IRC logs for #openttd on OFTC at 2010-03-17
⏴ go to previous day
00:00:22 <andythenorth> I could probably suppress it, but the other FIRS authors prefer that I don't for various good reasons
00:00:33 <Eddi|zuHause> i hate the forum... it doesn't display 90% of the images... and i can't find the reason why
00:01:14 *** Coco-Banana-Man has quit IRC
00:05:02 <DaleStan> Ah. yes. I remember that issue. (Actually, part of the plan for NFORenum is to have it catch that too. (Un)fortunately, renum doesn't quite know that 1A is constant.)
00:07:09 <andythenorth> DaleStan: I don't mind the workaround, it's fine
00:10:19 *** Eddi|zuHause has joined #openttd
00:12:07 *** Adambean has joined #openttd
00:12:46 *** roboboy has joined #openttd
00:13:55 *** Prof_Frink has joined #openttd
00:25:45 *** _Muddy is now known as Muddy
00:48:55 *** robotboy has joined #openttd
02:03:59 <PeterT> Is it true that cargo gets payed for the drop relative to the station sign's location
02:09:46 <Belugas> i guess i'll be hitting bed now
02:20:04 <Eddi|zuHause> a Belugas at this hour? how rare...
02:20:33 <Eddi|zuHause> PeterT: at least it used to be the station sign a while ago
02:21:14 <Eddi|zuHause> my information is more 0.5-ish :)
02:22:03 <PeterT> Eddi|zuHause: When was that last time you played OpenTTD?
02:22:32 <Eddi|zuHause> what does that have to do with knowing game internals?
02:23:44 <PeterT> I was asking randomnly
02:24:28 <Eddi|zuHause> i started a game a few weeks ago, but didn't get far
02:24:39 <Eddi|zuHause> i played a longer game about half a year ago or so
02:24:53 <Eddi|zuHause> cargodist-custom?
02:28:28 <Eddi|zuHause> inhowfar did that information help you now?
03:00:47 *** JVassie_ has joined #openttd
03:25:16 *** rhaeder1 has joined #openttd
04:36:09 *** DanMacK has joined #openttd
06:02:02 *** valhallasw has joined #openttd
07:10:51 <ccfreak2k> <PeterT> Is it true that cargo gets payed for the drop relative to the station sign's location
07:10:55 <ccfreak2k> It makes the most sense to me.
07:11:28 <ccfreak2k> At least if the algorithm doesn't go into too much detail.
07:20:08 *** einKarl has joined #openttd
07:39:39 *** TheMask96 has joined #openttd
07:44:42 *** Wizzleby has joined #openttd
07:56:24 <Rubidium> oh, it's that time of day again? Ah well, good morning then :)
07:57:04 <dih> hehe - someone did not go to bed last night :-P
07:57:11 <dih> at least that's what it sounds like
07:57:29 <Rubidium> just hoped it was still night so I could sleep some mor
07:58:26 <peter1138> hmm, somebody broke greek
07:59:47 <Rubidium> yes, the Greek translator did
08:00:40 *** Cybertinus has joined #openttd
08:02:01 <dih> good morning cybertinitus ^^
08:05:25 <peter1138> doesn't grasp that adding gameplay features as singleplayer only is a big no no
08:07:47 *** roboboy has joined #openttd
08:22:15 *** Terkhen has joined #openttd
08:25:31 <roboboy> is the unified login for the website handled by an LDAP server?
08:25:54 <roboboy> if not how is it implemented?
08:27:28 * Rubidium wonders how hard it is to search for that information
08:29:37 * Rubidium thinks roboboy needs a 101 in news searching
08:29:44 <peter1138> what, first hit on google for 'openttd ldap'? :)
08:33:37 <peter1138> though that wouldn't help if it had been changed later...
08:43:04 *** JVassie has joined #openttd
08:54:49 *** TheMask96 has joined #openttd
08:56:09 <dih> now all you need is openttd->ldap auth
08:56:15 <dih> and ldap stores your uniqu id :-P
08:56:39 <dih> perhaps that should be done via a webservice or some other (openttd style) network service
08:57:24 <dih> client->master master->client master->server
08:58:04 <dih> kick/ban results in server->master ^^
09:02:32 <andythenorth> don't suppose there's a string code for pluralising depending on a value?
09:09:29 <andythenorth> yes there is or yes there isn't?
09:11:35 <peter1138> at least, in the language files
09:11:42 <peter1138> don't think it'll work for newgrf text
09:12:28 <__ln__> what, there's something one can't do with newgrf??!?
09:13:16 <peter1138> well, the plural form may not match the currently selected language's form
09:16:09 <peter1138> also the codepoint to identify a plural isn't fixed
09:16:25 <peter1138> it's part of an enum list that could change
09:18:48 <andythenorth> I figured it would be tricky to pluralise and have multilingual support :)
09:19:17 <andythenorth> the industry window seems slow to update it's text
09:20:39 <andythenorth> the monthly production change cb doesn't appear to trigger cb 3A (or a window redraw - I'm guessing one of the two is required)
09:20:58 <andythenorth> Moving the window does trigger the cb / redraw etc
09:21:18 <andythenorth> that is sub-optimal :|
09:22:41 <andythenorth> that puts the kybosh on my power station :P
09:24:28 <andythenorth> I could run the production cb...that will trigger the redraw quite soon after the production change cb runs
09:26:22 <peter1138> or supply a patch to trigger the redraw
09:26:29 <andythenorth> worthy of a fs entry?
09:31:07 *** lobster has joined #openttd
09:33:09 <andythenorth> is there a string code for the little 'warning' triangle used in the newgrf list window?
09:36:51 <andythenorth> probably for the best
09:48:07 <Hirundo> Would it be possible to enable the newgrf text parameter stack for all strings, instead of in a few selected locations?
09:51:34 <Hirundo> It would allow newgrfs to properly display numbers, velocities, etc. without resorting to parameters to e.g. switch between kph/mph (NARS 2)
09:57:43 <andythenorth> Pikka missed all the excitement :o
10:08:21 <andythenorth> close the industry dammit 9.9
10:10:30 <andythenorth> if industry is type 00 (no closure), newgrf seems to not be able to over-ride that :o
10:11:55 <andythenorth> Yay! *All* my power stations disappeared from the map. Result :D
10:12:03 * andythenorth just thought of a refinement
10:12:11 <andythenorth> fricking refinements :(
10:12:14 <andythenorth> refinements = work
10:28:31 <Pikka> andy: have you experimented with the prospecting/building cost property? and is it just me or does it not work in OpenTTD? :P
10:29:14 <peter1138> andythenorth, "pixel creations for open transport tycoon" ???
10:29:43 <Pikka> what's an open transport tycoon? ;)
10:29:55 <peter1138> "transport tycoon" or "openttd", please?
