Config
Log for #openttd on 31st October 2015:
Times are UTC Toggle Colours
00:04:43  *** sim-al2 [~sim-al2@108-221-158-76.lightspeed.mmphtn.sbcglobal.net] has joined #openttd
00:05:39  <fjb> Moin
00:06:07  *** KouDy [~koudy@188.75.190.58] has quit [Ping timeout: 480 seconds]
00:12:10  <Eddi|zuHause> i'd say "Mahlzeit", but it doesn't really apply :p
00:13:42  <fjb> It's always time to eat.
00:14:47  *** KouDy [~koudy@188.75.190.58] has joined #openttd
00:18:35  <Eddi|zuHause> not for gremlins
00:19:08  <Wolf01> mmm, i have cookies right here... no they are on the left
00:23:56  <Eddi|zuHause> https://www.youtube.com/watch?v=7LYYoGpiq9Q <-- this'll be awome
00:26:10  <Wolf01> lol
00:30:06  <Wolf01> 'night
00:30:10  *** Wolf01 [~wolf01@0001288e.user.oftc.net] has quit [Quit: Once again the world is quick to bury me.]
00:31:03  *** KouDy [~koudy@188.75.190.58] has quit [Ping timeout: 480 seconds]
00:34:37  *** KouDy [~koudy@188.75.190.58] has joined #openttd
00:40:44  *** Hiddenfunstuff [~Geth@y32.ip1.anvianet.fi] has quit [Quit:  HydraIRC -> http://www.hydrairc.com <- Would you like to know more?]
00:59:40  *** FLHerne [~flh@dsl-217-155-24-22.zen.co.uk] has quit [Quit: There's a real world out here!]
01:07:51  <Eddi|zuHause> orudge: so, why does the forum always seem to freeze around backup-time?
01:22:59  *** fjb [~frank@000158aa.user.oftc.net] has quit [Ping timeout: 480 seconds]
01:27:27  *** tipsyTina [~tipsy@63.138.45.36] has joined #openttd
01:53:13  *** Biolunar [Biolunar@x4d08ef39.dyn.telefonica.de] has quit [Ping timeout: 480 seconds]
01:57:19  *** Tirili [~chatzilla@2a02:8109:680:910::2] has joined #openttd
02:01:43  *** Snail [~jacopocol@cpe-98-14-130-227.nyc.res.rr.com] has joined #openttd
02:35:29  *** JGR_ [~JGR@host86-135-162-94.range86-135.btcentralplus.com] has joined #openttd
02:37:26  *** JGR [~JGR@host86-135-49-206.range86-135.btcentralplus.com] has quit [Ping timeout: 480 seconds]
02:37:26  *** JGR_ is now known as JGR
02:48:51  *** glx_ [~glx@2a01:e35:2f59:c7c0:1fb:d12a:b94e:ea92] has joined #openttd
02:48:51  *** glx is now known as Guest7208
02:48:51  *** glx_ is now known as glx
02:55:14  *** Guest7208 [~glx@000128ec.user.oftc.net] has quit [Ping timeout: 480 seconds]
03:01:49  *** Tirili [~chatzilla@2a02:8109:680:910::2] has quit [Quit: ChatZilla 0.9.92 [SeaMonkey 2.38/20150928234557]]
03:21:29  *** smoke_fumus [~smoke_fum@188.35.176.90] has quit [Ping timeout: 480 seconds]
03:53:46  *** Supercheese [~Superchee@cpe-76-178-136-186.natnow.res.rr.com] has quit [Read error: Connection reset by peer]
03:54:20  *** Supercheese [~Superchee@cpe-76-178-136-186.natnow.res.rr.com] has joined #openttd
03:56:59  *** glx [~glx@000128ec.user.oftc.net] has quit [Quit: Bye]
04:16:21  *** liq3 [liq3@CPE-120-148-51-74.gdfw1.lon.bigpond.net.au] has joined #openttd
04:39:05  *** tokai|noir [~tokai@00012860.user.oftc.net] has joined #openttd
04:39:08  *** mode/#openttd [+v tokai|noir] by ChanServ
04:42:00  *** GenevaMode [~Adium@90.212.89.0] has quit [Quit: Leaving.]
04:45:22  *** tokai [~tokai@00012860.user.oftc.net] has quit [Ping timeout: 480 seconds]
05:11:04  *** Speedy [~speedy@the.wrong.domain.name] has quit [Remote host closed the connection]
05:15:31  *** Snail [~jacopocol@cpe-98-14-130-227.nyc.res.rr.com] has quit [Quit: Snail]
05:30:00  *** Snail [~jacopocol@cpe-98-14-130-227.nyc.res.rr.com] has joined #openttd
05:56:01  *** Eddi|zuHause [~johekr@p5DC676C3.dip0.t-ipconnect.de] has quit []
05:56:16  *** Eddi|zuHause [~johekr@p5DC670F7.dip0.t-ipconnect.de] has joined #openttd
06:25:48  *** Snail [~jacopocol@cpe-98-14-130-227.nyc.res.rr.com] has quit [Quit: Snail]
07:11:04  *** andythenorth [~Andy@cpc10-aztw26-2-0-cust867.18-1.cable.virginm.net] has joined #openttd
07:14:58  <andythenorth> moin
07:15:34  <tipsyTina> Meow
07:18:42  *** Alberth [~alberth@2001:981:c6c5:1:be5f:f4ff:feac:e11] has joined #openttd
07:18:45  *** mode/#openttd [+o Alberth] by ChanServ
07:19:26  <Alberth> moin
07:21:24  <Flygon> Menta
07:24:11  <andythenorth> lo Alberth
07:57:10  *** roidal [~roland@62-46-143-231.adsl.highway.telekom.at] has joined #openttd
08:08:30  *** Progman [~progman@p57A19B20.dip0.t-ipconnect.de] has joined #openttd
08:16:50  * andythenorth needs a better Heart of Darkness
08:18:24  <andythenorth> or maybe just remove it
08:18:31  *** gelignite [~gelignite@mue-88-130-83-076.dsl.tropolys.de] has joined #openttd
08:21:48  <Alberth> ?
