttd
06:28:51 *** raimar3 has joined #openttd
07:09:50 <cjk> I am seeing rdtsc() in the game's code, but is not that not-so-good when the TSC is not constant?
07:13:22 <Rubidium> the number of processor ticks is imo a better way to check how long a short piece of code takes than the number of seconds
07:13:35 <Rubidium> it's after all only used for crude profiling-ish stuff
07:14:38 <cjk> yes, but I was referring to a change in frequency with which the TSC increases
07:15:28 <Rubidium> The instruction returns a 64-bit value in registers EDX:EAX that represents the count of ticks from processor reset.
07:15:56 <Rubidium> so with the frequency change an operation starts taking more processor ticks?
07:18:30 <Tefad> TSC count is sometimes used for random number generation
07:18:46 <Tefad> however on pentium 4s the result seems to always be even (or odd, i forget)
07:19:25 <Rubidium> maybe every operation takes a multiple of 2 processor ticks?
07:20:18 <Tefad> it screwed with the guy that codes foobar2000
07:20:57 <Tefad> i don't think he had any netburst machines to test his code with, took a while to figure out why there was a bug in shuffle mode with some computers and not others
07:21:03 <Tefad> common denominator was P4
07:21:13 <Tefad> perhaps all netburst.. i forget
07:22:12 <cjk> Rubidium: Well I am referring to constant_tsc in /proc/cpuinfo; on some CPUs, the TSC increases constantly even if the frequency changes - this is good use for timekeeping, but when the TSC changes with frequency, it is not that suitable. How does it affect openttd?
07:23:32 <Rubidium> so constant_tsc is absolutely unuseable for the use of rdtsc in OpenTTD
07:24:03 <Rubidium> as I said before, we use the op count to determine how long a certain function takes in crude profilings
07:24:37 <cjk> ah so it's just for profiling, and not for timekeeping?
07:24:52 <Rubidium> it's actually not even been used at the moment
07:25:04 <Noldo> but it has no gameplay effect
07:25:20 <Tefad> wouldn't the system clock be best for this
07:26:03 *** Brianetta has joined #openttd
07:26:04 <Rubidium> what do you think OTTD uses?
07:26:15 <Tefad> probably system clock eh?
07:26:25 <Rubidium> something very similar yes
07:26:38 <Tefad> (heh, i typoed, causing me to do that)
07:27:08 <Rubidium> for sdl it's the OS clock (number of ms since 1970 % UINT32_MAX)
07:27:27 <Rubidium> for windows it's the number of ms since boot
07:28:51 <Rubidium> for the null video driver we don't need it (fast forwards always)
07:29:54 <Rubidium> dedicated uses the same as sdl for unixes and same as windows for windowses
07:30:01 <Rubidium> and for the cocoa driver I've got no idea
07:31:16 <Rubidium> so now, what was the problem with rdtsc?
07:34:19 <Rubidium> but you could've found out where rdtsc was used quite easy with a simple grep
07:34:33 <Rubidium> which your system would most likely have installed
07:34:44 <cjk> greps do not seem to work on the source very well :p
07:34:57 <cjk> like, trying to get rid of all the "queried from x.x.x.x" messages
07:35:38 <cjk> there's just a larger part of preprocessor magic (which is ungrepable by nature)
07:35:59 <Rubidium> grep -R 'queried from' src <- how hard can it be?
07:36:13 <cjk> yeah but what to actually _change_ without just ripping out the DEBUG() call
07:36:31 <Rubidium> the debuglevel at which it prints?
07:36:31 <cjk> in the end, I settled with removing level==0 from debug.h
07:37:32 <cjk> anyway, in a different problem "min_players=0" on the console does not work (variable not found)
07:39:13 <cjk> min_players from [network] from openttd.cfg
07:41:15 <Rubidium> as stated in the help
07:41:25 <Rubidium> min_players = 10 works
07:43:11 *** GoneWacko has joined #openttd
07:43:13 <cjk> argh.. space handling really needs to be improved (I tried "min_players=10")
07:43:36 <Rubidium> the console sucks anyways
07:46:05 <cjk> and the debug levels are inverted :(
07:46:42 <Rubidium> 0 no debugging, 9 lots of debugging
07:46:59 <cjk> hm, then -d net=x is screwed up; because "queried from" is tagged with level 2, but both -d net=1 and -d net=3 let it appear
07:47:16 <Rubidium> because dedicated overrides the debuglevel
07:47:40 <cjk> does it figure out by looking at $DISPLAY?
