IRC logs for #openttd on OFTC at 2008-10-24
⏴ go to previous day
00:00:12 * Belugas just gave a test run on OpenGfx
00:00:22 <Belugas> one comment : why is it so dark??
00:00:43 <Eddi|zuHause> nobody could ever properly answer that question...
00:00:43 <Belugas> another comment : hell of a job been done!
00:01:04 <Belugas> ln, not in OpenTTD. and this is stupid
00:02:00 <Eddi|zuHause> ln: well, that is one explanation... others involved people in foreign countries, or foreign people in germany, asking for something that could mean direction to the station
00:02:45 <Eddi|zuHause> basically, any time, someone asks something that you can't understand, you give him direction to the staion, as that's the most likely place he would get what he wants
00:08:00 <ln> this nerd guy in prison break is damn annoying.
00:08:41 <Eddi|zuHause> i have not actually watched prison break
00:09:02 <Eddi|zuHause> i once recorded season 1, but i somehow lost it...
00:09:09 <ben_goodger> can anyone advise me about pan-european value added tax law? :P
00:09:15 *** welshdragon has joined #openttd
00:09:45 <Eddi|zuHause> Mehrwertsteuer...
00:09:51 <welshdragon> whoops... crashed xchat :(
00:10:19 <Eddi|zuHause> in germany, it is illegal to give legal advise if you are not a certified lawyer
00:11:03 <ln> Eddi|zuHause: so you can't say "you should call the police" when someone breaks your neighbor's windows?
00:11:27 <Eddi|zuHause> ln: that's not legal advise
00:11:43 <eMJay> Belugas: I gave it a shot just now, the colours are very saturated I think
00:11:46 <ben_goodger> well, I don't want actual legal advice
00:11:53 <eMJay> Belugas: OpenGFX that is
00:12:24 <ben_goodger> my accountants are stumped, however :(
00:12:50 <ln> Eddi|zuHause: oreally. :) hm, but is that like ... if you do give legal advice, the one you adviced can sue you, or the bundesrepublik can sue you?
00:13:23 <Eddi|zuHause> yeah, something like that...
00:13:36 <eMJay> ln: I would assume that the one you gave advice could sue you
00:13:40 *** [com]buster has joined #openttd
00:14:18 <Eddi|zuHause> eMJay: not only that, also someone witnessing that action can sue, or call the police...
00:16:26 <ln> i don't know the legal terms in english, but there are basically two kinds of crimes over here; ones where the victim has to actively do something to get the suspect sued (such as theft), and ones where the government will sue the criminal regardless of what the victim wants (e.g. murder)
00:16:53 <ben_goodger> suing isn't related to crime here
00:16:54 <ln> errr.. murder is not a particularly good example of *what the victim wants*
00:16:58 <eMJay> when you say "sue" do you mean "prosecute"?
00:17:02 <ben_goodger> things you sue for are called torts
00:17:10 <ln> actually yes, i probably mean "prosecute".
00:17:23 <ben_goodger> I have no idea what our terms for those are
00:17:56 <Eddi|zuHause> well, it's actually not much different from other laws, like you have to be a certified plumber to fix the plumbing
00:18:27 <Eddi|zuHause> or at least be employed by or employ one
00:18:38 <ben_goodger> ...but does it actually criminalise it or just make you liable for it?
00:19:20 <ben_goodger> you can be sued for giving faulty advice here [liability], but you can be fined for giving driving instruction for money without a license [crime]
00:19:27 <eMJay> In OpenTTD related discussion, what does sample.cat contain?
00:19:47 <Eddi|zuHause> eMJay: the sound effects
00:20:03 <Eddi|zuHause> like the train whistle and stuff
00:20:37 <Eddi|zuHause> do you want it to be more?
00:20:53 <eMJay> no, just wanted to be sure
00:21:34 <eMJay> so sample.cat contains the sounds for [open]ttd
00:21:51 <ben_goodger> actually, the opposite of that usage of "right" is "I don't understand"
00:22:34 <Eddi|zuHause> you can safely assume that i used the "wrong" usage on purpose :p
00:23:02 <ben_goodger> very well, carry on
00:23:19 <Eddi|zuHause> Bjarni is sober? since when?
00:23:47 <Eddi|zuHause> or do you mean "saufen", which is kind of the opposite? :p
00:24:06 <ln> Eddi|zuHause: ich meine "sova = schlafen"
00:31:13 *** Digitalfox has joined #openttd
00:33:06 *** Eddi|zuHause has joined #openttd
00:33:48 <Eddi|zuHause> no german would ever say that :p
00:34:37 <Eddi|zuHause> in the best case, they would say "webää" :p
00:38:01 *** thingwath has joined #openttd
00:42:09 <ln> now let's see, the international openttd meeting yet to be arranged, isn't it?
00:42:46 <Sacro> lets have it in England
00:43:02 <ln> Sacro: but Bjarni doesn't travel to England.
00:43:45 <ln> well, england's fine for me.
00:46:11 <Bjarni> <ln> Bjarni: will you arrange an international openttd meeting in copenhagen? <-- no. I wouldn't know where it could be placed
00:46:41 <Bjarni> there are some ok places it could be, but they all cost money :(
00:47:42 <Bjarni> to be honest I wouldn't know there to host such a thing
00:47:57 * Sacro offers the use of his living room
00:48:02 <Bjarni> maybe in the train repair shop... it's big enough
00:48:13 <Bjarni> but it lacks both computers and an internet connection
00:48:23 <ln> who needs computers and internet connection
00:48:29 <eMJay> Is there any use of strings or templates in openttd?
00:48:42 <Sacro> you can't really output without strings
00:49:01 <Bjarni> then what is the meeting about?
00:49:04 <eMJay> i mean string as in the class, not char*
00:49:10 <ln> Bjarni: will you travel to england or germanyland for such a meeting?
00:49:29 <Bjarni> I'm not convinced I would walk to the house next door yet
00:49:33 <ln> Bjarni: it's about meeting the people behind those nicks.
00:49:40 <Bjarni> since I haven't got a clue to what you are trying to do
00:50:10 <Bjarni> I know what Sacro intends to do
00:50:34 <Bjarni> I don't think I'm willing to leave the country
00:50:35 <Sacro> Bjarni: lots of cuddles
00:50:42 <Bjarni> at least not at the moment
00:51:37 <ln> Bjarni: surely you have time to leave the country for 2 or 3 days.
00:52:30 <ln> dunno. Q1/Q2 next year, or so.
00:53:01 <Bjarni> but at least I can't this year
00:53:13 *** ChanServ sets mode: +v tokai
00:57:04 <ln> let's meet at Tron's front yard but let's not tell him in advance.
