IRC logs for #openttd on OFTC at 2010-12-29
⏴ go to previous day
05:20:48 *** roboboy has joined #openttd
05:47:17 *** Kurimus has joined #openttd
05:56:18 *** Eddi|zuHause has joined #openttd
07:18:32 *** DayDreamer has joined #openttd
07:31:26 *** Cybertinus has joined #openttd
07:43:56 *** andythenorth has joined #openttd
07:56:55 *** |Jeroen| has joined #openttd
08:40:54 <Terkhen> wow, now spambots talk to each other
08:47:20 * andythenorth got all excited about 'new terrain generator' thread
08:47:28 <andythenorth> followed by crushing disappointment :P
08:54:30 <andythenorth> who wants new FISH?
08:57:49 <ccfreak2k> New FISH old FISH red FISH blue FISH?
08:59:10 <Terkhen> looking nice on 1.0.5 :)
09:00:10 * andythenorth had a question for Terkhen about FISH, but has forgotten it :o
09:00:23 <Terkhen> which ships had smoke?
09:01:12 <andythenorth> it was something about buy menu / ship info
09:01:23 <andythenorth> only a few ships have smoke
09:01:25 <Terkhen> the large coaster's colours look different to me in each rotation, but since it is a brown-green change it's probably "caused" by my colour blindness
09:01:51 <andythenorth> different in what respects?
09:02:37 <Terkhen> completely green, completely brown
09:03:16 <andythenorth> between FISH versions, or when you change CC?
09:04:09 <Terkhen> brown when the ship is facing a diagonal direction, green otherwise
09:04:51 <andythenorth> that's interesting
09:05:02 <andythenorth> the shading on the hatches is different for diagonal
09:05:08 <andythenorth> it uses a different colour shade
09:05:17 <andythenorth> must be just the other side of brown for your eyes :o
09:05:32 <andythenorth> visual effects are quite fascinating
09:06:44 <Terkhen> this also happens to me for certain hair colours; depending on the direction of the light I see their hair brown or green
09:07:27 <andythenorth> FISH release done :)
09:07:34 <andythenorth> done, done, onto the next one
09:08:17 *** LordAro has joined #openttd
09:08:27 <Terkhen> which one is next? :P
09:10:10 <andythenorth> probably more FIRS animation
09:10:37 <andythenorth> but....could I interest any of you people in ponies?
09:11:02 <Terkhen> depends on the pony, I guess
09:12:14 <andythenorth> the cb28 fix is possibly relatively simple
09:12:29 <andythenorth> frosch thought that the behaviour I requested was what already happened
09:12:46 <andythenorth> seems to just be sub-optimal use of functions
09:13:09 <andythenorth> don't think anything new is needed, just moving / cleaning up
09:13:40 <Terkhen> those ponies are kind of big for me :P
09:14:07 <andythenorth> the smoke one means a spec change, which is a bit more complicated
09:14:15 <andythenorth> cb28 is just broken in my view
09:14:41 <andythenorth> I have no other small ponies :(
09:14:52 <Terkhen> yes, my problem with cb28 is that I don't know it
09:15:15 <andythenorth> well if you use it for the oil refinery, you'll know it better :)
09:15:28 <Terkhen> I could also continue unifying more road vehicle and train code
09:21:40 <LordAro> something to do with the double station, probably (perhaps a signal inbetween?)
09:22:47 * planetmaker has no idea, but it probably depends on the orders
09:23:05 <Wolf01> hello LordAro, maybe if you place signals between one station and the other it would be better since PBS used to reserve the path through all which is not a signal
09:23:09 <planetmaker> though it *shouldn't* in this case
09:23:45 <Wolf01> e.g. station are not safe waiting points
09:23:58 <Terkhen> it might go faster with signals between both stations, but I don't think that should stop trains
09:24:23 <Terkhen> hmm... should I add signals at the beginning of the first station?
09:25:27 <Wolf01> no, at first you need to separate the loading and unloading parts, or use only one station
09:25:50 <planetmaker> adding signals in between might do the trick
09:26:06 <Terkhen> yes, with signals between them they are going way faster
09:26:24 <Terkhen> I still don't understand why they did not use the free tracks, though
09:26:25 <Wolf01> yes, but if trains need to stop in the second part, theey'll block the first part for other trains
09:27:20 <planetmaker> Wolf01: they'll always do that
09:27:32 <Terkhen> why? there is a station between train stopping in the second station and the intersection
09:28:15 <planetmaker> well. with signals a train in the 2nd station doesn't block the 1st, sorry, yes
09:29:26 <Terkhen> hmm... if that's the case I don't understand why they usually would enter anyways even with trains in the first or second stations
09:30:04 <Wolf01> probably pathfinding penalties too high, I can't see anything else
09:30:06 <planetmaker> Terkhen: they need to reserve a path through to the next stop signal
09:30:29 <planetmaker> and as Wolf01 says, penalties are probably playing a trick here with you. Some nasty border-line case
09:30:44 <planetmaker> what happens if you turn and turn one train waiting?
09:31:21 * planetmaker also thinks that a savegame will help :-)
09:31:58 <Terkhen> wait, I'll upload it somewhere
09:31:58 <Wolf01> could I ask if is possible to have a "wait at PBS/Double way time" setting in the advanced options? I always forget about the right setting in console :/
09:32:31 <Terkhen> also I have never touched penalty configuration, I know that they exist but that's all :P
09:32:39 *** IchGuckLive has joined #openttd
09:32:46 <planetmaker> Wolf01: tricky :-)
09:33:07 <Terkhen> hmmm... never mind, I created enough trains to have trains waiting again and now they are picking all tracks
09:33:13 <Terkhen> thanks to the signal in the middle
09:33:15 <planetmaker> it's one of those settings which potentially break the fine-tuning of pf settings. But then it's the one with the least influence so-to-say
09:34:13 <planetmaker> there are probably hooks with that setting which I'm not aware of now
09:34:21 <Terkhen> thank you for your advice, guys :)
09:34:48 <Wolf01> gah, tried again to scroll the screenshot :|
09:35:07 <Terkhen> if you do want a savegame I can still upload it :D
09:35:20 <planetmaker> :-) I think it's not needed then
09:35:24 <Wolf01> no, I alreeady have too much problems with mine
09:35:42 <planetmaker> Terkhen: what might have been confusing: if a train doesn't find immediately a path, it only tries again after x ticks
09:35:54 <planetmaker> which might seem odd in such cases
09:36:46 <planetmaker> path_backoff_interval
09:37:02 <IchGuckLive> Hi all,i got a nother NOOB question: if i recive a offer from a city to transport to another,is this then for ONE way.Ore for both directions
09:37:10 <planetmaker> if you set that to 1 (20 or so is default) it is faster in these cases
09:39:00 <Wolf01> IchGuckLive, if you make trains do the return travel empty you'll lose money, so I let you think
09:39:37 <IchGuckLive> Thanks its on full load wait
09:40:24 *** Phoenix_the_II has joined #openttd
09:40:43 <IchGuckLive> ok, the test says its one Way
09:42:31 *** Alberth has joined #openttd
09:42:31 *** ChanServ sets mode: +o Alberth
09:47:15 <planetmaker> worst thing actually: it got through peer-review :S
09:48:23 <Alberth> I didn't need the hint :)
09:55:23 <Alberth> planetmaker: obviously, math in medical research is different from ordinary math
09:55:48 <planetmaker> of course. You can also apply the limit of big numbers for n=4 :-P
09:56:46 <Alberth> in the same way that bio-informaics is different from normal computer science :p
10:00:11 <Wolf01> it reminds me when Eddi tried to proof that all the odd numbers > 1 are prime
10:01:02 <Eddi|zuHause> what? 3 is prime. 5 is prime, 7 is prime, 9 is a measuring error, 11 is prime, 13 is prime, ...
10:01:42 <Alberth> measuring error!! ROFLAHTC !!
10:01:48 <LordAro> lol...:D measuring error?
10:03:02 *** Progman has joined #openttd
10:03:22 <Eddi|zuHause> there was a whole list of these... "proof of all odd numbers >1 being prime by <insert profession>"
10:04:09 <Eddi|zuHause> those are the experimental physicists ;)
10:06:03 *** tokai|mdlx has joined #openttd
10:09:19 <Wolf01> I found that (x+1)^2-x^2 is an odd number
10:10:19 <Eddi|zuHause> that's fairly trivial considering that each x^2 can be represented as sum[k=0..x](2k+1)
10:10:39 <Eddi|zuHause> x-1, or something
10:11:20 <Eddi|zuHause> or you can just put the binomial formula on the first summand, the x^2 goes away and 2x+1 is left
10:12:56 <Eddi|zuHause> [i'm not finding a translation immediately...]
10:14:33 <Wolf01> I'm not good at math :(
10:15:00 <Eddi|zuHause> Alberth: i don't see anything that wouldn't work with x=0
10:15:26 <Alberth> Eddi|zuHause: afaik 0 is not a non-even number
10:15:41 <Alberth> it is an odd number though :p
10:16:00 <Eddi|zuHause> Alberth: but (0+1)^2+0^2 = 1 = odd
10:16:28 * Alberth gives Eddi|zuHause a cookie
10:16:40 <Alberth> you are much more awake than me :)
10:18:24 <Eddi|zuHause> yeah... someone woke me demanding i hand out the car key...
10:19:00 <Alberth> you sleep with your car keys under your pillow?
