Log for #openttd on 19th November 2014:
Times are UTC Toggle Colours
00:02:14  *** ginko_ [] has joined #openttd
00:02:31  *** luaduck is now known as luaduck_zzz
00:09:06  *** ginko [] has quit [Ping timeout: 480 seconds]
00:31:27  *** b_jonas [] has quit [Ping timeout: 480 seconds]
00:33:50  *** pxr [] has quit [Read error: Connection reset by peer]
00:34:26  *** b_jonas [] has joined #openttd
00:40:43  *** mordant [~mordant@] has quit [Remote host closed the connection]
00:45:33  *** mordant [] has joined #openttd
00:56:32  *** b_jonas_ [] has joined #openttd
00:56:32  *** b_jonas [] has quit [Read error: Connection reset by peer]
01:03:02  *** Flygon_ [] has joined #openttd
01:10:41  *** Flygon [] has quit [Ping timeout: 480 seconds]
01:23:04  *** JacobD88 [] has quit [Quit: JacobD88]
01:50:12  *** ginko_ [] has quit [Ping timeout: 480 seconds]
01:58:45  *** FLHerne [] has quit [Ping timeout: 480 seconds]
02:22:29  *** SHOTbyGUN [] has quit [Read error: Connection reset by peer]
02:39:26  *** Myhorta[1] [] has quit [Remote host closed the connection]
02:44:20  *** Klanticus_ [~quassel@] has joined #openttd
02:47:46  *** Klanticus [~quassel@] has quit [Ping timeout: 480 seconds]
02:48:11  *** DDR [] has joined #openttd
03:12:26  *** looptrooper [] has quit [Ping timeout: 480 seconds]
03:25:46  *** Pikka [] has joined #openttd
03:44:30  *** slonik [] has joined #openttd
03:50:51  *** KWKdesign [] has quit [Ping timeout: 480 seconds]
03:54:32  *** mordant [] has quit [Ping timeout: 480 seconds]
04:12:23  *** glx [] has quit [Quit: Bye]
04:29:25  *** Hazzard_ [] has joined #openttd
04:35:57  *** Hazzard [] has quit [Ping timeout: 480 seconds]
05:15:06  *** Flygon__ [] has joined #openttd
05:21:56  *** Flygon_ [] has quit [Ping timeout: 480 seconds]
05:56:01  *** Eddi|zuHause [] has quit []
05:56:20  *** Eddi|zuHause [] has joined #openttd
05:59:39  *** Flygon [] has joined #openttd
06:01:35  *** Flygon_ [] has joined #openttd
06:06:45  *** Flygon__ [] has quit [Ping timeout: 480 seconds]
06:07:11  *** Flygon__ [] has joined #openttd
06:07:51  *** Flygon [] has quit [Ping timeout: 480 seconds]
06:10:06  *** Flygon_ [] has quit [Ping timeout: 480 seconds]
06:25:05  *** killertester [] has joined #openttd
06:31:04  *** Hazzard [] has joined #openttd
06:31:23  *** Hazzard_ [] has quit [Remote host closed the connection]
06:33:10  *** HerzogDeXtEr1 [~flex@] has joined #openttd
06:38:17  *** HerzogDeXtEr [~flex@] has quit [Ping timeout: 480 seconds]
06:40:01  *** supermop [] has joined #openttd
06:52:52  *** Flygon [] has joined #openttd
06:57:26  *** Flygon_ [] has joined #openttd
06:59:16  *** Flygon__ [] has quit [Ping timeout: 480 seconds]
06:59:16  *** DDR [] has quit [Read error: Connection reset by peer]
07:01:19  *** Flygon [] has quit [Ping timeout: 480 seconds]
07:03:19  <George> Hi
07:03:52  <George> since I do not see the resulting GRF. Where is it now?
07:17:56  *** TomyLobo [] has joined #openttd
07:34:37  *** HerzogDeXtEr1 [~flex@] has quit [Quit: Leaving.]
07:46:01  *** TomyLobo [] has quit [Quit: Standby mode...]
08:01:00  *** andythenorth [] has joined #openttd
08:01:04  <andythenorth> Pikka: flooded?
08:01:38  <Pikka> bit wet here and there.
08:02:09  <V453000> yoyo
08:02:15  <Pikka> it does
08:08:12  <andythenorth> Pikka: floating buses?
08:09:23  <Pikka> just a car, apparently.
08:41:22  <planetmaker> George, it is there where it always has been: on the bundles server
08:41:39  *** andythenorth [] has quit [Quit: andythenorth]
08:41:45  <planetmaker>
08:41:53  *** Pikka is now known as Pikka|dindins
08:43:49  <George> I know, but previously it was on the jenkins too
08:47:02  <planetmaker> yes. But we cannot afford to store everything twice
08:47:13  <planetmaker> jenkins is our build server. Not our storage server
08:49:08  <planetmaker> and it wouldn't build anything anymore due to out-of-disk space, had I not disabled storage of build results there
08:50:46  <V453000> which I know nothing about
08:50:47  <V453000> (:
08:52:23  <planetmaker> lalalala ;)
08:53:04  <planetmaker> George, I can add you a link in the navbar on the left to the bundles server, if it helps you
08:53:53  <planetmaker>
08:54:22  <George> no, thank, you, I've already fixed that page
08:54:23  <planetmaker> V453000, is going to make a nice link icon there today, I heard. Or maybe you make us one :)
08:54:45  <George> I just got a report that there is no GRf on the link
08:54:59  <planetmaker> aye
08:55:43  <planetmaker> I also would to have liked to keep it. But I don't see how to keep a link in both places without duplication
08:56:10  <George> So now fixed. But it would be nice next time something is changed to send some report to projects managers to make them to fix the links
09:00:00  <planetmaker> Sorry for the inconvenience. We always advertized the bundles server as the download portal for builds by the devzone. So changes the compiler itself which don't affect its operations didn't seem worth making a fuzz about
09:00:28  *** andythenorth [] has joined #openttd
09:01:02  <planetmaker> but you're right
09:10:04  *** Marshy [~oftc-webi@] has joined #openttd
09:10:40  <argoneus> ayy
09:15:30  *** andythenorth [] has quit [Quit: andythenorth]
09:17:09  <Marshy> ayyy
09:18:49  <peter1138> You could always request a refund
09:19:28  <FUZxxl> foamers gonna foam.
