IRC logs for #openttd on OFTC at 2007-06-02
⏴ go to previous day
00:12:29 *** Digitalfox has joined #openttd
00:28:01 *** mataamad has joined #openttd
00:39:24 *** Progman has joined #openttd
00:47:26 <CIA-1> OpenTTD: rubidium * r10018 /website/ (26 files in 2 dirs): -Update: the website so it can handle the new flags/languages introduced in r10017.
01:04:35 *** Apocalipsys has joined #openttd
01:23:53 *** roboboy has joined #openttd
01:30:58 *** Eddi|zuHause3 has joined #openttd
01:51:25 *** robotboy has joined #openttd
02:10:40 *** Apocalipsys has left #openttd
02:18:57 *** Digitalfox has joined #openttd
03:50:18 *** Progman has joined #openttd
04:02:28 *** ChanServ sets mode: +v tokai
04:18:37 <Jay> I finished my magazine column for Railroads Illustrated
04:18:49 <Jay> I have to move my uncle to a new house early in the morning
04:18:56 <Jay> but all I want to do is play openttd
05:40:49 *** boekabart has joined #openttd
07:06:44 *** Draakon has joined #openttd
07:44:56 *** XeryusTC has joined #openttd
07:47:58 *** Thomas[NL] has joined #openttd
08:19:31 *** peterbrett has joined #openttd
08:38:14 *** dihedral has joined #openttd
08:40:41 *** Maedhros has joined #openttd
08:50:23 *** setrodox has joined #openttd
08:52:33 *** boekabart has joined #openttd
08:52:36 *** ChanServ sets mode: +o Bjarni
08:56:56 <CIA-1> OpenTTD: peter1138 * r10019 /trunk/src/newgrf.cpp: -Codechange: Implement GRM for newcargos
09:00:15 *** RamboRonny has joined #openttd
09:05:36 *** Vikthor has joined #openttd
09:08:21 <stillunknown> peter1138: that smatz person did some optimizations related to the hash map, were they bad?
09:09:02 *** Progman has joined #openttd
09:09:44 <Maedhros> grf resource management - designed to avoid conflicts between grf files
09:11:21 <CIA-1> OpenTTD: maedhros * r10020 /trunk/src/vehicle.cpp: -Fix [FS#824]: GetNextVehicle() is invalid for anything that isn't a train.
09:18:58 *** Unaimed has joined #openttd
09:19:50 *** skidd13 has joined #openttd
09:21:00 <Unaimed> Does anyone know how to use the {P} (Plural Modifier) in strings?
09:23:52 <Unaimed> Thank you, i was looking in the strings help section
09:24:41 <stillunknown> peter1138: ping-pong
09:26:15 <CIA-1> OpenTTD: rubidium * r10021 /trunk/src/newgrf_gui.cpp: -Fix (r9560): memory "corruption" that could lead to a failure to load newgrfs.
09:26:15 *** ThePizzaKing has joined #openttd
09:26:33 <Rubidium> robotboy: is the problem with ausland.grf now solved?
10:03:41 *** skidd13 has joined #openttd
10:34:08 *** peterbrett has joined #openttd
10:45:53 *** Sleepie has joined #openttd
11:10:24 <robotboy> i SHALL CHECK FOR YOU NOW RUBIDIUM
11:10:48 <robotboy> will I have to wait for the nightly?
11:11:30 <Rubidium> if you can compile yourself you don't need to wait
11:12:47 <Unaimed> Where/how does one submit a patch?
11:13:29 <robotboy> ill go get buildottd
11:15:00 <boekabart> Unaimed: post it on the forum, get people to like it, talk to devs here
11:15:44 <Unaimed> so i shouldn't submit it to flyspray?
11:16:02 <CIA-1> OpenTTD: maedhros * r10022 /trunk/src/train_cmd.cpp: -Fix (r10007): Set the type of multiheaded trains before the subtype.
11:17:56 <robotboy> what version of .net does build ottd need?
11:36:03 *** Nigel_ is now known as Nigel
11:38:13 *** skidd13 has joined #openttd
11:42:28 <Wolf01> mmmh, bug, you can remove roads also if you don't have money
11:44:13 *** TheJosh has joined #openttd
11:46:20 <TheJosh> Hey could someone point me in the right direction I wanted to add some code to the game (a patch) that would draw some lines and stuff on the main viewport. I want to be able to specify co-ords in isometric and have them worked out. what file should I look in?
11:46:20 *** dihedral has joined #openttd
11:47:17 *** DJ_Mirage has joined #openttd
11:48:33 <TheJosh> basically i want to build a system that i call 'routes' where you can define lines that are drawn on the ground, and are basically a sign, but it covers an area (a line) rather than just a point
11:49:25 <Eddi|zuHause3> if you want to draw stuff on the map, you should probably add a sprite
11:49:49 <Eddi|zuHause3> like what the tile highlight code does
11:55:50 *** roboboy has joined #openttd
12:02:42 <Bjarni> stillunknown: I think we need to work on your ping time :D
12:08:31 <valhallasw> TrueBrain: where is the party? :P
12:09:29 <Bjarni> it could improve communication if we knew who is so paranoid that they have to use different names all over the place
12:09:39 <Bjarni> I bet you have a 3rd name on flyspray xD
12:09:57 <stillunknown> I don't think so.
12:10:17 <stillunknown> I don't actually remember why i choose a different irc name.
12:10:31 <Bjarni> also you picked 2003 in 2004 :p
12:11:00 <stillunknown> I still use madman2003, just plug into a search engine ;-)
12:11:20 <Bjarni> "This concept cannot be applied to road, because there is no space in the map array there" <-- can't the road vehicle check if there is a train in the crossing using the train counter?
12:11:35 <boekabart> ah, best thing about saturday: pizza for lunch
12:11:59 <stillunknown> Bjarni: I forgot that road vehicles do not collide.
12:12:14 <Bjarni> they collide with trains
12:13:02 <stillunknown> I will also look at the roadveh collision stuff.
12:13:31 <stillunknown> Bjarni: Do you like the idea?
12:14:08 <Bjarni> however I'm still not sure that it's bug free :s
12:14:27 <stillunknown> Hence the: it could use some testing
12:14:39 <TrueBrain> valhallasw: at mine place
12:14:49 <boekabart> Bjarni: how many times did has a _committed_ patch been bugfree ;)
12:14:58 <Eddi|zuHause3> hmm, technically, road vehicles collide with other road vehicles, just they slow down instead of crash
12:15:03 <valhallasw> The AHC regatta is 6th of june, too
12:15:12 <boekabart> 273 out of 10000 ;)
12:15:37 <boekabart> TrueBrain: Why is the party in the middle of a working day :(
12:15:56 <valhallasw> TrueBrain: 'zeilwedstrijd'? :p
12:15:58 <TrueBrain> boekabart: becaue it means our OV is valid :p
12:21:10 <stillunknown> TrueBrain boakabart: You expect him to know what an OV is (it's a card which can be either used during weeks or weekends, to travel to and from university/college/whatever)
12:21:44 <stillunknown> They apparently have a week card, so they would have to pay for travel in the weekend.
12:23:01 <TrueBrain> stillunknown: and why would you assume boekabart doesn't know what it is?
12:23:24 <TrueBrain> stillunknown: I personally assume anyone living in The Netherlands knows what an OV is :p
12:24:19 <valhallasw> TrueBrain: where is it, kamp duwo? :p
12:24:44 <TrueBrain> valhallasw: haha, no, not really :p Around 15 minutes from Huygens by bike
12:31:11 <stillunknown> Bjarni: The roadveh collision thing was easy.
12:31:17 <stillunknown> and it works ;-)
12:31:56 <Bjarni> stillunknown: "if (GetPrevVehicleInChain(u) == NULL) {" <-- we have a front engine flag that's faster to check
12:32:11 <Bjarni> I even added a function to read it
12:32:20 <Bjarni> IsFrontEngine() or something
12:33:07 <Bjarni> GetPrevVehicleInChain() checks if a vehicle has next set to u. It's way faster to just read a bit in subtype in u :)
12:37:05 <stillunknown> Bjarni: any other comments?
12:37:17 <Bjarni> loading old savegames...
12:38:01 <Bjarni> all the caches will be 0 when loading, so it can load ghost train issues
12:38:10 <Bjarni> trains passing though each other
12:38:53 <Eddi|zuHause3> should be just a simple loop on loading
12:39:23 <Bjarni> also it's preferred to add {} after if(), even if it's only one line following. While it works without it, it's easier to read and (specially) to prevent future bugs
12:39:35 <boekabart> stillunknown: I can remember the days I had an OV... Now I always have to PAY. (so i go by car most of the time)
12:39:45 <Bjarni> Eddi|zuHause3: something like that
12:40:59 <Bjarni> but the loop should only be when loading old savegames
12:41:15 <Bjarni> since the map (and therefore cache) is saved
12:42:09 <Eddi|zuHause3> if (check savegame version) { loop } ... that one is pretty straight forward :p
12:42:41 <Bjarni> this would (at least in theory) also reduce the need for optimising the load loop since it will be called rarely and never when time matters (like when joining multiplayer games)
12:43:06 <Bjarni> Eddi|zuHause3: well, the loop thing is the interesting part
12:43:16 <Bjarni> the question is how to do it best
12:43:46 <Eddi|zuHause3> i would not put a lot of thought in that...