08:24:05  <andythenorth> I think it fails
08:24:21  <andythenorth> it was intended as some kind of ‘colonialism is bad, mmmkay?’ lesson
08:24:28  *** tipsyTina [~tipsy@63.138.45.36] has quit [Ping timeout: 480 seconds]
08:24:33  <andythenorth> but I think it just glorifies resource exploitation
08:25:06  <andythenorth> and repeats tired stereotypes about ‘Africa, the dark continent'
08:25:24  <andythenorth> accepting that most players won’t think about it that far, but still
08:26:04  <Alberth> I think it's an approximation of how it was
08:26:25  <Alberth> players can draw their own conclusions
08:26:55  <andythenorth> I see so much dumb misogyny and racism from gamers on the internet, I don’t want to feed that :P
08:27:05  <Alberth> and the economy makes people aware this history  even existed
08:28:09  <Alberth> gaming has nothing to do with, imho
08:28:26  <Alberth> just like religion has nothing to do with IS
08:29:46  <andythenorth> the thing that tipped my opinion
08:30:21  <andythenorth> I have been researching present-day Congo and neighbouring countries
08:30:26  <Alberth> but euhm, well, you can reverse the thing, make western world export stuff :)
08:31:04  <andythenorth> it’s kind of miserable making an economy that features food scarcity, when that’s still the real situation in these countries (most of the population have severe food scarcity)
08:31:15  <Alberth> quite
08:31:27  <andythenorth> if gaming is separate, maybe gaming should just be fun
08:31:33  <andythenorth> and not try to teach history, badly
08:32:11  <Alberth> well, lots of people mistake openttd for a train simulation game :p
08:33:00  <andythenorth> I think I replace it with something obviously fun, and stop trying to teach people stuff :)
08:33:41  <Alberth> fair enough
08:34:31  <andythenorth> hmm, thanks for discussion, helpful :)
08:34:46  <Alberth> np
08:34:54  <Alberth> and yw :)
08:47:04  * andythenorth will think on
08:47:47  <andythenorth> bbl
08:47:48  *** andythenorth [~Andy@cpc10-aztw26-2-0-cust867.18-1.cable.virginm.net] has quit [Quit: andythenorth]
08:52:27  *** fjb [~frank@000158aa.user.oftc.net] has joined #openttd
09:00:22  <fjb> Moin.
09:02:03  <Alberth> o/
09:04:57  *** Hiddenfunstuff [~Geth@y32.ip1.anvianet.fi] has joined #openttd
09:10:27  <planetmaker> moin
09:26:54  <Alberth> hi hi
09:40:04  <fjb> Moin planetmaker
09:41:39  *** Biolunar [Biolunar@x4d086a34.dyn.telefonica.de] has joined #openttd
09:51:05  *** George [~George@185.43.94.91] has joined #openttd
09:51:20  <George> Hello
09:52:05  <George> Need help with nmlc ver 5780 (but the error is the same on 5655 too)
09:52:27  <George> nmlc ERROR: nmlc: An internal error has occurred:
09:52:27  <George> nmlc-version: unknown
09:52:27  <George> Error:    (IndexError) "pop from empty list".
09:52:27  <George> Command:  ['nmlc', '--grf=xussr.grf', '-c', '--nfo=xussr.nfo', '--nml=xussr_optimized.nml', '-M', '--MF=xussr_dep.txt', 'xussr.nml']
09:52:27  <George> Location: File "nml\free_number_list.py", line 74, in pop
09:52:29  <George> [ERR]
09:53:41  <Alberth> add a -s option, and run again, please use a pastebin to paste output
09:54:22  <Alberth> hmm, free number list eh?   probably you are using too many items of something
09:56:16  <George> But how can I find what are they?
09:58:35  <Alberth> I hope the full stack trace gives that information
09:59:17  <Alberth> another option is to compile one of the newest versions that does compile, and have a look at how many items are used of each type
10:02:58  <George> https://paste.openttdcoop.org/phjott3qp
10:03:47  *** andythenorth [~Andy@cpc10-aztw26-2-0-cust867.18-1.cable.virginm.net] has joined #openttd
10:04:58  <Alberth> looks like an actionD expression that is too complicated
10:05:21  <Alberth> or you have too many, not sure
10:05:33  <George> How can I test it?
10:06:40  <Alberth> replace some expression by something simple, and see if it compiles
10:06:48  <Alberth> "1" would work, I think?
10:07:27  <Alberth> obviously, it won't give a useful result if you try to run the result, but that's not the intention
10:08:18  <George> The previous commit provides: https://paste.openttdcoop.org/p9kqqueyq
10:08:31  *** andythenorth [~Andy@cpc10-aztw26-2-0-cust867.18-1.cable.virginm.net] has quit []
10:09:11  <Alberth> line 7 of that paste seems like a problem
10:09:36  <Alberth> and you even get a line number
10:10:16  <Alberth> 126/127 means you have 1 Action2 register left
10:10:27  *** Pensacola [~quassel@88.159.51.52] has joined #openttd
10:11:42  <George> Yes, but that limit was achieved on the other wagon
10:12:26  <George> The wagon I code now should not have that amount of variants
10:12:51  <Alberth> Action2 registers are a shared resource, what you use at one place can't be used at another place
10:13:35  <Alberth> maybe you use a result further down or so?  that would mean the compiler has to keep the value in a register
10:13:46  <George> AFAIR that is not correct. 127 is a limit for them at once, but ones from a different cain does not count
10:14:01  <Alberth> I am just guessing here, I don't know enough of it
10:14:21  <George> I
10:14:50  <George> 've changed to "1" as you suggested, but it does not compile anyway
10:15:24  <Alberth> I have no idea how nml allocates registers, sorry
10:16:17  <Alberth> but as it is a shared resource, many expressions are linked together, I don't think you can pinpoint the one expression
10:16:43  <Alberth> as there isn't one that is wrong
10:17:00  <Alberth> it's like a bucket of water that gets filled drop by drop
10:17:21  <Alberth> at some point it overflows, but you cannot easily blame a single drop
10:17:57  <George> I've cut out one more part and no it compiles
10:18:07  <Alberth> (where drops vary in size, different expressions use different numbers of registers)
10:18:13  <George> Have to cut off smaller parts
10:19:00  <Alberth> try to simplify expressions, perhaps compute some part once, and use it several times
10:19:16  <Alberth> at least I hope that's possible in nml
10:24:59  <George> This code causes it
10:25:00  <George> https://paste.openttdcoop.org/p2fuiynhw
10:25:27  <George> I'm confused, because it is only a list of named values
10:25:46  <Alberth> it doesn't
10:26:00  <Alberth> it's the combination of all code
10:26:13  <Alberth> lots of drops, and at some point the bucket overflows
10:26:26  <Alberth> take out one random drop, and it doesn't overflow
10:26:37  <Alberth> but that holds for pretty much any drop you pick
10:27:47  <Alberth> people have any chance of using all the engines and wagons you added in a game?