10:19:44 <Eddi|zuHause> no, but generally i taught people to not just enter my room and search my pockets while i'm sleeping :p
10:20:37 <Alberth> you did that well, apparently :) too bad, you have to pay a price for it by people waking you up
10:23:56 <Eddi|zuHause> Wolf01: anyway, this kind of proof is typical for a 9th grade student ;)
10:24:59 *** Progman has joined #openttd
10:28:20 * andythenorth laughs at people who think of sleep
10:35:43 <Rubidium> andythenorth: interesting, as now you're basically laughing at everyone... but especially yourself for bringing it up
10:38:57 *** lasershock has joined #openttd
10:40:09 <andythenorth> sleep / uninterrupted sleep /s
10:41:37 *** JVassie_ has joined #openttd
10:45:06 *** DayDreamer has joined #openttd
10:46:46 <Alberth> andythenorth: you'll keep some non-smoking industries? If they all smoke, it is so boring
11:10:51 *** JVassie has joined #openttd
11:13:43 <andythenorth> I think most industries with chimneys should smoke
11:13:49 <andythenorth> but some only occasionally
11:14:04 <andythenorth> there is a subtle smoke puff from the tropic copper minor
11:14:52 <planetmaker> the new industrial symphony by andythenorth: tropic industries writtein in copper minor
11:15:00 <planetmaker> coming next: a variation in copper major
11:15:12 <andythenorth> that's what happens when you don't sleep for 10 months
11:15:53 * planetmaker hugs andythenorth
11:24:37 *** Adambean has joined #openttd
11:35:21 *** andythenorth has joined #openttd
11:36:08 *** ZirconiumX has joined #openttd
11:37:16 <ZirconiumX> ah, hello Alberth, and Lord Aro
11:39:22 <ZirconiumX> I had possibly the wierdest dream last night
11:40:27 <ZirconiumX> I dreamt that the only way to fix FS4345 was to remove the display options menu altogether, and wondered where you'd place the scattered settings
11:41:43 <ZirconiumX> I was up all night making *one half* of a model concorde
11:44:39 <ZirconiumX> I'm going to go bug exploring, see you in a bit
11:44:43 *** KenjiE20 has joined #openttd
11:46:25 <ZirconiumX> meh, that's a point
11:47:56 <ZirconiumX> When you go to another application - OpenTTD to Firefox for example - when I click on OpenTTD again, it stays as if I hadn't touched it.
11:48:51 <ZirconiumX> I can move it around, but the only way to get OpenTTD's 'attention' is to shrink it to dock, then activate it from dock
11:53:48 <ZirconiumX> Ah, update on FS4345
11:54:28 <ZirconiumX> It has sat around othing for 5 minutes now, and has only just given serious fault condition
11:56:19 <planetmaker> just don't open the adv. settings->interface->display options for now
11:57:08 <Rubidium> interestingly this bug has not been noticed *or* reported for a very long time
11:57:25 <Rubidium> even though it likely must have been there for years or so
11:57:30 <planetmaker> it's ppc only... so it's <=10.4.11 only, too
11:57:45 <planetmaker> and it only shows with that one particular setting
11:57:54 <planetmaker> Even though others use the exact same code...
11:58:00 <ZirconiumX> Rubidium: By that standard, I doubt it will be fixed
11:58:46 <Rubidium> planetmaker: given it is ppc, thus old Mac OS X it is becoming increasingly more unlikely that these kinds of bugs do get noticed and reported
12:00:03 <planetmaker> Rubidium: sure. That's the argument I tried to make ;-)
12:00:32 <planetmaker> -Fix: Discontinue support for PPC :-P
12:00:43 <ZirconiumX> I have a feeling a line will have to be drawn between Intel and PPC
12:01:14 <Rubidium> ZirconiumX: if that's the case I'd favour dropping Mac OS X Intel support
12:01:20 <ZirconiumX> -Fix of Fix: Lose some 1/3 of the users on Mac OS X
12:01:36 <planetmaker> ZirconiumX: hardly
12:01:53 <ZirconiumX> Intel macs are IMHO a tad overpriced
12:01:56 <Rubidium> in any case, have you tried bisecting the bug already?
12:02:06 <Rubidium> ZirconiumX: s/Intel//
12:02:44 <Rubidium> all macs are overpriced
12:03:04 <Alberth> s/a/b/ means find 'a', and replace with 'b'
12:03:12 <ZirconiumX> If you hunt around the PPC ones are cheaper
12:03:30 <Rubidium> just because they are old/ancient
12:03:39 <Rubidium> and generally unsupported
12:03:51 <planetmaker> Rubidium: there's a fix for it... glx suggested it last night and it makes sense as it even increases code consistency. The question actually is: why is it a bug as it is?
12:04:18 <planetmaker> I just wonder whether I should commit that and stop wondering ;-)
12:04:25 <Alberth> a crash is desirable behavior?
12:04:47 <ZirconiumX> They dropped support because PPC couldn't keep up with the speed of intel
12:05:38 <ZirconiumX> PM, what is the fix
12:05:46 <Alberth> are you not asking why the crash is a bug?
12:06:16 <ZirconiumX> Why hasn't it got worse?
12:06:22 <planetmaker> I'm asking why the fix is a fix or why it crashes in the first place
12:06:29 <planetmaker> it's not really obvious
12:07:18 <Rubidium> expenses_layout = uint8 that is read/written as bool
12:07:42 <ZirconiumX> yet it's an Integer
12:08:43 <Rubidium> planetmaker: that's not fixing it I fear, just hiding it better
12:09:08 <ZirconiumX> I'm going with Rubidium's fix
12:09:17 <Rubidium> after all, the thing is stored as a uint8 whereas it's read as a bool
12:09:37 <planetmaker> hm... didn't see that yet :-)
12:10:25 *** Devroush has joined #openttd
12:10:53 <Rubidium> so, who has donated the most? That way we can determine how to weigh the votes
12:11:29 * planetmaker surely would have many votes, if the DevZone counts as dontation ;-)
12:11:30 <LordAro> :D smallfly has angered quite a few people with that definitely...
12:11:32 <dihedral> to have had a happy x-mas to you Rubidium ;-)
12:12:09 <dihedral> planetmaker, only if it's run under dev.openttd.org or something like that ;-)
12:12:11 <ZirconiumX> Right all those who vote in favour of planetmaker say aye
12:12:29 <ZirconiumX> all those in favour of Rubidium's fix say nay
12:12:51 <planetmaker> ZirconiumX: do you realise how stupid this 'vote' is?
12:13:09 <planetmaker> it's like "please vote for pi=3 or pi=4"
12:13:11 <dihedral> hehe - changing uint to bool ;-)
12:13:43 * roboboy votes for the most correct fix
12:13:56 <dihedral> are you american by any chance?
12:13:58 <Rubidium> let me see... 576k lines changed, lets assume one line changed 1 contribution point. That's mean I've donated 5.793 Euros
12:14:15 <Terkhen> I'm pretty convinced that if we had voting, linux and mac support would be removed just because most people play on windows
12:14:16 <Rubidium> (although that excludes all branches)
12:14:39 <Rubidium> plus another 507 euros for grfcodec/nforenum ;)
12:14:41 <dihedral> Rubidium, chanced lines or one line removed and one line added (diff)
12:14:54 <Rubidium> dunno; using ohloh's stats
12:14:56 <Terkhen> but with Rubidium's system things would probably remain the same :P
12:15:09 <ZirconiumX> someone has to compile
12:15:58 <dihedral> Terkhen, what about those developing ;-)
12:16:48 <dihedral> ZirconiumX, i'll have a look at the 2 fixes once i am back home and have access to my mac ;-)
12:16:53 *** JVassie has joined #openttd
12:16:59 * LordAro nods his head sagely :p
12:17:00 <Wolf01> (*(bool*)var)... I can't read that, planetmaker's one looks better, but if Rubidium's one fixes it then it's better :P
12:17:05 <planetmaker> yes, that does the trick, too, Rubidium
12:17:14 <planetmaker> I prefer to combine both patches, though
12:17:31 <dihedral> bool on = (*(bool*)var); <- would it not make sense to check for > 0 ?
12:17:35 <planetmaker> because it avoids code duplication of some sense
12:17:45 <dihedral> incase that could happen somewhere else too
12:17:53 <planetmaker> dihedral: that's effectively what my patch does
12:18:18 <planetmaker> just using the proper function to read those values
12:18:52 <dihedral> perhaps it would make sense to have a month of QA :-P
12:18:56 <planetmaker> dihedral: no worries, I can test it here, too ;-)
12:19:15 <dihedral> just does not segfault on you ;-)
12:19:15 <Yexo> planetmaker: your patch has the problem that if a compiler reads *bool as 4-byte value, it reads 3 other bytes which might be !=0, so it might read true even when the uint8 is 0
12:19:41 <Eddi|zuHause> <dihedral> bool on = (*(bool*)var); <- would it not make sense to check for > 0 ? <-- in any case, shouldn't that be (bool)(*var)?
12:19:41 <planetmaker> ppc has indeed bool as 4-byte value
12:20:08 <CIA-2> OpenTTD: yexo * r21654 /trunk/ (5 files in 2 dirs): -Add: [NoAI] AITown::IsCity() so AIs can find out which towns grow faster than others (Lord Aro)
12:20:44 <planetmaker> but the point is, it's only important when the value is initialized from the config file. Correct values are used when I change the value once ingame
12:20:48 <ZirconiumX> Once dicided, does this mean you *won't* drop support for PPC, and potentially mac?
12:21:01 * ZirconiumX begs PM not to drop support
12:21:21 <Eddi|zuHause> ZirconiumX: the least mac problems are actually related to ppc
12:21:53 <planetmaker> Eddi|zuHause: there's one missing feature left... which is CJK character support. And SDL which is actually not really a mac issue either
12:21:59 <dihedral> ZirconiumX, mac support has only just been picked up again
12:22:00 <Eddi|zuHause> a propos ppc? what happened to the morphos user?
12:22:01 <planetmaker> so what do you talk about?
12:22:26 <dihedral> by the way pm: i always have two mice images in OpenTTD on my mac
12:22:49 <planetmaker> dihedral: with the current version(s)?
12:22:54 <dihedral> "the morphos user" <- LOL
12:23:05 <dihedral> planetmaker, yesterdays testing ;-)
12:23:09 <dihedral> did not touch my mac for a long time
12:23:23 <ZirconiumX> @ Dihedral: As in the cursor, then the system cursor, I've had that before
12:23:57 <dihedral> planetmaker, don't worry! it's not important
12:24:30 <dihedral> it's a graphical glitch, probably again only for PPC macs
12:24:39 <Eddi|zuHause> @openttd log 4691
12:24:41 <dihedral> how many ppc mac users are there
12:24:52 <Eddi|zuHause> @openttd commit 4691
12:24:52 <DorpsGek> Eddi|zuHause: Commit by peter1138 :: r4691 /trunk (newgrf_engine.c newgrf_spritegroup.h) (2006-05-02 18:56:07 UTC)
12:24:52 <dihedral> and how many will there be in a years or two years time
12:24:53 <DorpsGek> Eddi|zuHause: - Codechange: don't use unnamed unions as GCC 2.95.ancient won't compile it. Needed for our MorphOS user ;)
12:25:08 <dihedral> i do not think the effort of trying to fix that is worth the hassle
12:25:21 <dihedral> nice one Eddi|zuHause
12:25:22 <ZirconiumX> oooh, *optimistic* four?