10:32:23 <__ln__> "transport tycoon" is known to be a registered trademark, so talking about "open transport tycoon" is like asking for unnecessary trouble and attention.
10:33:45 <Ammler> why don't you want attention?
10:33:54 <tokai> Also it would be Open Transport Tycoon *Deluxe* anyway, no? :)
10:34:18 <Ammler> Sounds like you have to hide something...
10:34:50 <Rubidium> Ammler: because the kind of attention you get means all your work on openttdcoop and it subprojects is pointless as there is no OpenTTD to play anymore
10:35:12 <Pikka> and is it just me or does it not work in OpenTTD? :P <- aww shoot, it helps if I get the industry ID correct in the action 0... don't mind me
10:35:46 <peter1138> don't see a change :p
10:36:50 <Pikka> it says openttd in the title :P but still "open transport tycoon" in big letters at the top of the page...
10:37:56 <__ln__> Ammler: i meant trademark/copyright holder's attention
10:38:22 <andythenorth> Who put that there :P
10:38:55 * andythenorth is somewhat sleep deprived
10:39:06 <Rubidium> I think "Not suitable for use in ways that violate good taste." makes them DFSG non-free
10:39:18 <Pikka> andythenorth: blame it on the boogie
10:39:45 <Pikka> Rubidium: it says not suitable, not not permitted...
10:39:53 <andythenorth> Rubidium: it's a disclaimer not a licensing term :)
10:40:18 * Pikka wonders if people will hate me if I incorporate pb_build into my toyland replacement ;)
10:40:35 <andythenorth> does that make you *more* likely to do it :P
10:40:56 <andythenorth> Which is better? Simple to explain, or fairer to player?
10:41:06 <andythenorth> The case would be industry closure rules
10:41:11 <Rubidium> Pikka: just wonder whether people already hate you now. If that's the case then they'll most likely hate you if you incorporate it and as such nothing changes
10:42:00 <andythenorth> when the game starts, five years protection from closure is probably not enough for an industry.
10:42:13 <andythenorth> once cargo has been delivered for the first time, five years is way too generous
10:42:48 <Ammler> __ln__: if "someone" really fears attention in that matter, it might be time to change the name ;-)
10:42:51 <Pikka> andy: do you need to explain the exact mechanics to the player?
10:43:19 <dih> that'll just confuse them even more
10:43:28 <dih> they don't know the exacts of the game anyway :-D
10:43:38 <Pikka> then go for fairer and don't explain it ;)
10:44:20 <Pikka> Ammler: I propose changing the name of OpenTTD to "Open Largely Unrelated Transport Game"
10:44:36 <andythenorth> hmmm I'll just diddle the probability of closure, not the length of time. Easier
10:44:43 <__ln__> Ammler: dunno whose name you are talking about, but OpenTTD's name in the title was changed something like 5 years ago for this precise reason.
10:44:48 <andythenorth> And means I don't have to rewrite my texts (again)
10:44:59 <Ammler> Rubidium: maybe you should start a "name competition" in tt-forums ;-)
10:46:10 <Ammler> __ln__: I am sure, andythenorth isn't the first and the last who does refer TTD of OpenTTD to Tranport Tycoon Deluxe
10:47:09 * andythenorth deletes his website, as it seems to be controversial
10:47:31 * andythenorth undeletes his website. Phew!
10:47:54 <JVassie> I need inspiration to finish these sprites :x
10:48:18 <PeterT> JVassie: Do it, or else
10:48:28 <JVassie> British Modular Stations Set
10:48:46 * andythenorth blew up *OpenTTD*
10:48:51 <Pikka> JVassie: finish those sprites or I'll beat you around the head with a piece of wood
10:49:08 * andythenorth definitely didn't blow up anything to do with a trademarked name
10:49:38 <andythenorth> JVassie: don't we all need a coder?
10:49:39 <Rubidium> where's "NewGRF coding for dummies"?
10:49:52 <JVassie> andythenorth, probably
10:50:03 <JVassie> hopefully though, a complete spritesheet will persuade one
10:50:17 <Pikka> my wiki page is headed "Writing NFO for smarties. Because you're not as dumb as you think you are," Rubidium
10:50:51 <Pikka> trying to help dummies is generally unproductive. ;)
10:51:33 <Pikka> JVassie: stations are the one thing I've never properly attempted to code. I'll be doing some of my own soon though...
10:51:41 <andythenorth> ha ha this is evil
10:51:54 <andythenorth> so if you *never* service the power plant, it will eventually have a low chance of closure
10:51:55 <Terkhen> unless they buy your "for dummies" books
10:52:03 <JVassie> Ive attempted trying to code stations myself
10:52:12 <JVassie> however I get lost in some of the vasics
10:52:16 <andythenorth> if you start delivering cargo, then don't service it, the chance of closure is somewhat higher
10:52:33 <Pikka> oh man, someone's gone and converted a big block of <pre> on my wiki to a proper wiki table...
10:53:14 <Pikka> now if I want to change it I'll have to edit it properly instead of just overwriting it from the text dump D:
10:53:27 <andythenorth> wiki formatting is major suck :(
10:53:35 <peter1138> or you can just overwrite it all from the text dump :D
10:53:45 <Pikka> I'd feel guilty, peter1138 :D
10:53:54 <Rubidium> that's why I leave dumping the changelog on the wiki to others :)
10:54:19 <Pikka> must have taken him ages... D:
10:58:31 * andythenorth now feels power plants are sufficiently evil
10:58:43 <andythenorth> mostly "FIRS is not evil"
10:58:51 <andythenorth> but there has to be something...
11:01:05 <Rubidium> FIRS is really sinister industry replacement set is really sinister industry replacement set ... :)
11:01:31 <andythenorth> FINE - FIRS Is Not Evil
11:02:31 *** DanMacK has joined #openttd
11:06:22 <andythenorth> 81 18 00 \b50 //randomise
11:06:26 <andythenorth> 1 in 50 chance right?
11:06:34 <andythenorth> or do I need to be more bit savvy
11:06:56 <andythenorth> cos all my fricking power stations just announced closure :|
11:07:42 <andythenorth> this is inside cb 3A
11:07:49 <andythenorth> if you see what I mean
11:08:31 <andythenorth> ah there are only 32 random bits available :)
11:09:08 <Pikka> 32 random bits isn't a problem :P
11:09:20 <Pikka> 32 random bits is a dword...
11:09:46 <andythenorth> am I nibbling var 18 wrong?
11:09:49 <peter1138> what happened to your varadjust?
11:10:07 <andythenorth> that is a good question
11:10:22 <andythenorth> this appeared to be working code elsewhere :)
11:10:26 <andythenorth> probably wasn't then?
11:10:43 <Pikka> andy: does it need to be 1 in 50?