07:49:46 <peter1138> if you compiled with video drivers, it will fail to start if they can't be used
07:50:01 <peter1138> if no drivers were compiled, it will always be dedicated
07:50:16 <cjk> There is no page titled "config file". You can create this page.
07:50:59 <cjk> so now I've got to find out what variable in openttd.cfg makes it a dedicated
07:51:27 <peter1138> there is no setting in openttd.cfg to make it dedicated
07:52:10 <cjk> yes, self-compiled, but how would it know in advance whether I wanted a dedicated one or the GUI
07:52:21 <Rubidium> because you failed to install libsdl-dev
07:52:30 <Rubidium> and configure said that
07:53:11 <cjk> now I see... if I start /usr/bin/openttd instead, I get the GUI in aalib...
07:55:00 *** Slowpoke has joined #openttd
07:55:18 <peter1138> yes, when you run configure it tells you
07:55:30 <peter1138> but i guess most people ignore that
07:55:51 <cjk> then don't call it configure and don't make its output look like configure :)
07:56:08 <cjk> ok, makes sense. openttd -D for the gui version..
07:56:30 <cjk> I just compiled from source because I wanted to change something to suit my needs, but I could have also used a precompiled openttd..
07:57:09 <Rubidium> well, you could've read readme.txt...
07:57:16 *** Slowpoke_ has joined #openttd
07:57:22 <cjk> see, most people don't even do that
07:57:50 <GoneWacko> Can someone go to my house and wait for my UPS package to arrive?
07:58:15 <GoneWacko> If the online tracking thingie is to be believed, yes.
07:58:51 *** stillunknown has joined #openttd
08:29:22 *** Progman has joined #openttd
08:32:05 *** Vikthor has joined #openttd
08:41:20 <Alberth> Station and vehicle windows disappear, seemingly on a bankrupcy (news window?)
08:45:19 <Alberth> Euh, sorry, I was bankrupt :(
08:48:46 <Alberth> ah, well, I am not French afaik
08:49:27 <Alberth> and my French is more than non-eixstent
08:49:40 <cjk> Solve[x > nonexistant, x]
08:49:59 <Eddi|zuHause> did you mean "mais-tu"? being an inversion and all
08:50:33 <Eddi|zuHause> well, i'm baad at french ;)
08:50:50 <cjk> mais is food in german :)
08:50:56 <Alberth> Gekz: Solve[x > terrible, x] plz
08:51:07 <cjk> Alberth: OutOfBoundsError.
08:52:15 <Alberth> Gekz: If more than non-existant ~ existant, then i'd like to know what's more than terrible :)
08:52:58 <cjk> "what's more than nothing? \"something\". So more than non-existant is existant"
08:53:11 <cjk> nom de dieu de saloperie de connard!
08:54:54 <Eddi|zuHause> anglais, s'il vous plaît
08:56:10 <Tefad> ouais anglais, n'est pas?
08:56:15 <Gekz> Eddi|zuHause: Deutsch bitte.
09:05:10 *** dR3x4cK has joined #openttd
09:11:03 *** Brianetta has joined #openttd
09:16:43 *** GoneWacko has joined #openttd
10:02:35 *** Wezz6400 has joined #openttd
10:53:37 <cjk> openttd needs a better way to detect version mismatches
10:53:52 <cjk> the curent method of comparing strings (?) is horrible.
10:54:15 <cjk> I can't join a server running svn-r13032 because uh... configure decided my revision is "g92385022"
10:54:31 <cjk> though it's - code-wise - r13032 too
10:54:45 <SmatZ> then don't bitch about it
10:54:54 <cjk> that's why I said it should not use a string comparison for version checking
10:55:00 <Eddi|zuHause> ./configure --version=rXXXX?