00:57:06 <eMJay> Ok, so are there any generics/templates in use in openttd?
00:57:21 <eMJay> like vector or hashmap or something?
00:57:35 <Bjarni> ln: while it could be cool to do, I would prefer to survive
00:58:02 <Bjarni> and in order to survive, I need sleep
00:58:27 <Sacro> ln: what is "what time is it" in french?
00:58:45 <Sacro> but I'm sure uhr is german
00:58:52 <ln> sounds a mixture of german and french
00:59:27 <ln> well, doesn't matter, probably the correc question is pronounced the same way in any case.
01:05:42 <ln> feel free to adjust spelling to what seems best.
01:06:04 <Sacro> 273 6/1/44 60 Clerk instructed to write to the Commanding Officer of the U.S. Forces stationed in the village drawing attention to the excessive speed with which the coloured troops drove and failure to comply to traffic signals.
01:15:50 *** [alt]buster has joined #openttd
01:15:53 *** [alt]buster is now known as [com]buster
01:20:37 <welshdragon> bloody impossible to build a machine, once you've done it, post a screenie
01:29:45 * Sacro watches Woof on youtube
01:45:08 <eMJay> welshdragon: I finished higher
02:03:20 *** TrueBrain has joined #openttd
02:13:17 *** Phoenix_the_II has quit IRC
02:17:41 *** [alt]buster has joined #openttd
02:17:43 *** [alt]buster is now known as [com]buster
02:37:01 <Belugas> eMJay, i don't know if it's the color saturation, but yes, it's darkish
02:37:52 <NukeBuster> you are still able to turn trains while the game is paused...
02:39:44 *** DJGummikuh has joined #openttd
02:40:12 <NukeBuster> wouldn't that lead to desyncs when one does that while a new player is joining?
02:42:19 <NukeBuster> if you'd turn it on the right moment... would that matter for the game?
02:42:21 <Belugas> i think (correct if i'm wrong) that commands are still passed
02:42:29 <Belugas> try it, you will see :)
02:42:36 <NukeBuster> also to the incoming client?
02:42:51 <NukeBuster> next time someone joins i'll turn a train around
02:44:13 <NukeBuster> probably easy to fix?
02:48:20 <Belugas> truely, i have no clue.
02:48:30 <Belugas> i'm not in networking logic
02:48:51 <Belugas> but, to say if it's easy to fix, one has to prove that it is broken ;)
02:52:04 <glx> IIRC commands are queued during the download
02:53:35 <Belugas> mmh... true... i think i remember that
02:58:20 *** Brianetta has joined #openttd
02:59:43 *** [alt]buster has joined #openttd
02:59:59 *** [alt]buster is now known as [com]buster
03:03:15 <Belugas> and if it takes too long, the game cannot empty the queue fast enough and you're rejected
03:03:18 <Belugas> or something like that
03:07:46 <eMJay> NukeBuster: I think you would have to be damn fast...
03:22:24 *** [alt]buster has joined #openttd
03:22:27 *** [alt]buster is now known as [com]buster
04:03:27 *** [alt]buster has joined #openttd
04:03:30 *** [alt]buster is now known as [com]buster
04:05:54 *** roboboy has joined #openttd
04:32:38 *** TrueBrain has joined #openttd
04:45:36 *** [alt]buster has joined #openttd
04:45:39 *** [alt]buster is now known as [com]buster
04:45:43 <eMJay> hmmmm, I hope my boss doesn't mind me trawling through the openttd code instead of coding for him all day :P
04:50:05 <eMJay> unfortunately, if he knew, he probably would mind
04:53:16 <eMJay> segmentation fault is when I try to access memory that isn't allocated to my program right?
05:11:43 *** roboboy has joined #openttd
05:27:38 *** [com]buster has joined #openttd
05:29:05 *** mikegrb has joined #openttd
05:36:53 *** De_Ghosty has joined #openttd
06:37:21 *** De_Ghosty has joined #openttd
06:45:37 *** De_Ghosty has joined #openttd
06:49:06 *** Xerres is now known as Jaws
07:10:00 *** De_Ghosty has joined #openttd
07:13:53 *** Yeggzzz is now known as Yeggs-work
07:54:53 <dih> i just realized something
07:55:14 <dih> as it's not called 'patches' but 'settings'
07:55:23 <dih> would it not make sense to change some console commands?
07:55:38 <dih> e.g. list_patches, patch <setting> [<value>]
07:55:53 <dih> list_settings and set <setting> [<value>]
07:56:25 <dih> kinda some consistency missing in that place :-D
07:58:34 *** Brianetta has joined #openttd
08:01:02 *** Celestar has joined #openttd
08:01:02 *** ChanServ sets mode: +o Celestar
08:04:03 *** Vikthor has joined #openttd
08:05:58 * Celestar can't help but notice that his 2.5" external USB driver manages to choke the 16-disk Fibre-Channel RAID6 array of his fileserver.
08:40:55 <Gekz> slow it down unbearably I'd imagine
08:43:11 <Gekz> I CAN TRANSFER AT THE SPEED OF LIGHT
08:43:19 <Gekz> and the USB is like "Slow down, I'm old"
08:43:45 <Eddi|zuHause> "'n alter Mann ist doch kein D-Zug"
09:10:28 <Eddi|zuHause> why the hell, when i search in google for "newgraphicsspecs", i get any page, except the newgraphicsspecs wiki page?
09:16:38 <Celestar> Gekz: the thing is, it's the other way round. the USB drive is fast, the RAID seems slow
09:27:41 *** [com]buster has joined #openttd
09:37:51 *** NukeBuster|laptop has joined #openttd
09:38:19 <NukeBuster|laptop> What information do you need for debugging a desync error?
09:38:36 <NukeBuster|laptop> At least a save, what else?
09:40:03 <Eddi|zuHause> desync errors are the most evil to debug, as it is necessary for server and clients to join at different times
09:40:24 <Eddi|zuHause> so complete instructions to reproduce are a must
09:40:44 <NukeBuster|laptop> dih's nightly server is currently constantly desynching
09:40:49 <NukeBuster|laptop> got a save game already
09:40:58 <NukeBuster|laptop> and when you join instant desync
09:41:08 <Eddi|zuHause> like "start the server, let the game run for exactly 10 days, then start the client, wait for exactly 1 day, and you see a desync"
09:41:26 <NukeBuster|laptop> hmm should try that with the save when I'm home...
09:41:29 <Eddi|zuHause> joining a running server will not help
09:41:49 <NukeBuster|laptop> you need to compare debug data?
09:41:50 <Eddi|zuHause> you need to reproduce it by starting a new server, and then desyncing
09:42:11 <NukeBuster|laptop> loading a game won't work?