12:44:05 <Bjarni> most likely by looping though all front engines and then walk though each train and increase the counter for each tile it passes
12:44:15 <Bjarni> just found a bug in the diff
12:44:56 <Bjarni> when two trains crash into each other, they will... crash (didn't expect that one coming :p ) and after a while the crashed train is cleared, yet it will not decrease the counter
12:45:30 <stillunknown> That's intentional.
12:45:49 <Bjarni> even after it's cleared?
12:46:18 <stillunknown> When train number 3 comes by, it will force a full check when leaving.
12:46:21 <Bjarni> if a tile contain 2 trains that crashes, the counter will be 2, even after the trains are gone
12:46:29 <stillunknown> This will reset the counter to 0.
12:46:39 <Bjarni> what about the rear of both trains?
12:46:59 <Bjarni> the counter will be 1, then grow to 2 and after the trains leave, it will be 1 again
12:47:04 <Eddi|zuHause3> which counter? after the vehicles are cleared, you do not know which tiles where occupied by the train
12:47:17 <stillunknown> Above 3 trains, it will do a full check.
12:47:37 <stillunknown> So it will find 1, minus one leaving.
12:47:59 <Bjarni> but the rear end of a crash train will be on tiles where the counter is 0
12:48:08 <Eddi|zuHause3> that is really dirty implementation...
12:48:29 <stillunknown> I will improve that.
12:49:16 <Bjarni> when clearing a crashed vehicle, you have to loop though all vehicles to count them on the tile in question
12:49:36 <Bjarni> it's a rare event (or at least should be), so it's ok
12:49:41 <stillunknown> I know, just looking for the place were that happens.
12:50:05 <Bjarni> train_cmd.cpp has some function called clear crashed train or something
12:50:15 <Eddi|zuHause3> hm, when clearing a crashed vehicle, you should only have to check wether the next vehicle in the train is on the same tile
12:50:22 <Eddi|zuHause3> if not, decrease counter
12:50:36 <Eddi|zuHause3> hm... the train could be self looping
12:50:41 * Bjarni decides to actually open the source files instead of just the diff
12:50:41 <Eddi|zuHause3> that won't work...
12:51:05 <Eddi|zuHause3> so loop through all vehicles of the train
12:51:18 <Bjarni> no, it was called something else
12:51:55 <Bjarni> the last lines will delete the train when the counter reaches a certain point
13:05:21 <Bjarni> <Eddi|zuHause3> hm... the train could be self looping <-- I already thought about this. It will leave the tile and then come back to it. In such a case the train will be counted twice, so it's ok
13:05:29 <Bjarni> but that would leave another problem
13:06:07 <Bjarni> if it enters the same tile say 5 times and then leaves, it will count the number of trains in the tile, which will be only 1, but it will leave 4 times
13:06:43 <Eddi|zuHause3> then you have to disallow this multiple counting
13:07:09 <Eddi|zuHause3> i.e. if a train enters a tile, check, that it is not already on it, and if it leaves a tile, check also
13:07:45 <Eddi|zuHause3> looping through a train should not be that expensive
13:11:17 <Bjarni> that's not a huge issue
13:11:41 <Bjarni> it will be an issue to check if the train in question is already on the tile when leaving though
13:12:20 <Bjarni> except if we make the trains double linked lists so it has a prev counter to increase speed when reading the previous vehicle
13:12:20 *** BobingAbout has joined #openttd
13:19:51 <stillunknown> Bjarni: I can see one issue, if a train crosses itself (pretty rare) it will do collision checking, The problem is when a train crosses itself and another train is on the tile, this triggers the >=3 trains, which will only detect 2 trains, so when the crossed train leaves, it is set to zero. Another train could ride through that other train.
13:19:58 <stillunknown> Very unlikely i suppose.
13:20:02 *** DJ_Mirage has joined #openttd
13:21:13 <Bjarni> the train can cross itself more than once on the same tile if somebody tries to break the game
13:21:19 <Bjarni> and as we know, people will do that
13:21:35 <Bjarni> people tries to break the game in all sorts of weird ways
13:21:46 <Eddi|zuHause3> Bjarni: doesn't a wagon have a link to the front engine?
13:21:55 *** redmonkey has joined #openttd
13:21:56 <Eddi|zuHause3> that removes the need for double linked list
13:23:23 <Bjarni> I suppose that the first pointer could be used here
13:23:38 <Bjarni> Vehicle *next; // next <-- now that's a wonderful comment :p
13:24:55 <stillunknown> I wonder, can a train cross and still have another train on the tile.
13:24:56 <Eddi|zuHause3> don't you love that kind of comments?
13:25:31 <Eddi|zuHause3> Some_Type variable_that_does_something // variable of some type that does something
13:25:31 <Bjarni> stillunknown: highly unlikely, but we should be prepared for user stupidity
13:25:51 <Bjarni> also the counter should be correct when the train leaves
13:25:58 <Tobin> "we should be prepared for user stupidity" <-- Damn skippy.
13:26:32 <Tobin> Stupidity knows no bounds and people will push things if they think they can break them.
13:26:48 <Bjarni> let me put it this way: "user stupidity can figure out how to break the game and the stupid user will then do that over and over in multiplayer games"
13:27:04 <Bjarni> so it's a protect MP games issue
13:27:42 <Eddi|zuHause3> <stillunknown> I wonder, can a train cross and still have another train on the tile. <- sure, just have a double-track tile, and have the train loop on only one half, have the other train pass (or even loop) on the other half
13:28:06 <Bjarni> user stupidity always find a way :p
13:28:30 <stillunknown> The only option i see, is going through the entire train when encountering >1 trains on a tile
13:28:41 <redmonkey> which version of openttd is the closest to the original ttdlx?
13:28:47 <stillunknown> And checking if another piece of the train is on the same tile.
13:29:05 <stillunknown> minus the first 2 wagons, ofcource.
13:29:36 <Bjarni> redmonkey: 0.1.1, but then again it's buggy, so maybe not
13:29:47 <Bjarni> but why do you want TTD?
13:30:26 <Eddi|zuHause3> most additional features can be disabled or ignored...
13:30:26 <redmonkey> because there is no ttdlx for unix systems :)
13:30:46 <Bjarni> didn't really answer my question
13:30:59 <Bjarni> why do you want TTD when you can have OpenTTD?
13:31:19 <Tobin> redmonkey: Checkout the earliest revision from SVN and see if it compiles for you.
13:31:51 <Tobin> redmonkey: And be prepared to be disappointed. TTDLX was good but it's _very_ frustrating after using OTTD.
13:32:26 <redmonkey> because i loved the old TTD game just the way it was. sure, OpenTTD is great too. but sometimes i just wanna play the original one
13:32:28 <stillunknown> People would have been stunned in '94 if they got ottd instead of ttd.
13:32:53 <Eddi|zuHause3> people would have been frustrated of ottd, because it runs much slower...
13:33:05 <Bjarni> <stillunknown> People would have been stunned in '94 if they got ottd instead of ttd. <-- specially by the resolutions we play in today xD
13:34:06 <redmonkey> could you bugfix OTTD 0.1.1 without adding new features as a sideproject? that'd be great! and i know a lot of other people who'd love it too :)
13:34:23 <Kjetil> Playing the original is just annoying. All the cool features are missing
13:34:35 <Bjarni> I don't think anybody will spend ages doing that
13:34:46 <Bjarni> also 0.1.1 is windows only
13:35:06 <Bjarni> you really has to use 0.2.2 or something before the ports became any good
13:36:02 <Bjarni> I really wonder by so many people downloaded 0.1.4 for OSX and nobody bug reported that it had hardcoded absolute paths to my 3rd party libs
13:36:23 <Bjarni> tells you something about most users :p
13:36:29 <Tobin> Bjarni: Hardly anyone gives good bug reports.
13:36:41 <Bjarni> I didn't say good, I said *any*
13:36:55 <Bjarni> I didn't even get "this shit fails to start" or anything
13:38:31 <redmonkey> it was just an idea :) but you're right. if i wanna play the original, i can play it in an emulator or something. OTTD is way more advanced and i'm looking forward to the new 32bit graphics.. they just look awesome!