10:55:08 <Rubidium> cjk: then use what else?
10:55:19 <cjk> a number that is dependent on the actual binary interface, not the program's version
10:55:26 <cjk> much like .so files in /usr/lib
10:55:56 <Eddi|zuHause> the problem is, how do you define what the "interface" is?
10:55:57 <Rubidium> so we are going to update that for *every* codechange?
10:56:06 <cjk> Eddi|zuHause: just look at ELF
10:56:21 <cjk> or rather, the info page entry for libtool about binary numbering
10:56:25 <Rubidium> cause when the interface doesn't change a codechange can still cause desyncs
10:56:28 <cjk> i can paste it to a bin if you like
10:56:34 <Eddi|zuHause> cjk: for network play, every client has to have the exact same codebase, not the same interface
10:57:05 <Noldo> cjk sounds like the young me
10:57:17 <cjk> For example I changed the industries table and people are successfully connecting with 0.6.0
10:57:27 <cjk> it's not the "exact" code base, just an approximation
10:57:48 <Eddi|zuHause> every operation in a network game must yield the exact same result
10:58:14 <cjk> but production rates are specified by the map, so I guess that is safe
10:58:19 <Eddi|zuHause> there are some parts that are multiplayer agnostic, like stuff that is GUI-only
10:58:27 <Eddi|zuHause> or stuff that happens before game creation
10:58:40 <Eddi|zuHause> but those are very limited
10:59:00 <cjk> explains why I could not limit the destruction field (landscape.cpp ClearArea)
10:59:42 <cjk> Rubidium: it does not change the interface per se, but if it is not synchronized, it can lead to problems
11:00:07 <Noldo> should it bump version?
11:00:53 <Eddi|zuHause> [newgrf-static] section is multiplayer-agnostic
11:01:17 <Eddi|zuHause> it is restricted to grfs that basically only do sprite replacement
11:01:46 <Eddi|zuHause> like catenary, or trees
11:01:57 <Rubidium> cjk: so you want us to manually update some number for like 80% of the commits?
11:02:02 <Eddi|zuHause> stuff that has no gameplay value
11:02:09 <cjk> well I am sure you can automate that :p
11:02:42 <Eddi|zuHause> it gets automatically generated from rev.cpp.in ;)
11:02:44 <cjk> svn/git hook, but I guess that closes the loop of going back to using the rev number
11:03:18 <SmatZ> maybe if git/hg could update some number somewhere to svn revision while updating...
11:03:34 <Rubidium> cjk: problem with git is that it is (for me) almost impossible to tell whether a git checkout is not tainted by a non-trunk-commit to it
11:03:59 <Rubidium> or do you want a script to go over 10000+ commits each compile cycle?
11:04:30 <cjk> why, you just have to check if the top commit is a trunk or not
11:04:52 <Rubidium> no, because then there could be other commits with changes that are not in trunk in that repository
11:04:54 <SmatZ> sorry, I don't know git...
11:05:00 <Rubidium> that would make it at least XXXXM
11:05:47 <cjk> something along the lines of if `git-rev-parse HEAD` != `git-merge-base trunk/master HEAD`
11:07:33 <cjk> if [ `git-rev-parse HEAD` == `git-merge-base trunk HEAD` ]; then echo I am an official revision; fi;
11:10:30 <Rubidium> argh... git is slooooow
11:10:57 <cjk> but downloading revisions with svn is slower
11:11:13 <cjk> hg is in python, it can't be
11:12:14 <Alberth> cjk: only if you have a really fast FS and a slow CPU
11:13:16 <Rubidium> hg clone takes... 30 secs, git clone 5-ish minutes
11:13:27 <cjk> depends on what you clone and from what server
11:14:42 <cjk> Receiving objects: 100% (38825/38825), 17.93 MiB | 6155 KiB/s, done.
11:14:56 <Alberth> Almost all VCS operations have bottle-neck in IO and not in CPU. It really doesn't matter where you program it in.
11:15:12 <cjk> see, my clone was even faster than your hg.