12:25:30 <planetmaker> it means that basically commits r21572-r21578 are somewhat flawed
12:26:08 <dihedral> ZirconiumX, you can, but if you are the only one seriously in deep love with that architecture and combined with that OS, i doubt it's worth giving you a personal bug fixer
12:26:31 <dihedral> planetmaker, i am happy to consider some setup with my mac so that you could access it
12:26:56 <peter1138> not sure anyone cares about 2.95 anymore?
12:27:07 <planetmaker> I wonder whether ssh-x would work
12:27:35 <ZirconiumX> Even I'm using 4.0.1
12:27:49 <Eddi|zuHause> peter1138: wasn't there announced an updated morphos using GCC4?
12:28:09 <Eddi|zuHause> i haven't really followed it ;)
12:28:17 <peter1138> not sure announced means much :)_
12:29:02 <SmatZ> ZirconiumX: I wouldn't call that a win over 2.95 :P
12:29:31 <ZirconiumX> *snip* ppc-mos-gcc.lha 2.95.3 dev/gcc 671 5.2M 2002-06-23 m68k-amigaos ppc-morphos GCC v2.95.3 - MorphOS cross compiler (68k hosted) - (readme)
12:30:00 <dihedral> planetmaker, only downside :-) compiling takes like 20 mins but i have ccache setup so one could use that
12:30:05 <ZirconiumX> I thought they would have something slightly more recent
12:30:18 <dihedral> + it's only a 2Mbit dsl connection :-P
12:30:48 <dihedral> \o/ lets support OS 9
12:30:57 <dihedral> uh - i think i still have a copy of 6.2
12:31:14 <ZirconiumX> I have to go, best of luck with those patches, and I hope that you fix the problem
12:31:57 <Yexo> planetmaker: it might be good to have that too, but it doesn't reliable solve the problem (a least not when bool is 4-byte long)
12:33:11 <peter1138> typedef bool to a char, heh
12:34:46 <Eddi|zuHause> last reference to GCC 2.95 i see in the logs is r13985
12:35:06 <Eddi|zuHause> which is over two years ago
12:35:09 <Yexo> gcc < 3 couldn't handle the templates for NoAI, which was r15xxx
12:35:40 <planetmaker> Yexo: even if it then picks up garbage, and (incorrectly) reads a value of 'false' as 'true', it will get converted to either 'true' or 'false' and not cause an assertion or out-of-bounds memory access then.
12:35:56 <Eddi|zuHause> Yexo: but that's not mentioned in the commit log, at least not for trunk
12:37:03 <Yexo> Eddi|zuHause: I'm quite sure there has been a cleanup commit later which got rid of some references to gcc 2.95 in the configure logs, not sure if it explicitly mentioned that in the log though
12:38:00 <DorpsGek> SmatZ: Commit by smatz :: r16492 /trunk (15 files in 5 dirs) (2009-06-01 11:49:46 UTC)
12:38:01 <DorpsGek> SmatZ: -Remove: support for gcc2. It hasn't been able to compile OTTD for months. All attempts to do another workaround failed.
12:38:09 <Eddi|zuHause> Yexo: at least it doesn't show up when grepping for 2.95...
12:38:24 <Yexo> planetmaker: true, but it's better to fix it properly (rb's patch). Having said that it'd be good to apply your patch too :)
12:38:41 <planetmaker> Yexo: no doubt. I just like to combine both :-)
12:40:37 <dihedral> i am sitting in starbucks - had lunch and a coffee and am enjoying my laptop :-D
12:40:39 <planetmaker> oh, ok :-) I thought you might have access to your ppc right now
12:40:47 <dihedral> don't you just love mobile interwebs? :-D
12:41:12 <dihedral> nah, but i think i'll set up something when i am at home
12:41:15 <CIA-2> OpenTTD: yexo * r21655 /trunk/src/ai/api/ai_changelog.hpp: -Add: mention AITown::IsCity in the NoAI changelog
12:41:17 <dihedral> i want a vpn access anyway :-D
12:41:32 <Eddi|zuHause> [29.12.2010 13:14] <dihedral> Rubidium, chanced lines or one line removed and one line added (diff) <-- technically new lines should be worth more than changed lines
12:41:45 *** andythenorth has left #openttd
12:42:07 <dihedral> Eddi|zuHause, depends on the new line :-P
12:42:33 <Eddi|zuHause> dihedral: i mean on average
12:42:50 <Yexo> when changed lines means fixing bugs and new lines mean adding new bugs (with features), than changing lines is worth more :)
12:42:59 <Alberth> removed lines are worth much more
12:43:09 <Alberth> adding new lines is easy
12:43:27 <Alberth> ie don't fix bugs, just add a special case for it
12:43:28 <dihedral> docs is worth a bunch too
12:43:43 <Eddi|zuHause> "perfection is not when nothing more can be added, but when nothing more can be removed"
12:43:58 *** ChanServ sets mode: +v tokai
12:44:05 <dihedral> Eddi|zuHause, i like that quote :-) who is it from?
12:44:19 <Eddi|zuHause> it's some line in civ4
12:44:28 <Eddi|zuHause> i'm sure it was an important person ;)
12:44:46 <Alberth> probably from the same person that wrote "sorry, I did not have time to write a short letter to you" :)
12:45:44 <ABCRic> any luck with yesterday's bug?
12:45:54 <Ammler> heffer: are there any rh specs for openttd too?
12:46:59 <Eddi|zuHause> "Perfektion ist nicht dann erreicht, wenn es nichts mehr hinzuzufügen gibt, sondern wenn man nichts mehr weglassen kann." - Antoine de Saint-Exupéry
12:47:19 <Eddi|zuHause> so the guy who wrote the little prince...
12:51:51 <dihedral> yes, but "redhat specs for openttd" got me off track :-P
12:52:12 *** PhoenixII has joined #openttd
12:52:19 *** Phoenix_the_II has quit IRC
12:52:24 *** JVassie has joined #openttd
12:59:29 <Eddi|zuHause> dihedral: .rpm files are built from .spec files?
13:00:10 <Rubidium> planetmaker: my diff is definitely needed to properly fix the issue, although your diff seems fine as well; less point magic is good :)
13:00:47 <planetmaker> I'll wait for dih to confirm it fixed, even though I'm quite confident as it works for me and makes sense :-)
13:03:09 <dihedral> i tried glx patch yesterday and it fixed the matter
13:03:16 <dihedral> i can already confirm that
13:03:22 <dihedral> i did not try Rubidiums patch though
13:03:37 <dihedral> but by the looks of it, that need not be tested, unless you insist
13:03:42 <planetmaker> Hm. Well. As it's the right thing [TM] combining won't hurt...
13:04:08 <planetmaker> ok, then it's not needed more :-)
13:05:36 <Eddi|zuHause> i remember an incident back in school when in pascal i had an array of bool, and ord(array[n]) was either 0 or 1, except for the last array entry
13:06:04 <Eddi|zuHause> where it was 10.
13:06:09 <Eddi|zuHause> for some obscure reason
13:06:17 <Eddi|zuHause> which i never found
13:06:34 <planetmaker> seems to be the same here. It's 10 here for the last bool entry in the gui.* settings ;-)
13:06:35 <Alberth> ord() on a bool? Pascal even allowed that?
13:06:51 <Eddi|zuHause> ord() works on all enum types in pascal
13:09:12 *** ABCRic_ has joined #openttd
13:09:15 <Alberth> ah, it defines false < true. Sneaky, Wirth :)
13:09:23 *** ABCRic is now known as Guest2601
13:09:24 *** ABCRic_ is now known as ABCRic
13:10:17 <dihedral> when sending command packet details into the admin network... is there anything specific i could drop ?
13:10:42 <dihedral> because if i rely on the code that is already there for appending command packet data to the network packet, there will be issues should that change
13:11:12 <Eddi|zuHause> in my case, i had a chess board for knight movements. to prevent out-of-bound access the array ran from -2 to 9, with the outer two rows/columns being marked as "visited"
13:11:44 *** Kurimus has joined #openttd
13:12:40 <Alberth> My brother once programmed a checkers board with unusable squares around the board. Much the same idea.
13:12:46 <Eddi|zuHause> we tried out several algorithms for solving the problem: "given a random starting location, find a path through the chess boards visiting all fields, and no field twice"
13:13:39 <Eddi|zuHause> Alberth: well, it's not much different from openttd's "void" tiles at the edge
13:13:52 <dihedral> planetmaker, see my last message
13:14:14 <Alberth> Eddi|zuHause: yes, although we seem to be missing a few iirc :)
13:14:26 *** andythenorth has joined #openttd
13:16:34 <planetmaker> dihedral: I'm not sure... what do you want to change and go for?
13:18:05 <dihedral> i want to make sure that the admin network is stable - as in, no changes to the protocol happen that could break existing bots
13:18:39 <dihedral> seeing as we need the client id, i was considering a packet structure which first contains the client id, and is then sent to NetworkGameSocketHandler::SendCommand
13:18:56 <dihedral> however, if this should ever change ....
13:20:01 <dihedral> but of course i do not want to unnessesarily duplicate code
13:23:37 <planetmaker> I know little about the networking code...
13:26:21 <dihedral> it's not about networking code, it's about making sense :-P
13:26:21 <dihedral> anyway - i shall head home
13:26:43 <dihedral> at least you know what should be ready in the next few hours / days :-P
13:26:43 <dihedral> have you not heard of it? :-P
13:27:07 <dihedral> defaults to tcp port 3977
13:27:26 <dihedral> check src/network/core/tcp_admin.h
13:31:29 <peter1138> how do i access the in-game developer tools these days?