09:20:26  <argoneus> how many planets have been made today?
09:22:04  <Supercheese> Arguing from mass conservation, about as many as have been destroyed.
09:22:07  <Supercheese>  :P
09:22:54  <planetmaker> Supercheese, that's unlikely to be true :)
09:23:13  <Supercheese> Yeah you could destroy/dismantle some other stuff
09:23:20  <planetmaker> some mass will remain in black holes ;)
09:23:38  <Supercheese> but he, it sounded cool
09:23:41  <Supercheese> eh*
09:23:42  <planetmaker> which is conserved. But never again will make new planets :P
09:23:56  <argoneus> are black holes confirmed to be true anywya?
09:24:03  <argoneus> or is it a concept that explains some things that are happening
09:24:12  <Supercheese> Astronomy is all one big black hole
09:24:13  <V453000> haha refunds peter1138
09:24:37  <argoneus> we need someone to land a lander in a black hole
09:24:51  <argoneus> though it'd be hard to get reception there
09:25:15  <planetmaker> black holes are as real as your house exists while you don't look at it
09:25:39  <argoneus> "while you don't look at it"
09:25:43  <argoneus> reminds me of a video about quantum mechanics I saw
09:26:12  <Supercheese> probably was a cat video
09:26:14  <Supercheese> most videos are
09:26:28  <argoneus> ono
09:26:30  <argoneus> nono
09:26:39  <argoneus> it was about blasting some particles through a slit
09:26:50  <argoneus> and if you looked at it it created diffrent results than if you didn't
09:26:54  <argoneus> or something like that.
09:27:03  <V453000> we need more pixels
09:27:04  <V453000> everywhere
09:27:16  <Supercheese> need to fire up the Large Pixel Collider
09:27:20  <Supercheese> look for the Higgs Pixel
09:27:32  *** JacobD88 [] has joined #openttd
09:28:02  <V453000> xd
09:28:32  <V453000> supermop the community DEMANDS your buildings to be done now
09:29:04  <planetmaker> did anyone say cat? miow!
09:29:47  <FUZxxl> planetmaker: that cat is made of cute
09:29:52  <FUZxxl> and a little derp
09:29:59  <planetmaker> yep. Totally
09:30:02  <FUZxxl> 9.5/10 would view again
09:30:07  <planetmaker> :D
09:31:02  <FUZxxl> planetmaker: for your effort, observe this:
09:31:07  <FUZxxl>
09:33:54  <FUZxxl> ☑ productivity killed
09:34:33  <Supercheese> Welcome to the Internet; porn aside, it's mostly cat .gifs
09:34:46  <planetmaker> damn you!
09:34:54  <FUZxxl> I also have dog gifs:
09:35:03  <FUZxxl>
09:35:24  *** JacobD88 [] has quit [Quit: JacobD88]
09:36:15  <Supercheese> bwahahaha
09:37:13  *** ginko_ [] has joined #openttd
09:37:13  <FUZxxl> nobody cares about dog gifs :-(
09:37:32  <LordAro>
09:37:38  <LordAro> file is empty :(
09:42:56  *** itsatacoshop247 [~itsatacos@2601:9:1180:b9c:f16d:3c7a:3ff5:b6bc] has quit [Ping timeout: 480 seconds]
09:45:48  *** andythenorth [] has joined #openttd
09:46:52  *** Jomann [] has joined #openttd
09:48:55  *** andythenorth [] has left #openttd []
09:50:51  <V453000> XD
09:51:05  <V453000> #openttd, the place to be productive
09:55:30  <peter1138>
09:55:58  <V453000> not ok :D
09:56:06  <peter1138> epileptics, don't look
09:57:41  *** looptrooper [] has joined #openttd
09:58:48  *** SHOTbyGUN [] has joined #openttd
09:59:07  *** Pikka|dindins is now known as Pikka
10:01:36  <Marshy> I just learnt there is such a thing called 'Maxillofacial Prosthetics'
10:02:36  <Marshy> Fake eyes, chins, and whatnot
10:02:37  <Pikka> how exciting
10:02:40  <Marshy> Indeed
10:02:45  <Marshy> I'm easily pleased when at work
10:02:58  <Marshy> snore
10:15:19  <FUZxxl> If somebody has new further cat gifs, please tell me
10:15:23  <FUZxxl> I'll add them to my collection.
10:16:06  *** sla_ro|master [] has joined #openttd
10:19:55  *** Marshy [~oftc-webi@] has quit [Quit: Page closed]
10:29:03  *** sla_ro|master [] has quit []
10:32:59  *** sla_ro|master [] has joined #openttd
10:35:36  <__ln__>
10:35:59  *** Quatroking [] has joined #openttd
10:55:28  *** Yotson [~Yotson@2001:980:6ac8:1:adbd:2865:f19f:6537] has joined #openttd
10:57:38  *** FLHerne [~flh@] has joined #openttd
10:59:57  *** zeknurn [] has quit [Read error: Connection reset by peer]
11:05:45  *** zeknurn [] has joined #openttd
11:15:40  *** looptrooper [] has quit [Read error: Connection reset by peer]
11:19:02  *** DabuYu [DabuYu@] has quit [Ping timeout: 480 seconds]
11:19:37  *** DabuYu [DabuYu@] has joined #openttd
11:23:20  <V453000> hm
11:23:26  <V453000> I think I will try to re-code all of NUTS wagons XD
11:23:29  <Pikka> hmmmmm
11:23:37  <Pikka> sounds nuts
11:23:39  <planetmaker> why that, V453000 ?