11:15:22 *** buffoon has joined #openttd
11:15:40 <Alberth> you should look at wall-clock time, not CPU seconds
11:15:50 <cjk> this *is* the wall clock time.
11:16:25 <Alberth> sorry for the confusion
11:18:00 <buffoon> hi! since the update to 0.6.0 (tried also 0.6.1 RC1) i have several sync errors causing disconnections with my friend. It is a Debian Linux and a Windows XP System (tried launching the server on both, normally on the debian machine). I didn't find anything on the BTS so i do you have some suggestions?
11:18:07 <cjk> perhaps git just takes longer because the compression is more efficient
11:24:21 <peter1138> buffoon: yeah, wait for RC2
11:24:55 <buffoon> peter1138: ah, ok... are there bugfixes planned?
11:26:31 <Eddi|zuHause> buffoon: most known bugs are already fixed in trunk/nightlies, they need backporting to 0.6 and then a release of RC2
11:27:18 <buffoon> thx to all developers of openttd btw... Great Job ;)
11:27:59 <Eddi|zuHause> you can try running a nightly server, but there are new features which might decrease stability
11:31:20 <peter1138> but hydrairc's crap
11:32:07 <buffoon> Hmm, I prefer to wait for RC2 then.
11:33:20 *** Ridayah has joined #openttd
11:43:35 *** GoneWacko has joined #openttd
11:43:52 <Eddi|zuHause> well, new features are usually totally awesome ;)
11:44:10 <Noldo> just because they are new
11:44:54 <Eddi|zuHause> well, how many statistics about introducing old features have you studied?
12:13:01 <cjk> is there a way to define waypoints for trams?
12:15:57 <SmatZ> build a drive-thru station, set order "Go via"
12:18:43 *** DaleStan_ has joined #openttd
12:18:43 *** DaleStan is now known as Guest570
12:18:43 *** DaleStan_ is now known as DaleStan
12:20:54 *** DaleStan is now known as Guest571
12:20:55 *** DaleStan has joined #openttd
12:44:57 *** Singaporekid has joined #openttd
13:32:41 *** thgergo has joined #openttd
13:51:02 <Belugas> [08:11] <cjk> is there a way to define waypoints for trams? <-- i read "farms" for a second...
13:56:10 <Lakie> But trams doesn't look like farms... and would a farm need a waypoint? -_-
13:56:46 <Belugas> thus me standing there and scratching head...
13:57:02 <Belugas> maybe he wants to use the farm as a waypoint?
13:57:21 <Belugas> not enough coffee yet in system
14:12:12 *** stillunknown has joined #openttd
14:52:54 *** GoneWacko has joined #openttd
14:55:51 <CIA-3> OpenTTD: glx * r13060 /trunk/src/ (industry_gui.cpp settings.cpp): -Codechange: update build industry window when raw_industry_construction setting is modified
15:20:35 *** |Jeroen| has joined #openttd
16:28:44 *** Boyinblue0 has joined #openttd
16:52:49 *** HerzogDeXtEr has joined #openttd
17:02:30 *** dR3x4cK has joined #openttd
17:07:14 <Ammler> Wouldn't it make sense to include ottd_grf to trunk/src and to make it on compiling like the language files?
17:08:04 <Ammler> hmm, I guess, nforenum and grfcodec would be the problem... :-/
17:19:05 <nicfer> anyone from here knows action 52?
17:19:34 <Rubidium> are there already that many?
17:20:18 <Ammler> no :-), nicfer, you do not mean NewGRF Actions?
17:21:15 <nicfer> action 52, oh well, it's for creating a graphic set with atari 2600 quality
17:22:56 <nicfer> well, it was a joke, action 52 is actually a horrendous collection of crappy games that you never forget from the moment you play it
17:29:22 <Roujin> okay it's holiday AND the nice devs are busy with window OO'ifying, so I won't nag about saying something about my drag&draw patch...
17:29:53 <Roujin> *trying to use reverse psychology*
17:38:09 *** ChanServ sets mode: +o Bjarni
17:43:54 *** Chicago_Rail_Authority has joined #openttd
17:53:00 <Belugas> new Nine Inch Nail ALBUM!!!
