IRC logs for #openttd on OFTC at 2008-07-27
⏴ go to previous day
00:01:57 <Eddi|zuHause> hm... i was never really interested in agatha christie stuff
00:07:08 <ln> "ten little niggers" aka "and then there were none" is good at least.
00:09:06 <ln> good night for now, though
00:10:29 <Eddi|zuHause> "zehn kleine jägermeister" :p
00:33:45 *** Eddi|zuHause2 has joined #openttd
01:14:22 *** Singaporekid has joined #openttd
01:26:49 *** Sir-Bob has joined #openttd
01:33:24 *** DaleStan has joined #openttd
03:07:50 *** elmex_ is now known as elmex
03:08:45 *** ProfFrink has joined #openttd
03:13:07 *** ProfFrink is now known as Prof_Frink
05:33:03 *** thingwath has joined #openttd
07:08:27 *** Artamir has joined #openttd
07:09:08 <Artamir> someone can give me a link of a compilation newer than r13501 please?
07:20:26 <DaleStan> Artamir: google:BuildOTTD
07:20:37 <Wolf01> there's a svn guru online? I've problems with the pre-revprop-change hook, the manual say to make it executable (renaming it to .bat on windows) but now when I try to change a commit log the log browser freeze and doesn't change anything
07:35:34 <Rubidium> Wolf01: is it still called hooks/pre-revprop-change?
07:36:40 <Wolf01> uhm, wait, maybe I *should* not make a copy of the .tmpl file and rename the copy, but rename the .tmpl directly
07:39:17 <Rubidium> it should be .bat or .exe according to some website
07:40:04 <Wolf01> now the log browser don't popup a warning message, it just freezes
07:50:01 *** plakkertjes has joined #openttd
07:57:05 *** Frostregen has joined #openttd
08:03:15 *** frosch123 has joined #openttd
08:04:08 *** Doorslammer has joined #openttd
08:52:21 *** stillunknown has joined #openttd
09:03:26 <Celestar> peter1138: I've just got to get a single line to work and then I'm set
09:03:31 <Celestar> (minus tons of debugging :P)
09:04:34 <Celestar> peter1138: but the route network is built up for existing savegames
09:12:31 <CIA-5> OpenTTD: frosch * r13845 /trunk/src/ (station_cmd.cpp water_cmd.cpp): -Fix (r13838): When adding a WaterClass to oilrig station tiles, also make use of it.
09:16:35 <CIA-5> OpenTTD: frosch * r13846 /trunk/src/water_cmd.cpp: -Fix (r13838): Do not draw water borders inside of industries.
09:16:49 <Celestar> why is ##c++ full of idiots :S
09:20:02 <Artamir> someone knows what is "queue aircraft at airports"????
09:22:34 <frosch123> maybe aircrafts landing in the order they enter the airport holding pattern
09:22:52 <peter1138> Yeah, I think it's a FIFO queue.
09:22:55 <Artamir> "track ageing speed"?
09:23:34 *** Dred_furst has joined #openttd
09:23:56 <frosch123> build some track and observe what time will do to it
09:23:57 <hylje> time needed for track degrading one level
09:24:33 <Artamir> i don't understand this
09:24:54 <Artamir> tracks spend real time
09:25:08 <Artamir> without this option tracks are instantaneous?
09:25:20 <hylje> building is instant no matter what
09:31:54 <Celestar> what kind of patches are those? are they in trunk?
09:33:26 <frosch123> IIRC the first one is called "aircraft queueing' and the second 'grass growth on unused track'
09:36:01 *** Vikthor has joined #openttd
09:36:04 <frosch123> yawn - why do you always encounter hundreds of other bugs when you just want to analyse a single one :x
09:36:36 <peter1138> Did you decide if that if articulated part check is unnecessary yet?
09:37:13 <frosch123> I guess it is unnecessary, but there are a lot other pieces in the code that perform a similiar test :p
09:40:51 <frosch123> I did not explore all, but one of the more evil ones was: When an articulated vehicle enters a electrified/not-electrified track, is it powered when the first part is energized, or is every part handled individually. (i.e. artic vehicles with mixed tracktion types)
09:44:32 <frosch123> but for the most current one: In CmdBuildRailWagon(): "(GetEngineProperty(engine, 0x17, rvi->base_cost) * _price.build_railwagon) >> 8", in EstimateTrainCost() used in CmdBuildRailVehicle "GetEngineProperty(engine, 0x17, rvi->base_cost) * (_price.build_railvehicle >> 3) >> 5". Which one is correct?