10:29:51  <George> :D that would be a hard task ... the set is ready for about only 10% :D
10:33:27  <planetmaker> George, it might be that the last paste you provided means that each definition 'eats' one actionD register... Maybe it can be circumvented by using #defines - but then means the nml needs some kind of preprocessing which makes it harder to compile under windows
10:33:40  <planetmaker> But I'm not sure whether I'm right with my assumption either :)
10:36:36  *** Wormnest [~Wormnest@s5596abd2.adsl.online.nl] has joined #openttd
10:44:01  <Eddi|zuHause> i ran into this years ago, but somebody didn't think anything needs to be done
10:45:34  <planetmaker> question is: what can be done about it?
10:47:39  <George> looks like "#define" may help me. I'm looking it at the moment
10:49:27  <Eddi|zuHause> i think my suggestion was along the lines of "[const|bit|byte|word|dword] <id> = <expr>;"
10:50:02  *** frosch123 [~frosch@x5f746a39.dyn.telefonica.de] has joined #openttd
10:50:09  <Eddi|zuHause> const will not reserve an actionD parameter, bit, etc. will be collected and stuffed
10:50:12  <George> seems like
10:50:12  <George> #define LV_RED_ARROW                 30
10:50:20  <planetmaker> maybe we need it afterall, Eddi|zuHause :)
10:50:21  <George> is what I need
10:50:35  <Eddi|zuHause> George: that would probably work
10:51:19  <planetmaker> that will work, if my assumption is correct, yes
10:51:34  <planetmaker> quak
10:51:42  <Alberth> quek
10:52:09  <planetmaker> quick quek quak :P
10:52:41  <frosch123> hola hi hoi
10:53:00  <Eddi|zuHause> quick quack quuck?
10:53:37  <frosch123> i like how companies send you an email these days when your package is about to get delivered
10:54:03  <frosch123> except when they send the mail at 2am that the package arrives at 9am :)
10:54:33  <frosch123> probably would have worked for eddi though
10:54:55  <Eddi|zuHause> 9am? what an ungodly hour to deliver anything :p
10:55:13  <Alberth> at worst, you are back to the pre-email situation :)
10:58:40  *** HerzogDeXtEr [~flex@i59F6A2B2.versanet.de] has joined #openttd
11:15:09  *** Pensacola [~quassel@88.159.51.52] has quit [Read error: Connection reset by peer]
11:18:51  *** Pensacola [~quassel@88.159.51.52] has joined #openttd
11:34:49  *** GenevaMode [~Adium@90.212.89.0] has joined #openttd
11:40:10  *** GenevaMode [~Adium@90.212.89.0] has quit [Quit: Leaving.]
11:56:39  *** GenevaMode [~Adium@90.212.89.0] has joined #openttd
12:22:25  *** George is now known as Guest7229
12:22:25  *** Guest7229 [~George@185.43.94.91] has quit [Read error: Connection reset by peer]
12:22:25  *** George [~George@185.43.94.91] has joined #openttd
12:30:53  *** GenevaMode [~Adium@90.212.89.0] has quit [Quit: Leaving.]
12:31:21  *** Supercheese [~Superchee@cpe-76-178-136-186.natnow.res.rr.com] has quit [Read error: Connection reset by peer]
12:31:54  *** Supercheese [~Superchee@cpe-76-178-136-186.natnow.res.rr.com] has joined #openttd
12:36:18  *** Wolf01 [~wolf01@0001288e.user.oftc.net] has joined #openttd
12:36:29  *** GenevaMode [~Adium@90.212.89.0] has joined #openttd
12:37:58  <Wolf01> o/
12:38:05  <Alberth> hi hi
12:38:55  *** GenevaMode [~Adium@90.212.89.0] has quit []
12:44:02  *** smoke_fumus [~smoke_fum@188.35.176.90] has joined #openttd
12:48:06  *** Pulec [~pulec@2a01:4f8:110:1463:127::2] has quit [Remote host closed the connection]
12:48:44  <Wolf01> http://img-9gag-fun.9cache.com/photo/aepqeAm_460s.jpg really, i could reduce the rendering variables names length and they should remain understandable
12:49:47  *** Pulec [~pulec@2a01:4f8:110:1463:127::2] has joined #openttd
13:07:26  <George> Thank's all, With #define I've solved my problem!
13:10:43  <Alberth> given that you are now at 10%, do the used numbers of various registers make sense?
13:11:40  <George> What do you mean?
13:12:43  <Alberth> eg "nmlc info: Concurrent spritegroups: 173/256 ("xussr.nml", line 4513)"   <-- would that grow to 1730 ?
13:13:08  <frosch123> no, everything with "concurrent" does not grow if you add independent stuff
13:14:43  *** Biolunar [Biolunar@x4d086a34.dyn.telefonica.de] has quit [Quit: leaving]
13:14:47  <Alberth> nmlc info: D0xx strings: 812/1024   <-- so that one would grow to 8120 ?