17:53:31 <Chicago_Rail_Authority> online?
17:53:57 <Chicago_Rail_Authority> <---- clicks...
17:56:07 <Prof_Frink> [18:56:07] <--- clocks
17:56:35 <Chicago_Rail_Authority> clicks again?
17:58:27 <Chicago_Rail_Authority> * kicks his lousy Comcast connection
17:58:36 * Belugas drags in WinAmp and byebye Queensrych
17:58:50 <Chicago_Rail_Authority> stupid IRC q -- how do you do the blue text with the "*" ?
17:59:30 * Prof_Frink wonders what you mean
17:59:42 <Chicago_Rail_Authority> exactly that
17:59:56 <Prof_Frink> It's all /me /me /me , isn't it.
18:00:27 <Chicago_Rail_Authority> pretty much
18:03:40 *** Chicago_Rail_Authority has quit IRC
18:06:35 *** Boyinblue0 has joined #openttd
18:07:26 *** Chicago_Rail_Authority has joined #openttd
18:15:04 * Chicago_Rail_Authority Finally gets the "it's all about /me" joke from earlier...
18:15:15 * Chicago_Rail_Authority ....mumble...
18:34:32 <Bjarni> ln: I know it's the language you use to pick up women but there aren't any women here
18:35:11 <Bjarni> it was a statement that they aren't here
18:35:38 <Bjarni> last time I saw a woman on IRC it turned out to be a guy
18:35:48 <Bjarni> but I guess that is to be expected
18:36:22 <Belugas> ln is in a linguistic identity crisis
18:39:50 <DorpsGek> Sacro: Karen_ was last seen in #openttd 1 week, 0 days, 21 hours, 57 minutes, and 3 seconds ago: <Karen_> jawohl!
18:40:25 <Bjarni> <Belugas> ln is in a linguistic identity crisis <-- he should talk about it in #someonewhocares
18:40:41 <Belugas> i've heard that the named Karen_ entity can be found on #openttd.riddle
18:40:56 <Eddi|zuHause> <Belugas> ln is in a linguistic identity crisis <- you mean лн ;)
18:41:22 <Belugas> if you wish, Eddi|zuHause :)
18:41:45 <Belugas> hooo.... Corona Radiata seems promising :D
18:45:07 <Bjarni> <Eddi|zuHause> <Belugas> ln is in a linguistic identity crisis <- you mean лн ;) <-- if we start to define terms for him then this one could do as well: 言語能力のアイデンティティクライシス
18:48:11 <Eddi|zuHause> "Die Fähigkeit zur Nutzung der Sprache Identitätskrise" <- that does not make too much sense...
18:48:21 <Belugas> The Four of Us are Dying : MESMERIZING!
18:49:29 <Eddi|zuHause> # Yes, yes, yaw, yaw! I'm a hell of a man!
18:49:29 <Eddi|zuHause> # I'm walkin', walkin' down the street and hear you, hear you saying: "Damn!
18:49:29 <Eddi|zuHause> # I wanna, I wanna be like Smudo, too,
18:49:29 <Eddi|zuHause> # makin', makin' rap records, saying, saying 'Fuck you'!"
18:50:27 <Bjarni> thinking about it I guess the sentence should actually be: アイデンティティクライシスの言語能力
18:50:45 <Bjarni> not that it would matter to most of you :P
18:52:50 <Bjarni> did you try the test from Crocodile Dundee?
18:54:31 *** blathijs has joined #openttd
18:55:10 <Bjarni> and something happened afterwards
18:55:21 <Bjarni> Wolf01: are you able to walk again?
18:55:45 <Eddi|zuHause> (such a great song)
18:56:05 <Sacro> Eddi|zuHause: isn't that an off topic youtube link?