11:10:44 <peter1138> well, you ANDed var18 with 50?
11:10:57 <andythenorth> I just need it to be 1 in 'some relatively large number' :)
11:11:09 <Pikka> you could do 1 in 50 with a modulo, but it's probably easier to do, say, 1 in 64 with the and mask
11:11:19 <andythenorth> 1 in 32 would be sufficient I think
11:11:40 <Pikka> for 1 in 32 you'll want to and with... 1F?
11:11:41 <andythenorth> what's a 'low chance of closure' mean to you in probabilty terms?
11:12:05 <Pikka> 1 in 64 = 3F, 1 in 128 = 7F, 1 in 256 = FF
11:12:17 <andythenorth> I can use escapes?
11:12:24 <Pikka> you can, but you wouldn't want to
11:12:46 <Pikka> it's easier to think in bits if you write in hex :)
11:13:54 * andythenorth tries again.....make make make
11:15:54 *** fonsinchen has joined #openttd
11:16:07 <andythenorth> hmmm they didn't all immolate themselves at once this time
11:16:19 * andythenorth leaves the game running for a bit...
11:17:57 <JVassie> Pikka, when is the first emu in UKRS?
11:19:29 <JVassie> AM2 is 1958 apparently
11:31:40 *** oskari89 has joined #openttd
11:44:00 *** KenjiE20 has joined #openttd
11:47:43 <TrueBrain> LOL @ Latest xkcd :)
11:49:01 <Forked> don't forget the mouse-over text
11:49:33 *** _newage_ has joined #openttd
11:50:01 <TrueBrain> I really like this one :)
11:50:01 <TrueBrain> the IQ graph is funny :)
11:53:58 <andythenorth> random bits have me baffled
11:56:11 <andythenorth> I don't get them
11:56:24 <peter1138> they're bits, and they're random
11:56:39 <andythenorth> but the result in a varaction 2 is 1 or 0?
11:57:43 <andythenorth> so anding random bits with 32 gets me anything between 0 and 32?
11:58:03 <Pikka> anding random bits with 0x1F gets you anything between 0 and 32
11:58:07 *** Chris_Booth has joined #openttd
11:58:36 <andythenorth> hmm. Well I have conceptual fail, but my code is correct
11:59:17 <andythenorth> I'm only checking for 0 in the varact 2 anyway
11:59:35 <Pikka> but it's significant because it's 11111 in binary... so it's checking the bottom 5 bits.
12:00:07 <andythenorth> ah I know have a mental picture of it
12:00:41 <andythenorth> when I think of little black and white squares instead of numbers, bit stuff makes more sense!
12:00:45 <Pikka> if you ANDED 0x20 (ie, 32 dec, binary 100000), there's only 1 bit it's checking, so if you checked for 0 you'd get 50% hits
12:08:00 <andythenorth> Pikka: any chance you could review my code? The only way to test this is to run a long game with a lot of power stations!
12:13:46 <Pikka> when NUM_MONTHS_INDUSTRY_CLOSURE_PROTECTED is reached, that ticker stops increasing?
12:16:03 <Pikka> you want it to have a high chance of shutting down if cargo /has/ been delivered?
12:17:01 <Pikka> that seems conceptually to be the wrong way round, but otherwise the code looks good :)
12:17:10 <andythenorth> it is conceptually the wrong way round :)
12:17:34 <andythenorth> but it might just work
12:17:49 <andythenorth> I am trying to prevent that wave of closures five years after the game starts
12:18:29 <Pikka> so it waits for a period, then shuts down all the industries you're serving ;)
12:18:46 <Pikka> well, a quarter of the industries you're serving every month.
12:20:31 <Pikka> you're also not setting "HAS_CARGO_EVER_BEEN_DELIVERED_FLAG" in there anywhere, does that happen somewhere else?
12:20:55 <andythenorth> in the production cb for cargo delivered
12:21:11 <andythenorth> which also resets LOCAL_DELIVERY_TICKER_STORE
12:23:07 <andythenorth> so I haven't screwed up the random closure code?
12:24:58 <Pikka> I don't know, I didn't check what the callback results mean, but if you say so :)
12:26:12 * andythenorth checks the results
12:26:28 <andythenorth> that's probably that done then :o
12:26:44 <andythenorth> dunno if it's a good idea
12:46:44 *** Eddi|zuHause2 has joined #openttd
12:46:57 *** Coco-Banana-Man has joined #openttd
12:57:00 *** Progman has joined #openttd
13:02:00 <TrueBrain> and this is the moment you find out we hav eno ubuntu users ;)
13:04:33 <Rubidium> glx: it's not DST here yet! :)
13:06:19 <blathijs> Rubidium: You've added a ubuntu builder to the compile farm?
13:06:59 <Rubidium> cause I had hopes Squeeze would install on Lucid, but that's not going to be the case in a few weeks anymore
13:08:16 <blathijs> I can hijack $gf's Ubuntu system tonight, if I remember to :-)
13:12:19 <Rubidium> that would be lovely
13:17:22 *** robotboy is now known as roboboy
13:25:08 *** Eddi|zuHause2 is now known as Eddi|zuHause
13:35:38 * Pikka toddles off to make dinner
13:36:04 *** fjb is now known as Guest1301
13:41:55 *** Chruker has joined #openttd
13:58:20 <Eddi|zuHause> if all mad fools would get out of here, there would be noone left to develop the game...
14:01:40 *** Grelouk has joined #openttd
14:23:37 <Belugas> nope, since we'd all be outside having fun with those fools
14:23:51 <Belugas> you know what they say: the more fools, the more fun!
14:30:46 <Eddi|zuHause> <TrueBrain> the IQ graph is funny :) <-- the question is, are the people using the internet really smarter than average, are they fooled by "fake" intelligence tests, or are they plain lieing?
14:31:55 <Eddi|zuHause> besides, IQ numbers are useless, as there are multiple incomparable scales, that only have in common that 100 is the average...
14:32:26 <Rubidium> Eddi|zuHause: if you take a test (with scaling) of 10 years ago your IQ will be higher
14:32:51 <Rubidium> so I assume that they were using a test from before the whole copyright stuff, i.e. early 1900s
14:33:49 *** ChanServ sets mode: +v tokai
14:33:52 <Rubidium> nevertheless, the tests aren't quite accurate about 160 or so (so they just cap it there)
14:34:38 <Eddi|zuHause> there are scales that don't spread much beyond 130
14:35:00 <Rubidium> and the tests on TV from the last few years make it even impossible to get more than like 150
14:35:58 <Eddi|zuHause> i wouldn't believe any number higher than 120 anyway...