13:32:14 *** jonty-comp has joined #openttd
13:32:46 *** jonty-comp has joined #openttd
13:32:48 <CIA-2> OpenTTD: planetmaker * r21656 /trunk/src/ (settings_gui.cpp settings_type.h): -Fix [FS#4345] (r3722, r15301): Use a bool instead of uint8 to store a bool and use the dedicated accessor function when reading boolean settings (glx, rubidium)
13:33:03 <SpComb> router chaos no router chaos
13:34:30 <Alberth> peter1138: set a developer flag to true in the config file
13:35:25 <Rubidium> @calc (2**63)/(1.2**170)
13:35:25 <DorpsGek> Rubidium: 319211.036769
13:40:24 <peter1138> doesn't show what i wanted anyway :)
13:48:05 <CIA-2> OpenTTD: smatz * r21657 /trunk/src/saveload/afterload.cpp: -Cleanup (r17107): remove superfluous semicolon
14:01:23 <peter1138> i wonder if perhaps 10000 -> 500 is a bit... drastic?
14:05:15 <Rubidium> so... resistance is twice TTDPatch's
14:05:45 <Rubidium> so it's been increased too much
14:06:22 <Eddi|zuHause> michi_cc said he set the rolling resistance slightly higher than realistic.
14:06:35 <Eddi|zuHause> but i haven't checked how much
14:06:41 *** ABCRic_ has joined #openttd
14:06:46 *** ABCRic is now known as Guest2604
14:06:46 *** ABCRic_ is now known as ABCRic
14:07:07 <Rubidium> all graphs start with resistance somewhat around 0, but the new OpenTTD realistic starts more like at 25
14:07:56 <peter1138> which isn't airdrag
14:08:00 <Rubidium> reaching 100 @ 45, whereas TTDP's realistic doesn't reach 50 @ 45 or OpenTTD's old that reaches 50 @ 60
14:09:12 <Rubidium> on the other hand... "simple physics" is closest to the old OpenTTD realistic
14:10:02 <Eddi|zuHause> the orange lines don't look proper...
14:10:06 <peter1138> static resistance used to be 13 * mass / 10
14:10:12 <peter1138> now it's just 10 * mass
14:10:34 <peter1138> Eddi|zuHause, what do you mean "proper"?
14:10:41 <peter1138> they're the defined max speed
14:10:56 <Eddi|zuHause> ah. the scales are different
14:11:04 <Eddi|zuHause> makes them difficult to compare
14:11:35 <peter1138> set the max speed to 60
14:11:39 <peter1138> then they'll be on the same scale
14:11:44 <Eddi|zuHause> might be useful to make the scales align
14:11:46 <peter1138> (but then ignore the orange line)
14:11:58 <peter1138> it would, but there's no state between the graphs
14:12:43 <peter1138> all the talk about sets not setting airdrag, heh
14:13:02 <Eddi|zuHause> what's "simple physics" anyway?
14:13:19 <Eddi|zuHause> peter1138: it's a totally silly thread.
14:13:45 <Eddi|zuHause> the DBSet's VT95 is the only one with a real air drag problem ;)
14:14:07 <Eddi|zuHause> its top speed is 90km/h, but in tunnels it slows down to 70km/h :p
14:15:37 *** ZirconiumX has joined #openttd
14:15:58 <peter1138> Rubidium, the killer is the new rolling resistance
14:16:02 <Rubidium> Eddi|zuHause: but given the overall resistance seems to high, a small effect in extra air drag will show up substantially
14:16:04 <peter1138> mass * 30 * (512 + speed) / 512
14:16:18 <peter1138> even at 0, there is rolling resistance?
14:17:04 <Eddi|zuHause> rolling resistance is independent of speed
14:18:12 <Eddi|zuHause> and what's the black line?
14:20:30 <Rubidium> what happens when the rolling resistance gets changed back?
14:22:01 <ZirconiumX> Good to see that the bug has been fixed, thank you!
14:22:29 <Rubidium> peter1138: that seems to fix the initial offset difference
14:23:02 <Rubidium> though it's still too steep
14:23:22 <peter1138> if you change the 500 to 1000 (for airdrag) you get this (reload)
14:24:11 <peter1138> i should put in figures for the VT95 too :)
14:25:01 <Rubidium> that's be even faster than before the change, but still slower than TTDP
14:26:21 <peter1138> what you want is for the train to reach max speed slower
14:26:27 <peter1138> not for max speed to be unattainable
14:26:41 <peter1138> s/slower/over a longer time period/
14:28:07 * peter1138 checks what the black line is :p
14:38:01 <peter1138> acceleration output value, but it's scaled to the graph
14:40:57 <Eddi|zuHause> that's what i said earlier. the rolling resistance is the initial offset, the air drag is the steepness
14:41:30 *** DanMacK has joined #openttd
14:53:46 * Rubidium wonders whether acceleration is ran twice as often for OpenTTD than TTDPatch
14:53:51 *** andythenorth has joined #openttd
14:55:33 <Rubidium> given TrainLocoHandler (and thus UpdateSpeed) is called twice per train tick
15:02:04 <michi_cc> peter1138: If you want some real-world data, the actual resistance of an ICE 1 train set (with 12 carriages) is given as F = 6.12 + 36.2*10^-3*v + 0.893*10^-3*v^2, v in km/h and F in kN
15:02:07 <Rubidium> michi_cc: not sure; I'm first trying to investigate whether TTDPatch calls UpdateSpeed twice a tick as well
15:03:23 <peter1138> not sure it makes much difference Rubidium
15:03:24 <michi_cc> peter1138: I increased the the rolling friction to get a "slower" acceleration, but 0.15% is more like the real-world friction.
15:04:42 <peter1138> what is our rolling friction, if eddi says that rolling resistance is not affected by speed...
15:04:46 <michi_cc> Point is, a BR 01 from DBSetXL now reaches top speed, but the acceleration still uses a long enough time period.
15:05:35 <Rubidium> peter1138: calling "UpdateSpeed" once a tick instead of twice will essentially slow down acceleration by a factor 2, while not breaking the actual acceleration
15:06:14 <michi_cc> peter1138: Rolling friction is slightly affected by speed, but not much for steel on steel (more for rubber tires). The patch makes it so that it starts at 0.15% at 0 km/h, linearly grows to 0.3% at 512 km/h, 0.45% at 1024km/h etc.
15:06:50 <michi_cc> So a slight increase with speed, as more speed means more elastic deformation in the rails.
15:06:52 <Rubidium> people tell us that the old realistic acceleration is "way too fast" compared to TTDPatch, however peter1138's graphs show that OpenTTD's pre-r21106 is slower than TTDPatches
15:08:30 <Rubidium> but then my assembly isn't that good to quickly understand TTDPatch's acceleration code
15:09:27 <dihedral> "Senior Systemadministrator" :-)
15:09:44 <peter1138> Systemadministrator is clearly not meant to be one word
15:09:57 <dihedral> it is in this country
15:12:20 *** Kurimus has joined #openttd
15:13:36 <Yexo> TrueBrain: the noai.openttd.org svn doens't work (502 Bad Gateway)
15:13:39 <Alberth> but not in a English channel :p
15:14:41 <Eddi|zuHause> i'm really not the administrator kind of person...
15:15:28 <andythenorth> matching acceleration to TTDPatch is a requirement? or just easiest?
15:16:09 <Rubidium> andythenorth: not a pure requirement, but having trains behave similarly in both is better for consistency
15:16:40 <andythenorth> will I be resetting HEQS TE, weight and power values again?
15:16:52 <andythenorth> it was fun the first time, less so the second time
15:16:56 <Rubidium> after all, it's implementing the NewGRF specs so you should have similar implementations between both
15:17:09 <andythenorth> oh well, third time's the charm :P
15:17:30 <andythenorth> it's fun when the implementation detail keeps changing :)
15:17:38 <andythenorth> at least my commit count increases
15:18:12 <peter1138> for VT-95: r21106 is better
15:18:36 <peter1138> oops, parts should be 1, not 0
15:19:00 <peter1138> i dunno if dbsetxl sets airdrag for the vt95 though
15:19:15 <Rubidium> so it can't reach max speed in TTDPatch either?
15:19:34 <peter1138> with airdrag=0, yeah
15:19:40 <Eddi|zuHause> it went 88km/h on level track, 90km/h downhill and 70km/h in tunnel, last time i checked
15:20:07 <DorpsGek> Rubidium: 24.4444444444
15:20:50 * Rubidium better stays away from the vehicle acceleration...
15:21:09 <Eddi|zuHause> i never could make heads or tails out of the acceleration code
15:21:31 <peter1138> just the units were in imperial
15:22:07 <Eddi|zuHause> whenever i changed something, there was always an unsurmountable discrepancy between the logic, the physics and the behaviour.
15:25:10 *** Biolunar has joined #openttd
15:29:21 <JVassie> Does the ECS Houses .grf need parameters to run properly? No readme supplied with download, cant find a specific thread for ECS Houses either, anyoen got an idea?
15:30:01 <Yexo> it'll work without parameters, it might just not have the exact behavior you want
15:30:23 <peter1138> michi_cc, hmm, it was * 4 originally, and was changed to * 2 in your changes
15:30:33 <JVassie> Ok, more specifically then I guess, does anyone know where a list of the paramters can be found pleasE?
15:30:47 <peter1138> or previous changes
15:31:24 <michi_cc> Was it? Well, it's still some random, non-physical constant
15:31:43 <peter1138> ah, i changed it :p
15:31:48 <Eddi|zuHause> running ttdpatch creates a segfault in mciseq...
15:32:18 <DorpsGek> michi_cc: Commit by peter1138 :: r15320 trunk/src/train_cmd.cpp (2009-02-02 21:38:36 UTC)
15:32:19 <DorpsGek> michi_cc: -Codechange: Increase 'realistic' acceleration 'resolution' by one bit by reducing a division and removing a multiplication: 3/4*2 = 0; 3/2 = 1. And a smidgeon less CPU usage, hah.
15:32:59 <peter1138> so yeah, changed for a different reason
15:34:29 <JVassie> that page has nothing but a picture on though it would seem
15:35:14 <Alberth> there is also a thread about ECS in one of the openttd forums
15:35:25 <peter1138> we could just avoid it all by using ttdpatch's algorithms...
15:35:26 <Alberth> and it has a search button too :p
15:35:59 <Alberth> but otherwise, sorry, no idea
15:36:07 <JVassie> in the .grf description it mentiones a wiki page entitles ECSHouses
15:36:12 <JVassie> but it doesnt exist :(
15:36:22 <dihedral> wow - i have never seen something like this - wget works, firefox does not? :-P
15:37:07 <Alberth> there is ECS town iirc
15:37:15 <Yexo> JVassie: best to ask George in the "ECS problems" forum topic
15:37:53 <George> <JVassie> Does the ECS Houses .grf need parameters -> currently 5
15:38:23 <JVassie> is there a list of parameters anywhere please George ?