09:42:47 <Eddi|zuHause> loading a game can work, but you have to let the server run without client until the desync reason occurs
09:43:31 <Celestar> NukeBuster|laptop: with or withouth cargodest?
09:44:14 <Celestar> NukeBuster|laptop: there's quite some information on the wiki on how to debug a desync problem
09:45:02 <NukeBuster|laptop> latest nightly
09:45:12 <NukeBuster|laptop> shouldn't have cargodest
09:45:18 <NukeBuster|laptop> or didn't have it turned on
09:45:30 <Eddi|zuHause> NukeBuster|laptop: the problem is, to debug a desync, you need to restart the server many times over with the same savegame, and the desync must happen always at the same time
09:46:02 <NukeBuster|laptop> are all actions timestamped?
09:46:36 <NukeBuster|laptop> or tick stamped perhaps?
09:48:55 <Celestar> normally, they're not recorded at all (=
09:49:04 <Celestar> NukeBuster|laptop: the first thing you need to understand is what a desync is
09:50:06 <Celestar> so NukeBuster|laptop , what is a desync?
09:51:59 <NukeBuster|laptop> a desync is a client that does something else then what the server does
09:52:18 <NukeBuster|laptop> some where in time the client took a wrong turn the
09:53:01 <NukeBuster|laptop> so i thought to compare actions and check for differences
09:54:54 <Celestar> NukeBuster|laptop: a desync is when the random number generator on the client returns a different result than the one on the server.
09:55:17 <Celestar> thus causing the game states to diverge
09:55:50 <NukeBuster|laptop> so the random seeds are usually given out by the server?
09:56:05 <Eddi|zuHause> yes, every 100 ticks or so
09:56:17 <Celestar> Eddi|zuHause: they're compared every so-and-so ticks
09:56:25 <NukeBuster|laptop> so in a 100 tick timespan a desync could occur
09:57:00 <Eddi|zuHause> no, the reason for the desync is often very long before the desync actually occurs
09:57:15 <NukeBuster|laptop> what if you would simply rejoin the game on a desync
09:57:15 <Celestar> but can be milliseconds ago.
09:57:26 <NukeBuster|laptop> or would that be annoying?
09:57:26 <Celestar> NukeBuster|laptop: on join, the seeds are exchanged
09:57:43 <Celestar> NukeBuster|laptop: we've opted not to do this, because we'd not get any desync reports then
09:58:02 <Celestar> which needs to be fixed
09:58:09 <Celestar> rejoining is a workaround, not a solution
09:58:11 <NukeBuster|laptop> yeah thats true
09:58:21 <Eddi|zuHause> many desyncs cannot be resovled by just rejoining the server
09:58:33 <NukeBuster|laptop> lets hope dih saves seperate autosaves :)
09:58:54 <NukeBuster|laptop> its probably due to 1000 trians
09:59:27 <NukeBuster|laptop> and signals every tile
10:00:09 <Celestar> if the client or the server is overwhelmed, that would cause a disconnect
10:00:44 <NukeBuster|laptop> what about the ability to turnaround trains while in pause?
10:01:00 <NukeBuster|laptop> there were alot of traffic jams on the server
10:01:03 <NukeBuster|laptop> would that matter
10:01:25 <Celestar> it shouldn't cause the gamestates to differ
10:01:29 <Eddi|zuHause> "it shouldn't" is why it's a bug ;)
10:01:35 <NukeBuster|laptop> if someone joins and in pause trains turn around because they waited to long for a signal
10:01:39 <NukeBuster|laptop> and not 1 but alot
10:01:58 <Celestar> then the game states on all clients (and the server) should be identical
10:02:06 <Eddi|zuHause> NukeBuster|laptop: throwing around wild theories won't help debugging
10:02:24 <NukeBuster|laptop> I know, but I just wanted to know if that could lead to a desync
10:02:39 <NukeBuster|laptop> if it happened on join
10:02:47 <Eddi|zuHause> everything could lead to a desync
10:03:13 <Eddi|zuHause> no, desyncs that happen on join are due to the game state not saved properly
10:03:40 <NukeBuster|laptop> and once you have a poisoned server it will remain that way?
10:03:48 <Eddi|zuHause> i.e. the server has some information that is not saved, and when the client loads that savegame, the game state is different
10:04:06 <Eddi|zuHause> yes, until the server itself loads its own savegame
10:04:29 <NukeBuster|laptop> does that happen often?
10:04:36 <NukeBuster|laptop> and could one force that?
10:04:44 <Eddi|zuHause> the server owner can do that
10:04:52 <Celestar> NukeBuster|laptop: the server is never poinsoned per definition. the server is law, all clients must have the same state as the server.
10:05:28 <NukeBuster|laptop> choose the wrong word for it as said the server save was incorrect
10:06:23 <Celestar> the savegame might miss some information, or some cache (i.e. non-saved data that can be reconstructed) is faulty, or marked clean while it isn't.
10:06:45 <Celestar> or: some buffer overflow might have occured overwriting variables
10:08:52 <Gekz> at some time in the past
10:09:34 <Celestar> better than being kicked in the future :P
10:10:51 <Eddi|zuHause> well, you could go back into the past to find your kicker and then kick him yourself :p
10:10:52 <Gekz> Celestar: now I'm superstitious
10:11:13 *** Gekz was kicked by petern (gekz)
10:11:19 *** Gekz was kicked by petern (gekz)
10:11:26 *** Gekz was kicked by petern (gekz)
10:12:21 *** mucht_work has joined #openttd
11:19:39 *** NukeBuster|laptop has quit IRC
11:22:36 *** Progman has joined #openttd
11:25:09 *** FloSoft has joined #openttd
11:25:25 <FloSoft> hmm path-based presignals do not work correctly?!?
11:26:56 <Rubidium> there's no concept of path based presignals
11:27:04 <Rubidium> not in the world of OpenTTD at least
11:30:21 <FloSoft> hmm why are they placeable then?
11:32:02 <Rubidium> because they might be useable in some (obscure) situations
11:32:45 <FloSoft> okay? in which situations are they usable? if they dont work?
11:33:15 <Brianetta> FloSoft: Describe the signal you believe to be a path based presignal
11:33:38 <Brianetta> we'll then be able to tell you what it actually is.