13:39:15 <Tobin> Heh, 32bit graphics. Been there played that. :)
13:39:48 <Tobin> And back before any of you youngin's had a chance to get your hands on it too. :P
13:40:04 <Tobin> I don't think anything cam of the early work though.
13:40:22 <Tobin> By early work I mean Mek's.
13:41:13 <stillunknown> What the easiest (and cheapest) way to check if a tile has branches?
13:42:06 <redmonkey> but please don't overload OpenTTD with too much new stuff. because then it wouldn't be TTD anymore.
13:43:01 <stillunknown> They'll only ruin gameplay in favor of eye candy.
13:44:00 <Maedhros> stillunknown: TracksOverlap(), probably
14:29:49 *** Calamarain has joined #openttd
14:42:48 <CIA-1> OpenTTD: maedhros * r10023 /trunk/src/train_cmd.cpp:
14:42:48 <CIA-1> OpenTTD: -Fix (r3218): When selling trains, if there were no wagons between multiheaded
14:42:48 <CIA-1> OpenTTD: engines the rear part could be checked despite having already been deleted,
14:42:48 <CIA-1> OpenTTD: triggering an assert.
14:47:00 *** Apocalipsys has joined #openttd
15:06:54 <nairan> i fell down near bd at the lont road up to the bd fort and now im fown at the ground stuck in the floor
15:08:40 <nairan> bad to write in the wrong cha
15:22:46 *** thgergo has joined #openttd
15:32:16 *** peterbrett has joined #openttd
15:41:49 <CIA-1> OpenTTD: bjarni * r10024 /trunk/src/build_vehicle_gui.cpp: -Feature: [build windows] trains, road vehicles and ships can now be sorted by cargo capacity (planes already had this option)
15:47:24 <_Mist_> aggregated cargo capacity?
15:48:00 <CIA-1> OpenTTD: bjarni * r10025 /trunk/src/build_vehicle_gui.cpp:
15:48:00 <CIA-1> OpenTTD: -Fix: [build windows] sorting planes for capacity didn't check mail capacity. Now mail capacity is used if passenger capacity is the same
15:48:00 <CIA-1> OpenTTD: The sort will still resort to EngineID if both mail and passenger capacities are the same
15:48:21 <Bjarni> now that part should be ok
15:48:55 <peter1138> git, i just compiled ;(
15:49:03 <Bjarni> it's just one cpp file
15:49:07 <Bjarni> no headers or lang files
15:49:19 <Eddi|zuHause3> how do trains with multiple cargos get handled?
15:49:32 <Bjarni> it's in the build window
15:49:44 <Phazorx> coopers server crashes again :/
15:50:15 <Bjarni> so it just reads the same number as the one printed in the cargo capacity in the details part of the window
15:50:52 <Bjarni> so if it sorts cargo incorrectly, then it's because it's also printed incorrectly
15:51:07 <Bjarni> anyway it shouldn't cause any problems
15:51:39 <Bjarni> dualheaded engines are sorted with their combined capacities
15:52:06 *** Rexxars has joined #openttd
15:52:58 <Maedhros> Phazorx: what's the problem this time?
15:54:44 <peter1138> Maedhros: though i'm investing this multihead/wagon bug
15:55:46 <peter1138> investigating, of course
15:55:52 <Phazorx> peter1138: did i misunderstoof that? #10020 does fix /src/train.h:234: Vehicle* GetNextVehicle(const Vehicle*): Assertion `v->type == VEH_TRAIN' failed.
15:56:12 <Phazorx> investing time = investigating
15:56:35 <Maedhros> Phazorx: it fixes one of them, yes...
15:56:53 <Maedhros> was anyone selling an aeroplane when it happened?
15:57:10 <Maedhros> alternatively, do you have disasters turned on?
15:57:28 <Phazorx> disasters are off for sure
15:57:51 <Phazorx> selling a chopper apparently was the trigger
15:58:12 <Phazorx> is it shared orders related?
15:58:19 <Maedhros> ok, then it was fixed by r10020
15:58:50 <Phazorx> heh coop should get renamed to beta test team
15:59:03 <Maedhros> it's not to do with shared orders - it's calling GetNextVehicle which is only valid for trains on aircraft
15:59:35 <Phazorx> i figured that, but i'm not familiar with what that function does
16:00:05 <Eddi|zuHause3> Bjarni: example, i have a train with 1 mail wagon, and 4 passenger wagons, then a train with 4 passenger wagons, and 1 mail wagon (same capacity, different order), how does this get compared?
16:00:54 <Maedhros> vehicles are linked lists - aircraft have 2 parts, trains have as many as you want
16:01:06 <Maedhros> but trains also have articulated parts which are skipped by GetNextVehicle
16:01:11 <Eddi|zuHause3> other example, i have a train with 2 livestock wagons, and 3 grain wagons, then one with 3 livestock wagons, 2 grain wagons (one capacity higher, one lower)
16:01:28 <Maedhros> but to do that it needs to check v->subtype, which holds different values depending on the vehicle type
16:01:30 *** skidd13 has joined #openttd
16:01:48 <Bjarni> Eddi|zuHause3: it's the build window, so it's not the list of actual trains, but the available engines
16:01:48 <peter1138> i'm blaming bjarni on this one ;p
16:02:07 <peter1138> reverting r9973 makes the crash go away
16:02:13 <Eddi|zuHause3> ah... then i misunderstood things... :)
16:02:33 <Eddi|zuHause3> !openttd commit 9973
16:02:34 <_42_> Commit by bjarni :: r9973 /trunk/src/train_cmd.cpp (2007-05-29 10:35:52 UTC)
16:02:36 <_42_> -Cleanup: removed some code that kept both ends of a dualheaded engine in the same train (when moving wagons in a depot)
16:02:38 <_42_> NormaliseTrainConsist() is called later in the same command and it will take care of this issue
16:03:37 <Bjarni> great, I fix one issue and then another one appears
16:03:48 <Bjarni> but I have yet to figure out what went wrong this time :/
16:04:27 <peter1138> Bjarni: triggered by building a multiheaded engine, putting a couple of wagons on it, then ctrl-moving all the wagons off, and then ctrl-moving them all back again
16:04:57 <peter1138> possibly other combinations, but that does it always, for me
16:09:41 <Bjarni> looks like the next pointer isn't set up like it should
16:11:11 *** lolman_ has joined #openttd
16:11:16 <Eddi|zuHause3> is the normalise function called for both consists?
16:31:23 <Phazorx> multihead bug strikes coopers server again :/
16:31:52 <Bjarni> Eddi|zuHause3: yes, but that's not the problem
16:35:44 <Maedhros> Phazorx: what happened this time?
16:36:38 <Phazorx> selling turbotrain crashed it apparently
16:37:19 <Maedhros> was it a multiheaded train without any carriages inbetween?
16:37:33 <Phazorx> turbotrain is dualhead
16:38:09 <Maedhros> did it have any carriages? and was someone selling it?
16:38:21 <Phazorx> someone was selling it
16:38:25 <Phazorx> cghecking on carriage status
16:38:39 <Maedhros> it was probably fixed by r10023 :)
16:40:48 <Phazorx> i guess we have to wait for nighty to figure it out
16:42:32 <Bjarni> hmm, one step closer to what goes on
16:43:17 <Bjarni> some cache isn't updated when moving the wagons OUT of the train and when it relies on the cache when adding them again, it asserts
16:58:47 <Wolf01> is here a developer available to help me fixing the main title viewport scroll patch?
17:04:25 <Wolf01> look nice, but i prefer vertical signals at road sides :)
17:04:37 <Noldo> The arrows look nice but the road it self not so much
17:04:41 <Maedhros> Thomas[NL]: looks nice
17:04:42 <Eddi|zuHause3> these arrows look really bad...
17:04:58 <Eddi|zuHause3> rather put an arrow on each lane...
17:05:48 <Eddi|zuHause3> and on the sloped roads, they are off center
17:07:15 <Thomas[NL]> they use the same graphics as the flat road, also in the "original" one-way arrows
17:08:17 <Bjarni> finally. Now I know what went wrong
17:08:25 <Bjarni> now the big question is how to fix it...
17:08:39 *** UndernotBuilder has joined #openttd
17:08:40 <peter1138> revert that commit ;p
17:09:13 <Bjarni> because then we will be back to the bug that the previous commit fixed
17:13:59 <UndernotBuilder> -Fix (r10008): Add articulated_vehicles.h to the project files.