14:36:02 <peter1138> intelligence is limited to knowing when to google
14:37:54 <Eddi|zuHause> i do believe that people have a 23cm penis, though :p
14:39:42 *** Biolunar has joined #openttd
14:41:08 <Rubidium> Eddi|zuHause: let me say that people who are smart enough to know how IQ work probably don't need to brag about their IQ, the rest just thinks they're smarter than the rest
14:44:38 <lennard> since in nl we use . to seperate the decimals, I'll just truncate that to 4 :P
14:45:17 <aber> infact it is 18.446.744.073.709.600.000, i don't know. Is my computer broken?
14:47:21 <peter1138> that's a lot of zeros at the end
14:49:03 <Eddi|zuHause> that's the broken part :)
14:49:53 <Eddi|zuHause> 18.446.744.073.709.551.616 <- actual 2^64
14:50:31 <peter1138> don't be silly, you can't have that many decimal points in a number
14:51:01 <Eddi|zuHause> better than the swiss people, they have apostrophes in their numbers...
15:00:27 *** Chris_Booth is now known as Guest1311
15:00:29 *** Chris_Booth has joined #openttd
15:03:40 *** Chris_Booth is now known as Guest1314
15:03:42 *** Chris_Booth has joined #openttd
15:07:39 *** Chris_Booth is now known as Guest1316
15:07:41 *** Chris_Booth has joined #openttd
15:11:15 <Eddi|zuHause> hm... this error message "inflate failed", could that be made more descriptive so that it mentiond "possibly incomplete download" or something?
15:14:02 <__ln__> the DarkVater approach would be not to show the error message at all.
15:14:23 *** Belugas has joined #openttd
15:14:23 *** ChanServ sets mode: +o Belugas
15:18:26 *** Chris_Booth is now known as Guest1319
15:18:28 *** Chris_Booth has joined #openttd
15:22:09 *** Rhamphoryncus has joined #openttd
15:24:08 *** PeterT is now known as Guest1320
15:24:27 *** Guest1320 is now known as PeterT
15:27:49 *** Chris_Booth has joined #openttd
15:36:08 *** Adambean has joined #openttd
15:51:12 <andythenorth> I haven't a clue how to do date maths in nfo
15:53:12 <Eddi|zuHause> anything in particular?
15:53:59 *** Phoenix_the_II has joined #openttd
16:09:22 *** ChanServ sets mode: +v tokai
16:10:01 * andythenorth design dichotomy
16:10:05 <andythenorth> makes my brain ache
16:10:15 <andythenorth> normally comes out in the wash though :)
16:10:53 <andythenorth> question: why should industries close at all? Gameplay reasons?
16:11:39 <Belugas> cough ** rea ** cough * lism*** cough... ARGH!!!!
16:11:45 <Rubidium> depletion of resources?
16:12:15 <andythenorth> I'll narrow the question: processing industries
16:12:21 <andythenorth> is it just irritating?
16:12:22 <Rubidium> dangerous situations, e.g. a fireworks depot in the middle of a town isn't such a good idea
16:12:41 <Rubidium> or something petrochemical in a town... same idea
16:12:52 <andythenorth> in my experience only two things happen when an industry closes:
16:12:56 <Rubidium> you'll see that such companies get relocated out-of-town over time
16:13:06 <andythenorth> (a) you've just built a route to it, and you're screwed. so game over
16:13:11 <Rubidium> (in the real world ofcourse)
16:13:16 <andythenorth> (b) or you have enough money that you don't care
16:14:20 <andythenorth> so basically it's either "stupid game ruined my game" or "meh"
16:15:37 <Rubidium> if you go bankrupt the move happens generally in the first year
16:15:39 <Eddi|zuHause> andythenorth: only reasons imho should be (a) annoy the player [this should be disable-able], (b) unserviced, (c) dry out a cargo chain after a certain date [like oil wells in temperate]
16:15:55 <andythenorth> (a) I don't like any more
16:16:05 <andythenorth> (b) is the one I am having trouble with
16:22:24 * andythenorth has yet another idea
16:24:09 <andythenorth> Rubidium: I have to get it released before the nightlies stop compiling on OS X
16:26:56 <andythenorth> There is currently a 5 year 'grace period' for industries to close due to lack of supply.
16:27:19 <andythenorth> it's not enough, and it leads to a mass extinction of secondary industry 5 years into the game
16:27:24 <Rubidium> FYI: the compile farm is still capable of building OS X binaries; it's just still horribly slow
16:27:36 <andythenorth> And I could still send you an intel mac
16:28:03 <andythenorth> anyway, back to problems that *can* be solved....a grace period of twenty years seems more appropriate
16:28:13 <andythenorth> large industrial plants are mothballed, not just destroyed
16:28:20 <Rubidium> you could, but I won't burn my fingers on Apple's APIs
16:28:37 <andythenorth> and then I need some way to spread out the closing of industries so there is no mass exctinction
16:29:51 <andythenorth> I could store a random seed in persistent storage at each industry - this would be used to spread closures out
16:31:30 <andythenorth> or I could just prevent closure of secondary industry (which is very easy)
16:33:41 <Eddi|zuHause> andythenorth: i'd try to limit the number of industries of the same type that can close within a certain time (year?)
16:33:57 <andythenorth> that would need a patch
16:34:05 <Eddi|zuHause> so industries will not close all at the same time, but instead one by one
16:34:35 <andythenorth> I could keep running the industry count var, and do something with comparing that I suppose
16:35:22 <andythenorth> Eddi|zuHause: I think it's a good suggestion. I don't think the implementation is trivial
16:35:27 <Eddi|zuHause> you have global variables, right? so you just set a bit for each industry type, if you close one such industries, when the bit is already set, you skip closing the industry. at the end of the year, you reset the variable?
16:35:57 <andythenorth> depends what you mean by 'global' variables
16:36:24 <Eddi|zuHause> one that is accessible from any place in the grf
16:36:31 <andythenorth> not as far as I know
16:36:44 <Eddi|zuHause> iirc you can use parameters for that
16:38:29 <andythenorth> Eddi|zuHause: I can't see how a varaction 2 is going to write to one of the grf parameters :o
16:38:35 <andythenorth> if it could, it would be....handy
16:38:51 <Eddi|zuHause> hm... i'd have to read up on that...
16:42:11 <Eddi|zuHause> but Action D only seems to be run on grf activation, not during callbacks and stuff...
16:44:05 <andythenorth> Eddi|zuHause: I can see how to do it with persistent storage at the industry....
16:44:35 <andythenorth> I would use the industry count var
16:45:10 <andythenorth> I can forsee a strange whack-a-mole side effect though!
16:45:39 <Eddi|zuHause> what do these "store" operators of varaction2 do?
16:45:54 <andythenorth> stick values in registers or persistent storage
16:46:12 *** devilsadvocate has quit IRC
16:49:35 <Eddi|zuHause> how do you propose to do this with an industry-local storage only?