11:23:58  <V453000> increasing amoutn of reasons :d
11:24:06  <V453000> 1. duplicate sprites of empty wagons
11:24:20  <V453000> 2. when engines have built-in wagons, it would be nicer if wagons could adapt to them
11:24:38  <V453000> 3. making Ultimate Wagon that would adapt to engine, no more "this engine can haul this", would be awesome
11:25:23  <V453000> when e.g. goods have 10 various spritesets, with 10 same empty sprites, the filesize increases a lot more than necessary, kind of considerable with 32bpp
11:25:29  <V453000> the problems are:
11:25:36  <V453000> 1. HELL with re-aligning all templates
11:25:51  <V453000> 2. re-balancing whole NUTS because different length of wagons = different curve speed
11:26:00  <V453000> 3. like 40k lines of code to care about
11:26:22  * V453000 is considering  the options
11:26:46  <V453000> I need Pikka or andythesouth to tell me if REDUCING the amount of wagons is the way to go :P
11:26:56  <Pikka> almost certainly :P
11:27:09  <V453000> but what if complaints
11:27:41  <Pikka> what if complaints?
11:27:47  <V453000> THE COMMUNITY
11:28:49  <planetmaker> not announcing it on tt-f didn't solve the problem?
11:29:01  *** Klanticus [~quassel@] has joined #openttd
11:31:17  <V453000> probably? :P
11:31:53  *** FLHerne [~flh@] has quit [Ping timeout: 480 seconds]
11:32:11  *** Klanticus_ [~quassel@] has quit [Ping timeout: 480 seconds]
11:33:47  <V453000> might write a super important blog about it
11:34:07  <Pikka> you should
11:40:12  <V453000> nice, pineapple finally getting yeti cargo :D
11:40:16  <V453000> on the same day as NUTS :P
11:40:33  <V453000> ok I was 1 day late
11:41:40  <FUZxxl> Anybody of you want to get a postcard? I have this huge stack of cards and I want to get rid of them.
11:43:07  <FUZxxl> see ya!
11:43:16  <V453000> xd
11:44:42  *** Klanticus_ [~quassel@] has joined #openttd
11:47:50  *** Klanticus [~quassel@] has quit [Ping timeout: 480 seconds]
12:05:18  <Quatroking> is it possible to put trains on some sort of "ignore list" regarding their yearly incomes?
12:05:45  <planetmaker> yes. It's called "sell it" ;)
12:06:20  <V453000> XD
12:06:21  <V453000> XD
12:06:22  <V453000> XD
12:06:28  <V453000> planetmaker bringing the bombs
12:07:32  <planetmaker> no no. the fjords
12:08:08  <V453000> flooding the dutch?
12:09:03  <Quatroking> planetmaker, teehee, but no, I have two trains that make year-long trips so one year they earn like 750m while the other year they run on -20k income
12:09:19  <Eddi|zuHause> something tells me the dutch will be faster at building new dykes than the sealevel will be rising
12:09:33  <Quatroking> Eddi|zuHause, we probably are
12:10:05  <Quatroking> I live close by one of our delta works
12:10:06  <Eddi|zuHause> countries like the maledives will have much bigger problems
12:10:06  <V453000> Quatroking: if your company is making money in total, why care?
12:10:07  <Quatroking> they're nice
12:10:19  <Quatroking> V453000, because it still whines in the news messages
12:10:40  <V453000> disable these messages?
12:10:42  <Eddi|zuHause> V453000: filtering the news between irrelevant and relevant is somewhat tricky
12:10:47  <V453000> vehicles -> warn about negative profit?
12:10:56  <Eddi|zuHause> V453000: sometimes they show actual problems
12:11:01  <Quatroking> Yeah
12:11:06  <V453000> well sure, build without problems? :P
12:11:45  <Eddi|zuHause> he who builds with no problems, may throw the first message
12:11:45  <Quatroking> I'm nearing 200 trains, building without problems doesn't exist
12:12:17  <V453000> I dare disagree :P
12:16:48  <Quatroking> you wouldn't!
12:17:31  <V453000> WOULD
12:17:36  <V453000> and DO
12:17:47  <Quatroking> no
12:18:21  <V453000> just sort your vehicles by profit in the vehicle list and see if they have some problem occassionally
12:18:31  <V453000> no need to be informed about neg profit in the messages
12:20:35  <Eddi|zuHause> maybe it should be changed into "train made over x% less profit than last year"
12:25:47  <Quatroking> that would solve the problem though, seeing how there's a 100%+ difference between 1m income and -20k income
12:25:53  <Quatroking> wouldn't*
12:27:18  <Eddi|zuHause> you mean for trains arriving every other year?
12:27:24  <Quatroking> Yeah
12:27:38  <Quatroking> one year it earns over a million, the other year it's 20k in the red
12:28:02  <Quatroking> so in reality it's doing 500k a year
12:28:23  <Eddi|zuHause> that problem should be solved in a different way
12:33:22  *** sla_ro|master [] has quit [Ping timeout: 480 seconds]
12:34:29  *** Flygon__ [] has joined #openttd
12:35:57  *** Flygon_ [] has quit [Ping timeout: 480 seconds]
12:38:25  *** Supercheese [] has quit [Ping timeout: 480 seconds]
12:43:23  *** killertester [] has quit [Quit: Leaving.]
12:44:45  *** Xrufuian [] has quit [Quit: Quit]
12:53:13  *** sla_ro|master [] has joined #openttd
13:00:35  *** supermop [] has quit [Remote host closed the connection]
13:18:31  *** Klanticus_ [~quassel@] has quit [Ping timeout: 480 seconds]
13:20:03  *** Klanticus [~quassel@] has joined #openttd
13:20:55  *** gelignite [] has joined #openttd
13:40:14  *** Klanticus_ [~quassel@] has joined #openttd
13:42:33  <argoneus> I think instead of looking at individual trains
13:42:43  <argoneus> you should look if your line is over or undersaturated, Quatroking
13:42:56  *** Klanticus [~quassel@] has quit [Ping timeout: 480 seconds]
13:43:07  <argoneus> if it's saturated just enough, you're making a profit most likely 100% of the time
13:45:24  *** Klanticus [~quassel@] has joined #openttd
13:48:35  <Quatroking> oh I am making a profit, I know that
13:48:36  *** Klanticus_ [~quassel@] has quit [Ping timeout: 480 seconds]
13:48:57  <Quatroking> I have 6 trains running the line, the line is just really long
14:07:41  *** tokai|noir [] has joined #openttd
14:07:45  *** mode/#openttd [+v tokai|noir] by ChanServ
14:10:56  <Pikka> obviously the answer is smaller map sizes :)
14:12:32  <V453000> ^
14:13:44  *** tokai|mdlx [] has quit [Ping timeout: 480 seconds]
14:14:41  *** slonik [] has quit [Ping timeout: 480 seconds]
14:33:57  *** slonik [] has joined #openttd
14:58:41  *** Pikka [] has quit [Quit: Leaving]
15:01:38  <argoneus> sorry for not browsing through the code, but just a quick question if anyone knows
15:01:51  <argoneus> if someone places a tunnel, is there any information in the tiles between "there's a tunnel here"?