17:23:44 <CIA-1> OpenTTD: bjarni * r10026 /trunk/src/train_cmd.cpp:
17:23:44 <CIA-1> OpenTTD: -Fix: unlinking a rail vehicle didn't clear the first pointer
17:23:44 <CIA-1> OpenTTD: After r9973, this could cause some asserts as some vehicles could end up with first pointers to chains they didn't belong to
17:26:01 <Bjarni> Phazorx: this should solve your issue and it's in time for the next nightly build (in half an hour)
17:30:18 <Wolf01> yes, i really like that
17:31:07 <Bjarni> remember roadsides should apply ;)
17:31:20 <Bjarni> it's not everybody, who drives on the left
17:33:22 <Bjarni> damn, what a stupid bug... I spent ages debugging, backtracing and source reading just to add one simple line >_<
17:36:08 <Bjarni> spending 90 minutes on entering 16 chars of code... that's one char every 5,6 minute o_O
17:37:42 <Wolf01> i own you all, 4 hours to make one line of code like "if (st->owner != player) {"
17:38:01 <Eddi|zuHause3> Thomas[NL]: the line on the front sign should have an angle. like the lines of the road
17:38:37 <Bjarni> well, my line was just var = NULL;
17:38:49 <Thomas[NL]> Eddi|zuHause3, your right
17:39:04 <Eddi|zuHause3> Thomas[NL]: and you probably get issues with vehicles driving "over" the sign
17:39:17 <Eddi|zuHause3> unless you handle them like rail signals
17:40:51 <Wolf01> so.. this feature requires a patch
17:41:27 <Bjarni> usually new features has a strange demand for modifying the code ;)
17:43:04 <Phazorx> heh as i justr said to coopers - i hate trams already w/o even seeing them
17:43:27 <Phazorx> snowball effect is not very motivating
17:43:32 * Sacro thinks he should have commit rights
17:43:33 <Eddi|zuHause3> if you introduce code to handle road signs like rail signals, you also have no problem with traffic side, because you can just switch it
17:43:50 <Maedhros> Phazorx: those bugs had nothing to do with trams :p
17:43:54 <Phazorx> btw anyone looked into stillunknown idea?
17:44:22 <Phazorx> Bjarni: you change little thing that is supposed to be fine and discover whole chain of events tstarting from it
17:44:56 <Bjarni> I broke dualheaded engine because of some bug
17:44:58 <Phazorx> Maedhros: exactly... but i figure implementing trams caused revamping od some common parts that uncovered old issues
17:45:04 <Eddi|zuHause3> hm... not enough, i think, Thomas[NL]
17:45:05 <Bjarni> and the fix uncovered another bug
17:45:10 <Bjarni> and that fix caused this
17:45:14 <Wolf01> Bjarni, you have 5 minutes to fix them :P
17:45:26 <Phazorx> Bjarni - will cacghing make it into truck soon
17:45:36 <Phazorx> i kinda liked performance boost of that
17:46:01 <Eddi|zuHause3> Phazorx: there are still issues with some corner cases
17:46:20 <Bjarni> ahh yes, the original bug that I fixed was autoreplacing trains with 100 wagons
17:46:23 <Phazorx> Bjarni: that fix for a fix of a fix reminds me of MS bugs -"there was an error diplaying an error for unknownn error"
17:46:28 <Bjarni> and that broke dualheaded engines
17:46:30 <Wolf01> i wonder if caching will be extended to ships too
17:46:40 <Bjarni> fixing that broke something else regarding dualheaded engines
17:46:54 <Phazorx> but i still hate trams
17:47:09 <Maedhros> you should probably hate articulated road vehicles instead ;)
17:47:31 <Bjarni> actually I didn't cause new bugs, I just made it easier to trigger bugs we already had in the code
17:47:38 <moe> hmm where to make the one way roads active?
17:47:39 <stillunknown> Phazorx: the main issue currently is that weird cases like trains going trough itself need to be handled.
17:47:43 <Phazorx> "... I found the fork later. But we feelings are gone now"
17:47:43 <Eddi|zuHause3> i love trams, hopefully, there will be a complete tram set...
17:48:01 <stillunknown> trams made it into ottd?
17:48:47 <moe> how to buld one way roads?(latest nightly (10015)
17:49:01 <Maedhros> control click with the road building tool
17:49:18 <Phazorx> stillunknown: but there is hope right?
17:49:58 <stillunknown> Phazorx: Ofcource there is, but i'm still considering various solutions.
17:49:58 <Phazorx> how about one way bridges?
17:50:16 <Bjarni> <Phazorx> stillunknown: but there is hope right? <-- well, it depends if I can figure out some outer rare cases where it will fail :p
17:50:20 <Phazorx> stillunknown: i see... your efforts are gretly appreciated
17:50:41 <Phazorx> Bjarni: coopers are good for figuring out rare cases when soemtihng that usualy works - fails
17:51:01 <Bjarni> we better get this one right from the start as it affect savegames and buggy data is saved
17:51:37 <Bjarni> Thomas[NL]: the right one is still wrong :p
17:51:51 <Thomas[NL]> yeah, don't look at that one :)
17:53:33 <Bjarni> hurry up and fix it before the nightly builds
17:53:39 <Maedhros> peter1138: you're good at finding these :)
17:54:05 <peter1138> basically, have a 'built-up' road with trees on it
17:54:19 <peter1138> build a tram track perpendicular across it, with no roads
17:54:27 <peter1138> the trees are still there, right on the tracks
17:54:59 <stillunknown> One of the solutions i'm considering, is during VehicleEnterTile check for crossing track and occupancy > 1, then iterate trough train, if wagon is on same tile --> add a flag to vehicle (which can be increased when more crossings happen), the last wagon does the same and will remove the flag when the crossing ends.
17:55:08 *** Frostregen__ has joined #openttd
17:55:20 <Eddi|zuHause3> Thomas[NL]: the left one looks ok now
17:55:27 *** Frostregen__ is now known as Frostregen_
17:56:00 <Bjarni> stillunknown: not good enough. Imagine a circle line where the train is 3 times as long as the tracks. It will drive on top of itself even on tiles without crosses
17:56:44 <Maedhros> why is that trains can't crash into themselves again? they could in tto, as far as i remember...
17:56:56 <stillunknown> Bjarni: But the train still has an end, which can detect if another section is on that tile
17:56:58 <Eddi|zuHause3> no, they never crashed
17:57:30 <peter1138> damn, cs trams are too long
17:58:21 <Phazorx> stillunknown: i'm not very familiar with the code - but what is the special case with train going thorugh itself?
17:58:33 <Phazorx> and should that be handled in same fasion as any other colision?
17:58:37 <stillunknown> Bjarni: so the last wagon and the first check each tile, if they are entering a tile that isn't a simple track and if a train is already present on that tile
17:58:57 <stillunknown> phazorx: trains can ride through themselves, in circles if you want
17:59:17 *** Frostregen_ is now known as Frostregen
17:59:34 <Bjarni> trains can drive though itself on tiles that are simple tracks
17:59:36 <Phazorx> stillunknown: is that on purpose?
17:59:43 <Eddi|zuHause3> stillunknown: the problem with a "flag" is, that it can run in circles multiple times
17:59:53 <stillunknown> that flag would be a counter
18:00:12 <Maedhros> Phazorx: well, it was done deliberately, yes
18:00:21 <stillunknown> But i agree a circle situation could be problematic.
18:01:23 <Eddi|zuHause3> but i have a suggestion
18:01:37 *** DJ_Mirage has joined #openttd
18:01:38 <Eddi|zuHause3> if you count the multiple train multiple times
18:02:00 <Eddi|zuHause3> and you have to recheck if the counter is >= 3
18:02:08 <Eddi|zuHause3> you can walk through the train
18:02:12 <Phazorx> wasnt that supposde to be 1bit counter?
18:02:26 <stillunknown> the counter in the map is 2bits
18:02:43 <stillunknown> Since common situations do include 2 trains on one tile.
18:02:43 <Phazorx> stillunknown: is that 1 flag 4 cases or 2 flags ?
18:02:48 <Eddi|zuHause3> and if the tile of the next wagon is different than the tile on the current wagon
18:02:48 <Phazorx> and what they would represent ?
18:02:57 <Eddi|zuHause3> you can increase the counter again
18:03:10 <Eddi|zuHause3> that way, the train will be counted multiple times even on recheck
18:03:13 <Bjarni> stillunknown: notice the last circle... it's normal tracks (no switches or crossings), yet the train is on top of itself
18:03:35 <hylje> silly example is sillyyyyy
18:03:49 <stillunknown> normal track is track that doesn't have branches
18:03:58 <Bjarni> yeah, but it's a matter of not leaving incorrect cache behind
18:04:03 <stillunknown> But i see the problem of this situation.
18:06:16 <Eddi|zuHause3> so you can have a function: RecheckTrain(front engine) { for each (wagon) {if (previous tile != this tile) { increase counter } } }
18:06:54 <Eddi|zuHause3> Thomas[NL]: but still the issue is, that vehicles will not drive "behind" the sign, but "over" the sign
18:07:39 <stillunknown> Eddi|zuHause3: I think so too, but i'm not 100% sure.