15:39:38 <George> I could not do ECSHouses before new year - I'll do them next year :)
15:40:10 *** DanM is now known as DanMacK
15:40:18 <JVassie> So how do we know which parameters to use? :s
15:40:41 <George> In fact it is a bug to be fixed
15:40:47 <JVassie> Have just updated my openttd to today's nightly
15:40:49 <George> Should work without any
15:40:59 <JVassie> and sortign through my large amount of grfs
15:41:07 <JVassie> im hopefulyl intending it to work like TTRS3 did
15:43:59 *** DayDreamer has joined #openttd
15:44:01 <JVassie> is that possible with ECS Houses George?
15:51:09 <JVassie> the ttrs3 r24 nightly seems to work better
15:57:09 *** lasershock has joined #openttd
15:57:36 <JVassie> When did the 'Build in Pause' get removed from the cheats menu?
15:58:12 <JVassie> ah found it in advanced settings
15:58:36 <dihedral> SmatZ, your definition of time is interesting
15:58:59 <dihedral> your week comes out to about 1 - 2 months in my world :-P
15:59:16 <Alberth> he is in manager mode :p
15:59:26 <planetmaker> it only proves that he's much faster than you :-P
16:01:51 <planetmaker> @calc sqrt(1 - 0.125**2)
16:01:51 <DorpsGek> planetmaker: 0.992156741649
16:03:57 *** Dante123 has joined #openttd
16:04:25 <planetmaker> yes. Actually he has to be damn fast to obtain a factor of 8 in time dilatation ;-)
16:04:30 <Eddi|zuHause> looks like a time dilation factor ;)
16:05:15 <planetmaker> the result is the required speed in units of the speed of light
16:05:41 <Eddi|zuHause> i was just wondering, if you had the formula the wrong way around, but it should be symmetric
16:06:28 <Eddi|zuHause> @calc sqrt(1-0.992156**2)
16:06:29 <DorpsGek> Eddi|zuHause: 0.125005886517
16:06:35 <JVassie> Is there a way to start other companies without havign an AI installed?
16:06:45 <Eddi|zuHause> JVassie: start a multiplayer server
16:07:08 <Yexo> JVassie: just install IdleAI, it won't do anything
16:07:32 <Yexo> be sure to take the newest though, IdlePlus or somethng like that
16:07:49 <Yexo> the older version keeps paying back the loan so it's quite annoying if you actually want to use the company
16:08:22 <Yexo> you can also start a server, start a few companies, save game, load in singleplayer and play them all via cheats
16:08:47 <JVassie> IdleMore i believe it is
16:13:02 <dihedral> Rubidium, network_command.cpp:317 void NetworkGameSocketHandler::SendCommand(...
16:13:12 <dihedral> may this be turned into a static?
16:13:32 <dihedral> seeing as it's actually not even accessing 'this'
16:15:16 *** HerzogDeXtEr has joined #openttd
16:20:03 <dihedral> spot the fun translation :-P
16:21:55 <Eddi|zuHause> dihedral: a propos 'fun translation': do you know what a "Reißverschlußkode" is? :p
16:22:51 <planetmaker> hm, what 'license' do screenshots of a game have? Any ideas or pointers what the general concensus about that is, in this community as well as in a broader sense?
16:23:23 <dihedral> Eddi|zuHause, i know what a Reissverschluss is, but not with the 'kode' suffix
16:23:36 <Eddi|zuHause> planetmaker: typically, they don't inherit the license of the game
16:23:51 <planetmaker> that's what I think, too.
16:24:16 <planetmaker> but an IRC quotation of Eddi|zuHause is not the reference I really look for :-P
16:24:20 <Eddi|zuHause> planetmaker: in american law, screenshots (of games or movies) fall under 'fair use'
16:24:31 <Eddi|zuHause> the german law is... more complicated...
16:24:53 <planetmaker> he... as usual :-(
16:25:06 <Eddi|zuHause> dihedral: it's someone translating "zip code" into german :p
16:25:35 <dihedral> i like "wer nase" for "who knows" :-P
16:26:23 <planetmaker> Eddi|zuHause: does 'fair use' require something like CC-BY?
16:26:26 <Eddi|zuHause> my father found this in a portuguese brochure (on madeira)
16:26:57 <Eddi|zuHause> planetmaker: no, the concept of "fair use" is over 300 years old.
16:30:34 <ABCRic> I like that bread they make in madeira.
16:31:05 * peter1138 wonders where ttdp's magic 683 value comes from
16:32:15 <Eddi|zuHause> planetmaker: the problem is, the german UrhG does not know the spongy-foamy concept of "fair use", but has instead very clearly defined "Schranken" (boundaries) of copyright
16:32:47 <ABCRic> I don't like the regional president of madeira though
16:32:59 <ABCRic> or whatever it's called
16:33:38 <ABCRic> he was once interviewed about some topic related to continental portugal and he actually said "f*** them"
16:34:22 <Eddi|zuHause> planetmaker: especially (3). but it's not trivially applied to digital copies.
16:34:40 <Eddi|zuHause> which is currently the worst flaw of the entire law.
16:35:11 <planetmaker> well, a screenshot is not a copy of the programme nor its graphics. It's... something else.
16:36:14 <ABCRic> a screenshot is just a picture of the screen. it's not the author's fault if the program was there when he took it :P
16:37:16 <planetmaker> something less. Gah... but the (German) law doesn't seem to know that... or I miss it :-)
16:37:39 <planetmaker> ABCRic: same with photos. But still you may not publish every photo you took yourself
16:38:18 <__ln__> however... is it fair use if the graphics files were not legally yours in the first place
16:38:22 <Eddi|zuHause> ABCRic: the content of the screen may have very significant influence on what you may do with the screenshot
16:40:26 <Eddi|zuHause> __ln__: german law originally had the concept of "the work must have been legally published", meanwhile they extended it to "the copy must not be obviously illegally created", and since they noticed that this doesn't apply to filesharing, also "the copy must not obviously illegally have been made available"
16:41:03 <Eddi|zuHause> although, it still is not clear in how far this actually applies to filesharing
16:41:03 <planetmaker> __ln__: de-facto I'm interested in: under which terms could anyone use a screenshot of a vanilla OpenTTD game (OpenGFX graphics)
16:41:49 <planetmaker> "do what you want" might be a bit too "laisser-fair". But "list all people who ever worked on OpenTTD & OpenGFX" seems overdoing it
16:42:30 <planetmaker> what if the screenshot is integrated in another are ensamble? If it's a picture hanging on the wall in another picture?
16:42:47 <Eddi|zuHause> planetmaker: if you write a text (scientifical or fictional), and print it out using a font, does the designer of the font suddenly have any rights to your text?
16:43:37 <planetmaker> Eddi|zuHause: indeed he could
16:44:05 <planetmaker> just most songs which you have allow you to use them for whatever purpose
16:45:13 <ABCRic> 79 1. You may copy and distribute verbatim copies of the Program's
16:45:15 <ABCRic> 80 source code as you receive it, in any medium, provided that you
16:45:16 <ABCRic> 81 conspicuously and appropriately publish on each copy an appropriate
16:45:18 <ABCRic> 82 copyright notice and disclaimer of warranty; keep intact all the
16:45:20 <ABCRic> 83 notices that refer to this License and to the absence of any warranty;
16:45:21 <ABCRic> 84 and give any other recipients of the Program a copy of this License
16:45:23 <ABCRic> 85 along with the Program.
16:45:36 <Eddi|zuHause> ABCRic: this doesn't apply here.
16:46:37 <Eddi|zuHause> ABCRic: also, the question is whether screenshots are generally independent of _ANY_ game license
16:47:34 <planetmaker> s/songs/fonts/ of course :-P
16:47:46 <ABCRic> Eddi|zuHause: they're not if the license says something about them
16:48:01 <planetmaker> wikipedia seems to assume that screenshots are under the license of the graphics / the game
16:48:53 <ABCRic> planetmaker: no one respects that, then :P
16:50:36 <ABCRic> I suppose that means that game companies own the rights to screenshots of their games
16:53:30 <Eddi|zuHause> ABCRic: the question is not whether screenshots are a derived work of the game graphics, but whether these derived works are covered by any of the exceptions to copyright law, so that the original license generally doesn't apply to them anymore.
16:57:00 <Eddi|zuHause> so when the actual "creative work" is your game world, then using the graphics is "insignificant"
17:15:43 <Alberth> Eddi|zuHause: fixed several days ago, we have already 5-6 issues in FS about it
17:16:10 <Alberth> FS#4327 is the original one
17:16:12 <Eddi|zuHause> just making sure
17:16:53 <Alberth> people don't seem to know how to add a bug report, any more :(
17:35:55 <dihedral> how important is the callback in the command packet?
17:40:18 <Alberth> dihedral: it is used :)
17:40:33 <Terkhen> ABCRic: I'm trying to reproduce it right now
17:40:46 <Alberth> are you going to solve it ABCRic ? :)
17:41:36 <ABCRic> In my testing, the lines do get a little 'broken', but not as much as in the picture
17:41:57 <Alberth> try adding more movement behind the windiw
17:42:00 <ABCRic> Alberth: I gotta study some more C++ first :P
17:42:17 <Rubidium> dihedral: I guess it can
17:42:33 <Terkhen> yes, I think I saw a dot moving a few pixels up, but that's all
17:42:48 <dihedral> Rubidium, i am assuming that's referring to making it static? :-P
17:43:07 <dihedral> else i'd just drop the callback from being sent to the admin network for now
17:43:19 <Rubidium> ... making it a method of Packet is a second possibility
17:44:07 <dihedral> that would be a bit odd though, as the logic for packet structure is (so far) never in Packet
17:45:13 <dihedral> how important can the callback be to logging?