16:51:54 *** DJNekkid has joined #openttd
16:54:24 *** snorre_ has joined #openttd
16:55:08 <andythenorth> Eddi|zuHause: for each industry: once per year, count and store the number of industries of type x; when the closure cb runs, test if the count has changed. If it is the same, allow closure. If it is different, prevent closure
16:56:07 <Eddi|zuHause> so you can end up with repeated open/close industries in some corner cases
16:56:38 <andythenorth> Currently my preference is to prevent closure entirely
16:56:45 <Eddi|zuHause> i don't think that's going to be very dramatic
16:57:13 <Eddi|zuHause> no closure at all? that sounds silly...
16:58:50 *** frosch123 has joined #openttd
16:59:30 <andythenorth> Eddi|zuHause: something smells wrong about doing this in nfo
16:59:36 <andythenorth> it seems like it should be in game
16:59:51 <andythenorth> *so* many players have reported this as an irritant
17:00:27 <Eddi|zuHause> andythenorth: i have only ever heard of complaints about ECS, never about the default industries
17:01:19 <andythenorth> on large maps there will often be a mass extinction of industries after five years
17:01:54 <andythenorth> the solution you described would solve that, but it would be better patched for...the nfo route is baroque
17:04:58 * andythenorth has spent most of today watching industries close on fast-forward
17:05:07 <andythenorth> not the most exciting day in recent times
17:11:41 * andythenorth oblique strategies time
17:11:59 <andythenorth> "Faced with a choice, do both"
17:14:07 *** Rexxars has joined #openttd
17:15:22 *** devilsadvocate has joined #openttd
17:24:29 <andythenorth> so lets say I want to randomly choose one value from 4 ,8, 12, 16
17:24:36 <andythenorth> how do I do that cleanly in nfo?
17:32:19 <Eddi|zuHause> mask with 0x3, add 1, shift by 2
17:33:49 <Eddi|zuHause> or mask with 0xB (=0xF-0x3) and add 4
17:34:44 <Eddi|zuHause> 0xB is 1100 (i believe), so you get values 0,4,8,12
17:35:07 <andythenorth> does this come in a pre-written nfo flavour :)
17:35:22 <Eddi|zuHause> isn't that a simple varadjust?
17:35:24 <andythenorth> sorry, normally I try, but babies steal my sleep
17:35:51 <andythenorth> actually just the one
17:36:10 <glx> Eddi|zuHause: actually 0x1100 is 0xC
17:36:26 <Eddi|zuHause> glx: possibly, i can't count :=
17:36:36 <Eddi|zuHause> especially not the alphabet backwards :)
17:38:42 <Eddi|zuHause> andythenorth: so you set bit 6 in shift-num (=0x40?), then 0xC as and-mask, 0x4 as add-val and 0x1 as divide-val
17:40:19 <Eddi|zuHause> results in 40 0C (00 00 00) 04 (00 00 00) 01 (00 00 00)
17:40:42 <Eddi|zuHause> where the numbers in () are dependent on the bit-width
17:41:26 * andythenorth thinks of a way to test that
17:41:38 * andythenorth needs a cb that runs when an industry is built
17:41:56 <andythenorth> cb 14B might do that
17:44:02 * andythenorth is too tired for nfo!
17:47:00 <andythenorth> ha ha completely different idea, flash game style coding
17:47:16 <andythenorth> I however realise this channel is becoming a monologue, by me
17:47:32 <andythenorth> I unusually have time on my hands :o
17:49:35 *** |Jeroen| has joined #openttd
17:50:54 *** Eddi|zuHause has joined #openttd
17:59:25 <Belugas> mind you opening your hands, i might grab a few of that precious time!
18:00:09 *** Chris_Booth has joined #openttd
18:01:11 <andythenorth> sometimes doing things 'wrong' is just nicer
18:10:06 <Belugas> yeah.. like keep playing that false note for a while and it will become right :D
18:17:31 <DJNekkid> is the grfID in little endian?
18:20:34 <Eddi|zuHause> DJNekkid: why does that matter?
18:20:55 <Eddi|zuHause> it's never converted to a decimal number anywere
18:21:09 <Eddi|zuHause> or used as a number at all...
18:21:38 <glx> except to compare it with another GRFID
18:21:39 <DJNekkid> how i would write an action 7/9 that skips if this or that ID is active :)
18:22:13 <DJNekkid> and i know the grfID
18:22:15 <Eddi|zuHause> just write it the same way as it appears in action 8
18:22:38 <Hirundo> grfID or vehicle/station ID?
18:23:32 <Eddi|zuHause> DJNekkid: it is generally safe to assume that anything in NFO is little endian
18:23:58 <Ammler> I thought, everything except the GRFID :-)
18:24:54 <DJNekkid> Hirundo: i figuered out what to do...
18:25:28 <DJNekkid> it were the grfID in were AABBCCDD or \dxAABBCCDD :)
18:26:41 <Ammler> [19:25] <DJNekkid> it were the grfID in were AABBCCDD or \dxAABBCCDD  <-- isn't that big endian?
18:28:45 <Eddi|zuHause> Ammler: the first one is little endian, the second one is big endian, and he asked which one is right
18:28:57 *** lobster has joined #openttd
18:29:27 <andythenorth> renum is stumbling on this \2* 1A 20 \d04 //multiply a bit for luck
18:29:30 <Eddi|zuHause> (and the answer is: the first)
18:29:43 <andythenorth> that's supposed to multiply previous value by 4
18:30:06 <andythenorth> 1A is the var to create constant, the 20 is there because I'm going to store it in the next operation
18:30:33 <Eddi|zuHause> i have never used renum
18:30:36 <andythenorth> do I need to mask it?
18:31:10 <Hirundo> What is the error message?
18:32:52 *** De_Ghosty has joined #openttd
18:34:31 <Hirundo> \2* 20 \d04 //multiply a bit for luck <- missing variable (1A) here?
18:36:40 <andythenorth> doesn't fix it though :)
18:37:24 <blathijs> DaleStan: Is Makefile.common in grfcodec still relevant? A comment in there says it is used for both the normal Makefile and Makefile.win, but the normal Makefile includes stuff from Makefil.common already and Makefile.win doesn't seem to exist?
18:38:27 <Hirundo> 18 20 07 //randomise <- no dword-sized and-mask here, I think
18:39:08 <blathijs> DaleStan: Seems Makefile.win got removed in r2167, so I guess Makefile.common should go as well?
18:42:27 <andythenorth> are there any permanently available random bits lurking in an industry anywhere?
18:43:33 <Eddi|zuHause> put them in the persistent storage?
18:45:31 <Eddi|zuHause> on a quick glance i find random bits for industrytiles...