15:02:01  <argoneus> for easier detection if tunnels cross or such
15:05:07  <Eddi|zuHause> no
15:06:01  <Eddi|zuHause> because detecting whether tunnels cross is not a time-critical task
15:06:33  <Eddi|zuHause> this is stored for bridges, because it's needed for drawing
15:09:43  <peter1138> And tunnels can overlap.
15:11:38  <planetmaker> via cheat :)
15:12:13  <V453000> I think peter1138 means on a tile
15:13:15  <Eddi|zuHause> there can be multiple parallel tunnels on top of each other and stuff
15:13:49  <Eddi|zuHause> so caching this stuff and keeping the cache up to date would probably not save a lot of processing
15:17:18  <argoneus> how do you tell
15:17:20  <argoneus> if tunnels cross then
15:19:25  <peter1138> Test.
15:19:53  <planetmaker> Pong
15:20:02  <peter1138> As in, test the map  :p
15:20:09  <NGC3982> Pung.
15:21:51  *** Myhorta [] has joined #openttd
15:22:11  <peter1138> Listen son, you're wasting your time, there's a future for you in the fire escape trade.
15:22:51  <argoneus> I'm not your son, buddy
15:25:23  <Eddi|zuHause> argoneus: the test is really not that complicated. for every step you take in the tunnel, you go left and/or right until you find a tile of the same height as the tunnel entrance. if there's a tunnel there->conflict, otherwise->good
15:25:57  <Eddi|zuHause> assuming that mountains are usually uniformly long as wide, that works out to an average runtime of O(n^2)
15:25:58  <argoneus> so in the worst case
15:26:02  <Eddi|zuHause> where n is the tunnel length
15:26:05  <argoneus> it's O(kn)
15:26:20  <Eddi|zuHause> in the worst case, it's n*map_x or n_map_y
15:26:29  <Eddi|zuHause> n*map_y
15:26:46  <argoneus> how did you get n^2
15:26:50  <argoneus> you have a map of size n
15:26:53  <argoneus> and tunnel of size k
15:26:55  <argoneus> for each k, you go max n
15:26:58  <argoneus> O(kn)
15:27:02  <argoneus> no?
15:28:50  <Eddi|zuHause> on a perlin-generated map, you hardly ever will find such a map slice where it could reach the worst case runtime
15:29:00  <argoneus> yes
15:29:03  <argoneus> but kn is better than n^2
15:29:26  <Eddi|zuHause> no
15:29:30  <argoneus> yes it is
15:29:32  <argoneus> k is always < n
15:29:33  <Eddi|zuHause> because you didn't listen
15:29:36  <argoneus> you can't have a tunnel longer than the map
15:29:46  <Eddi|zuHause> i said n is the tunnel length
15:30:06  <argoneus> oh
15:30:07  *** Hazzard [] has quit [Quit: Goodbye.]
15:30:19  <argoneus> it's still kn and not k^2 though :<
15:30:45  <Eddi|zuHause> in the worst case, yes
15:31:09  <argoneus> well, O is usually used for the worst case
15:31:59  <Eddi|zuHause> only if you're in your 2nd year of computer science :p
15:32:32  <Eddi|zuHause> and haven't got to the good parts yet about probabilistic estimates and average values :p
15:32:47  <argoneus> well
15:32:51  <argoneus> quicksort is O(n^2)
15:33:02  <peter1138> I never learnt all this O stuff :p
15:33:04  <argoneus> but more often than not it's n log n
15:34:27  <Eddi|zuHause> argoneus: yes, so if you sum up over the range of all possible inputs the values runtime(input)*probability(input), you get the expected (average) value
15:34:45  <argoneus> but yeah, you are right
15:34:55  <Eddi|zuHause> like if you throw a dice, you get 1*1/6+2*1/6+...*6*1/6 = 3,5
15:35:02  <argoneus> O(...) is used for average cases and amortized cases etc too
15:35:12  <argoneus> it's just that when someone says "algo X is O(n^2)" it usually means the worst case
15:35:15  <argoneus> without context
15:35:28  <Eddi|zuHause> but that's why i did give context.
15:35:35  <argoneus> right, sorry :<
15:35:52  <argoneus> I never really got into this theory
15:36:03  <argoneus> like, I know that mergesort is O(nlogn) and I understand why
15:36:06  <argoneus> but I can't show/prove it
15:36:07  <argoneus> it just.. is..
15:36:40  <Eddi|zuHause> heapsort is better
15:37:07  <Eddi|zuHause> it's also nlogn, but it can be done in-place. with mergesort you have to make copies
15:37:14  <argoneus> there is inplace mergesort
15:37:23  <argoneus> but it's ugly as fuck
15:37:39  <argoneus> also
15:37:44  <argoneus> you only need 2n memory
15:37:45  <argoneus> for mergesort
15:38:02  <Eddi|zuHause> yes, but if you have 2GB of input
15:38:18  <Eddi|zuHause> then using 4GB may be troublesome
15:38:35  <argoneus> how does heapsort work anyway
15:38:42  <argoneus> do you just heapify an array and then call extractMax() until empty?
15:39:13  <argoneus> so n to get all the items in the structure, and log n to get them out?
15:39:17  <argoneus> since it's a binary tree p much
15:39:24  <Eddi|zuHause> you treat the array as a heap, and call raise and sink on each element
15:40:40  <Eddi|zuHause> raise is a log(n) operation, and sink is a log(n) operation
15:40:48  <argoneus> what is raise and sink?
15:40:52  <argoneus> maybe I know but I never heard these terms
15:41:13  <argoneus> is that the process of adding elements to the heap?