13:15:06  <Alberth> https://paste.openttdcoop.org/p9kqqueyq  btw
13:15:07  <frosch123> yes, if the strings cannot be reused
13:16:17  <George> https://paste.openttdcoop.org/poiys8rbs
13:34:00  *** gelignite [~gelignite@mue-88-130-83-076.dsl.tropolys.de] has quit [Quit: http://bit.ly/1kso8Ta]
13:38:32  <Eddi|zuHause> i assume there's room for optimization there, like using dynamic composition via textstack instead of static composition through nmlc
13:40:03  <frosch123> considering the amount of strings in eints, it does not look like nml is composing anything
13:48:01  *** GenevaMode [~Adium@cpc25-stok15-2-0-cust157.1-4.cable.virginm.net] has joined #openttd
13:49:08  <Eddi|zuHause> well, if every vehicle gets its own cb23 string, you're probably running into trouble
13:49:23  <Eddi|zuHause> the vehicle names themselves don't count
13:58:15  <Eddi|zuHause> frosch123: i mean things like this: "additional_text: return string(STR_PURCHASE_HINT_ENGINE,  string(STR_PURCHASE_HINT_ENGINE_TYPE_PASSENGER_STEAM),  string(STR_PURCHASE_HINT_SECTIONS_1),  string(STR_PURCHASE_HINT_FROMTO, 1906, 1910),  string(STR_PURCHASE_HINT_FACTORY_PUTILOVSKY));"
13:58:40  <Eddi|zuHause> if you turn this composition into textstack, then it should significantly reduce the amount of strings
14:00:33  <Eddi|zuHause> compare with https://dev.openttdcoop.org/projects/cets/repository/entry/scripts/write_engine.py#L190
14:00:46  <frosch123> Eddi|zuHause: 1447 strings in eints, minus 323 train names, minus 97 cargo strings: remain 1027 strings
14:01:04  <frosch123> i don't know how many a14 strings though
14:01:22  <frosch123> but 1027 strings is quite beyond the reported 810 d0xx strings
14:03:06  *** Progman [~progman@p57A19B20.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
14:05:28  <Eddi|zuHause> frosch123: the way i see it, every vehicle has a short and long name, which can be switched somewhere
14:05:45  <Eddi|zuHause> so that's 2 non-strings per vehicle
14:06:19  <Eddi|zuHause> there are probably also a bunch of unused things in there
14:07:14  <Eddi|zuHause> anyway, still, the suggested replacement would reduce string usage by another #vehicles
14:09:02  <Eddi|zuHause> in general, it's difficult to predict though, as nmlc does all sorts of messy optimizations
14:09:35  <Eddi|zuHause> it may very well be that this will increase the number of strings
14:10:12  <Eddi|zuHause> (but it then should be safer against growing with the number of vehicles included)
14:10:27  <Alberth> you'd need an optimization target for better control
14:19:02  <Eddi|zuHause> i guess hardly anybody uses the text stack, because nmlc doesn't abstract it away
14:19:19  <Eddi|zuHause> so it's difficult do wrap your mind around it
14:19:48  <Eddi|zuHause> also, it's not actually a stack, you can't "push" to it independently from what is already on there
14:20:05  <Eddi|zuHause> but that's rather an underlying NFO problem
14:20:37  <Alberth> nml isn't finished by a long shot
14:20:49  <Alberth> not sure, it will ever move much though
14:21:32  <Eddi|zuHause> these tools usually grow with the people that use them
14:22:39  <Alberth> very few users know about compiler construction :(
14:22:56  <Eddi|zuHause> that is probably true
14:23:04  <Eddi|zuHause> and the ones that do, are totally lazy. :p
14:25:46  <Alberth> it's probably a misconception that people with commit rights care for all areas of the program
14:28:21  *** roidal [~roland@62-46-143-231.adsl.highway.telekom.at] has quit [Ping timeout: 480 seconds]
14:28:45  <frosch123> oh, eddi knows about compiler construction?
14:28:58  <Eddi|zuHause> yes.
14:29:09  <Eddi|zuHause> i wrote my diploma thesis in this area
14:31:30  *** roidal [~roland@62-46-143-231.adsl.highway.telekom.at] has joined #openttd
14:54:30  *** andythenorth [~Andy@cpc10-aztw26-2-0-cust867.18-1.cable.virginm.net] has joined #openttd
14:55:00  <Alberth> hi hi
15:00:26  <LadyHawk> http://ladyhawk.flawlesscorruption.net/screens/depotwaypoint.png << i was kinda happy when i did that and figured out it flowed right (yes apparently im using the wrong pbs signals but hey)
15:00:36  <LadyHawk> couldn't handle the traffic anymore so now it's http://ladyhawk.flawlesscorruption.net/screens/depotwaypoint2.png
15:01:05  <andythenorth> o/
15:01:28  <Wolf01> ho
15:08:02  <Alberth> you probably don't need the crosses after the way point, with high enough traffic they will cause more harm than good
15:08:43  <Alberth> also, a signal at every tile is useless, trains can never get that close after each other, unless you don't have any junction
15:09:08  <Alberth> (and without junction, you can't have two trains on the same track :p )
15:09:40  <Alberth> but double depots are very useful, it's a good setup
15:10:30  <LadyHawk> the only reason i stuck signals so close together is at junctions where they filter so when they slow down, they can stay closer together and have more chance to remain at some kind of speed even if it's slow
15:10:42  <LadyHawk> just looked nicer to copy it everywhere lol
15:11:00  <LadyHawk> hmm
15:11:06  <LadyHawk> about those crosses
15:12:15  <LadyHawk> they're there to allow the traffic from the center to cross over 1 track when required, the outher 2 tracks are allowed to filter out to the outter depot lines as required as a result
15:12:46  <LadyHawk> because they reserve a track before the waypoint... they wont cross unless they are required to (both depots or the track to them in front of their nose = blocked)
15:13:05  <LadyHawk> that's the function of the waypoint
15:13:09  <LadyHawk> stop them choosing at all
15:13:41  <LadyHawk> having 2 depots on a single track to force serve the trains can't handle the traffic
15:14:10  <LadyHawk> so i cant really see how i can work without the crosses at all, without giving each line 2 dedicated depot tracks.. which would make the whole thing bigger
15:15:28  <LadyHawk> every once in a while the trains are required to cross, and with this setup the rest can move around them so nothing stops.. just seems to flow better to me