18:47:24 <Eddi|zuHause> the Industry struct has a "uint16 random; ///< Random value used for randomisation of all kinds of things"
18:48:08 <andythenorth> Eddi|zuHause: yep could put them into persistent storage. I'd need to use cb 14B and I can't be bothered to figure that out today :o
18:48:24 <andythenorth> unless there is another cb that runs when industry is built
18:48:44 <andythenorth> If I get cb 14B wrong...I screw the cargos for the industry :)
18:49:26 *** Alberth has joined #openttd
18:53:24 <Eddi|zuHause> there certainly are random bits for industries, but i don't find how you access them...
18:54:15 <frosch123> via randomaction2 or var 5f
18:56:01 <CIA-6> OpenTTD: translators * r19440 /trunk/src/lang/ (10 files in 2 dirs): (log message trimmed)
18:56:01 <CIA-6> OpenTTD: -Update from WebTranslator v3.0:
18:56:01 <CIA-6> OpenTTD: catalan - 2 changes by arnau
18:56:01 <CIA-6> OpenTTD: czech - 7 changes by TheLamer
18:56:01 <CIA-6> OpenTTD: frisian - 97 changes by Fopper
18:56:02 <CIA-6> OpenTTD: greek - 4 changes by
18:56:02 <CIA-6> OpenTTD: icelandic - 34 changes by baldur
18:56:24 <Eddi|zuHause> case 0x5F: return (object->GetRandomBits(object) << 8) | object->GetTriggers(object); <-- that's probably the place
18:56:47 <andythenorth> frosch123: is var 5F consistent, i.e. will it always return the same random bits for a given industry?
18:56:59 <andythenorth> (that would seem undesirable generally, but it's what I am after)
18:57:24 <CIA-6> OpenTTD: rubidium * r19441 /trunk/src/lang/greek.txt: -Fix (r19437): broken Greek translation.
18:57:36 <CIA-6> OpenTTD: rubidium * r19442 /trunk/ (known-bugs.txt readme.txt): -Update: readme about openmsx and make it easier to find the pulseaudio-ubuntu-slowness bug in the readme/known-bugs
18:57:41 <Eddi|zuHause> i don't know what a trigger is, but otherwise it seems to remain constant
18:57:47 <PeterT> CIA-6: You're slow today
18:57:58 <CIA-6> OpenTTD: frosch * r19443 /trunk/src/industry_cmd.cpp: -Fix [FS#3701]: Mark industry windows dirty more often.
18:57:59 <frosch123> andythenorth: it is only randomised on creation or by using an randomaction2 with triggers
18:58:49 <andythenorth> so it could change during the game?
18:59:41 <Eddi|zuHause> andythenorth: only when you want it to (i believe)
18:59:42 <frosch123> only if you tell them to change
19:03:38 *** rhaeder has joined #openttd
19:05:40 *** rhaeder2 has joined #openttd
19:08:21 *** De_Ghosty has joined #openttd
19:15:44 <CIA-6> OpenTTD: rubidium * r19444 /trunk/bin/ (8 files in 2 dirs): -Update: base set translations for Polish, Slovak and Turkish.
19:16:05 *** Chris_Booth has joined #openttd
19:19:21 <peter1138> andythenorth, persistent random bits are very desirable, heh
19:19:42 <peter1138> think randomised liveries, for example
19:20:04 <Eddi|zuHause> industry producing random cargos :)
19:32:28 *** ajmiles has joined #openttd
19:33:10 <andythenorth> renum: //!!Warning (86): Offset 4: Testing nonexistant variable 5F.
19:33:26 <andythenorth> from 89 5F 20 \dx07
19:36:30 <Zuu> Yexo: Is it a design desicion or a bug that AI libraries can't be cyclic dependent on eachother?
19:37:47 <Eddi|zuHause> andythenorth: did you try ingame? maybe it's a renum bug
19:38:13 <andythenorth> seems to work. I'm getting signed values though
19:38:45 <Eddi|zuHause> whether a value is signed or unsigned is purely interpretation dependent
19:38:45 <andythenorth> so if I want random 0-7, how do I mask for that
19:39:11 <andythenorth> It's a dword "Feature-specific random data: triggers in low byte, bits in other three bytes. Bits of the variable not associated with random or trigger bits are reserved"
19:39:44 *** Nite_Owl has joined #openttd
19:39:54 <Eddi|zuHause> a shift of 8 and mask of 0F 00 00 00, probably
19:46:41 <andythenorth> line 4 is my attempt to get random 0-7 from var 5F
19:48:17 <Eddi|zuHause> no, not \dxBLAH... just BLAH
19:48:35 <Eddi|zuHause> shift 8, to mask out the trigger bits
19:49:03 <frosch123> or \dxFF ? do you want values 0-7 or bits 0-7 ?
19:49:14 <Eddi|zuHause> \dx0F, i presume
19:49:37 <Eddi|zuHause> yeah, \dx07 is for 0-7
19:51:33 *** ajmiles2 has joined #openttd
19:58:11 *** fonsinchen has joined #openttd
20:02:32 <Yexo> <Zuu> Yexo: Is it a design desicion or a bug that AI libraries can't be cyclic dependent on eachother? <- where is that not possible? In the game or on bananas?
20:02:54 <Yexo> but imo if two libraries are cyclic dependent you should merge them in one library
20:03:25 <Zuu> That is what I'm trying to do or rather trying to keep them organized in classes but in the same big library.
20:03:31 <andythenorth> there's a fast-forward bug on OS X right?
20:04:21 *** ChanServ sets mode: +v tokai
20:04:25 <andythenorth> that is a very useful bug today
20:04:29 <Zuu> Trying to break out common stuff from CluelessPlus and PAXLink will most likely result in a big library on bananas that has cyclic dependencies inside it.
20:05:10 <Zuu> Btw, using require inside a library is that allowed? OpenTTD do not complain, but the classes seam to appear at global scope in the AI that uses the library.
20:05:25 <Zuu> In my case that is not a problem, but for someone else it could become.
20:05:33 <Yexo> each library can only have one class
20:05:51 <Yexo> if it has more then one class the others will end up in the global scope currently, but that is bad
20:06:03 <Zuu> Yep, but you can create static members in that class with instances of other classes.
20:06:56 <Zuu> I agree that having them in global scope is bad, still I'm not really happy with merging all my utilities into just a single class.
20:07:21 <Yexo> you could use subclasses
20:07:37 <Yexo> ehm, nut subclasses, but child classes
20:08:12 <Zuu> If I can figure out how that will work I'll might make a script that compile the main.nut file from files with the subclass files.
20:09:15 <Zuu> Child classes is basically a nested class definition?
20:09:48 <peter1138> 20:03 < andythenorth> there's a fast-forward bug on OS X right?