15:41:21  <Eddi|zuHause> yes, somewhat
15:41:36  <argoneus> the idea of heaps was psychotic to me at first
15:41:41  <Eddi|zuHause> you move upwards the tree until you find the place where the heap condition is invalid, and then go down again
15:41:44  <argoneus> "just randomly fill this binary tree from the left and swap values"
15:42:11  <argoneus> ah
15:42:12  <Eddi|zuHause> after that, the heap condition is valid up to the n-th element
15:42:26  <argoneus> so raise is when you add a new item, and you raise it until it meets the h-condition
15:42:31  <Eddi|zuHause> which implies that your array is now sorted up to this element
15:42:48  <argoneus> and sink is when you remove your root, replace it with the last item, and sink it down until h-condition applies?
15:43:40  <Eddi|zuHause> not sure i get your explanation. but something of that sort, yes
15:43:52  <argoneus> well
15:43:57  <argoneus> you can only remove the top item from the heap, no?
15:44:05  <argoneus> or rather, it's built for that
15:44:11  <argoneus> it's not built for searching inside it or such
15:44:36  <Eddi|zuHause> not quite. this is about following the path your new element takes
15:45:04  <argoneus> well
15:45:07  <argoneus> when you insert a new element
15:45:11  <argoneus> you just put it at the end of the array
15:45:13  <Eddi|zuHause> first it moves up the tree, until this branch of the tree complies to the heap-condition, and then it goes down another branch, until the whole tree is valid
15:45:17  <argoneus> and then you raise it up through its parents
15:45:25  <argoneus> until your condition no longer applies
15:45:31  <argoneus> wait it goes down?
15:45:36  <argoneus> I thought it's enough if it just goes up
15:45:42  <Eddi|zuHause> no
15:45:57  <Eddi|zuHause> it could still be inbetween some values in the other branch
15:46:06  <argoneus> wait
15:46:10  <argoneus> are binary heaps the same thing as heapsort?
15:46:15  <argoneus> like, is the structure and operations the same?
15:46:20  <Eddi|zuHause> yes
15:46:26  <Eddi|zuHause> hence the name
15:46:27  <argoneus> we were told
15:46:31  <argoneus> that it doesn't need to go down anymore
15:46:35  <argoneus> because it's certain the h condition applies
15:46:52  <Eddi|zuHause> i don't see how that is
15:47:09  <argoneus> well
15:47:17  <argoneus> if you have number 7, with sons 3 and 4
15:47:19  <argoneus> and you add 9
15:47:25  <argoneus> then it swaps with 3
15:47:30  <argoneus> and then swaps with 7
15:47:36  <argoneus> and since 9 > 7
15:47:41  <argoneus> it will be > than whatever is the other son
15:47:46  <argoneus> automatically
15:47:46  <argoneus> no?
15:48:19  <Eddi|zuHause> and if you add 3,5?
15:48:26  <argoneus> then it won't swap with 7
15:48:27  <argoneus> it will swap with 3
15:48:29  <argoneus> and done
15:48:38  <Eddi|zuHause> i think this example is too small
15:49:06  <argoneus> my point is
15:49:07  <argoneus> the idea is
15:49:10  <argoneus> that any given node
15:49:13  <argoneus> is of larger value than the two sons
15:49:14  <argoneus> right?
15:49:30  <argoneus> or smaller, depending on the kind
15:49:37  <Eddi|zuHause> also, i think this 7->(3,4) is not satisfying the heap condition
15:49:44  <Eddi|zuHause> it must be 4->(3,7)
15:49:51  <argoneus> that's not a heap
15:49:54  <argoneus> that's a binary tree
15:49:56  <Eddi|zuHause> all things left of the root must be smaller, and all things right must be larger
15:50:02  <argoneus> that's a binary tree
15:50:02  <argoneus> not a heap
15:50:12  <Eddi|zuHause> hm, no
15:50:17  <Eddi|zuHause> it was still different
15:50:18  <argoneus> I'm talking about maxHeap
15:50:22  <Eddi|zuHause> but i probably misremember
15:50:38  <argoneus> in a binary search tree (BST)
15:50:40  <argoneus> for any node
15:50:43  <argoneus> its left son is smaller
15:50:45  <argoneus> and right son is higher
15:51:17  <argoneus> in a maxheap
15:51:24  <argoneus> the node just needs to have a higher value than both the sons
15:51:35  <argoneus> I think so, at least
15:51:54  <Eddi|zuHause> i'm probably missing some condition, but i don't find my script now
15:53:18  <argoneus> Eddi|zuHause:
15:53:21  <argoneus> this is what I meant
15:58:54  <Eddi|zuHause> so, i found the script, and yes, sink is when you remove the root
15:59:15  <Eddi|zuHause> first you add all the elements into the heap by raise, and then you remove all elements by sink
15:59:25  <Eddi|zuHause> and then your array is sorted
16:00:52  <Eddi|zuHause> you store the removed elements in the places that are not used up anymore in your tree
16:01:36  <argoneus> yes
16:02:01  *** planetmaker [] has left #openttd [Verlassend]
16:02:19  <Eddi|zuHause> hence it can be done in-place
16:03:16  *** Suicyder [~Suicyder@] has joined #openttd
16:03:18  <argoneus> well
16:03:19  <argoneus> basically
16:03:23  <argoneus> you just call extractMax() over and over
16:03:33  <argoneus> because the root is the maximum always
16:03:41  <argoneus> and what extractMax do
16:03:50  <argoneus> is return the root value, put the last value in the root, and sink it down until it's in place
16:03:57  <Eddi|zuHause> yes
16:04:00  <argoneus> thus you end up with the second max in root
16:04:06  <argoneus> etc
16:04:25  <argoneus> so basically
16:04:31  <argoneus> you fill your array from end to beginning
16:04:34  <argoneus> sorted ascending
16:04:35  <argoneus> right?
16:04:46  <Eddi|zuHause> yes
16:05:37  <Eddi|zuHause> so the order of functions is: (add, raise)*n, (swap, sink)*n and you're done
16:06:01  <Eddi|zuHause> add and swap are O(1), raise and sink are O(logn)
16:08:50  <argoneus> raise/sink is log n
16:08:55  <argoneus> add is O(1)
16:09:00  <argoneus> but you have n items
16:09:06  <argoneus> so add is O(n)
16:09:09  <argoneus> so n log n
16:09:23  <Eddi|zuHause> yes
16:09:46  <argoneus> and then there's radix sort
16:09:51  <argoneus> O(n) because LOL
16:10:06  <argoneus> or rather
16:10:08  <argoneus> O(kn)
16:10:12  <argoneus> k being the amount of digits
16:10:25  <Eddi|zuHause> that's only if you have an additional condition
16:10:40  <Eddi|zuHause> that the numbers are bounded by some maximum value
16:10:49  <argoneus> why?