17:47:02 <Rubidium> it's probably not important
17:48:12 <dihedral> i have the impression that you are about to use the word 'but' :-P
17:48:54 <Terkhen> I'll just ask for a savegame in which the effect is noticeable
17:57:14 <Rubidium> dihedral: actually, I was about to use the word actually: actually, not showing the callback means one less randomly changing variable sent to the admin bots
17:57:35 <Rubidium> although the whole command stuff is inherently random given there is no stable protocol
18:13:49 <Rubidium> michi_cc: I'm not sure whether that's the solution for the problem, but we could definitely make a build with that patch and let some those most vocal about the problems try it
18:15:50 *** fjb is now known as Guest2617
18:20:42 <CIA-2> OpenTTD: rubidium * r21658 /trunk/src/economy.cpp: -Fix [FS#4351]: cargo payment window wasn't properly invalidated when payment changed
18:30:10 *** andythenorth has left #openttd
18:40:55 *** ZirconiumX has joined #openttd
18:43:33 <ZirconiumX> hello guys, how are you all, and how's it going with whatever you're doing
18:44:33 <Terkhen> fixing my brother's computer, boring but not bad
18:45:08 <ABCRic> making my first patch :D
18:45:34 <CIA-2> OpenTTD: translators * r21659 /trunk/src/lang/ (6 files): (log message trimmed)
18:45:34 <CIA-2> OpenTTD: -Update from WebTranslator v3.0:
18:45:34 <CIA-2> OpenTTD: hungarian - 6 changes by IPG
18:45:34 <CIA-2> OpenTTD: indonesian - 9 changes by fanioz
18:45:34 <CIA-2> OpenTTD: italian - 3 changes by lorenzodv
18:45:36 <CIA-2> OpenTTD: malay - 24 changes by rionix88
18:45:36 <CIA-2> OpenTTD: norwegian_bokmal - 5 changes by Utvik
18:46:01 <ABCRic> a guide on how to add new settings would be nice though
18:46:09 <michi_cc> Rubidium: Well, it's my version of a solution. Making a build with it seems like a good idea. If people don't like it, they don't like it, but then somebody else has to come up with something instead.
18:46:15 <ZirconiumX> @Terkhen, how the heck did he break it?!?! @ABCRic, best of luck!
18:46:56 <ZirconiumX> ah, hello, Lord Aro
18:47:44 <Terkhen> ZirconiumX: nothing, my guess is that the battery died
18:47:44 <ABCRic> LordAro & Yexo: thanks :)
18:47:46 <Alberth> ah, the good old days when it was called a patch :)
18:49:10 <LordAro> oh yes! i remember them as well...:D
18:49:32 <LordAro> oh for the heady days of 0.4.8
18:50:44 * ZirconiumX sort of remembers, but not quite
18:50:56 <LordAro> i believe the first version i had was 0.4.5... you(read:everyone else)?
18:51:39 <ZirconiumX> 0.6 had it changed
18:51:47 <Terkhen> but I did not play nor fiddle with settings much back then
18:52:53 <Alberth> ABCRic: we skipped 0.8 and further :)
18:53:38 <ABCRic> it is the most recent of the stated nonetheless :P
18:54:00 <Alberth> that is good, we need new players :)
18:54:12 <Terkhen> I remember that day, the omission of 0.8.x was very surprising :)
18:55:08 <Alberth> too bad we never released 0.10 :p
18:55:33 <ZirconiumX> 0.114.2, if you hadn't changed it
18:55:41 <bb10X> Why isn't the windows 1.1.0 beta 1 version not labeled as such?
18:55:49 <LordAro> heh, lots of confused bug reporters....
18:56:17 <LordAro> bb10X: it was broken slightly, will be fixed in beta2
18:57:25 <ZirconiumX> I went and played an old patched version, and it appears that the display options bug was still there
18:57:31 <bb10X> also, setting server_port doesn't seem to work
18:57:45 <ZirconiumX> but as it's fixed, I need worry no longer
18:59:17 <ZirconiumX> BRB, going to play with AroAI
19:01:13 <Alberth> ZirconiumX: 'going to play with' means 'do some hacking in' with me :)
19:03:19 <ZirconiumX> heh, I think I'll have to - AroAI is not being detected, but I think that's to do with the fact it uses 1.1 API
19:18:06 <JVassie> Does anyone happen to have an old version of NuTracks lyign around anywhere?
19:18:20 <JVassie> a version from before they introduced the new graphics?
19:24:54 <Wolf01> nutracks is really cool, but if it could be split into rail, 3rd rail, subway and planning it would be more cool
19:26:24 <JVassie> now have to try and fidn the version which had the default track graphics for all speeds
19:32:48 <ABCRic> meh. the wiki article on adding settings is outdated
19:33:28 <ABCRic> what does SDTC_VAR() argument 8 do?
19:34:40 <Alberth> do you have a line number and filename of SDTC_VAR #define ?
19:35:15 <Alberth> (and the name of parameter 8 there)
19:35:37 <ABCRic> I'll check if MSVC can tell me though
19:36:12 <Alberth> src/table/settings.h:187
19:37:32 <Alberth> interval if I am counting correctly
19:38:16 <ABCRic> correct, but I can't figure out what it does
19:38:30 <Yexo> it's only for integer settings, it's used in the gui
19:38:43 <Yexo> one press on the left/right buttons will decrease/increase the setting with that amount
19:39:33 <ABCRic> how come it is 0 on some settings?
19:40:06 <ABCRic> gui.drag_signals_density, for example
19:40:49 <Yexo> that one cannot be changed from the advanced settings window IIRC
19:41:51 <ABCRic> I suppose gui.max_num_autosaves can't be changed from there either?
19:42:19 <ABCRic> thanks for clearing that up :)
19:42:50 <Alberth> uint32 step = (sdb->interval == 0) ? ((sdb->max - sdb->min) / 50) : sdb->interval;
19:43:54 <Alberth> src/settings_gui.cpp, line 1614
19:50:12 <ABCRic> ok, so... the setting uses the multi-string gui flag. How do I tell it what strings to use?
19:54:34 <Alberth> you have a string for the description (ie shown in the advanced settings)
19:54:53 <Alberth> below it is a sequence strings for values
19:55:35 <ABCRic> it uses the strings below the description?
19:56:03 * Zuu looks forward to the dictator AI :-p
19:56:54 <Alberth> ABCRic: not really :)
19:57:30 * SmatZ should test PAXLink once again :)
19:57:33 <ABCRic> then what do I have to do?
19:58:19 <Alberth> not really convenient :)
19:58:35 <Alberth> look for STR_CONFIG_SETTING_SMOKE_AMOUNT in src/lang/english.txt for an example
19:58:58 <Alberth> you get duplicate strings if settings use the same values
20:00:11 <ABCRic> but it makes it a lot easier to add new settings
20:00:31 <Alberth> eg values of STR_CONFIG_SETTING_TRAIN_ACCELERATION_MODEL and STR_CONFIG_SETTING_ROAD_VEHICLE_ACCELERATION_MODEL are the same, which means all translators translate those strings more than once, have to keep them consistent, etc
20:01:06 <ABCRic> maybe a pointer string system could be implemented
20:01:34 <Alberth> I have a patch for it, but it is not good enough to merge to trunk
20:02:11 <Alberth> ie add the StringID of the first value to the setting as well
20:07:46 <ABCRic> almost done with my patch :) I'm going to compile and see if I added the setting right. After that, I'll dig into the unknown to make the game use the setting...
20:09:26 <Alberth> and change the savegame version, and allow loading old save games
20:11:35 <ABCRic> Not needed :) see FS#4115
20:13:47 <Alberth> sounds as a local option indeed :) makes life much easier :)
20:14:50 <ABCRic> compiler says an error and a warning: error was a typo, but the warning was caused by the fix to yesterday's issue...
20:15:34 <ABCRic> bool on = (bool)ReadValue(var, sd->save.conv); is causing warning C4800: 'int64' : forcing value to bool 'true' or 'false' (performance warning) (MSVC 10)
20:16:07 <ABCRic> looks like we need another "make MSVC 2010 happy" commit...
20:18:24 <planetmaker> what makes msvc happy here? It already has an explicit cast...
20:18:30 <glx> I don't remember MSVC complaining about that when I tried yesterday
20:18:51 <planetmaker> another msvc version maybe?
20:19:12 <Rubidium> 2k8 (32/64) complain as well
20:20:52 <CIA-2> OpenTTD: rubidium * r21660 /trunk/src/timetable_cmd.cpp: -Fix [FS#4354]: autofill timetable had side effects in test mode, possibly causing desyncs in MP
20:20:54 <CIA-2> OpenTTD: smatz * r21661 /trunk/config.lib: -Codechange: use -fwhole-program when linking in LTO mode so compiler can do more optimisations
20:21:29 <glx> hmm indeed 2010 complains
20:25:13 *** LordAro has joined #openttd
20:33:02 <LordAro> a question for developers: what software do you use? msvc, mercurial, etc.? (i know they have different uses ;) )
20:34:43 <ABCRic> meh, skipped a string at english.txt :P
20:34:44 *** fmauneko has joined #openttd
20:35:06 <glx> mingw, msvc, svn, mercurial
20:36:16 <LordAro> i find that svn has the best .diff format (and version naming ;) ) but hg has the best committing system, agreed?
20:36:30 <planetmaker> hg, xcode, svn, gcc...
20:36:39 <planetmaker> LordAro: vice versa ;-)
20:36:54 <planetmaker> git diffs are IMHO 'best'
20:37:18 <planetmaker> for certain definitions of 'best'.
20:37:33 <LordAro> by 'best' i mean most user friendly (read:newbie friendly)
20:37:38 <planetmaker> Disussions about 'best' in this context highly depend on the exact context and are quite mute
20:38:13 <planetmaker> then... what is the difference between svn diff, hg diff and git diff?
20:38:28 <LordAro> btw, why does mercurial give things such strange version numbers, i know it's hex (?), but why?
20:38:50 <LordAro> planetmaker: it's easier to read (IMHO)
20:39:07 <ABCRic> LordAro: Mercurial uses SHA-1 hashes to identify revisions
20:39:24 <Rubidium> kate, sed, awk, gimp, make, gcc (4.3-4.6) x86_64-gnu-linux, MSVC 2k[5,8,10] (IA32/x64), gcc 4.4 i586-dos-msdosdjgpp, gcc 4.3 i486-linux-gnu, gcc 4.5 i586-mingw32msvc, tortoisesvn, subversion, git, mercurial, ... and more
20:39:50 <LordAro> definitely everything :)
20:39:56 <planetmaker> LordAro: what is easier to read there?