15:15:49  <LadyHawk> unless you can give me more of an idea as to what it would look like? i'll be happy to give other things a try =)
15:17:20  <Alberth> let me see if I can find an example
15:21:33  <LadyHawk> and apparently signals that close together does allow more traffic on the line http://ladyhawk.flawlesscorruption.net/screens/closeup.png
15:21:53  <LadyHawk> front train was slowed down during filtering, the train behind sped on at full speed and caught up haha
15:22:54  <Alberth> http://devs.openttd.org/~alberth/4_to_2.png  4 entry -> 2x2 exits, so each of the 4 entry lines can go in either direction, which then gets merged
15:23:25  <Alberth> I do have better ones, but much less findable, I am afraid
15:24:02  <Alberth> your more traffic is only temporary, the second train will block on the first junction that comes along
15:25:55  <LadyHawk> it's a pbs entry station where that line has an immidiate choice to divert to a 2nd & 3rd line in... designed so that they can enter that close together since the track is unreserved right behind the train
15:27:25  <LadyHawk> i'm finding that setup of yours incredibly hard to understand for some reason... i've studied similar examples on the wiki over and over but i just dont understand those
15:29:00  <LadyHawk> http://ladyhawk.flawlesscorruption.net/screens/stationentry.png
15:29:50  <LadyHawk> i keep having the nagging feeling that desperately needs changing but i don't know how and keep thinking to myself that it works fine the way it is.. because it does work perfectly
15:30:34  <LadyHawk> the queue is due to the station being too small... trying to filter in as many trains as it can deal with, takes a few round trips
15:32:14  <Alberth> I would have a waiting space in front of each platform, so the next train is waiting already
15:32:51  <LadyHawk> hmm
15:33:38  <Alberth> there are about 4 platforms not used there
15:34:13  <LadyHawk> how likely is the situation that for example in a 4 track station (2 tracks occupied 2 free), you'll have 1 or 2 trains in the waiting space just sitting around?
15:35:08  <Alberth> depends on the amount of traffic that you feed it
15:35:26  <Eddi|zuHause> you should split the traffic earlier, and have no crossovers before the platform
15:35:48  <Eddi|zuHause> so you go 3->6->12 or something, without interference
15:36:02  <Eddi|zuHause> or simply 3->9
15:36:17  <Eddi|zuHause> with each line having access to 3 platforms separated
15:37:07  <Eddi|zuHause> one platform a train is loading, one platform is empty and ready to take a train in, and one platform a train is leaving
15:37:32  <LadyHawk> the issue i get with early traffic splits is a train's preference for the inner track (the whole layout is basically a circle, inner track = shorter = always preferred)... your idea would be brilliant though Eddi|zuHause if there's a way to stop this inner track preference
15:37:40  <Alberth> http://devs.openttd.org/~alberth/3k9_goods.png   8 platforms
15:37:55  <LadyHawk> it ends up with the inner station tracks having queues, hence i have the split as late as possible
15:38:01  <Eddi|zuHause> having more choices only means you get more chances to block the other train's route
15:38:18  <Eddi|zuHause> make a waypoint.
15:39:53  <LadyHawk> use the waypoints how? put them in places so at the train's time to choose they litterally want to go straight ahead?
15:40:53  <LadyHawk> (sorry for all the questioning, just trying to understand in detail ><)
15:41:45  <Eddi|zuHause> if the distance between order points is shorter, you can make sure better that the differences in length even out
15:42:12  <Eddi|zuHause> to solve this problem of "the inner track is always a tiny bit shorter, and thus preferred"
15:42:39  <LadyHawk> aha gotcha
15:42:43  <Eddi|zuHause> this emphasises the penalties for occupied tracks, red signals and stuff
15:43:13  <LadyHawk> yeah
15:43:53  <Eddi|zuHause> so you make a waypoint short before, and short after there is a choice
15:44:54  <LadyHawk> ironicly though my setup atm has 3 entry tracks and 9 station tracks, and the station cant handle the traffic.. i try to give the trains the minimum amount of options but because the station is too small, the free tracks are erratic which causes the need to cross more than i'd like to allow them
15:46:08  <LadyHawk> im going to have a little think about your early split idea Eddi|zuHause and give it a try, i may set 3 tracks for 1 line & a 4th track shared between 2 or something
15:47:23  <LadyHawk> and stick a waypoint down somewhere before it too
15:50:03  <Alberth> https://devs.openttd.org/~alberth/junctions.png    some intermediate steps in my selection junction
15:50:08  <LadyHawk> i kinda wish there was a way to take train preference out of the picture but i can imagine that's quite impossible to do
15:51:23  <Alberth> it shouldn't matter, the path finder will re-balance, although with your signal setup it may not work very well
15:51:45  <Alberth> what you want is free flow no matter where trains go
15:51:53  *** Snail [~jacopocol@cpe-98-14-130-227.nyc.res.rr.com] has joined #openttd
15:52:03  <LadyHawk> lol yeah i guess i do
15:52:41  <LadyHawk> from that last link it looks like it's the queueing  that's making it so confusing to me
15:52:43  <Alberth> you have to always give trains a choice to switch to a more quiet track
15:53:49  <Alberth> it's just like your station. Now the train stops before the junction, if all platforms are blocked, and then traverses the junction when a platform becomes free
15:54:05  <Alberth> the queue removes that delay to a minimum
15:54:33  <Alberth> at the cost of choice, so you only do this at places where the train in front will be quick to leave
15:54:46  *** Pensacola [~quassel@88.159.51.52] has quit [Remote host closed the connection]
15:54:47  <Alberth> or if it doesn't matter
15:55:14  <Alberth> ie one train can be blocked for a longer time, but other trains will flow around it