11:34:06 <Rubidium> FloSoft: a pbs junction followed by a priority merger
11:34:23 <FloSoft> aah okay, the one with the yellow "presignal" sign on it
11:35:05 <petern> that's probably a one-way pbs signal
11:35:15 <Brianetta> It is if it has three lights
11:35:17 <FloSoft> okay i read the tooltip
11:35:26 <Brianetta> if it has only two, it's a non-path presignal
11:36:17 <FloSoft> its not very intuitive, i thought its a path-based presignal
11:36:27 <FloSoft> because of the yellow sign
11:36:35 <dih> wel - appart of the word 'pre' it's true :-P
11:38:22 *** welshdragon has joined #openttd
11:38:33 *** grumbel has joined #openttd
11:39:59 <Rubidium> FloSoft: then come up with a better design for the signal ;)
11:49:18 <FloSoft> green sign instead of a yellow one?
11:58:48 <FloSoft> FauxFaux: yea but its not intuitive
11:58:58 <FloSoft> you see yellow sign -> "ah presignal"
12:03:10 <fjb> Be aware of the tab completion. It is the dark side of the IRC. Easy to use but doing odd things.
12:10:12 <dih> beware - the tab-completion is coming over you!
12:20:42 <FauxFaux> I used to MUD a lot, and the clients for that tended to tab complete words; I got rather lazy. ¬_¬
12:21:31 <Gekz> no need to spell words anymore
12:21:48 <Gekz> n tab ne tab tab to sp tab tab tab wo tab tab any tab tab
12:22:26 <keyweed_> no nero to splice womanly anyway?
12:23:33 <Brianetta> OpenOffice.org has enter to complete words
12:23:48 <Brianetta> which is so completely brain-dead that I stare at it in horror whenever I use a new install
12:30:47 *** HerzogDeXtEr has joined #openttd
12:49:20 <dih> Brianetta: do you know erlang?
12:56:19 <Celestar> Description: 4-year on site service. In Stock? No (delivery in 3-4 weeks).
12:57:26 <Celestar> on can "service" be in stock?
12:59:28 <Celestar> yeah and it takes 3-4 weeks to print a form?
13:04:53 <mikegrb> new delivery tech is due to be born in 3-4 weeks
13:20:55 <Brianetta> dih: big nope there.
13:31:51 *** Wezz6400 has joined #openttd
13:32:19 <dih> interesting - let me put it that way
13:42:28 *** stillunknown has joined #openttd
13:42:53 *** HerzogDeXtEr1 has joined #openttd
13:45:03 *** Aylomen has joined #openttd
14:04:42 <fjb> Is the improved time table patch worth a try?
14:06:10 <Celestar> rumor has it that it rocks
14:06:18 <Celestar> maybe Eddi|zuHause can tell you more
14:06:58 <Sacro> Celestar: i doubt it, Eddi|zuHause talks nowt but rubbish :p
14:09:55 <dih> you are in a room with 2 guards and 2 doors, one door leads you to freedom, the other door leads you to the lions. you also know that one of the guards always tells the truth, the other guard always lies - for both facts you know, you do not know which is the door to freedom nor which is the guard that lies... you may ask one guard one question (not each guard) - find your way into freedom :-)
14:11:54 <Celestar> dih: "What would your comrade say if you ask him which door leads to freedom"
14:16:15 *** De_Ghosty has joined #openttd
14:20:09 <Rubidium> now what if everything you "know" is told by the lieing guard?
14:25:29 <yorick> then it will be one lies and one tells the truth instead of one tells the truth and one lies?
14:26:51 <De_Ghosty> the liar will say yes
14:26:55 <De_Ghosty> the truth will say no
14:27:10 *** |Jeroen| has joined #openttd
14:28:23 <glx> both will point to the other door
14:30:01 *** XeryusTC is now known as Xeryus|bnc
14:31:02 <yorick> you can only ask one question ;)
14:31:29 <Celestar> hence you need to ask A a question that involves B
14:39:01 <Celestar> man building paraview is a pita :S
14:39:20 <Celestar> i've started about 5 hours ago
14:40:09 <Celestar> but as it won't build with qt4.4 (requires qt4.3) this involves a qt rebuild as well
14:40:50 <glx> does that mean all other stuff using qt4.4 qill break?
14:42:12 <Celestar> glx: no, I have QT4.4 installed in /usr and qt4.3 in /usr/local now
14:42:32 <Celestar> and I've got to reboot in a minute because I got a kernel update too
14:43:07 <Celestar> man why do computers in 2008 require more babysitting than 20 years ago :S
14:43:41 <keyweed_> more complexity, more possibility for fail
14:50:21 <CIA-5> OpenTTD: rubidium * r14524 /trunk/src/ (6 files in 2 dirs): -Add: stop-in-depot as part of orders (PhilSophus)
14:50:25 *** Celestar has joined #openttd
14:50:25 *** ChanServ sets mode: +o Celestar
14:50:40 <Celestar> at least the system has survived a kernel update
14:50:47 <Celestar> Linux rivendell 2.6.25.16-0.1-default #1 SMP 2008-08-21 00:34:25 +0200 i686 i686 i386 GNU/Linux
14:50:54 <Celestar> I wonder how long a version string may be :P
14:52:11 <Celestar> machine, processor, hardware platform
14:52:31 <petern> Linux lachesis 2.6.26-1-vserver-686 #1 SMP Wed Sep 10 19:51:11 UTC 2008 i686 GNU/Linux
14:53:11 <Celestar> Linux andromeda 2.6.16.21-0.8-smp #1 SMP Mon Jul 3 18:25:39 UTC 2006 x86_64 x86_64 x86_64 GNU/Linux
14:53:41 <Rubidium> playing the: who has the most ancient kernel game?
14:54:28 <Celestar> Rubidium: 2.6.16 is the kernel for SLES10
14:54:38 <Celestar> so my server is stuck with that
14:54:59 <Rubidium> "my" VPS is running 2.6.9-023stab046.2-enterprise ;(
14:55:37 <Celestar> can I change the current directory in irssi?
14:56:11 <Celestar> but not of the underlying shell
14:56:20 *** Celestar has joined #openttd
14:56:20 *** ChanServ sets mode: +o Celestar
14:56:30 <Eddi|zuHause> Sysinfo for 'johannes-i': Linux 2.6.25.11-0.1-default running KDE 3.5.9 "release 49.1", CPU: AMD Athlon at 1628 MHz (3260 bogomips), HD: 1762/1784GB, RAM: 996/1010MB, 186 proc's, 7.4d up
14:57:35 * Celestar ponders dumping the output of "hwinfo" into the channel
14:59:41 <Eddi|zuHause> yay for r14524 \o/
15:02:08 <Celestar> !openttd commit 14524
15:03:11 <Rubidium> Celestar: you NEVER learn, do you? At least in this instance of knowledge gathering you seem to be failing quite badly ;)
15:04:48 <Eddi|zuHause> @openttd commit 14524
15:04:48 <DorpsGek> Eddi|zuHause: Commit by rubidium :: r14524 /trunk/src (6 files in 2 dirs) (2008-10-24 14:49:45 UTC)
15:04:49 <DorpsGek> Eddi|zuHause: -Add: stop-in-depot as part of orders (PhilSophus)
15:05:06 <Celestar> Rubidium: hence the "sigh"
15:05:40 <Eddi|zuHause> or you could have just scrolled up two pages, if you didn't kill your client ;)
15:06:07 <NukeBuster> Eddi... I stopped the trains and desyncs stop
15:06:26 <NukeBuster> with the global stop all trains thingy
15:06:59 <NukeBuster> started the trains again and desyncs are back
15:07:02 <Eddi|zuHause> that only means that the desync reason does not surface anymore, not that it is gone
15:07:25 <Rubidium> Celestar: I like the Top Gear vehicles more ;)
15:07:46 <Eddi|zuHause> Celestar: looks very... italian...