16:10:56  <argoneus> all you need to know is how many digits max there are
16:11:01  <Eddi|zuHause> yes
16:11:07  <argoneus> so
16:11:11  <Eddi|zuHause> which you can't know always
16:11:12  <argoneus> if you have like 2 GB of 3 digit numbers
16:11:14  <argoneus> it sorts really quick
16:11:17  <argoneus> it's 3n
16:11:21  <argoneus> that's as fast as it gets
16:11:39  <argoneus> counting sort would be just 2n
16:11:44  <argoneus> but take up a loooot of memory
16:11:47  <Eddi|zuHause> if you have 2GB of floating point numbers
16:12:37  <Eddi|zuHause> basically, k is ~ log(n) when you don't know the shape of the input data
16:12:55  <Eddi|zuHause> so you're back to nlogn
16:13:47  <Eddi|zuHause> there's also "lossy sort", which is O(1) :p
16:13:58  <Eddi|zuHause> for every input, it returns [1,2,3,4,5,6,7,8,9,10]
16:14:13  <Eddi|zuHause> which obviously is a sorted list :p
16:14:46  <argoneus> lol
16:14:57  <argoneus> bogosort is O(1)
16:15:00  <argoneus> in the best case
16:15:01  <argoneus> :P
16:15:36  <argoneus> or is it
16:15:39  <argoneus> hm
16:15:44  <argoneus> it can't be faster than O(n)
16:15:50  <argoneus> since it needs to check that every item is in place
16:15:52  <argoneus> which takes at least n
16:16:00  <argoneus> actually, no sorting algo can be faster than n for this reason
16:16:00  <argoneus> :D
16:17:05  <Eddi|zuHause> no operation can be faster than O(n) if you assume that the complete input is read at least once
16:17:45  <argoneus> yeah
16:17:52  <argoneus> but it's proven that sorting can never be faster than O(n)
16:18:54  <Eddi|zuHause> sorting in genera cannot be done faster than O(nlogn) in the worst or average case
16:19:05  <Eddi|zuHause> *general
16:19:23  <argoneus> well
16:19:29  <argoneus> maybe it can
16:19:39  <argoneus> but right now heapsort/mergesort/radixsort/quicksort are like the top 4
16:19:44  <argoneus> depending on input data
16:27:16  <Eddi|zuHause> there's a theoretical algorism called "FJA-sort"
16:27:29  <Eddi|zuHause> that tries to reach the minimal upper bound
16:28:19  <peter1138> HI
16:28:47  <V453000> NO
16:28:55  <Eddi|zuHause> hello internet?
16:30:38  <horazont> hello eddi
16:35:14  *** Alberth [~hat@2001:981:c6c5:1:be5f:f4ff:feac:e11] has joined #openttd
16:35:17  *** mode/#openttd [+o Alberth] by ChanServ
16:36:37  *** Progman [] has joined #openttd
16:36:53  *** ntx [] has quit [Ping timeout: 480 seconds]
16:40:45  *** ntx [] has joined #openttd
16:41:39  *** liq3 [] has quit []
16:55:32  *** TomyLobo [] has joined #openttd
16:56:35  *** Myhorta[1] [] has joined #openttd
16:57:42  *** frosch123 [] has joined #openttd
17:01:12  *** juzza1 [] has quit [Quit: Lost terminal]
17:11:57  <Eddi|zuHause> oooh... i found the TTO savegame with the "one-way" system
17:15:25  <Alberth> TTO already had a single direction block signal, didn't it?
17:16:04  <Eddi|zuHause> no
17:16:09  <Eddi|zuHause> that was introduced in TTD
17:16:16  <Alberth> ah, ok
17:16:21  <Eddi|zuHause> which i never had
17:17:03  <peter1138> "the alpha where the tiles meet doesn't quite sum to 1 so you get some marks between the tiles"
17:17:10  <peter1138> ^ seems to be a misconception on how alpha works?
17:17:23  <peter1138> (That alpha is somehow "summed")
17:18:16  <Alberth> sounds like it
17:18:53  <Eddi|zuHause> i'd expect alpha to be multiplicative
17:19:48  <Alberth> you draw several sprites at the same time?
17:20:49  <Alberth> hmm, when you draw a sprite on top of a non-opaque one, you mean
17:21:29  <peter1138> Once it's drawn, it's no longer transparent.
17:21:37  <Alberth> it's still wrong, a transparent pixel on top of an opaque one is not transparent
17:23:54  <Eddi|zuHause> of all the sprites that are drawn on one particular pixel, at least one has to be completely opaque, otherwise a fraction of the original pixel will shine through
17:24:05  *** DDR [] has joined #openttd
17:24:45  <Eddi|zuHause> and the fraction of the orgiginal pixel that is remaining is (1-product(alpha))
17:24:48  <Alberth> isn't that the idea of non-opaque pixels?
17:25:17  <Eddi|zuHause> yes, but it's not the desired outcome for ground tiles
17:25:55  <Alberth> so how exactly is  1*0 == 1 ?
17:26:46  <Eddi|zuHause> hm, then my expectation is probably wrong
17:34:07  *** killertester [] has joined #openttd
17:34:18  *** pxr [] has joined #openttd
17:36:04  *** Hazzard [~Hazzard@] has joined #openttd
17:36:08  *** oskari89 [] has joined #openttd
17:51:52  *** TheMask96 [] has quit [Ping timeout: 480 seconds]
17:52:52  *** mordant [] has joined #openttd
17:54:14  *** sla_ro|master [] has quit []
17:54:29  *** sla_ro|master [] has joined #openttd
17:55:54  *** TheMask96 [] has joined #openttd
17:58:43  *** FLHerne [] has joined #openttd
17:59:21  *** glx [] has joined #openttd
17:59:24  *** mode/#openttd [+v glx] by ChanServ
18:01:36  *** itsatacoshop247 [] has joined #openttd
18:11:16  *** itsatacoshop247 [] has quit [Ping timeout: 480 seconds]
18:14:05  *** andythenorth [] has joined #openttd
18:15:40  <argoneus> evening my friends
18:15:48  <argoneus> train friends even
18:16:15  <NGC3982> dial-a-pirate!