20:10:27 <andythenorth> peter1138 in the first couple of months...super speedy. I'm sure Rubidium mentioned it in a discussion somewhere :o
20:10:38 <andythenorth> kind of useful for testing industry closure though
20:11:05 <peter1138> fast-forward goes as fast as your system can
20:11:36 <Yexo> it the first few months go very fast and then it slows down an AI probably started at that time
20:12:24 <andythenorth> thanks, that could save me a lot of time staring at industry windows
20:17:27 <Zuu> The Road Pathfinder, v3 has a class "Road.Cost" defined after the main class (after the ending } ), this class is not a subclass right?
20:17:59 <andythenorth> woah that's strange industry generation on big map
20:18:26 <Zuu> (sorry, for my bad food examples)
20:19:05 <andythenorth> 2578 coal mines, 1517 power stations, 963 sawmills. 1 of everything else :o
20:19:14 <andythenorth> I don't *think* that's my fault
20:19:17 *** devilsadvocate has quit IRC
20:19:27 *** De_Ghosty has joined #openttd
20:20:13 <CIA-6> OpenTTD: frosch * r19445 /trunk/src/ (rail_gui.cpp road_gui.cpp): -Change: No need to drag an area when overbuilding bridges.
20:20:18 <Zuu> no, now it says "expected '=' on the same row.
20:23:34 <andythenorth> does \2^ raise to a power?
20:23:56 <Eddi|zuHause> i'd assume bitwise xor
20:25:02 <Eddi|zuHause> power is usually only available for floating point arithmetics
20:25:17 <Eddi|zuHause> as integer power is quite calculation intensive
20:26:02 <frosch123> you can only multiply with powers of two by shifting
20:26:39 <Eddi|zuHause> pow(a,b) can be calculated by the "square and multiply" algorithm
20:27:03 <Eddi|zuHause> but for float values, pow(a,b) is exp(b*log(a))
20:27:37 <Eddi|zuHause> where exp and log translate to FPU commands
20:28:00 <andythenorth> some var \2* some var again does the trick :)
20:30:24 <Eddi|zuHause> if you want integer power for arbitrary values, the algorithm is: pow(a,b) { if (b==0) return 1; result := pow(a,b/2); result*=result; if (b is odd) result*=a; return result; }
20:31:46 *** devilsadvocate has joined #openttd
20:31:47 <andythenorth> doesn't look much like nfo to me
20:32:09 <Eddi|zuHause> turning that into NFO is easy, just have to read it backwards :)
20:32:47 <Zuu> Yexo: After experimenting a bit I came up with this: http://paste.openttd.org/225288 - but I have not yet figured out how to not need to define the functions in the same place as they are declared.
20:33:29 <Zuu> It's basically an un-named class so I guess it's hard to not define the functions straight away in the declarations.
20:33:33 <Eddi|zuHause> you can probably also turn that into an iterative version
20:33:52 <Yexo> Zuu: yes, s/hard/impossible/ even
20:34:21 <andythenorth> oh. Mass industry extinctions still happening :|
20:35:34 <Eddi|zuHause> for(res=1;b!=0;r=b&1,b>>=1) { res*=res; if (r) res*=a }
20:35:39 <Eddi|zuHause> or something similar...
20:35:59 <Zuu> A problem with not being able to split the classes into files is that you can't solve cyclic dependancies then. Unless squirrel support adding functions after a class has been declared.
20:36:20 <Eddi|zuHause> hm, that's probably not correct
20:37:19 <Yexo> Zuu: you can call a function before it is defined, as long as the function definition is encountered before the call is actually executed
20:37:36 <Eddi|zuHause> b must be unrolled from the other end
20:38:30 <Eddi|zuHause> so b<<=1,r=get_overflow_flag_somehow
20:38:41 <Zuu> Hmm, got the problem that it didn't find classes when I just inlined all my code in main.nut at global scope.
20:39:18 <Yexo> the main class is renamed by openttd
20:39:58 <Zuu> Yep, but the main class was in this case rather empty. They used the global-scope names of the sub-libraries.
20:42:12 <Zuu> Ah, sorry, it was the main class that didn't find the global scope classes, since I didn't move the main class to the bottom.
20:42:43 <Zuu> Doing that, the subclasses can be interdependent with cyclic dependencies (at global scope) in a single file.
20:44:24 <Zuu> So mayby using the child classes will work technically, but would produce a main.nut that hardly could be used as doccumentation.
20:45:08 <Yexo> you could work in multiple files and cat them together before releasing
20:45:13 <Yexo> a bit of a hack, but that could work
20:47:47 <Zuu> That was my plan. Though I was planing to use some more logic in the script to not have move around the code, but that might just make a mess of it.
20:48:19 <andythenorth> code random is never quite as random as it ought to be :o
20:48:28 <Zuu> I usually put the documentating comments about how to use a function at the declaration rather than at the definitions.
20:48:43 <Zuu> A such file can't be cated togeather.
20:59:16 * andythenorth is going to fix industry closure if it kills him (possible!)
21:00:01 *** Grelouk has joined #openttd
21:02:01 *** Terkhen has joined #openttd
21:03:39 <Alberth> andythenorth: just remove it from the newgrf specs, so the program code can be fixed :p
21:04:05 <andythenorth> Alberth: tempting
21:04:16 <andythenorth> not my call though :)
21:05:26 * frosch123 wonders why petert still did not become coder of cets
21:06:26 * andythenorth suspects some law of large numbers
21:07:20 <andythenorth> if 10 industries each sum a lot of random numbers, they'll probably get the same result?
21:08:21 <frosch123> then you need different frequencies
21:09:17 <frosch123> small noise which changes often, larger noise which only changes over years
21:10:06 <andythenorth> frosch123: here is my current code
21:10:37 <Eddi|zuHause> andythenorth: it can only work if the industries have another protection period after an industry closed, or if the closing probability depends on the number of closed industries
21:10:38 <frosch123> instead of checking every month for 1% closure, compute a random number of months when to check the next time or so
21:10:42 <Belugas> freedom freedom, we will not obey
21:10:57 <Belugas> freedom freedom take them all away!
21:11:12 <Eddi|zuHause> frosch123: there seems to be no global storage
21:11:22 <Eddi|zuHause> only industry-local
21:11:32 <frosch123> yup, what's the problem with that?