15:07:49 <Celestar> it only has a range of 400m :P
15:07:52 <NukeBuster> as soon as i start all the trains again within 10 secs desync
15:08:30 <NukeBuster> now to turn on trains 1 by one
15:08:33 <Eddi|zuHause> NukeBuster: then start only single trains until you see a pattern
15:08:53 <NukeBuster> saved the game first
15:08:53 <Eddi|zuHause> then try to reproduce that pattern in a different server
15:09:25 <Sacro> Linux desktop 2.6.27-ARCH #1 SMP PREEMPT Fri Oct 17 07:35:10 UTC 2008 i686 AMD Athlon(tm) 64 X2 Dual Core Processor 6400+ AuthenticAMD GNU/Linux
15:09:40 <NukeBuster> have got exactly 300 trains
15:09:58 <Eddi|zuHause> Rubidium: i really hope that's the "before" picture :p
15:10:11 <NukeBuster> hmm... seems to be ok
15:10:43 <NukeBuster> they are all trying to get a free path at the same time that can't be good
15:12:37 <Celestar> nice find Eddi|zuHause
15:12:55 <Eddi|zuHause> "Höchstgeschwindigkeit: angeblich 100 km/h" :p
15:13:20 <Celestar> yeah ... with a decent kick in the butt, on a downward slope with tailwind
15:13:37 <Celestar> Eddi|zuHause: Any Police-Wartburg pictures? (=
15:13:42 <Eddi|zuHause> yeah, something like that ;)
15:13:49 <Rubidium> Celestar: a few inches behind a truck going 100 km/h ;)
15:14:23 <Celestar> I once got stopped by a hungarian police squad, driving an M5 :)
15:14:28 <Celestar> at around 280km/h or so
15:15:25 <Eddi|zuHause> "Son, do you know how fast you were driving?"
15:16:00 <Celestar> but I wasn't driving
15:16:25 <Celestar> my friend was "They won't catch us anyway at this speed"
15:16:36 <Celestar> and then they zoomed past us
15:16:54 <Celestar> and told us we were only to go 130km/h or something
15:17:01 <Celestar> and he wanted 50 EUR or so :P
15:18:08 <NukeBuster> What things influence the random seed from the client or server?
15:18:41 <Rubidium> almost everything except rendering the screen and playing the music
15:18:46 <Eddi|zuHause> the server gives out the random seed on client join, each random action is based on that seed, and yields a new seed
15:19:06 <Celestar> the seed is changed every time Random() is called, is it not?
15:19:44 <Celestar> well, I have to go anyway
15:19:55 <Rubidium> so do I, so I'm not going to wait on you ;)
15:20:22 <NukeBuster> halfway through starting the trains
15:26:22 *** frosch123 has joined #openttd
15:28:55 <NukeBuster> got 75% running again without desync
16:03:57 *** Brianetta has joined #openttd
16:28:14 *** Dred_furst has joined #openttd
16:33:15 <bravo> i got a little problem, i thought you could help me here ... i am trying to try cargodest branch ... went along instructions in openttd wiki, however, when i enter command hg clone (...) to obtain code, thing gets stucked at "adding file changes" ... i cant see no progress for some 10 minutes :( (currently using debian testing unstable) ... do you not know what should i do ?
16:34:58 <Eddi|zuHause> the initial checkout process takes quite a while
16:35:41 <Eddi|zuHause> hg does not just transfer a single checkout, it transfers the entire commit history
16:36:21 <bravo> thanks for explanation Eddi|zuHause
16:49:39 *** Frostregen has joined #openttd
16:51:07 *** avdg[BEnl] has joined #openttd
16:51:41 <avdg[BEnl]> i need some help with r14508
16:52:07 <avdg[BEnl]> i have an error while loading my execute
16:53:10 <bravo> avdg[BEnl], i am not an expert but it might help to say what kind of error
16:53:18 <bravo> avdg[BEnl], on what operating system you use
16:53:49 <bravo> avdg[BEnl], people dont have the crystal balls to see your screen
16:53:55 <ln> no, no, that would be going too much into details.
16:54:37 <avdg[BEnl]> error while loading shared libraries: libicui18n.so.38: cannot open shared object file: No such file or directory
16:54:56 <bravo> avdg[BEnl], you need to install that library oboviously :P
16:55:16 <avdg[BEnl]> why does my other versions works well?
16:55:36 <petern> it is part of support for rtl languages, which was only recently added
16:55:41 <Belugas> because it weas recently introduced
16:55:51 <avdg[BEnl]> hum... how can i get it
16:56:08 <NukeBuster> can we get a silly english mode too?
16:57:57 <avdg[BEnl]> :/ many problems... my windows is not under control :/
16:59:07 <Eddi|zuHause> avdg[BEnl]: look if you can find a library called "icu"
16:59:27 <avdg[BEnl]> the problem is that i am on a livecd
16:59:31 <Eddi|zuHause> in your favourite packet manager?
17:00:01 <frosch123> avdg[BEnl]: compile ottd yourself, it will disable icu then
17:00:48 <avdg[BEnl]> i guess there is no compile tool on it
17:01:17 <Eddi|zuHause> well then you should think about getting a real installation for your distribution
17:01:39 <avdg[BEnl]> my windows is hacked by scareware :/
17:02:10 <Eddi|zuHause> that's what they invented fdisk for :p
17:02:53 <Eddi|zuHause> i learned that i don't need windows for anything...
17:03:14 <Eddi|zuHause> wine has done amazing improvements in the past few years
17:03:15 <avdg[BEnl]> yeah, but my most programs will only work at windows
17:03:28 <avdg[BEnl]> i need to buy a big hd :/
17:03:48 <avdg[BEnl]> and a new version of kubuntu...