18:18:25  <V453000> death train infidel
18:19:42  *** sla_ro|master [] has quit []
18:22:42  <Eddi|zuHause> i have a dial-a-pirate here
18:24:45  <andythenorth> cat problem is now solved
19:06:07  *** sla_ro|master [] has joined #openttd
19:06:53  *** Klanticus [~quassel@] has quit [Remote host closed the connection]
19:07:06  <frosch123> oh, the forums is pure gold today
19:09:14  <Alberth> :D
19:10:56  <andythenorth> really? :|
19:11:04  <andythenorth> did I miss?
19:11:51  <Alberth>   this one, perhaps
19:11:57  <frosch123> two topics where i do not know who is trolling who
19:12:13  <Alberth> oh, max loan size :p
19:12:22  <frosch123> nah, i ignore that
19:12:57  <frosch123> i meant the one which explains eddi to google information loss, and the one with imaginary numbers
19:13:27  <Alberth> oh, "imaginary numbers" are great indeed :D
19:13:52  <andythenorth> are they really imaginary numbers
19:14:02  <andythenorth> ?
19:14:25  <frosch123> andythenorth: everything is measureable with imaginary numbers
19:14:27  <andythenorth> like sqrt(-1) ?
19:14:34  <frosch123> even intelligency
19:15:18  <Eddi|zuHause> some people argue that the "real" numbers aren't as real as they sound, either
19:15:19  <frosch123> the absolute amount of intelligence can be huge, with no impact on the real part
19:15:59  <Alberth> andythenorth: yes, you define   i*i == -1      sqrt(-1) is thus i
19:16:23  <frosch123> sqrt(-1) is {-i, i}
19:16:25  <frosch123> :p
19:16:44  <Alberth> hmm, right :)
19:17:20  <Eddi|zuHause> frosch123: you can define either of those as "i", it'll give the same result
19:17:36  <andythenorth> imaginary numbers were about the last bit of fun I had doing engineering maths
19:17:42  <andythenorth> after that it got way beyond me
19:18:11  <andythenorth> mostly ever more complex differentation :P
19:18:14  <andythenorth> and equation solving
19:18:15  <andythenorth> boring
19:18:47  <frosch123> Eddi|zuHause: maybe write a book using j := -i
19:19:05  <andythenorth> imaginary numbers are quite fascinating when you’re ~18
19:19:21  <andythenorth> there’s nothing ‘imaginary’ about them for starters
19:19:25  <andythenorth> as real as any other number
19:19:27  <Eddi|zuHause> what? imaginary numbers were fun when i was 12
19:19:47  <andythenorth> yes, but you live in a society that doesn’t see mathematical ability as shameful
19:19:53  <andythenorth> whereas I live in England
19:20:07  <frosch123> andythenorth: psst, just say that you got your mental state by considering the axiom of choice
19:20:21  <frosch123> eddi will believe you everything after that
19:20:31  <andythenorth> yes, but it’s using words I don’t understand
19:20:38  <andythenorth> or at least, in that combination
19:20:44  <andythenorth> :P
19:20:54  <Eddi|zuHause> the axiom of choice isn't that difficult to understand
19:21:02  <frosch123> actually it is
19:21:12  <andythenorth> google probably knows
19:21:13  <Eddi|zuHause> it just says that for any set, you can pick a value from that set
19:21:14  * andythenorth looks
19:21:39  <Eddi|zuHause> it's just that it cannot be proven or disproven using "constructive" maths
19:22:12  <Eddi|zuHause> thus creating a new level of maths, where things can now be proven
19:22:19  <frosch123> every student is explained it like eddi just did, but i would claim that 99% do not understand what that really means
19:22:20  <Eddi|zuHause> which couldn't be proven before
19:22:53  <Eddi|zuHause> frosch123: maybe it's like quantum mechanics. if you have claimed to understand it, you didn't understand it :p
19:22:56  <frosch123> i only understood it in a lecture specifically about set theory, and axiom stuff
19:23:15  <andythenorth> if the students are anything like the (well-educated) students at the university I went to
19:23:19  <andythenorth> they learn the formula
19:23:23  <andythenorth> and the rote definition
19:23:31  <Eddi|zuHause> anyway, there's loads of maths you can do without the axiom of choice
19:23:39  <andythenorth> and they put those two together to the socially acceptable 63% exam mark
19:23:48  <andythenorth> which shows they aren’t trying too hard
19:24:14  <frosch123> Eddi|zuHause: you can say it more strict: you can do all math which have the slightest practical relevance without it :p
19:25:01  <frosch123> that axiom basically separates the theoretical from the applied maths
19:25:39  <frosch123> but ok, i guess number theory is not much affected by it
19:26:21  <Eddi|zuHause> the only place where it was really needed was functional analysis
19:26:44  <frosch123> yup
19:26:50  <frosch123> well, and set theory
19:26:50  <Eddi|zuHause> at least, among the fields of maths i came across
19:27:44  <Eddi|zuHause> another place where it was used was to create non-borel sets
19:28:00  <Eddi|zuHause> but nobody needs those :p
19:28:21  *** andythenorth [] has quit [Read error: Connection reset by peer]
19:29:22  <Eddi|zuHause> i don't quite remember: can it be used to prove or disprove whether there is a cardinality between countable and continuum?
19:29:34  *** andythenorth [] has joined #openttd
19:39:28  <andythenorth> hmm
19:39:37  <andythenorth> learn how to write a GS in Squirrel?
19:39:44  <peter1138> No.
19:39:46  <andythenorth> or learn how to draw charts in Raphael.js?
19:44:30  <andythenorth> [c] none of the above?
19:50:10  *** Supercheese [] has joined #openttd
19:51:50  <Alberth> feels like a deja-vu, but extend an existing one?