18:07:42 <Bjarni> the thing is: whatever we do, it shouldn't kill the performance boost or the whole idea is gone
18:07:58 <Eddi|zuHause3> Bjarni: circling trains should be rare
18:08:06 <Thomas[NL]> Eddi|zuHause3, yes I know, but I don't think I'm the one to solve that :)
18:08:33 <Bjarni> but the conditions where we should check if they are circling isn't, unless we figure out how to do this right
18:08:45 <Eddi|zuHause3> so if we could find an efficient check, if the train is circling
18:08:45 <Maedhros> Eddi|zuHause3: only if they're drawn as ground tiles
18:09:27 <Bjarni> I think I have an idea
18:09:46 <Bjarni> just let it count each time a train passes
18:10:39 <Bjarni> and solve the issue when the counter reaches 3. When calculating what the counter should be when decreasing, find the front engine and loop though the train to count how many times it passes though the tile in question
18:10:54 <Bjarni> and make some clever device to ensure that each train is only counted once
18:11:02 <stillunknown> You can keep a crossing counter in the train, were you could use a byte.
18:11:48 <Bjarni> I was thinking about keeping the crossing counter in a function, so it only lives as long as it counts the number of trains on the tile
18:11:51 <Eddi|zuHause3> you could have a flag in the front engine "circling", each time the engine enters a tile with count > 0, you check if it's circling (i.e. loop through all wagons), and each time the last wagon leaves a tile, and circling flag is set, recheck the flag
18:12:11 <Bjarni> and only call that function whenever the tile should be decreased and the count is 3
18:12:22 <UndernotBuilder> what means 'GRM' in -Codechange: Implement GRM for newcargos
18:12:40 <Bjarni> then the only time we will really waste CPU cycles is when trains circle or when the count is 3, both rare cases
18:12:53 <Eddi|zuHause3> [2007-06-02 11:09] <Maedhros> grf resource management - designed to avoid conflicts between grf files
18:13:06 <hylje> i think we can even assert out on circling trains
18:13:19 <Bjarni> that can break MP games
18:13:40 <hylje> yeah well, its nice for dos :)
18:13:45 <Eddi|zuHause3> hylje: asserts are for situations that (theoretically) can never occure
18:13:47 <Phazorx> hmm... can someone englighten me on what the flag actualy represents?
18:13:49 <stillunknown> Bjarni: what you want would require each car to check almost all other cars for matches.
18:13:52 * Tefad asserts his EXPLICTIVE-DELETED
18:14:11 <stillunknown> More efficient to let the first do the check and the last, and store in the train.
18:14:35 <Kjetil> What kind of a exiting problem av we discussing ?
18:14:42 <Eddi|zuHause3> hylje: e.g. train driving on a tile that has no rail
18:14:53 <Bjarni> stillunknown: it's in the case where the count is 3 and a reduction is requested, then it should count the trains to see if it should still be 3 or just 2
18:15:13 <Eddi|zuHause3> but train circling is a situation that can occure by the game logic
18:16:00 <stillunknown> Bjarni: It's a bad idea to let the same train count twice, better to let the train register once and unregister it's completely gone.
18:16:05 <Bjarni> stillunknown: actually how do you plan to figure out if it should be 3 or 2 when the count is 3 and a decrease is requested?
18:17:17 <kaan> I think that counting the same train twice is a nice way to get to check for self collision
18:17:25 <stillunknown> I made a function that uses the hash map, to count when needed.
18:17:56 <Bjarni> stillunknown: but in order to save CPU cycles, what would we gain most of? making a train count twice in this rare event or add some checks whenever a train enters or leaves a tile?
18:18:33 <Eddi|zuHause3> kaan: actually, the self collision has no game effect, so we don't need to check it, but it is a corner case that kills the cache idea
18:18:34 <kaan> actually, what ever the solution we should make a spin-off snake game out of it ;)
18:19:17 <stillunknown> It is my understanding that the hashmap only gives one entry for a train, but i will have to check if that's an easy way out.
18:20:15 <Bjarni> the easy way out is not to do anything. We want the bugfree and fast way out of this one ;)
18:20:45 <kaan> Bjarni, the conservative spirit personified :D
18:21:24 <stillunknown> Bjarni: known corner cases need to be handled, preferably in a way that is not expensive the rest of the time
18:21:38 <kaan> But you are right, a change like this should only be made once it is proven to have no ill effects at all
18:22:58 <Bjarni> stillunknown: and that's why I would put the extra CPU load on the case where the counter is 3 (rare) and don't care about the extra collision tests when the train crosses itself (even more rare)
18:23:20 <Bjarni> and prevent tests in more common cases
18:23:33 <Bjarni> like train entering/leaving tile
18:23:45 <stillunknown> But this would still require some awareness of crossing itself.
18:24:11 <kaan> that awareness is not that hard to acheive is it?
18:24:18 *** Sacro|Laptop has joined #openttd
18:24:29 <stillunknown> kaan: I have an idea, which i'm trying.
18:24:32 <kaan> you know the start of the train and you know when the end is exiting
18:24:39 <Phazorx> hmm.. different issue - why is there a train window oppening for many/all players when new train is created??
18:25:48 <kaan> so if the start of the train is on the same tile as a wagon that is not the end then we know that a crossing is in progress
18:26:17 <Eddi|zuHause3> stillunknown: have you considered my last idea?
18:26:30 <stillunknown> You mean count the number of crossings?
18:27:11 <Eddi|zuHause3> counting is unnecessary
18:27:56 <kaan> Phazorx: i havent got a clue, is there a bug report on it?
18:28:13 <stillunknown> Eddi|ZuHause3: good point
18:28:29 <Eddi|zuHause3> the case is so rare, instead of saving the count, you can just recalculate it
18:28:50 <Phazorx> i'm clarifiying if this is expected behavior or not
18:29:11 <Phazorx> if any of devs will join coopers game- you'll see it after 5 min or so
18:29:36 <peter1138> yeah, problem is the window popping up is based on company, not network client
18:29:55 <Phazorx> so i take it is unplanned then
18:30:57 <peter1138> unless it's a new one :p
18:32:16 <Phazorx> peter1138: pops up when trains are manualy upgraded
18:32:51 <peter1138> new bug, not new vehicle
18:33:25 <Phazorx> i guess it can have "feature" status since it is not fatal :)
18:33:39 <peter1138> well afaik it's done that for a *long* time
18:33:53 <Eddi|zuHause3> Thomas[NL]: looks fun, especially if it is animated :)
18:34:37 <Phazorx> peter1138: are you on cooop game?
18:34:39 <Maedhros> Thomas[NL]: that'll look very weird for rvs driving through from the other side ;)
18:35:15 <Thomas[NL]> Maedhros, ooh didn't think of that...
18:35:59 <Wolf01> do competitors vehicles ignore the allowed direction of the road?
18:36:24 <UndernotBuilder> updated to r10026 again by bugfixes?
18:36:25 <Phazorx> peter1138: part of company?
18:36:51 <UndernotBuilder> what means 'GRM' in -Codechange: Implement GRM for newcargos
18:37:11 <peter1138> that same as it meant last time
18:39:47 <Phazorx> peter1138: i see it when other do it
18:39:55 <Phazorx> and when i do it -they see it :/
18:40:38 <kaan> Could we use UndernotBuilders collision system for PBS?
18:41:31 <Tefad> i remember the second svn reset
18:41:46 <Tefad> the first one was a while ago
18:41:52 <stillunknown> Anyone know of function which gives the trackbits for a specific tile number?
18:42:33 <kaan> sorry, not UndernotBuilder, but stillunknown of course *blush*
18:50:18 <Phazorx> peter1138: it looks like it happens when moving orders to reassigned car which beciome active engine ?
18:50:29 <Phazorx> cuz if i sell 2nd engine 1st it does not happen
18:50:42 <Phazorx> but if i seel primary - it pops up for most/all players
19:00:55 *** Sleepie has joined #openttd
19:02:18 <Eddi|zuHause3> <stillunknown> Anyone know of function which gives the trackbits for a specific tile number? <- i'd search in rail_map.h :)
19:19:47 <izhirahider> How can I test trams to be able to see how they work? Do I have to download some extra thing?
19:20:49 <izhirahider> If I get a tram grf file, will it always be available to construction, or do I need to activate an ooption somewhere
19:21:54 <Bjarni> it will appear in the game when the grf is active
19:22:13 <Bjarni> like there are multiple rail types
19:24:21 <izhirahider> why isn't a tram grf included in the repo?
19:24:41 *** UndernotBuilder has quit IRC
19:25:15 <Eddi|zuHause3> izhirahider: same as why there are no alternative railsets included
19:26:50 <izhirahider> electrical, monorail, maglev
19:27:21 <Maedhros> those are railtypes, and that's because they were in the original game :p
19:28:38 <izhirahider> ah, by railsets you mean same trains but with different namings and image representation
19:29:00 <peter1138> well no, different trains...