17:04:08 <avdg[BEnl]> need to try all linuxdistr again :/
17:04:43 <avdg[BEnl]> installed all icu... doesnt work
17:15:27 <avdg[BEnl]> is there no option to start without icu?
17:15:30 <Rubidium> avdg[BEnl]: maybe try the etch debian package instead of the lenny one?
17:18:44 <avdg[BEnl]> im using adept atm
17:20:26 *** grumbel has joined #openttd
17:20:32 <avdg[BEnl]> :/ no windows, and no openttd...
17:21:14 <avdg[BEnl]> i hate computers :p
17:24:24 <Rubidium> the debian-etch binary is linked against libicu 3.6
17:24:41 <Rubidium> so you could install that one instead of one of the other unixy binaries
17:25:14 *** welshdragon has joined #openttd
17:25:43 <avdg[BEnl]> :/ still hopeless, its getting from mad to very bad...
17:30:09 * Rubidium got the idea he isn't heard
17:31:20 <Rubidium> nothing important for me ;)
17:31:41 <avdg[BEnl]> i hate windows, i hate my old kubuntu :/
17:32:43 <Aali> ooh, stop in depot is in trunk now?
17:36:50 <Rubidium> fjb: looks too much like linux, I recommend Bob. Who can hate a dog?
17:40:58 <Eddi|zuHause> what's this "zeta" that they always advertise in some homeshopping tv shows?
17:45:01 <CIA-5> OpenTTD: translators * r14525 /trunk/src/lang/unfinished/indonesian.txt:
17:45:01 <CIA-5> OpenTTD: -Update: WebTranslator2 update to 2008-10-24 17:44:48
17:45:01 <CIA-5> OpenTTD: indonesian - 412 fixed by dnaftali (412)
17:46:13 *** sigmund_ has joined #openttd
17:47:48 <Eddi|zuHause> is that useful for anything?
17:48:24 <fjb> Yes, you can study moders OS design and some cool ideas.
17:49:05 <fjb> Sadly you don't get the source.
17:50:16 <Eddi|zuHause> i wonder if i should apply for a job at microsoft :p
17:50:51 <Eddi|zuHause> "rinn in die organistaion und von innen uffmischen"
17:50:51 <De_Ghosty> when is signal in tunnels!?!?1
17:50:59 <De_Ghosty> it is the holy grail of ottd
17:51:16 <petern> have you written it yet?
17:51:18 <Eddi|zuHause> De_Ghosty: stardate 129242.2, and no, it's not
17:51:48 <fjb> Eddi|zuHause: They are too big for that plan.
17:52:13 <De_Ghosty> thinking about it tho
17:52:16 <Belugas> signals just do not want to enter the tunnels. they are too big
17:52:28 <Belugas> i tried to push them in, but i borke the light
17:52:30 <petern> and scared of the dark
17:52:30 <fjb> De_Ghosty: And one of ! or ? per sentence is enough.
17:52:49 <De_Ghosty> more makes it comical
17:52:51 <Eddi|zuHause> except if you are spanish
17:52:52 <Belugas> so, i tried shortening the posts, the signla was not been transmitted
17:53:05 <Eddi|zuHause> then you are allowed for an additional ¡ or ¿
17:53:46 <De_Ghosty> is it possible to impliment rail/road building like roller coaster tycoon?
17:54:02 <De_Ghosty> you get a magic box that goes up and down it would solve so much
17:54:42 <Eddi|zuHause> you mean ignore all those people who bitch about locomotion for using that stupid roller coaster tycoon build system?
17:54:54 <NukeBuster> and let it act as a bridge going into the ground?
17:54:57 <Eddi|zuHause> not that i have any idea how that even works
17:55:41 <Eddi|zuHause> anyway, i don't think the build system has anything to do why signals in tunnels are not implemented
17:55:45 <Belugas> De_Ghosty, this is openTTD, not Open_CLONE_EVERYTHING
17:56:28 <De_Ghosty> we should makes some tanks too :D
17:56:55 <NukeBuster> why don't you just use trains to crash into each other...
17:56:59 <Eddi|zuHause> you can already do that, De_Ghosty.
17:57:25 <De_Ghosty> the city have no unpaved streets
17:57:25 <Belugas> hey, shold we kick/ban this warpig, guys?
17:58:09 <Eddi|zuHause> i'm afraid if i vote 'yes', it'll come back and haunt me. wouldn't be the first time :p
17:58:15 *** Yeggs-work is now known as Yeggstry
18:00:49 <Eddi|zuHause> that's like one of the worst wiki pages i have ever seen...
18:01:27 <NukeBuster> it does provide the answer
18:03:51 <Eddi|zuHause> to which question?
18:04:50 <NukeBuster> hmm.... maybe not a question but "[19:56] <De_Ghosty> we should makes some tanks too :D"
18:05:07 <Eddi|zuHause> there isn't even a 42 on that page
18:06:26 <Eddi|zuHause> "In OpenTTD we consider certain options that affect gameplay to be patches." <- somebody should update this page :p
18:07:57 <NukeBuster> Go ahead... its wiki :P
18:21:40 <NukeBuster> but wiki is usually editable
18:22:06 <Eddi|zuHause> fine, you didn't get it... alright
18:23:19 <Eddi|zuHause> think again about what i said, and what you said. if you can tell me what i meant, we can speak again ;)
18:24:14 <NukeBuster> I probably took it the wrong way as in owner. You probably ment who wiki is...
18:25:00 <Eddi|zuHause> no, i'd have said "who is wiki" then...
19:00:49 *** HerzogDeXtEr has joined #openttd
19:03:48 <thingwath> Grammar police is coming for you.
19:05:24 * Belugas hides, as he would be in their target too
19:05:35 *** Aylomen has joined #openttd
19:24:36 *** grumbel_ has joined #openttd
19:28:46 *** ben_goodger has joined #openttd
19:45:27 *** sigmund has joined #openttd
19:58:15 *** Nite_Owl has joined #openttd
20:08:20 *** HerzogDeXtEr1 has joined #openttd
20:10:26 *** bravo is now known as sierra1024
20:31:59 <Belugas> maybe it's the only words he knows in english
20:32:00 *** Frostregen_ has joined #openttd
20:35:52 <ben_goodger> good evening, fellow humans
20:36:26 <ben_goodger> tonight's obligatory quote: "The question about Obama now is whether the result will be a landslide, or whether the CIA will find some other way."
20:36:40 *** Frostregen_ is now known as Frostregen
20:39:22 <Belugas> in Montreal, Quebec, there was a poll : "If you were an american, who would you vote for" 88% choose Obama
20:41:00 <frosch123> "if you were an american"? so as canadian you would decide different?