09:45:08 *** Progman has joined #openttd
09:46:49 <peter1138> The >> 3) >> 5 versus >> 8 difference is minor.
09:49:00 <peter1138> Unless I've missed the point.
09:49:10 <frosch123> ah, I did not notice the 'railwagon' vs 'railvehicle' difference :)
09:49:44 <frosch123> So the >> 3) >> 5 part is still weird, but is not incosistent
09:51:21 <Eddi|zuHause2> those should really use / instead of >>
09:51:48 <jordi> blathijs: well, it's standard practice
09:52:28 <jordi> blathijs: to some people it helps knowing there's a new version of their package
09:52:38 <jordi> that's not too useful for openttd of course
09:52:47 <jordi> as you always know in advance what's going on
09:53:02 <ln> Eddi|zuHause2: do not underestimate the 0.000000000000000000000000000000000000000000000001 per cent speed improvement that bit-shifting gives.
09:53:22 <jordi> but for example, it helps me retrieving the tarball with just one command, I won't have t
09:53:30 <jordi> to look for the tarball at sf and so on
09:53:32 <Eddi|zuHause2> ln: it doesn't... the compiler generates the same code
09:53:43 <Eddi|zuHause2> or at least it should...
09:55:06 <Eddi|zuHause2> besides, last time i checked, the code style said to use / for arithmetic calculations, not >>
09:55:29 <Rubidium> but is it arithmetic?
09:56:03 <Eddi|zuHause2> it is, obviously...
09:56:08 <Rubidium> it's just chopping of the insignificant bits
09:57:59 <Rubidium> bits used for inflation
10:00:48 <Eddi|zuHause2> if it really "masks" bits, it should then be the logical consequence to use GB()
10:23:59 <Celestar> peter1138: I'm basically there
10:26:34 *** Recimin has joined #openttd
10:31:12 <blathijs> jordi: But, I think it requires the download location to support file listings or something?
10:31:55 <blathijs> jordi: Or does uscan just try a few increments? (ie, when current version is 0.6.1, the next could be 0.6.2, 0.7.0 or 1.0.0 or whatever)
10:33:09 <frosch123> Gekz: there is still the sunos/solaris option, i.e. drop the first number somewhen
10:35:50 <peter1138> The AI has built a Diesel Shunter (27mph) along with 4 3 * freightliner container rakes
10:36:03 * peter1138 ponders teaching it about articulated wagons...
10:40:37 <Celestar> peter1138: I've got a little summin for ya
10:41:57 <Celestar> or the abstract representation of it
10:43:07 *** Belugas_Gone has joined #openttd
10:43:07 *** ChanServ sets mode: +o Belugas_Gone
10:44:25 <Celestar> Abstract Base Class, sorry :P
10:44:58 <blathijs> Hmm, I was just about to ask that :-)
10:45:54 <blathijs> That would be redownload, since it didn't get served as text/plain and firefox can't display text/x-diff or whatever it is
10:47:15 <peter1138> Order *o = v->orders;
10:47:41 * peter1138 grabs a checkout to patch it to.
10:48:44 <frosch123> Is there anything that can be done wrt. conditional orders?
10:49:44 <blathijs> Celestar: I wouldn't indent RouteNetwork_t in its own typedef. I didn't see it at first :-)
10:50:46 <peter1138> Hmm, incomplete route for adding?
10:51:30 <Celestar> peter1138: that's quite normal
10:51:46 <Celestar> peter1138: for example if you only have one station in the order
10:51:52 <Celestar> it doesn't add loops and stuff
10:52:10 <Celestar> I need to distinguish between debug, warning and errors
10:53:00 *** ChanServ sets mode: +o Bjarni
10:53:35 <peter1138> I assume it knows about unload/noload orders?