19:29:30 <izhirahider> Where can I find the tram grf? The wiki doesn't say
19:29:40 <Eddi|zuHause3> where you find all other grfs
19:29:49 <peter1138> Sacro: draw it and i'll make it
19:30:26 <Bjarni> Sacro: while you are at it, can you draw narrow gauge as well?
19:30:38 <Maedhros> someone's already done that
19:30:38 <peter1138> well narrow gauge is already drawn
19:33:08 <Eddi|zuHause3> if i load a narrow gauge railset, will narrow gauge replace a railtype, or add a railtype?
19:33:21 <Maedhros> it's not possible to add railtypes yet
19:33:31 <Sacro> fine, i'll draw some railtypes
19:33:51 <Sacro> izhirahider: add it yourself :p tis a wiki
19:34:09 <Eddi|zuHause3> hm, i thought the system was prepared for this
19:39:20 <Maedhros> it's not easy to make out the arrow - maybe you could make the sign more rectangular?
19:39:26 <Eddi|zuHause3> hm, looks more like a cross rather than an arrow
19:40:20 <Eddi|zuHause3> pherhaps move the left and right white pixels a little more up
19:40:36 <Sacro> one way should be a rectangle, no entry should be octagonal
19:40:43 <Sacro> so you can differ when facing away
19:41:05 <Eddi|zuHause3> no entry is round
19:42:02 <Tefad> no entry is square with circle
19:42:20 <Sacro> Tefad: no entry should be round
19:42:25 <Tefad> road signs are going to be interesting ; )
19:42:33 <Eddi|zuHause3> Tefad: where do you live?
19:42:33 <Tefad> US has different signs than EU
19:42:59 <Tefad> no entry signs are rectangular with red circle and white horizontal bar in the middle
19:43:02 *** RamboRonny has joined #openttd
19:43:30 <Tefad> sometimes "DO NOT \ ENTER" is written in the red margins of the circle
19:43:46 <Tefad> yes, that's always on a rectangular sign
19:43:50 <Eddi|zuHause3> the sign is round, not rectangular
19:44:02 <Tefad> i've never seen it on a round sign in the US
19:44:06 <Thomas[NL]> yeah I'll write Do no enter on the sign :P
19:44:12 <Tefad> i don't think i've seen round signs in the US
19:44:17 <Tefad> other than RR crossing signs
19:45:16 <Eddi|zuHause3> most signs in europe are round
19:45:25 <Sacro> Eddi|zuHause3: that means no vehicles
19:45:47 <Eddi|zuHause3> Sacro: exactly, vehicles may not enter
19:46:30 <Bjarni> it's usually combined with a sign telling exceptions like busses, local residents or similar
19:48:16 <Eddi|zuHause3> it's difficult on that scale :)
19:48:29 <Sacro> though i'd prefer the arrow to be on a vertical rectangle
19:49:10 <Eddi|zuHause3> Sacro: that forbids powered vehicles, but allows unpowered vehicles (bikes etc.)
19:49:23 <Sacro> Eddi|zuHause3: yes i know
19:50:09 <peter1138> Sacro: re drawing railtypes, i do have a plan for adding them, see...
19:50:20 <peter1138> so it'll be useful...
19:50:28 <hylje> its not like we have bicycle traffic here
19:51:06 <Sacro> peter1138: oh? cos i know the LUSet would like 4 rail to run on
19:51:17 <Sacro> and i think BRSet would use 3rd rail
19:52:55 <Sacro> Eddi|zuHause3: surely you just render the track after the train?
19:53:04 <Eddi|zuHause3> although that is technically a tram
19:53:17 <hylje> Sacro: and reasonably higher
19:55:52 <Sacro> now shall i reboot into linux and do some graphics
19:56:00 <izhirahider> How do I know if I'm using DOS or WIN grfs?
19:56:05 <peter1138> that just looks like a hanging monorail to me ;p
19:56:20 <hylje> izhirahider: win grfs tend to have "w" appended to the filename before .grf
19:56:40 <Bjarni> izhirahider: if you see a lot of pink in the sprites, then it's the DOS ones
19:58:35 <Sacro> Thomas[NL]: it looks a bit better, yeah
19:58:52 <Eddi|zuHause3> gets confused with the questionmark, i assume
19:59:13 <Eddi|zuHause3> as questionmarks are reserved for dynamic content
19:59:44 <Eddi|zuHause3> i liked the previous one better
19:59:53 <CIA-1> OpenTTD: richk * r10027 /branches/NewGRF_ports/ (313 files in 21 dirs): [NewGRF_ports] -Sync: with trunk r9506-10026
20:00:13 <Eddi|zuHause3> that was a huge commit :p
20:00:17 <izhirahider> Bjarni, I see pink and brown
20:00:30 <izhirahider> Mostly brown road, full of dots
20:01:24 <Eddi|zuHause3> i have never seen a 500 revisions sync before :)
20:01:26 *** dihedral has joined #openttd
20:01:39 <Eddi|zuHause3> most of the time, people would do 5 times 100 revisions
20:03:13 <peter1138> richk is not known for his syncing skills :p
20:03:59 <Sacro> ooh, he's still alive then
20:08:36 <CIA-1> OpenTTD: peter1138 * r10028 /trunk/src/train_cmd.cpp:
20:08:36 <CIA-1> OpenTTD: -Fix (r2270/r2951): When deleting the first engine of a train with multiple
20:08:36 <CIA-1> OpenTTD: engines, only reopen the train window if the player had the original train
20:08:36 <CIA-1> OpenTTD: window open. This fixes 'random' windows opening for multiple players of the
20:09:13 *** peterbrett has joined #openttd
20:09:40 <hylje> i just had some random trainz popping up
20:09:54 <Sacro> them are some old stats
20:11:05 <Sacro> yeah, but the list of people too
20:11:15 <Sacro> Celestar, Hackykid, DarkSSH
20:11:31 <peter1138> oh, you're still here. shame ;p
20:12:17 <Bjarni> heh, in those stats, everybody talks about OpenTTD, except Eddi|zuHause3 :p
20:12:43 <peter1138> might've been talking about a track from ttd...
20:12:52 <Bjarni> <Sacro> what happened to LoTP? <-- he took a hike and never returned
20:13:05 <Eddi|zuHause3> "Statistics generated on Friday 26 January 2007" <- it's not exactly current
20:13:26 <Bjarni> you mean you like to be less off topic now?
20:13:40 <Bjarni> do anybody have an up to date one?
20:14:21 <peter1138> i talk the most? :o
20:14:32 <hylje> you just had three lines that
20:14:42 <hylje> its more likely than you think
20:15:12 <Sacro> ^_^ wrote an average of 46.00 words per line.
20:15:28 <Bjarni> "it's like denying you're human" <-- jez did that?
20:15:41 <hylje> by faking a nickname and going "a a a..."
20:15:50 <Eddi|zuHause3> Ailure talks to him/herself a lot. He/She wrote over 5 lines in a row 296 times! <- now that is impressive :p
20:15:52 <hylje> Sacro couldn't decide whether to stay or go. 842 joins during this reporting period!
20:16:36 <Bjarni> Bjarni spoke a total of 214766 words!
20:16:45 <Bjarni> heh, I use more words than peter1138
20:17:11 <Eddi|zuHause3> hm, someone should join the stats of TrueLight and TrueBrain
20:17:54 <guru3> i had no idea those were still updating
20:18:07 <guru3> i had no idea anyone was still using the guru3.sytes.net domain
20:18:10 <Bjarni> It seems that Xera's shift-key is hanging: 10.5% of the time he/she wrote UPPERCASE.
20:18:10 <Bjarni> For example, like this:
20:18:46 <guru3> good old massive waste of cpu power
20:19:04 <Bjarni> Poor Sacro, nobody likes him/her. He/She was attacked 32 times.
20:19:36 <Eddi|zuHause3> Bjarni is a very aggressive person. He/She attacked others 40 times.
20:19:49 <Eddi|zuHause3> 32 times of that Sacro, i believe :p
20:19:56 *** Sacro was kicked by Bjarni (there is no way that you should be allowed to attack me)
20:20:06 *** Bjarni was kicked by peter1138 (why not?)
20:20:06 *** ChanServ sets mode: +o Bjarni
20:20:40 <dihedral> how would i go about using tram roads in r 10002?
20:20:56 <Eddi|zuHause3> dihedral: load a tram grf
20:20:57 <dihedral> just thought i'd have a little peek at one of the nightleys
20:21:07 <guru3> well i'm glad someone looks at the irc stats everynow and then
20:21:15 <guru3> i'll just quietly fade into the backgrond again now
20:21:25 <moe> btw ther is a narrow gauge somewhere on grftracker
20:22:13 <Bjarni> Ailure talks to him/herself a lot. He/She wrote over 5 lines in a row 291 times!