15:55:27  <Alberth> prefering other platforms/directions instead
15:56:51  <Alberth> or with my load station, the thing must be loaded anyway, if there is no cargo for the train before, it doesn't matter that it waits, as there would not be cargo for it either
15:58:07  <LadyHawk> that makes perfect sense
15:58:37  <Alberth> it took quite a few years to realize that :p
15:58:59  <Alberth> and no doubt the openttdcoop people have even more advanced ideas about it :p
15:59:44  <LadyHawk> in my case though the station can't be emptied, there's 54 coal mines stuffing the place
15:59:57  <LadyHawk> it's more a case of getting as many trains loading up at any one time as possible
16:00:19  <LadyHawk> i have 9 tracks, if there's 6 filled and 3 empty with 3 trains waiting in a que somewhere.. that's bad news for my rating
16:00:30  <Alberth> same with my 3k9 goods loading, you can't afford to have an empty platform
16:01:19  <Alberth> :o  54 mines :)
16:01:28  <LadyHawk> so dumming it down to simple terms, if i add the queueing, i may need to add more traffic
16:01:50  <LadyHawk> so even when there's trains in a que sitting around, there'll be more to flow around
16:01:55  <Alberth> you optimize on station loading
16:02:15  <Alberth> so at all times, there must be a train loading, leaving or arriving
16:02:36  <Alberth> where the latter two should be as small as possible
16:03:01  <LadyHawk> yeah
16:03:03  <LadyHawk> hmm
16:03:10  <Alberth> minimal solution is a signal right before and after the station platform
16:03:27  <Alberth> and a train waiting for the platform when the previous train is leaving
16:03:49  *** George [~George@185.43.94.91] has quit []
16:03:51  <Alberth> if that takes 'number of platforms' additional trains, fine
16:05:21  <LadyHawk> so in theory it's better since the station is as short as it could possibly be.. even if i have to add extra traffic to fill the queues and the station.. it shouldn't create more traffic on the rest of the line
16:05:46  <LadyHawk> this is interesting
16:05:52  <Alberth> traffic is limited by loading or unloading speed, eventually
16:05:56  <LadyHawk> yeah
16:06:20  <Alberth> (multiplied by number of loading/unloading platforms)
16:06:45  <Alberth> note this is not optimal from a train point of view, it may have to wait here and there
16:07:31  <Alberth> but optimal train means you have always a platform available when a train arrives
16:07:51  <LadyHawk> you want it optimal from a station's point of view
16:08:05  <Alberth> in this case yes
16:08:41  <Alberth> but the goal completely changes how you look at it, and how you build
16:09:46  <Alberth> and optimal for station doesn't mean just throw as many trains at it as you can
16:09:54  <LadyHawk> lol
16:10:07  <Alberth> since that mostly just creates long queues and nothing else
16:11:01  <Alberth> it's a lot of fun experimenting build styles :)
16:11:07  <LadyHawk> exactly
16:11:51  <LadyHawk> im gonna go build the queueing thing
16:12:04  <Alberth> good luck :)
16:12:14  <LadyHawk> on a different note i'm sorry but i stand by my original statement about the improved loader
16:12:31  <LadyHawk> on a busy loading station like this.. you want it off and use the old one
16:12:57  <LadyHawk> there's ALWAYS a full train load on the station at the very least.. you can't empty it because of how it works
16:13:06  *** GenevaMode [~Adium@cpc25-stok15-2-0-cust157.1-4.cable.virginm.net] has quit [Quit: Leaving.]
16:13:59  <Alberth> euhm, you mean improved loading setting, I guess?
16:14:03  <LadyHawk> yeah
16:14:17  <Alberth> that does load multiple trains if cargo is available
16:14:32  <andythenorth> mmm
16:14:37  <andythenorth> FIRS economies
16:14:52  <LadyHawk> yeah but the train that just arrives, before it's anywhere near finished loading, has a full train's worth reserved on the station.. nobody will touch it
16:14:58  <Alberth> hi andy, nice idea, something between basic and full
16:15:00  <LadyHawk> which means it sits there
16:15:20  <Alberth> ah, ok
16:15:32  <LadyHawk> hence you cant empty the station, there's always at least a full train load there
16:15:43  <LadyHawk> since supply is faster than loading speed
16:16:06  <LadyHawk> i dont like that
16:16:11  <andythenorth> I can’t decide on themes
16:16:24  <Alberth> in that situation, it might not matter much
16:16:27  <andythenorth> whether to use real geographical locations (like the basic economies)
16:16:36  <andythenorth> or made up to suit gameplay (like Full FIRS)
16:17:03  <LadyHawk> the old loader didn't have that issue and with high supply it doesn't matter if all trains are greedy and take what they can see
16:17:38  <andythenorth> the thing I want to avoid is adding economies that are bland, because they’re just like Full FIRS, but with bits missing
16:17:54  <Alberth> more detailed topic?  many different forms of oily things?
16:18:02  <andythenorth> considered that
16:18:10  <Alberth> longer industry chains?
16:18:16  <andythenorth> if I had Dan here, we’d probably be planning a detailed steel economy
16:18:33  <andythenorth> furnace -> rolling mill -> pressing plant etc :P
16:18:46  <Alberth> steel plate 1mm thick, 2mm thick, 5mm thick  :)
16:18:52  <andythenorth> that sort of thing :)
16:19:02  <andythenorth> daft, and quite train-spotterish I guess :)
16:19:28  <andythenorth> hmm maybe I keep Heart of Darkness structure, but just rename it and adjust it
16:19:41  <andythenorth> the point was to have an economy focussed on export, not processing
16:19:53  <Alberth> export of high tech waste from the west to other areas :p
16:20:01  <andythenorth> shipbreaking :P
16:20:27  <andythenorth> I think I want to avoid history lessons or lessons in politics and ethics :)
16:20:28  <Alberth> also electronics etc get dumped into africa
16:20:33  * andythenorth sticking to the trains
16:20:48  <Alberth> /me unstickies andy
16:21:03  <Alberth> it's dangerous glueing yourself to trains :)
16:21:59  <Alberth> so export of asian goods to europe is out too, I guess?