20:41:45 <LordAro> umm, it is clear where the file diffs change(different files e.g. abc.cpp and def.cpp), the lack of the a/ / b/ at the beginning of file address
20:43:35 <Rubidium> LordAro: the developers don't really care, but if you want to be funny for hg/git diffs
20:46:05 <planetmaker> LordAro: the advantage of git or hg over svn is that you have the whole history locally and that you have a better system for patch queues
20:46:34 <LordAro> it's just a personal preferance... ;)
20:46:34 <planetmaker> that is as a patch writer I definitely would recommend to use one of those two.
20:46:49 <CIA-2> OpenTTD: smatz * r21662 /trunk/ (Makefile.src.in config.lib): -Codechange: use -flto=jobserver for GCC 4.6, it makes linking with LTO use number of processes passed to make -j
20:46:54 <planetmaker> the diff file format is really of no consequence
20:47:14 <planetmaker> git style seems to me the most comprehensive one as it also covers binary ones
20:47:48 <Zuu> SmatZ: It's more than a year or so since last PAXLink release. I have some unreleased changes but they are just a start of making it compare different investment so that different parts of the AI don't compte on bying the same thing often resulting in bying cheap buses every now and then and never get any money for aircrafts.
20:47:49 <glx> it's just tortoisesvn is stupid and knows only a part of diff format spec
20:48:22 <Terkhen> you can get binary files on mercurial diffs too easily
20:49:18 <Zuu> Problem is that PAXLink have quite a large number of actions that need to be rewritten so that they can be enqued and compared.
20:49:26 <planetmaker> Terkhen: yes :-) - I enabled git style there ;-)
20:49:28 <SmatZ> Zuu: :( well, I really like its concept :)
20:49:40 <glx> and svn hates to have different EOL in one file :)
20:51:57 <Zuu> I've been combating with CSV files from excel where some cells have line breaks (\n) and the file have \n\r for the table rows :-)
20:55:02 <Eddi|zuHause> Zuu: sed s/\n\r/\r/g ?
20:56:43 <LordAro> planetmaker: i really like the patch queues as well, just the svn file format i like, and i guess i could get used to that ;)
20:56:44 <Zuu> I don't know how to match for not a \n, so my solution was: s/\n/####/g s/\n\r/ /g s/####/\n\r/g
20:57:24 <Zuu> This is how it sholud be: s/\n\r/####/g s/\n/ /g s/####/\n\r/g
20:57:35 <Eddi|zuHause> might need to teach sed to not treat the file line based
20:57:51 <LordAro> [20:39] <ABCRic> LordAro: Mercurial uses SHA-1 hashes to identify revisions <-- but why? why not use a linear numbering system like svn?
20:57:58 <Zuu> I used (g)vim to do the changes.
20:58:29 <Zuu> For not too big files it works well. For really big files, sed is way faster then (g)vim.
20:58:50 <michi_cc> LordAro: How would you assign a linear number in a *distributed* version control system without a single central server?
20:59:16 <Eddi|zuHause> Zuu: i have never figured out how to do vim-style multi-line-regexp from the command line
20:59:43 <LordAro> michi_cc: i have no idea... but i also have very little idea of what you're talking about ;)
21:00:30 <Zuu> I've not really tried to do multi-line regexps other than if you define changing line ending style as multi-line regexps.
21:00:33 <Eddi|zuHause> michi_cc: make it a cloud ;)
21:01:13 <Eddi|zuHause> Zuu: i mean having \n and \r in the search pattern.
21:02:03 <Eddi|zuHause> Zuu: the only place i got it to work was in vim
21:02:44 <michi_cc> Eddi|zuHause: Just like in that Microsoft commercial where you need a cloud to be able to copy and paste a face between pictures? :)
21:03:14 <Eddi|zuHause> michi_cc: i haven't watched any commercials in years
21:03:53 <Rubidium> Eddi|zuHause: really? What about the p1sim thread?
21:04:37 <Eddi|zuHause> Rubidium: what's a p1sim?
21:07:05 <Xaroth> the real question is, who's smallfly :P
21:07:26 <DanMacK> Is plsim anything beyond a website and some sketches right now?
21:07:34 <Xaroth> from what I can see.. no
21:08:09 <Rubidium> yes, it has a donation plan (or probably more like an investment scheme)
21:08:37 <Xaroth> heh, yeh, mr smallfly is really keen on covering his 'expenses'
21:08:49 <Xaroth> how he manages to spend so much already on a 'free' forum, and a wiki, is beyond me
21:09:58 <Xaroth> he does seem to like the whole ottd interface though :P
21:10:38 <Xaroth> P1 is coded 2-dimensional to enable having huge maps and many vehicles. But the game objects can be edited and saved 3-dimensional plus textures.
21:10:56 <Xaroth> I don't see how using 2d-only is the key to being able to have huge maps...
21:17:33 <ABCRic> no wonder it costs 96 euro/year. It's got 3 domain names!
21:17:43 <Xaroth> .nl domains are ~5 euro a year
21:18:05 <Xaroth> so that leaves 80 euro unaccounted for :P
21:18:29 <ABCRic> that depends on the registrar...
21:18:34 <Xaroth> com is 10 euro a year or so
21:19:21 <LordAro> DanMacK: some coding work has been screenshotted(?) but apart from pictures of the tools to implement stuff, we've seen nothing :rolleyes: :)
21:19:35 <ABCRic> say, 20 euro in domain name expenses, that leaves us with 76 euro
21:20:03 <ABCRic> let's say, 29.99 on server expenses
21:20:14 <ABCRic> 46.01 for coffee and donuts
21:20:28 <Xaroth> probably just an expensive host
21:22:18 <ABCRic> hmm... the concept page says that one of the feature idea sources is "up to 256 players per server"
21:23:36 <Terkhen> why would so many players be needed?
21:24:19 <ABCRic> furthermore, how is that an idea source?
21:24:25 <LordAro> am i the only one wondering what the system requirements are going to be?
21:24:33 <Rubidium> Terkhen: given the "suggestion" I got this week, it's a bit small
21:24:54 <ABCRic> LordAro: 2TBit internet connection :D
21:25:12 <Terkhen> any reason for requiring even more clients?
21:26:31 <LordAro> ABCRic: what i would give for something anywhere near close to that... :D
21:26:42 <Rubidium> some "assistant professor" thinks it's easy to make OpenTTD a MMOG
21:27:00 <Rubidium> so he's throwing some students on it or something
21:27:29 <Xaroth> it just requires a rewrite of the netcode.. and the command system
21:27:30 <ABCRic> I think the stuff on the "Commonly Asked for Suggestions" thread should be put on the forum header
21:27:37 <Xaroth> so basically.. most of it :P
21:28:09 <LordAro> Rubidium: well hopefully they'll give us/you the source when they've finished/failed :D
21:28:31 <Rubidium> Xaroth: the net code and the command system aren't the real problem I'd say
21:28:43 <Xaroth> Rubidium: it's a good start
21:28:46 <Rubidium> having the clients trying to keep up with the server is going to be the problem
21:29:08 <Xaroth> soooooo much calculations have to be made -_-
21:29:15 <Rubidium> meaning you'd basically have to rewrite the game from scratch to send only the shown data to the clients
21:29:29 <ABCRic> besides, 256 players would likely mean loads more companies
21:29:36 <Xaroth> who cares about companies
21:29:43 <ABCRic> and more competition...
21:29:44 <Xaroth> think of the amount of trains that all need a path calculated :P
21:29:46 <Rubidium> yeah, more than 15 companies means massive changes to the map array
21:29:53 <Xaroth> that shit is massive on the CPU
21:30:12 <ABCRic> and more vehicles, and more pathfinding, and massive CPU usage
21:30:33 <Xaroth> I think trains are worst at that
21:30:34 <Rubidium> as well as extra company colours, which would imply something like 32bpp company colours and all shit related to that
21:30:51 <Rubidium> trains are best w.r.t. pathfinding (assuming YAPF)
21:30:58 <ABCRic> so, besides the 2TBit connection, put a FTL processor on the list
21:31:00 <Rubidium> they got extensive caching
21:31:10 <Xaroth> yeh, but you can dumb down rv pathfinding to a minimum
21:31:15 <Xaroth> you can't really do that with trains :P
21:31:36 <Rubidium> just like ships... place waypoints every 4 corners (or junctions)
21:32:35 <LordAro> ABCRic: or a quantum one... :)
21:32:41 <ABCRic> hey, and since you're at rewriting, how about you make it 3D too? kthxbai :D
21:32:58 <Xaroth> it just uses 2d graphics
21:33:11 <LordAro> and don't forget rotation :D
21:33:24 <ABCRic> Xaroth: I can't rotate stuff
21:33:36 <Xaroth> ABCRic: rotation has nothing to do with the amount of dimensions :p
21:34:00 <ABCRic> but I still can't rotate stuff!
21:34:13 <Xaroth> but yes, rotation would be a bit of an issue with the current graphics set
21:35:46 <Rubidium> what? The smallmap is a rotation!
21:36:39 * LordAro wonders if there is a decent phpbb forum url wiki template anywhere...
21:38:26 <ABCRic> wait, wait, wait... in p1sim you can have a transport company... AND be a mayor?!?!
21:38:32 <dihedral> Rubidium, would it make sense to change the handling of command packets and restructure that before implementing command packets finding their way into the admin network?
21:38:55 <ABCRic> next thing you know, it can read my email too...
21:41:05 <Rubidium> dihedral: what needs to change?
21:41:14 <ABCRic> so... can anyone tell me where in the source the company color is set at the start of a new game?
21:41:33 <Rubidium> I'd guess company_cmd.cpp?
21:41:37 <Xaroth> ABCRic; 2nd door on the left
21:41:53 <Rubidium> also, try when AE fails, try BE
21:42:05 <dihedral> <Rubidium> although the whole command stuff is inherently random given there is no stable protocol <- make it more stable, that's what i meant
21:43:00 <dihedral> would it at all be possible
21:43:36 <Rubidium> dihedral: yes, but it's totally not worth the effort
21:44:57 <Rubidium> as it means we can't reorder/change/remove stuff from commands, so we end up with either lots of unreuseable space in the commands, or with lots of now defunct commands that we can't trash completely because the command index has already been used
21:45:15 <Rubidium> furthermore it means we need to stabilise the strings, which is way more effort than it's worth
21:46:27 <dihedral> would it make sense to send strings rather than various id's then?