18:56:13 <Eddi|zuHause> no, it is on topic
18:56:42 <Eddi|zuHause> i have talked about the song above, and bjarni talked about women that are not women
18:56:50 <Eddi|zuHause> it's double-topic ;)
18:57:34 <Bjarni> which means it overflows and makes it very off topic :P
19:01:00 <Eddi|zuHause> (same band, different decade)
19:01:14 <Eddi|zuHause> that's basically the song they got famous with
19:01:21 <Bjarni> Sacro: does that video reflect your view on Germans?
19:01:53 <Belugas> ho... dancing monkeys...
19:02:15 *** mynetdude|Away is now known as mynetdude
19:02:22 <Eddi|zuHause> they have a trabbi in the video ;)
19:02:28 <Eddi|zuHause> those are antique by now :p
19:02:39 <Sacro> Bjarni: no, tis just a catchy tune
19:05:09 <Eddi|zuHause> damn, i don't have enough bandwidth for that much youtube...
19:05:53 <Eddi|zuHause> that is hardly a german that sings there :p
19:06:21 <Bjarni> it's voice synth generated
19:06:48 <Bjarni> and it's better to listen to than your links :P
19:07:06 <Eddi|zuHause> actually, the "queen of the night" should represent a really mad woman bitching around :p
19:08:36 *** Brianetta has joined #openttd
19:09:37 <Sacro> one, two, three, swanky swanky pigs
19:12:40 <Bjarni> can you guys explain the election in Serbia. The pro EU guys got 123 seats while the nationalists got 127 seats yet the pro EU guys have been declared the winner
19:12:59 <Eddi|zuHause> <Bjarni> you pay for each MB? <- no, that would be a traffic limit... i just have really low bandwidth (low enough that i can't watch youtube videos in "real time")
19:13:38 <Eddi|zuHause> Bjarni: they got that from the election in Kongo ;)
19:14:13 <Eddi|zuHause> i didn't even know there was an election in serbia
19:15:30 <Eddi|zuHause> i'm sure you do, but please spare us the details :p
19:16:58 <Bjarni> Sacro: we don't want to hear the result of your "use right or left hand" election
19:17:52 <Bjarni> how can you have that election at all? I mean all the voters would show up after the time you need your result
19:18:59 <Bjarni> now I think I know how they declared the winner in Serbia
19:19:34 <Bjarni> it would appear that election observers look at who has the largest party rather than who has 50+% of the votes
19:19:59 *** Eddi|zuHause2 has joined #openttd
19:20:16 <Bjarni> that would work in USA and other countries with only two parties
19:20:23 <Prof_Frink> Now there's a good idea
19:20:36 <Prof_Frink> Have elections based on which party can throw the best one.
19:23:04 <Bjarni> according to EU the goal is to gain as much power to EU as possible
19:54:40 *** Nite_Owl has joined #openttd
19:56:10 *** ChanServ sets mode: +v tokai
19:59:06 *** sickie88 has joined #openttd
20:20:18 *** dasprid__ has joined #openttd
20:21:23 *** Osai is now known as Osai`off
20:22:15 *** Nite_Owl has joined #openttd
20:22:56 *** Osai`off is now known as Osai
20:38:10 *** Zealotus has joined #openttd
20:46:59 *** lobster_MB has joined #openttd
21:08:07 *** sickie88 has joined #openttd
21:24:45 *** dR3x4cK has joined #openttd
21:29:57 *** dR3x4cK2313 has joined #openttd
21:45:53 *** Osai is now known as Osai^zZz
21:46:21 <CIA-3> OpenTTD: glx * r13061 /trunk/src/ (gui.h toolbar_gui.cpp tree_gui.cpp): -Codechange: make a class of the BuildTreesWindow.
22:09:30 *** Osai^zZz is now known as Osai
22:23:59 *** christian has joined #openttd
22:29:29 *** Osai is now known as Osai`off
22:37:53 *** SmoovTruck has joined #openttd
22:47:36 *** izhirahider has joined #openttd
22:57:07 *** blathijs has joined #openttd
23:21:44 *** Brianetta has joined #openttd
23:26:22 *** Singaporekid has joined #openttd
23:48:49 *** Chicago_Rail_Authority has left #openttd
23:49:26 *** Phoenix_the_II has quit IRC
continue to next day ⏵