16:22:10  <andythenorth> nah
16:22:21  <andythenorth> just no need to tie it down to one specific country
16:22:34  <andythenorth> https://www.youtube.com/watch?v=VZvm5H4F-aA <- glued to a train
16:24:19  <andythenorth> Wolf01: it’s a style http://www.brickshelf.com/cgi-bin/gallery.cgi?i=6330452
16:24:20  <Alberth> crazy guy
16:24:30  <Wolf01> :o
16:24:32  <andythenorth> quite intense colour use ^
16:24:54  <Alberth> looks good!
16:25:33  <Alberth> ever looked at the trains and trams posted by skidlow? also very nice colours
16:29:38  <LadyHawk> thanks very much for the queing thing at the station... i built it.. on the loading station only, unload i left the way it was.. the que is at the unload station, there's a very clear difference in the amount of traffic and flow speed of the station
16:30:42  <LadyHawk> much better now
16:30:43  <Alberth> http://ladyhawk.flawlesscorruption.net/screens/stationentry.png    reducing the number of reachable tracks from each entry will help here
16:31:42  <Alberth> you don't want one track to block entry of another track, better have the former track wait a while
16:34:20  *** Wormnest [~Wormnest@s5596abd2.adsl.online.nl] has quit [Quit: Leaving]
16:35:26  *** glx [~glx@000128ec.user.oftc.net] has joined #openttd
16:35:29  *** mode/#openttd [+v glx] by ChanServ
16:41:55  <LadyHawk> hmm
16:42:15  <Alberth> sharing a few tracks to handle overflow at one incoming track is often useful if you have bursts of trains coming in
16:42:52  <LadyHawk> i may do that..
16:43:25  <LadyHawk> not because the traffic is blocking eachother off because the trains arent crossing at all if they have the option.. and if they need to they only pick the shortest cross
16:43:40  <LadyHawk> but because 1 train at the pbs block before the queue just had some really strange behaviour
16:44:10  <LadyHawk> it had a free track to cross and enter a queue but instead decided to stop and sit there
16:44:21  <LadyHawk> for long enough for 9 trains to stop behind it
16:44:29  <Alberth> ah yes, pbs tends to do that
16:44:44  <LadyHawk> eventually it decided to go ahead and enter the queue
16:44:48  <Alberth> it's not good at seeing where trains before it go
16:45:26  <Alberth> I switched back to the entry/exit signals due to that, at heavy traffic
16:46:03  <LadyHawk> i wonder if this is one of the reasons why i didn't build queues at stations before.. to avoid the odd behaviour
16:46:11  <planetmaker> 64k is enough for everybody^W spacecraft. :D
16:47:40  <Alberth> ?
16:49:06  *** liq3 [liq3@CPE-120-148-51-74.gdfw1.lon.bigpond.net.au] has quit []
16:49:28  <planetmaker> http://www.popularmechanics.com/space/a17991/voyager-1-voyager-2-retiring-engineer/
16:49:52  <planetmaker> thus if you look for a job opportunity it might be for you :)
16:50:12  <planetmaker> onboard system ressources might be a *bit* limited, though :)
16:51:43  <andythenorth> so the challenge of an export/import economy
16:51:48  <andythenorth> I only have 3 kinds of ports
16:51:55  <andythenorth> so 9 accepted cargos / 6 produced
16:52:08  <andythenorth> hard to make it intermediate or advanced, rather than basic
16:53:24  <LadyHawk> may have an idea to stop that odd behaviour, i'll have to build it later.. gonna move the pbs split so at the time they filter train doesn't know about the station yet.. increasing queue size but keeping it to 1 train in max
16:54:01  <LadyHawk> atm they seem hell bent on stopping if eventually they can reserve a track to the empty station track, rather than carrying on on a free track
16:54:54  <LadyHawk> or i suppose a waypoint should fix that, concern the trains with that order during the filter
16:55:17  <Alberth> planetmaker: ha, might be fun at times :)
16:56:03  <Alberth> never done any fortran though
16:59:51  *** sla_ro|master [~sla.ro@95.76.27.245] has joined #openttd
17:02:38  <Alberth> andy, perhaps the ports should only produce some cargoes if you supply some other cargoes?
17:04:03  <Alberth> perhaps in some rotation between the ports, so you can always deliver everything, but perhaps less efficient or so get things in return
17:07:31  *** tokai [~tokai@00012860.user.oftc.net] has joined #openttd
17:07:34  *** mode/#openttd [+v tokai] by ChanServ
17:10:42  <Alberth> probably for a different economy, if you can look at the coordinates of building, perhaps build only at one side of the map?
17:10:55  <Alberth> right now all industries are mostly everywhere
17:12:59  *** GenevaMode [~Adium@cpc25-stok15-2-0-cust157.1-4.cable.virginm.net] has joined #openttd
17:13:57  *** tokai|noir [~tokai@00012860.user.oftc.net] has quit [Ping timeout: 480 seconds]
17:14:20  *** tokai [~tokai@00012860.user.oftc.net] has quit [Quit: c('~' )o]
17:19:44  *** Progman [~progman@p57A19B20.dip0.t-ipconnect.de] has joined #openttd
17:21:07  <LadyHawk> odd pbs behaviour seems gone.. no longer stopping and sitting, but going for the clear track.. they're ordered to a waypoint in the queue now instead of the station when they're filtering
17:22:40  <LadyHawk> eliminating choice or preference other than straight ahead
17:31:08  <Alberth> :)
17:38:56  <LadyHawk> it looked like the pathfinder's lookahead was seeing the station and adding all the penalties and preferences for free tracks on top of the pbs its own penalties... making things go a lil wonky
17:41:36  <LadyHawk> http://ladyhawk.flawlesscorruption.net/screens/pbsmadesmart.png
17:43:48  <Alberth> your trains aren't very fast?  otherwise you could remove the 90 degrees corners after the station
17:44:12  <LadyHawk> yeah im working on that ><
17:44:32  <LadyHawk> you werent supposed to see that :P
17:45:04  <Alberth> :)
17:45:29  <LadyHawk> i love showing things here.. people critisize and give me ideas to make it better :D
17:45:39  <DorpsGek> Commit by rubidium :: r27429 /branches/1.5/src/lang (16 files in 2 dirs) (2015-10-31 18:45:32 +0100 )
17:45:40  <DorpsGek> [1.5] -Backport: language updates from trunk
17:46:02  <Alberth> it was there in the previous picture too I think?