20:22:13 <Bjarni> Another lonely one was CIA-1, who managed to hit 221 times.
20:22:38 <guru3> for some reason i forgot to escape cia 1, 2, and 13
20:22:39 <hylje> ailure just punctuates with return
20:22:44 <guru3> that'll be fixed on the next update
20:22:55 <Rubidium> CIA is still learning; speak after me: "...." ;)
20:31:46 <peter1138> yay, all my towns start off huge
20:34:52 <Bjarni> wanna play OpenTTD 0.9.1?
20:34:54 <Wolf01> because i can't fix the bug
20:43:09 <Bjarni> bbl later. Power died :(
20:43:12 *** De_Ghosty has joined #openttd
20:43:23 * Bjarni is once again happy for his UPS
20:43:44 <De_Ghosty> bump max player to 16 please :)
20:54:09 <Thomas[NL]> hmm trams turn around on top of road-construction-sites.
20:55:12 <Wolf01> i don't like trams turn-around too, i like more a turn-around in the center of the tiles
20:56:55 <Rubidium> Wolf01: not going to happen when you want to be newgrf compatible
21:05:00 <peter1138> no aftermath pictures :(
21:05:47 <stillunknown> Why doesn't vehicle have both next and prev pointers?
21:07:26 <CIA-1> OpenTTD: rubidium * r10029 /trunk/src/roadveh_cmd.cpp: -Fix (r9923): trams should not/cannot turn on roadworks.
21:10:01 <Wolf01> they should go straight on the bulldozer or simply stop?
21:22:43 <De_Ghosty> they should run down the bulldozer
21:23:17 <Zuu> Nice thing with working with the hotkey-patch: learn all neat hotkeys I did not knew about.. :)
21:25:10 <Rubidium> yup, quite understandable
21:26:24 <Sacro> FUCK THEM AND THEIR LAW
21:26:34 * Sacro dances to the prodigy :D
21:26:39 <dihedral> Sacro: fuck the fucking fuckers
21:26:58 <peter1138> do do do de do do, do do do de do do, do do do do do do do do do do do de do do
21:27:37 <Sacro> breathe the pressure, come play my game i'll test you
21:29:47 <Thomas[NL]> How do you think about using signs as one-way road markers?
21:30:28 <Maedhros> i'd rather use them than the arrows, personally :)
21:31:31 <Eddi|zuHause3> i like the signs
21:31:49 <Thomas[NL]> So it may get into trunk some time :)?
21:32:17 <Eddi|zuHause3> i have no authority to decide that, unfortunately :p
21:33:18 <Maedhros> you can already change the arrows using action 5, so it shouldn't be too difficult to support signs instead
21:35:23 <Eddi|zuHause3> you need some information to decide if it should be drawn on the road (arrow) or in front of the road (sign), for correct sprite sorting
21:35:23 <Maedhros> but now 'tis time for bed
21:35:30 <Maedhros> see you all in a week :)
21:35:56 <Eddi|zuHause3> i don't know if the newgrf spec is prepared for that
21:47:37 <kaan> oh man, Danmark-Sverige footballmatch has gone all wrong
21:49:06 <Eddi|zuHause3> i don't remember this being a football channel...
22:22:41 <Ailure> [22:14] <Eddi|zuHause3> Ailure talks to him/herself a lot. He/She wrote over 5 lines in a row 296 times! <- now that is impressive :p
22:23:11 <Ailure> Heh, I have a tendancy to write severeal lines at once so i'm not really surprised.
22:24:11 <stillunknown> Let's improve that stats.
22:25:12 *** RamboRonny has joined #openttd
22:38:31 <OwenS> Guys, have you never considered using C bitflags instead of the GB/SB/HASBIT macros you have now?
22:41:12 <Eddi|zuHause3> what do you mean?
22:43:35 <OwenS> C has bitflag support in structures
22:43:53 <OwenS> unsigned (char, etc) blah: 2, second: 1, third: 3, etc
22:44:00 <OwenS> Packs them all into a byte
22:44:19 <OwenS> It's little known, but surprisingly well supported
22:44:46 <Rubidium> it's considered, but never actually used (in trunk)
22:44:55 <Rubidium> branches/map uses it
22:45:00 <OwenS> Shame, it makes things so much neater
22:45:27 *** ThePizzaKing has joined #openttd
22:46:09 <Rubidium> is there a guarantee in which order those 'bits' are stored (as per C++ specifications)?
22:46:42 <OwenS> Theyre stored in order AFAIK. I know several small OS projects use them that way though, for processor structures
22:46:51 *** Apocalipsys has left #openttd
22:48:10 <Rubidium> but do they care about binary compatability between big and little endian?
22:49:09 <Rubidium> and moving bits around gets (probably) impossible
22:50:44 <Rubidium> like you swapped bits 0..1 with 2..3 in your blah, second, third example
22:52:31 <OwenS> Dunno about endianness issues
22:52:32 <Rubidium> and using the same bits used in multiple tile types gets probably uglier too ( RT GetRailType(tile) { return IsTileType(tile, street) ? tile.street.rail_type : tile.rail.rail_type; } (also needs tunnels and bridges and stations)
22:52:39 <OwenS> Although that won't matter for bytes
22:53:09 <Rubidium> but not everything is bytesized
22:53:12 <OwenS> Use a union and multiple sets of bit flags? Have a tile.all.rail_type or something similar
22:53:43 <OwenS> It stores them in the system native order. I'm assuming it does already and you swap them?
22:54:12 <Rubidium> that would "waste" four bits in everything that is not a level crossing, rail tunnel/bridge/station
22:55:07 <Rubidium> but then you have problems with stations (they are a different tile type)
22:55:18 <Rubidium> and even per tile type you've got differences
22:55:47 <Rubidium> you've got "normal road", "level crossing" and "depot"
22:56:23 <Rubidium> in the case of "normal road" you need 4 bits for each roadtype, in the case of "level crossing" you only need 1 bit to give the direction of both the rail and all roadtypes
22:59:28 <Rubidium> and for "level crossings" you need to store 4 owners
23:02:09 <Eddi|zuHause3> road, tram, rail, what is the 4th?
23:02:11 <Rubidium> where the space that is used in "normal" roads fort the road owner, is the rail owner in a "level crossing" as it's way more important that the rail owner is always using the same bits than the road owner. This because road vehicles don't care about ownership, whereas train do care
23:02:34 <Rubidium> Eddi|zuHause3: a yet unspecified/unimplemented third roadtype
23:03:23 *** peterbrett has joined #openttd
23:04:33 <stillunknown> Anyone who understands the "coordinate" system of the hash map?
23:04:38 <stillunknown> (vehicle hash map)
23:10:17 *** dihedral has joined #openttd
23:10:31 <Rubidium> I think it's in the "on screen" coordinate system; where it would be shown on the giant screenshot
23:10:53 <_Ben_> Bit late, but well done for getting to r10000. really great stuff. Thanks to all who are working on it
23:11:00 <dihedral> could the way a map is sent from server to client not be used to send grf files?
23:11:28 <dihedral> or is the map data send in some paticular way?
23:11:55 <OwenS> dihedral: I'm not an expert, but AFAIK all it does is send a save file?
23:12:04 <OwenS> (Someone else can answer better)
23:12:11 <TrueBrain> OwenS: you are correct
23:12:20 <TrueBrain> dihedral: grfs can not be send, because of copyright issues
23:12:39 <dihedral> i mean the new grf files... not the copyrighted ones
23:12:47 <TrueBrain> I mean the newgrf files
23:12:51 <TrueBrain> that is why we need a FAQ :p
23:12:52 <Rubidium> those are copyrighted...
23:12:54 <OwenS> TrueBrain: The copyright issue wouldn't land with OpenTTD however, if it sent newgrfs. It would land with the server owner
23:12:58 <TrueBrain> that question comes around every.... 2 weeks now?
23:13:03 <OwenS> They could turn it off if they wanted
23:13:44 <Rubidium> OwenS: the problem is that most NewGRF authors have an explicit license that states that you cannot redistribute the NewGRF
23:13:47 <dihedral> i did not know that those new grf files were copyrighted
23:14:03 <Touqen> dihedral: The person who draws them owns the copyright
23:14:21 <TrueBrain> okay, I made a boo-hoo, copyright was the wrong word, license is the correct one :)
23:14:23 <Rubidium> allowing to do so would violate the license and harm the OTTD community more than the "good" you would have from autodownload
23:14:27 <Touqen> Just like this sentence is copyrighted by me. :D
23:14:52 <Eddi|zuHause3> dihedral: everything is copyrighted
23:15:06 <dihedral> yeah - i also meant licensed
23:15:09 <OwenS> I'd be surprised if anyone was that peed off about it sending automatically
23:15:11 <TrueBrain> Rubidium: I just never get why they used such a license....