21:46:52 <dihedral> i.e. send a colon separated string
21:46:54 <Rubidium> no, as strings can't really be translated
21:47:05 <Rubidium> whereas StringIDs can
21:47:36 <Rubidium> though it would be a really nice way to spoil lots of Luukland's and the likes command handling
21:47:46 <Rubidium> but then it would also be a huge waste of bandwidth
21:47:58 <dihedral> i mean send strings to the bots
21:48:13 <dihedral> i.e. send a ready prepared string for logging of commands
21:48:34 <Rubidium> those strings are (in fact) just the "default" error messages, so quite pointless
21:48:42 <dihedral> or very limited selection of information
21:48:45 <Rubidium> at least for the bots
21:49:13 <dihedral> no callback info, no p1 / p2 ?
21:50:28 <Rubidium> dihedral: basically "nothing"
21:51:07 <Rubidium> although, company and tile might be relatively safe to send
21:51:15 <Rubidium> the rest has an unstable API
21:51:31 <Rubidium> oh, the frame's stable as well
21:52:12 <dihedral> client id, company id, cmd, text frame
21:52:30 <planetmaker> iirc you said once something along the lines of client id/name had to be matched by a separate query and then matched to the p1/p2 information
21:52:33 <dihedral> but p1 and p2 are uint32, and could simply be logged as numeric values
21:52:49 <dihedral> and leave the parsing of that to the human reading the data
21:53:32 <dihedral> the client_id = origin of the command
21:53:47 <Rubidium> dihedral: I'm just saying that the data from cmd, p1 and p2 are totally unstable and that it's not documented and can change without prior notification or even a mention in the changelog
21:54:28 <Rubidium> client id, company id and frame are relatively safe "stability" wise
21:54:39 *** toasterdd has joined #openttd
21:54:53 <Rubidium> whereas text is somewhat in between; no stable API, but it can't really have much meanings
21:54:59 <dihedral> i could introduce further 'protocol' packets for command details
21:55:45 <dihedral> send a bunch of strings naming the cmd's
21:56:12 <Rubidium> dihedral: how? There's plenty of bitfucking, meaning the meaning of one bit might depend on a multitude of other bits
21:57:20 <dihedral> the name of a cmd is not dependant on "bitfucking" is it?
21:57:35 <Rubidium> true, but that'd only be the command name
21:57:42 <Rubidium> and it'd still be unstable
21:58:34 <dihedral> so if the entire Commands enum is sent to rebuild an enum on the other end?
21:58:40 <Rubidium> though requesting the name for a command would be relatively trivial
21:59:19 <Rubidium> it's just that those names might, again, without notification be changed
21:59:19 <dihedral> BUILD_RAILROAD_TRACK as first string in a ADMIN_PACKET_PROTOCOL_COMMAND
21:59:58 <dihedral> then instead of sending the cmd id to the bots, send a string in the first place
21:59:59 <toasterdd> good evening at all! can everybody tell me, how do I use the "No.x Crawler" vehikels? Like normal cars? Please forgive me ;-)
22:00:07 <Rubidium> dihedral: people shouldn't be surprised it gets changed to BUILD_RAIL_TRACK in the future though
22:00:37 <dihedral> that can be documented
22:00:48 <dihedral> i.e. for logging purposes only
22:01:00 <Rubidium> the whole command API will be that way
22:01:44 <dihedral> which is why i am trying to think of a safe something for logging this stuff
22:02:42 <dihedral> but sending 100 of those strings is not going to make it a nice thing either
22:05:35 <Rubidium> dihedral: what about sending them whenever you request logging commands to be sent?
22:06:16 <dihedral> because the logging depends on the servers log level
22:06:20 <planetmaker> sending them once with their ID?
22:06:32 <dihedral> there is not a separate log level for the admin port iirc
22:06:34 <planetmaker> and then re-sending only the ids?
22:06:56 <Rubidium> dihedral: why would the logging depend on the server's loglevel?
22:07:20 <dihedral> because logging is a console output forwarding iirc
22:07:21 <Rubidium> and what do you see as "log level"?
22:07:52 <Rubidium> commands should generally not be logged to the console
22:08:30 <Rubidium> so, what's the problem in sending the admin the names for the commands when it requests to get the commands?
22:08:43 <Rubidium> hmm, maybe that the bot doesn't need them
22:09:17 <Rubidium> and it'd be adding a special case to the requesting stuff to be sent "protocol", so requesting a string representation for a command might be better
22:10:58 <dihedral> would it have to log everything (commands originating from the server)? or just incomming commands?
22:11:46 <Rubidium> not logging the server's command seems kinda pointless, doesn't it?
22:11:52 <Rubidium> especially with a non-dedicated server
22:12:02 <Rubidium> either you log it properly or you dont't log it at all
22:17:14 <dihedral> so a update type named CMDLOG just to make sure, send client_id, company_id, cmd_name, p1 and p2, tile, text, frame <- i think this would be my preference, p1 and p2 may be odd values, but hey are there for logging only, interpretation of their meaning would be left to the person reading the logs
22:18:11 <Rubidium> I'd send command IDs as they are much shorter and thus use less bandwidth given the large amount of commands sent
22:18:31 <Rubidium> just the low words though
22:19:10 <dihedral> that means all commands have to be named in another packet when the bot requests updates for command logging
22:20:08 <Rubidium> yep, the client would just request all command names
22:53:22 *** andythenorth has joined #openttd
22:53:41 <andythenorth> I didn't know that FISH Technology Group were involved in TTD
22:54:14 <Eddi|zuHause> yes, they did the windows port, afair
22:56:02 <Eddi|zuHause> hm... i'm missing two "Do Crazy Shit" entries in Alühn 2...
23:05:18 <ABCRic> Patch testing results: two companies with the same colour *facepalm*
23:05:52 <__ln__> W-T-F, running Folding@home GPU client makes my computer output high-pitched sound from somewhere inside the chassis.
23:06:29 <Andel> does anyone know please how to get rid of an AI company in a networked game?
23:06:48 <ABCRic> Andel: all companies have a company id, iirc...
23:06:56 <Eddi|zuHause> __ln__: i only know high-pitched sounds from speakers attached to onboard-sound
23:07:05 *** andythenorth has left #openttd
23:07:27 <Eddi|zuHause> __ln__: otherwise it could be a chemical capacitor on the brink of blowing up
23:07:37 <ABCRic> Andel: that can't be right...
23:07:58 <Andel> let me do a screen dump
23:08:03 <Eddi|zuHause> Andel: and if you check the ai debug window?
23:09:00 <ABCRic> press the rightmost button on the toolbar
23:09:04 <Yexo> Andel: that's impossible, every copmany has a CompanyID
23:09:18 <__ln__> Eddi|zuHause: not coming from speakers, that i've confirmed... this is very odd.
23:10:08 <Andel> and where would you like this screenshot?
23:10:21 <Yexo> Andel: just try "stop_ai <num>" with all numbers from 1 to 15, you can't remove human companies that way so it's save
23:11:51 <Andel> do you want this screenshot?
23:12:30 <Yexo> if you want me to tell you the correct number, yes
23:12:35 <Yexo> otherwise I have no use for it
23:13:30 <Andel> just wondered if we'd borked the impossible :-D
23:13:34 <Andel> thanks for your helps :D
23:14:14 <Yexo> the id is the location in the internal array of companies. If the company isn't in the internal list of companies it doesn't exist. If it is the in that list is has a CompanyID
23:14:27 <Yexo> no CompanyID -> no company.
23:20:35 <CIA-2> OpenTTD: yexo * r21663 /trunk/ (6 files in 2 dirs): -Add: [NoAI]: AIRail::GetName() to get the name of a railtype
23:33:34 <dihedral> static const Command _command_proc_table[] <- how on earth can i get hold of that, defined in command.cpp and i need to access it in network_admin.cpp
23:34:13 <Xaroth> heh, local book store, 60 euro for a book, local online shop, 52 euro for that book.. amazon uk (so shipping internationally).. 34 euro for that same book
23:34:20 <Xaroth> and i thought us dutchies were cheap bastards
23:35:08 <dihedral> and amazon still makes a profit on that
23:35:26 <Yexo> dihedral: not possible, as long as it's declared static
23:35:42 <dihedral> so i declare it extern instead?
23:36:02 <Yexo> no, just remove the "static"
23:36:07 <Xaroth> dihedral: while at it, a poll command for loaded newgrfs, if rubi hasn't already put that in yet :P
23:39:27 <dihedral> i think found is correct
23:40:09 <dihedral> i do not even think Founded exists
23:40:49 <Xaroth> is what you were looking for i think
23:41:14 <Xaroth> I can be useful from time to time :P
23:42:17 <dihedral> Yexo, what do i declare that var then as in the header file?
23:42:21 <Xaroth> i take it it's also AIEventStationBuilt ?
23:42:23 <dihedral> i thought i used extern or something
23:42:36 <Xaroth> (or something similar)
23:42:37 <Yexo> extern const Command _command_proc_table[];
23:42:56 <Eddi|zuHause> dihedral: "found" is past tense of "to find", "founded" is past tense of "to found"
23:43:14 <Yexo> there is no event for a built station, but it's also AI_ET_VEHICLE_CRASHED, and AI_ET_SUBSIDY_EXPIRED
23:43:28 <Xaroth> Yexo: as i said, something similar
23:44:51 <CIA-2> OpenTTD: yexo * r21664 /trunk/src/ (6 files in 3 dirs): -Add: [NoAI] AIEventTownFounded
23:48:01 <dihedral> lengthof(_command_proc_table) <- gives me a: error: invalid application of sizeof to incomplete type const Command []
23:48:41 <Yexo> you can't use lengthof unless the length of the array is known
23:49:07 <Yexo> which is isn't in network_admin.cpp because you haven't declared the length in the header file
23:49:47 <dihedral> work around? apart from declaring the length in the header file
23:49:49 <Yexo> maybe adding const int NUM_COMMANDS = lengthof(_command_proc_table); in command.cpp and using that variable in network_admin.cpp might work
23:50:16 <Yexo> adding "extern const int NUM_COMMANDS;" in the header file
23:50:19 <dihedral> would it not make sense to add another value at the end of the Commands enum
23:58:13 *** TheMask96 has joined #openttd
continue to next day ⏵