17:46:14  <LadyHawk> yeah
17:46:22  <LadyHawk> maybe =(
17:46:36  <LadyHawk> it's just so much effort since i'll have to move my mergers over
17:46:36  <Alberth> it was, but then I missed it as a possible problem :)
17:59:59  *** Wormnest [~Wormnest@s5596abd2.adsl.online.nl] has joined #openttd
18:28:05  *** tipsyTina [~tipsy@63.138.45.36] has joined #openttd
18:33:13  *** Snail [~jacopocol@cpe-98-14-130-227.nyc.res.rr.com] has quit [Quit: Snail]
18:35:58  *** Snail [~jacopocol@cpe-98-14-130-227.nyc.res.rr.com] has joined #openttd
18:42:35  <Alberth> good night
18:47:01  <LadyHawk> goodn ight
18:48:50  *** Alberth [~alberth@2001:981:c6c5:1:be5f:f4ff:feac:e11] has left #openttd []
19:11:54  *** liq3 [liq3@CPE-120-148-51-74.gdfw1.lon.bigpond.net.au] has joined #openttd
19:29:35  *** Certes [~Certes@cpc1-sgyl15-0-0-cust201.sgyl.cable.virginm.net] has joined #openttd
19:31:48  <Certes> hi, it's my first visit, great to see so many people in here
19:35:05  <Certes> I'm using trunk 1.5.1, lots of great new features.  I moved from an old ChillPP and I really miss the departure board and the minutes timetable (trains at 08:15, 08:45...).  Please can anyone recommend a patch pack (Windows binary preferred but I can set up a compilation if I have to) that includes these?
19:43:14  <frosch123> https://www.tt-forums.net/viewforum.php?f=33 <- check the forums
19:43:23  <Certes> thanks frosch, looking...
19:43:31  <frosch123> the top two topics are about two patchpacks
19:43:34  <frosch123> just check what they have
19:44:12  <Certes> great, there should be enough links there to find what I'm looking for.  thanks for your help.
19:46:59  *** gelignite [~gelignite@mue-88-130-83-076.dsl.tropolys.de] has joined #openttd
19:54:32  *** lx16 [~oftc-webi@80-246-167-46.slunecny.net] has joined #openttd
19:55:24  <Certes> I now have Tim's pack up and running with exactly what I was hoping for.  Thanks again to frosch for such a quick and helpful reply, and to everyone who's contributed to such a wonderful game.
19:55:48  <lx16> hi, server list in openttd 1.5.2 isn't working? ..my game fetch the ip's but not game info on them
19:58:16  *** GenevaMode [~Adium@cpc25-stok15-2-0-cust157.1-4.cable.virginm.net] has quit [Quit: Leaving.]
19:59:26  *** Certes [~Certes@cpc1-sgyl15-0-0-cust201.sgyl.cable.virginm.net] has quit []
20:00:58  *** tipsyTina [~tipsy@63.138.45.36] has quit [Ping timeout: 480 seconds]
20:05:54  *** lukasz [~quassel@boj19.neoplus.adsl.tpnet.pl] has joined #openttd
20:06:25  *** lukasz [~quassel@boj19.neoplus.adsl.tpnet.pl] has quit [Remote host closed the connection]
20:08:38  *** Snail [~jacopocol@cpe-98-14-130-227.nyc.res.rr.com] has quit [Quit: Snail]
20:14:14  *** lx16 [~oftc-webi@80-246-167-46.slunecny.net] has quit [Quit: Page closed]
21:05:23  *** Zuu [~Zuu@h-114-162.a98.priv.bahnhof.se] has joined #openttd
21:10:47  *** GenevaMode [~Adium@90.212.89.0] has joined #openttd
21:40:34  *** andythenorth [~Andy@cpc10-aztw26-2-0-cust867.18-1.cable.virginm.net] has quit [Quit: andythenorth]
21:40:42  *** roidal_ [~roland@62-46-143-151.adsl.highway.telekom.at] has joined #openttd
21:46:54  *** tokai [~tokai@00012860.user.oftc.net] has joined #openttd
21:46:57  *** mode/#openttd [+v tokai] by ChanServ
21:47:34  *** roidal [~roland@62-46-143-231.adsl.highway.telekom.at] has quit [Ping timeout: 480 seconds]
21:47:35  *** frosch123 [~frosch@x5f746a39.dyn.telefonica.de] has quit [Quit: be yourself, except: if you have the opportunity to be a unicorn, then be a unicorn]
21:47:47  *** roidal_ [~roland@62-46-143-151.adsl.highway.telekom.at] has quit [Quit: WeeChat 1.2]
21:51:25  *** sla_ro|master [~sla.ro@95.76.27.245] has quit []
22:09:46  *** Progman [~progman@p57A19B20.dip0.t-ipconnect.de] has quit [Remote host closed the connection]
22:20:12  *** DDR [~David@S0106602ad0773a2e.vc.shawcable.net] has joined #openttd
22:33:36  *** Snail [~jacopocol@cpe-98-14-130-227.nyc.res.rr.com] has joined #openttd
23:15:51  *** Wormnest [~Wormnest@s5596abd2.adsl.online.nl] has quit [Quit: Leaving]
23:17:45  *** Biolunar [Biolunar@x4d086a34.dyn.telefonica.de] has joined #openttd

Powered by YARRSTE version: svn-trunk