23:15:28 <Eddi|zuHause3> the question is how far people will go to let copying occure anyway
23:16:09 <OwenS> Hmm, it seems that i've finished stage one of 3 of my project =)
23:16:22 <Eddi|zuHause3> Touqen: in germany, you cannot even voluntarily give away the copyright, you can just license it
23:16:49 <OwenS> Thats why I like BSD...
23:17:06 <dihedral> Eddi|zuHause3: but one can just not care and there aint an issue then either
23:17:24 <Rubidium> TrueBrain: to quote PikkaBird: "...what I don't want is outdated versions of my grfs spread all over the internet, with people complaining to me about bugs that were fixed years ago, or asking silly questions about features which are explained on my website or wiki. Basically, it's a support-minimisation ploy, as well as quality control (making sure that when people go and download one of my grfs, they're getting the latest version)."
23:17:59 <Eddi|zuHause3> dihedral: it does not matter what you care about, it's what the particular grf author care about
23:18:03 <Touqen> sounds reasonable to me
23:18:18 <TrueBrain> Rubidium: what was the result about the talk of adding a flag that indicates if distribution is allowed?
23:18:30 <dihedral> Eddi|zuHause3: "one" as in the author
23:18:45 <Rubidium> well, that "flag" would be violated within like 10 seconds from it being committed
23:19:05 <TrueBrain> true, but at least we did our best...
23:19:31 <Eddi|zuHause3> dihedral: but it is not for you to decide wether they should care or not
23:19:50 <OwenS> openttd now quits normally as soon as it starts
23:19:56 <dihedral> Eddi|zuHause3: just saying that they have the option to not care :-P
23:20:56 <Eddi|zuHause3> dihedral: but that has nothing to do with the distribution
23:20:58 <dihedral> how come one of these authers is able to say something is free and can be downloaded, just not redistributed
23:21:10 <OwenS> dihedral: License agreements
23:21:11 <dihedral> if something is free, it's free?
23:21:29 <TrueBrain> dihedral: besides Public Domain, nothing is really 'free'
23:21:33 <TrueBrain> just 'free of costs'
23:21:50 <TrueBrain> OpenTTD isn't really 'free' either
23:22:04 <TrueBrain> if you make a modification and distribute the binary, you HAVE TO publish your source too
23:22:13 <Eddi|zuHause3> it is not about free, read the above statement, it is about reading readmes and version control and stuff
23:22:21 <dihedral> TrueBrain: aint anything wrong with that
23:22:47 <TrueBrain> dihedral: and I am not debating that :)
23:22:48 <dihedral> although i though it was 'make available upon request'
23:22:52 <OwenS> Thought about including libcurl and libunzip? =P So you can point the clients to download the GRFs from the author
23:22:56 <Eddi|zuHause3> dihedral: there are enough people who violate the GPL
23:23:02 <TrueBrain> dihedral: it is a tiny bit stronger :)
23:23:25 <OwenS> TrueBrain: Actually, it does say make the source available on request (Legally without fail)
23:23:26 <dihedral> actually i did not want to start a legal discussion here
23:23:34 <Sacro> TrueBrain: there are several openttd related violations of the gpl
23:23:39 <TrueBrain> but okay, a license can disallow redistriution, where they allow the rest to be 'free'
23:23:55 <Eddi|zuHause3> it also says you have to give contact information to send the request to
23:24:00 <OwenS> Hmm.. Why the hell would it try to start up and not continue? =S
23:24:13 <Sacro> TrueBrain: UIQ3 port, GP2X port
23:24:32 <TrueBrain> Sacro: so someone should send them an email
23:24:36 <TrueBrain> we did with the PocketPC port
23:24:46 <dihedral> what about encoding the grf files from the server in the map when sending the data to the client so available for that connection?
23:25:02 <Rubidium> dihedral: encoding in what manner?
23:25:06 <OwenS> It would still legally be illegal..
23:25:25 <dihedral> but having the grf data available in the map file
23:25:34 <TrueBrain> dihedral: by the license, nothing and nobody is allowed to redistribute the grfs, in any form :)
23:25:45 <dihedral> i.e. grf is stored in the map
23:25:50 <TrueBrain> it is not like if you scan a book, you can freely distribute it
23:25:56 <TrueBrain> after all, you made it available in an other format (??)
23:25:58 <Rubidium> dihedral: means embedding the GRF in the map, which means you're redistributing it
23:26:04 *** skidd13 has joined #openttd
23:26:45 <Eddi|zuHause3> just live with the fact that this feature will not be implemented
23:26:50 <OwenS> Guys, this is REALLY strange
23:26:57 <Eddi|zuHause3> it has been discussed many dozen times
23:27:01 <OwenS> OpenTTD is exiting before printing any messages, cleanly
23:27:12 <dihedral> how about sending the client the source url from where it can be downloaded and the client gets it from there automatically?
23:27:15 <TrueBrain> OwenS: even with -dall=9?
23:27:26 <dihedral> redistributing the url in no way can be a legal issue
23:27:40 <OwenS> TrueBrain: OK, not I have a tonne of output
23:27:42 <Eddi|zuHause3> dihedral: that's what grfcrawler is for
23:27:43 <TrueBrain> dihedral: see grfcrawler
23:28:04 <TrueBrain> besides, PikkaBird won't have older versions available, so if you want to join a server which has
23:28:05 <dihedral> sorry for blasting out so many questions :-S
23:28:22 <TrueBrain> dihedral: write them down in a wiki page, so we can just point people to there next time :p
23:28:40 <OwenS> Has the old PBS GRF been removed? :(
23:28:43 <TrueBrain> I am serious in fact
23:28:46 <OwenS> I was using it trmporarily
23:29:01 <dihedral> TrueBrain: i would not believe it was anything else
23:29:52 <Eddi|zuHause3> there was never a PBS grf
23:30:08 <TrueBrain> btw, did you guys know that GPLv2 states this:
23:30:09 <TrueBrain> a) You must cause the modified files to carry prominent notices
23:30:09 <TrueBrain> stating that you changed the files and the date of any change.
23:30:12 <OwenS> Eddi|zuHause3: I mean the one with the PBS signal in
23:30:34 <Eddi|zuHause3> the PBS signal pictures are still there
23:30:43 <Eddi|zuHause3> just they are not used
23:31:32 <Zuu> TrueBrain: Does that include anything that is commited to svn?
23:31:47 <TrueBrain> Zuu: no, as we in fact own our own copyright to the code :)
23:31:50 <Sacro> i would assume that svn can cover that..
23:32:00 <Eddi|zuHause3> OwenS: the last build that had "working" PBS was the miniin
23:32:11 <OwenS> Eddi|zuHause3: I don't want PBS. I just want the graphic
23:32:11 <TrueBrain> and besides, SVN in fact does keep track of exactly that, but rather less explicit :)
23:32:12 <dihedral> and thanks for letting me bother you :-D
23:32:12 <OwenS> Can someone point me to two free signal bits? =)
23:32:21 <TrueBrain> but based on that rule, I think almost all patches and modifications out there violate GPLv2 :)
23:32:50 <Eddi|zuHause3> like i said, the graphics are still there
23:32:59 <Eddi|zuHause3> in nsignalsw.grf, or so
23:34:16 <Eddi|zuHause3> TrueBrain: why? the patch files clearly contain all changes, and usually the files have a date of last change
23:34:48 <TrueBrain> _stating_ that _you_ changed the files and the _date_ of any change
23:34:58 <TrueBrain> patches miss the 'you' part and a 'date' of change :)
23:35:06 <OwenS> Are M2 bits 3 and 7 free?
23:35:25 <Eddi|zuHause3> sure, you can be overly pedantic :)
23:35:35 <Rubidium> see landscape.html and landscape_grid.html
23:36:04 <OwenS> I forgot M2 was a 16-bit number =P
23:36:08 <TrueBrain> Eddi|zuHause3: it is what GPLv2 says :)
23:36:09 <OwenS> But, 3 and 7 are free =D
23:43:59 <OwenS> Aah! It's a dedicated build
23:51:55 <OwenS> Oh good, stage one of programmable signals complete =)
23:53:22 *** Progman has joined #openttd
23:59:07 <OwenS> Nah... Luas more common
23:59:18 <OwenS> And for how much codes involved, easy enough (They will look identical)
23:59:27 <OwenS> And Lua doesn't have the class overhead
23:59:42 <Rubidium> OwenS: look at the NoAI branch
23:59:44 <TrueBrain> you should install squirrel if you want this thing to ever hit trunk :)
continue to next day ⏵