21:11:56 <Eddi|zuHause> frosch123: so what happens with one industry cannot easily influence what happens with other industries
21:12:21 <frosch123> you talk about 5 year protection after building. but you should consider a yearly chance of 10% to start a 5 year protection period
21:13:16 <frosch123> don't make industries change production monthly by +- 5%, but make them decide to increase production by 20% over the next 5 years
21:13:37 <Eddi|zuHause> hm, that might sound interesting
21:13:42 <frosch123> make stuff random, but random in big scale. no continuous small stuff
21:14:32 <andythenorth> frosch123: sounds like I should learn to work with dates in nfo :|
21:14:55 <frosch123> dates are stupid :) count months in the monthly callback
21:15:46 <frosch123> that is more fool proof wrt. changing date by cheat, starting before 1920 or even using daylength
21:16:22 <andythenorth> so wait n months (random) before trying to do the closure....fewer checks, less chance that all industries will converge on similar values
21:18:36 <andythenorth> still smells like a game thing, I think it's a problem for all industry sets :)
21:19:41 <andythenorth> Instead of monthly production change cb, I could just wait for the random one to get called. Don't really know how that works though.
21:20:51 <frosch123> true, the random change might also be useful for that
21:21:17 <frosch123> then you could do smaller changes +-1% in the monthly thingie, and bigger scale decisions in the random one
21:22:09 <frosch123> e.g. decide in the random one to increase by +-1% every month until the next random change appears
21:22:41 <andythenorth> does the random one apply to secondary industries? The documentation around it is unclear
21:22:50 <frosch123> (that is a linear increase/decrease between the random change callback)
21:23:48 <frosch123> the callbacks are called the same for all industries
21:24:12 <Eddi|zuHause> andythenorth: i presume that it applies to all industries, but for secondary industries only for the cargo that gets produced without any input
21:24:36 * andythenorth contemplates watching yet another game for 10 years on fast forward :o
21:28:43 <andythenorth> does the random production change scale with map size, or number of industries in game?
21:35:09 <frosch123> if you start a 2048x2048 map with a single industry, it will change production every few days :p
21:36:04 <andythenorth> so if there are very few industries on that map, they'll get culled fast
21:36:17 <andythenorth> and as industries close, it's ever more likely that the remaining ones close?
21:36:57 <frosch123> well, about the closure ... no idea
21:37:33 <andythenorth> as industry count decreases, frequency they are chosen by random cb increases?
21:39:22 <frosch123> [22:38] <frosch123> if you start a 2048x2048 map with a single industry, it will change production every few days :p <- twice a day actually
22:00:33 *** De_Ghosty has joined #openttd
22:05:14 * andythenorth thinks upside down and comes out right side up
22:05:41 <Zuu> Yexo: will BaNaNaS complain if you include other files than main.nut and library.nut in a tar? Eg. can I include my source files there or do I need to refer people to somewhere else to get the source files?
22:06:17 <Yexo> I think you can include as many *.nut files as you want
22:06:38 <Zuu> Ok, then I'll try when it's time for BaNaNaS.
22:07:45 *** Progman has joined #openttd
22:07:59 <Zuu> I've done my ruby script now.
22:10:27 <Zuu> Now that I got you both here, the one of you Yexo and TrueBrain who can add projects to noai.openttd.org can I asks for "CluelessPlus" and "SuperLib"?
22:10:40 <TrueBrain> Yexo can do that :)
22:11:16 <Yexo> "cluelessplus" and "superlib" as identifiers?
22:11:22 <Yexo> the identifier is part of the url and can't be changed later
22:11:36 <Yexo> hmm, is "lib-super" also fine?
22:11:49 <Zuu> yea, that's probably better then.
22:12:02 <Yexo> and then and "ai-cluelessplus" then too
22:14:30 <Yexo> you might have to wait up to an hour for the svn repo to be created
22:14:42 <Yexo> it was *:14 or *:18 I think, but that might have changed
22:20:36 *** roboboy has joined #openttd
22:20:43 <Zuu> Ok, I'm off for bed anyways. Sleeping over at the office I need to make sure I'm up and ready before anyone else arrives in the morining. :-)
22:21:14 <Zuu> I've posted a zip with the library and the ruby script at tt-forums if you are interested.
22:21:17 <CIA-6> OpenTTD: rubidium * r19446 /branches/1.0/ (41 files in 9 dirs):
22:21:17 <CIA-6> OpenTTD: [1.0] -Backport from trunk:
22:21:17 <CIA-6> OpenTTD: - Fix: Mark industry windows dirty more often [FS#3701] (r19443)
22:21:17 <CIA-6> OpenTTD: - Update: Readme about openmsx and make it easier to find the pulseaudio-ubuntu-slowness bug in the readme/known-bugs (r19442)
22:21:17 <CIA-6> OpenTTD: - Language updates
22:31:21 *** De_Ghosty has joined #openttd
23:20:30 <CIA-6> OpenTTD: rubidium * r19447 /tags/1.0.0-RC3/ (5 files in 4 dirs): -Release: 1.0.0-RC3
23:31:56 *** deghosty has joined #openttd
23:39:18 <andythenorth> Eddi|zuHause: you're right about needing to check for other similar industry closing. Think it's the only way to prevent 'waves' of closure
23:39:31 <andythenorth> so now I'll have to code that :)
23:40:09 <Pikka> andy: or just add a random amount to the protected time ;)
23:41:04 <andythenorth> Pikka: you've missed all that fun :)
23:41:09 <andythenorth> That's kind of happening
23:41:25 <andythenorth> but put enough random numbers together, the effect stops being random :)
23:42:24 *** De_Ghosty has joined #openttd
23:42:37 <andythenorth> I'm going to have to count the buggers :o
23:44:03 <Eddi|zuHause> andythenorth: other option is: on creation you make an even distribution of protection periods between 5 and 30 years
23:45:25 <Eddi|zuHause> that will spread out the closing wave very far
23:45:57 <Rubidium> just create one industry of everything at map creation!
23:46:14 <Rubidium> then there won't be a big wave of closures :)
23:46:37 <Eddi|zuHause> Rubidium: there won't be an industry network either :)
23:48:39 <Eddi|zuHause> or make the closure chance dependent on game time, like 0.1% per year since game start
23:49:06 <Eddi|zuHause> so you have a progressing rate, instead of a protection period and a jump
23:49:13 <andythenorth> add enough inustries, they'll still end up in sync
23:49:51 <Eddi|zuHause> in late game you have like a 10% closure chance for new industries, so you'll have to incorporate them into your network quickly
23:50:54 <Pikka> andy: obviously if the number of industries that need closing is greater than the number of possible closing date, you're going to have some collisions?
23:51:45 <Eddi|zuHause> hm... either the cat has grown too big, or the empty area of my desk has grown too small
23:52:04 <Eddi|zuHause> no matter how the cat lies, it always ends up with some part on the keyboard
23:54:38 <Terkhen> hmmm... too many vehicles in game
23:57:53 <andythenorth> tomorrow...I make friends with industry var 67
23:59:03 <CIA-6> OpenTTD: rubidium * r19448 /extra/website/general/utils/agent_detection.py: [Website] -Add: detection of Ubuntu user agents
continue to next day ⏵