20:41:39 <Belugas> tricky question indeed
20:42:08 <fjb> You don't need the CIA if you are using voting machines.
20:43:42 <Belugas> frosch123, i hate you! I can't pull your question out of my mind now!
20:43:59 <DaleStan> Well, if you weren't an American, one would generally hope that you weren't planning to vote for either. Mess up our voter turnout.
20:44:09 <frosch123> one of my colleages told me, when he was voting in bavaria the pencil was tied using a very short string
20:44:40 <DaleStan> Oh, wait. Never mind. We've got that quite well messed up already, albeit in the other direction.
20:48:50 <Belugas> the thing is, canadians do feel a lot concerned about what happens on the other side of the border
20:49:02 <FloSoft> ah i found out that there is no need for pbps (path based pre signals) you can put for example for a 4 station with 2 inputs a pbs signal and he will fill the station or wait in front of it
20:53:13 <Belugas> too much music on my card :S
20:53:19 <Belugas> no place for a new repo
20:54:09 <CIA-5> OpenTTD: rubidium * r14526 /trunk/src/ (station_cmd.cpp train_cmd.cpp): -Fix [FS#2379]: make sure trains stop at the end of a station; a 3/8th length train did stop 2/8th of it's length too early causing a 63/8th long train not to fit in a 4 tile station.
20:54:28 <Rubidium> Belugas: looks like it's time to start using hg ;)
20:54:52 * Belugas is scared to death to learn it!
20:55:13 <Eddi|zuHause> there's the "'" that was missing earlier :p
20:56:26 <Belugas> Rubidium, it's that or it's time i make some serious cleaning on that card :)
21:00:39 <Belugas> and now, ladies and gentlemen, let jsut say byebye, have a nice evening and weekend and see you on monday
21:01:56 <Belugas> Scotland's Shame is on the cell phone ;)
21:26:23 *** Sacro|Laptop has joined #openttd
21:40:51 <vvv444> A question, is there a way finding specific waypoint by name?
21:41:59 <frosch123> when it is part of orders, you can ctrl click the order
21:44:08 <vvv444> No, I mean that for stations you have a list that you can find any station there by name. Is there such a list for way points?
21:44:18 <frosch123> no, neither for buoys
21:44:43 <Rubidium> and neither for depots
21:49:59 <vvv444> Rubidium: Do you think it's logical adding buoys/depots/waypoints as additional 'station' types at station list? (with separate buttons)
21:50:29 <frosch123> depots would be silly as long as you cannot rename them
21:50:45 <vvv444> frosch123: That can be done as well :)
21:51:11 <vvv444> I just found myself looking for some waypoint for quite a long time today, so I think this functionality can be nice to have.
21:52:02 <vvv444> The best place I found to place it is station list. If it looks logical for people, maybe I'll implement it.
21:52:42 <frosch123> station list is quite different as it shows ratings and waiting cargo and such
21:53:31 <vvv444> Well, any other ideas?
21:54:03 <frosch123> no, I hardly have any waypoints nor buoys in my games
21:55:35 <Nite_Owl> Searchable like the sign list patch
21:56:10 <Nite_Owl> of course you would need a list first
21:56:15 <vvv444> Nite_Owl: Is signlist searchable?
21:57:02 <Nite_Owl> I am fairly sure someone was working on a patch to make it so
21:57:30 <vvv444> Well, it indeed requires list first
21:57:52 <Nite_Owl> I think the current trunk version is just sort of scrollable
21:58:20 <Nite_Owl> I do not use a lot of signs so I am not 100% sure
21:59:27 *** trainboy2004 has joined #openttd
21:59:32 <vvv444> The best idea I came with was the station list, but since I haven't heard any supportive optinions yet, but only objection from frosch123, I will wait for more ideas before even thinking about implementing it.
22:01:48 *** Yeggstry is now known as Yeggzzz
22:12:20 <Eddi|zuHause> you really don't know how the world works :p
22:12:55 <Eddi|zuHause> the schwabians coined the phrase "not being yelled at is praise enough"
22:15:52 <Eddi|zuHause> (that's basically the opposite of the, in the anglo-saxon culture very common, phrase "if you don't have anything nice to say, don't say anything at all")
22:16:57 <Eddi|zuHause> but german efficiency cannot afford "sandwiching" every criticism...
22:19:00 <Brianetta> Wie sagt man dass, auf Deutsch?
22:19:20 <Eddi|zuHause> Brianetta: putting a critical statement between two pseudo-nice statements
22:19:53 <Eddi|zuHause> which are basically only syntactic sugar to not violate the above rule
22:19:56 <Brianetta> We have a phrase for that
22:19:59 <Aali> soo, do banks ever close down?
22:20:09 <Brianetta> it's related to sandwiches, in a way
22:20:20 <Eddi|zuHause> yes, every day at 17:00, or 13:00 on friday
22:20:26 <Brianetta> the first pseudo-nice statement is "buttering up"
22:20:54 <Aali> Eddi|zuHause: close down, not close
22:21:02 <Brianetta> Aali: Close of banking in the UK is usually at 15:30 or 16:00
22:21:13 <Rubidium> in Iceland they close down...
22:21:34 <Brianetta> In the UK they could, but it's unlikely that the current government would let them
22:21:42 <Eddi|zuHause> i'm just amazed how efficient we are at not giving the answer the user wants ;)
22:21:46 <Brianetta> The Northern Rock bank had a run from its customers, and the government nationalised it.
22:21:50 <Eddi|zuHause> i feel almost like i'm in #mathe ;)
22:22:14 <Aali> Eddi|zuHause: its IRC, its just how things work
22:22:46 <Rubidium> Eddi|zuHause: lile Q: Is there a chance? A: Yes there is a chance ranging from (including) 0 up to and including 1.
22:23:31 <Eddi|zuHause> Q: is this the channel where you can ask questions about maths? A: no, we are a CS clan
22:23:48 <Eddi|zuHause> and for extra amusement: A: here is our clanpage: [...]
22:24:08 *** amews_aj has joined #openttd
22:24:16 <Eddi|zuHause> (sadly, our "clanpage" is offline meanwhile)
22:24:43 <ln> i've found a certain #openttd @ oftc to be a pretty good place to ask about maths on irc.
22:25:05 <Brianetta> Aali: When you say "do banks ever close down?" you need to be more specific, anyway. Firstly, the country is very important. Then, the type of bank.
22:25:19 <Eddi|zuHause> i think he means ingame ;)
22:25:43 <Brianetta> ln: Any game dev channel is good for maths questions.
22:25:44 <ln> can't be, that would be on-topic.
22:26:08 <ln> Brianetta: this is the only of that sort i'm on.