19:53:01  <andythenorth> hack SV
19:54:48  <Alberth> NoCarGoal seems like an easier target to me, but it depends on what you want
19:56:40  <andythenorth> what I want is people to write more GS :D
19:56:51  <andythenorth> what I *get* is to do that myself :P
19:57:37  <Alberth> sorry :)
19:57:37  *** HerzogDeXtEr [~flex@] has joined #openttd
20:00:45  * andythenorth learns a charting library by reading source
20:00:53  <andythenorth> because github pages for manual are broken :P
20:02:59  <Alberth> GH is quite broken in itself, lots of edge cases that don't exactly work
20:05:36  <andythenorth> maybe I start cloning more repos for dependencies :P
20:05:48  <andythenorth> instead of getting the release bundle and assuming I can trust GH in future
20:16:42  <Alberth> you are generally getting the service you pay for :p
20:17:17  <Alberth> but that holds for everything :)
20:18:12  <andythenorth> I want my money back
20:18:13  <andythenorth> :P
20:21:44  *** Wolf01 [] has joined #openttd
20:21:54  <Wolf01> hi hi
20:30:00  <__ln__> in bocca al lupo
20:32:11  <V453000> anus
20:48:24  *** Alberth [~hat@2001:981:c6c5:1:be5f:f4ff:feac:e11] has left #openttd []
20:55:03  *** Hazzard [~Hazzard@] has quit [Quit: Goodbye.]
21:01:51  *** Flygon [] has joined #openttd
21:07:51  *** Flygon__ [] has quit [Ping timeout: 480 seconds]
21:18:34  *** Quatroking_ [] has joined #openttd
21:19:35  *** [1]Suicyder [~Suicyder@] has joined #openttd
21:22:32  *** argoneus [] has quit [Remote host closed the connection]
21:22:34  *** argoneus [] has joined #openttd
21:23:04  *** funnel [] has quit [Remote host closed the connection]
21:23:44  *** funnel [] has joined #openttd
21:25:19  *** Suicyder [~Suicyder@] has quit [Ping timeout: 480 seconds]
21:25:19  *** [1]Suicyder is now known as Suicyder
21:25:24  *** Quatroking [] has quit [Ping timeout: 480 seconds]
21:47:59  *** luaduck_zzz is now known as luaduck
21:49:16  *** Yotson [~Yotson@2001:980:6ac8:1:adbd:2865:f19f:6537] has quit [Quit: .]
21:54:01  *** liq3 [] has joined #openttd
21:58:27  *** TomyLobo [] has quit [Quit: Standby mode...]
21:58:54  *** frosch123 [] has quit [Quit: be yourself, except: if you have the opportunity to be a unicorn, then be a unicorn]
22:00:53  *** gelignite [] has quit [Quit:]
22:05:13  *** itsatacoshop247 [] has joined #openttd
22:06:59  *** Progman [] has quit [Remote host closed the connection]
22:14:20  *** dreck [] has joined #openttd
22:14:24  <dreck> hi
22:14:56  <V453000> yo
22:15:01  <dreck> was wondering about it a bit for awhile and mm how do you code an engine to run like normal on one railtype but lose some power (and still run) on another railtype?
22:15:16  <dreck> hi v453000 :P
22:15:27  <V453000> just with a switch detecting current_railtype
22:15:29  <V453000> see nuts code
22:22:44  <dreck> mm ok
22:23:20  <V453000> just ctrl f for that variable and you got it easy
22:25:40  <andythenorth> bye
22:25:41  *** andythenorth [] has quit [Quit: andythenorth]
22:30:27  *** Jomann [] has quit [Remote host closed the connection]
22:31:12  <Wolf01> 'night
22:31:14  <Eddi|zuHause> "for the next five years, yahoo will be the default search engine in firefox"
22:31:16  <Eddi|zuHause> wtf?
22:31:24  *** Wolf01 [] has quit [Quit: Once again the world is quick to bury me.]
22:32:12  <FLHerne> Eddi|zuHause: Well, Bing isn't quite as awful as it used to be
22:32:15  <V453000> 1st april owat :)
22:34:17  *** sla_ro|master [] has quit []
22:34:18  <Eddi|zuHause> apparently, google paid mozilla a large sum of money for being the default search engine, and now yahoo pays that...
22:34:34  <V453000> nice enough
22:35:47  <Eddi|zuHause> "in 2012, 280 million $ out of the 311 million $ income of mozilla came from the search engine deal"
22:37:27  *** jinks [~jinks@2602:ffe8:102:213::1c:34ac] has quit [Quit: ZNC -]
22:37:43  *** jinks [~jinks@2602:ffe8:102:213::1c:34ac] has joined #openttd
22:38:09  *** luaduck is now known as luaduck_zzz
22:39:53  <dreck> mm and correct me if this is wrong, wagons can be introduced but never obsoleted?
22:40:06  <Eddi|zuHause> they can
22:49:25  *** oskari89 [] has quit []
22:49:31  *** Suicyder [~Suicyder@] has quit [Quit:  I love my HydraIRC -> <-]
22:51:01  *** ginko_ [] has quit [Remote host closed the connection]
22:53:25  <dreck> have to eat now so
22:53:27  *** dreck [] has left #openttd []
23:03:41  *** Hazzard [] has joined #openttd
23:08:41  *** planetmaker [] has joined #openttd
23:08:44  *** mode/#openttd [+o planetmaker] by ChanServ
23:15:07  <argoneus> what came first
23:15:11  <argoneus> planets or planetmaker
23:18:51  *** Speedy [~speedy@] has quit [Ping timeout: 480 seconds]
23:35:19  <Eddi|zuHause> first there was a singularity, then there was hot plasma, then there was hydrogen, and it all went downhill from there
23:36:13  <Eddi|zuHause> and if there hadn't been some uneven wobbles in the fabric of the universe, nothing would have happened ever again
23:39:36  <ST2> so, "planetmaker" is the Big Bang itself, I presume ;)
23:40:57  <ST2> (itself, himself or herself, I couldnt decide what to use. so, came out that way :D)
23:57:48  *** Speedy` [~speedy@] has joined #openttd
23:58:00  *** Speedy` is now known as Speedy

Powered by YARRSTE version: svn-trunk