22:26:15 <Brianetta> Aali: Ah. Fictitious banks. They don't seem to close down.
22:26:38 <Brianetta> ln: Ah. I'm also on some Linux channels.
22:27:35 <Eddi|zuHause> but now for something completely different
22:27:51 <Eddi|zuHause> can i tell "svn diff" to not look for line endings?
22:27:59 <Brianetta> Aali: Frankly, it was a bit of a surprise for somebody to be talking about the risk of a bank closing in an in-game context.
22:28:32 <Brianetta> Aali: You can probably find out more from the wiki
22:28:39 <Brianetta> It has pages about game mechanics
22:29:02 <Brianetta> and newgrfs can alter the behaviour of a bank, of course
22:30:27 <Aali> the wiki doesn't say anything about banks so i'll just assume its like any other industry
22:30:46 <ln> Eddi|zuHause: fix the line-endings in the repository :/
22:31:05 <Eddi|zuHause> ln doesn't help me getting a proper diff
22:34:17 <ln> meine damen und herren, 100% of ICE-T trains will be out of operation from now on.
22:40:06 <ln> must say Treve was quite successful obfuscating his question.
22:40:33 <Eddi|zuHause> ln: i'm pretty sure i know what he wanted to ask ;)
22:43:57 <Treve> some 1 helpt me in my own language, but i want to link my server to irc
22:44:09 <Treve> reading about autopilot now
22:44:15 <Treve> looking for downloadlink ;)
22:51:20 <fjb> I'm thinking to object oriented for C...
22:52:04 <ben_goodger> fjb: the human _brain_is too object-oriented for C
22:52:39 <ben_goodger> this may be why people who spend large amounts of time with it eventually go mad, like a certain kernel developer (distributed source control, indeed)
22:52:46 *** DaleStan is now known as Guest206
22:52:48 *** DaleStan has joined #openttd
22:53:37 <Treve> can someone plz help me how to link my server to irc? and where to get autopilot? plz
22:53:38 <fjb> But is using C++ for programming a microcontroller really a good idea?
22:54:21 <vvv444> Depends on how much ROM/Flash you have and how good your compiler optimizations are.
22:54:37 <ln> Treve is privmsging me, not good.
22:55:06 <Treve> dont know who you are actually, dint know thats a prob
22:55:20 <Treve> but my google is teasing my i guess
22:55:35 <DaleStan> PM is always a problem unless you know that it is not. Not the other way around.
22:56:12 <Zuu> Treve: Also expect it to take som time, no rush.
22:56:44 <Treve> if thats the spirit here
22:57:13 <fjb> vvv444: 256k ROM and 64k RAM. Compiler is GCC.
22:58:29 <vvv444> architecture? 256k is quite ok. If you don't have many time critical tasks C++ can be fine (given G++ compiles well to this arch)
22:59:06 <vvv444> Also, it depends on how many object things you have.
22:59:18 *** avdg[BEnl] has joined #openttd
22:59:34 <fjb> Timecritical may be a big problem. I have to use an OS with cooperative multitasking...
22:59:58 <vvv444> I've used C++ for ARM, it was ok.
23:00:56 <vvv444> But I think I've used IAR compiler
23:01:31 <Zuu> Was I to hard on Treve? Just don't like when people expect that things that is completely new to them to get done instantly, without the patience to spend the time it takes.
23:01:42 <vvv444> Well, cooperative multitasking isn't a VERY timecritical thing. By time critical I mean interrupts etc.
23:02:03 <fjb> I guess I will stay with C for the moment and try to port some parts to C++ in a branch. I can compare the code size then.
23:02:48 <vvv444> As I remember one of C++ weak points is vtables size.
23:02:51 <fjb> vvv444: Cooperative multitasking makes the timing worse.
23:03:45 <fjb> Zuu: You were not too hard. He was a bit annoying and I found what he was lokking for with 3 clicks on the next best search engine.
23:04:07 <vvv444> fjb: These requiring ROM of course. Also, people tend factoring too much, link time optimizations are quite important.
23:04:26 <Zuu> Better they decide to abort or spend the time than whining about it takning to long time.
23:04:36 <vvv444> fjb: I think GCC still has problems with these.
23:05:11 <fjb> Thank you, I will have to experiment on a side branch then and stay with C for the main branch.
23:06:07 <vvv444> fjb: You're welcome, I'll be glad to hear the results :) I hope you know to use scope to measuring times? ;)
23:06:55 <vvv444> Although on ARM you probably have a normal timer that can be used for that...
23:07:15 <fjb> Yes, the goal is to do things in 1ms steps.
23:07:41 <fjb> Problem is the cooperative IP stack.
23:07:53 *** ChanServ sets mode: +v tokai
23:08:38 * Zuu wonders why he only can find documents saying that IRC is good for deaf/hard of hearing, but no actual IRC channel/server to join. :p
23:08:49 <vvv444> Well, that's not so terrible :) I was optimizing fixed point calculations from 20us to 3 us yestreday in my proj since these were done in interrupt :)
23:09:43 <vvv444> Not a big fan of assembly coding though...
23:10:03 <fjb> I fear I have to do most things in interrupt routines because that IP stack tends to hog the cpu for 3ms.
23:10:24 <Eddi|zuHause> i'm just so glad that i don't count in that kind of speed :p
23:10:38 <vvv444> Hmm, that's not a good thing. You better try reorganizing the IP stack code.
23:10:42 <fjb> I hope to get away with C. Assembler is not that much fun on RISC architectures.
23:11:01 <fjb> The IP stack is part of the OS.
23:11:01 <Eddi|zuHause> the only kind of time constraint i have is that the algorithms shouldn't be exponential :p
23:11:45 <fjb> Yes, source, but I would have to almost rewrite the IP stack and parts of the OS.
23:12:02 *** welshdragon has joined #openttd
23:12:24 <vvv444> fjb: That's your chice, but doing things in interrupts isn't very nice.
23:12:46 <vvv444> It's the shortest path to trouble.
23:13:37 <fjb> The OS was not my choice. It is nice if you don't have to do timecritical stuff. And rewriting half of the OS gets me way beyond the time frame.
23:13:46 <vvv444> Also, it adds access control problem, since you can have shared resources used on several priority levels.
23:14:52 <fjb> I will see how far I get.
23:15:22 <vvv444> Well, I hope you manage with it. Good luck!
23:19:54 *** welshdragon is now known as Guest207
23:20:06 *** Guest207 is now known as welshdragon2
23:24:41 <Nite_Owl> Time to feed - later all
23:31:51 *** ChanServ sets mode: +v tokai
23:47:24 *** Sacro|Laptop has joined #openttd
continue to next day ⏵