10:53:47 <Celestar> peter1138: yes it does
10:53:56 <Celestar> peter1138: about "go via" as well
10:54:03 <Celestar> peter1138: it does not know about intermediate stops
10:54:10 <Celestar> peter1138: because they're never guarenteed
10:54:18 <Celestar> peter1138: it also doesn't know (yet) about cargo types
10:54:26 <Celestar> peter1138: we could leave that to the user
10:54:35 <Celestar> (most will use it for pax/mail only anyway)
10:55:13 <peter1138> And valuables could do with it.
10:55:36 <Celestar> peter1138: heh .. apparently there's a graphviz interface in boost
10:56:08 <peter1138> So can we make a map? ;)
10:56:24 <Celestar> I'm planning to add some console debug stuff
10:56:25 <blathijs> Celestar: lHow does the reusing of vertex indices work? It seems that it will only work when the AddStationToVertexList method is called with incrementing StationID's?
10:56:34 <peter1138> I think that other patch actually draws it on the minimap.
10:56:43 <Celestar> peter1138: thought about that too
10:56:54 <Celestar> blathijs: the number of vertices is <= the number of stations
10:57:02 <Celestar> the nth vertex is associated to the nth station
10:57:16 <Celestar> otherwise I'd need to store a map and stuff
10:57:30 <blathijs> Celestar: But shouldn't that if be a while then?
10:57:42 <DorpsGek> peter1138: bjarni was last seen in #openttd 3 weeks, 3 days, 14 hours, 49 minutes, and 49 seconds ago: <Bjarni> I didn't have any
10:57:44 <ln> Bjarni: do you have any now?
10:57:58 <Celestar> I've forgotten to remove some old code
10:58:12 <Bjarni> at least I have a working internet connection now :)
10:58:27 <peter1138> Different or same provider?
10:58:37 <Bjarni> actually the problem was local
10:58:42 <Celestar> redownload/refresh peter1138 and blathijs
10:58:51 <Bjarni> the ISP didn't screw up (this time)
11:00:32 <Celestar> peter1138: blathijs: store all questions till tomorrow, k?
11:01:28 <blathijs> If the first station I add has ID 10, then afterwards the graph wil contain a single vertex that should really be the tenth index
11:01:48 <blathijs> Celestar: K, just had this one written already when my laptop power ran out :-)
11:02:27 *** KillaloT has joined #openttd
11:02:47 * peter1138 ponders making it show on the mini-map
11:03:31 *** GoneWacko has joined #openttd
11:03:56 <peter1138> Bjarni, we've got a load of bugs lined up for you ;)
11:05:16 <blathijs> peter1138: Would be a lot of data to cram in there, I guess
11:07:01 <peter1138> ^ that's what the other paxdest patch does
11:08:05 <jordi> blathijs: if the files are in the same dir, it's an easy wildcard
11:08:14 <jordi> if it's in different dirs, the regex gets more complicated
11:08:26 <jordi> blathijs: there are tons of examples for sourceforce hosted packages
11:14:03 *** Zealotus has joined #openttd
11:21:10 <peter1138> So stationid === vertex. hmm.
11:23:36 *** ChanServ sets mode: +v tokai
11:28:12 <peter1138> Hmm, I have no idea how to iterate this data :o
11:35:42 <Bjarni> now 0.6.2-RC2 failed to compile due to something that was last changed in rev 4105
11:37:30 <SmatZ> Bjarni: what compiler / OS ?
11:38:27 <Bjarni> and I tried t build the OSX binary (in case you wondered)
11:40:50 <Eddi|zuHause2> with an iterator, i presume ;)
12:07:23 <Bjarni> I would say 100 is pretty high
12:07:43 <Bjarni> do we really need to spend resources on more than say 12?
12:07:55 <Bjarni> (which by itself is pretty high)
12:08:37 <frosch123> IIRC it was 16 somewhen, but somebody complained about it
12:09:21 <SmatZ> you could use CountArticulatedParts , but it would slow things down...
12:09:44 <Bjarni> how can 16 be too little?
12:09:52 <peter1138> SmatZ, that still needs a limit...
12:10:06 <peter1138> It was 10 originally, which was too short.
12:10:40 <peter1138> I guess 65535 makes the loop a little long... :)
12:11:55 <Bjarni> in real life our trains could live with a limit of 4. If that makes the train too short then they connect two sets ;)
12:12:32 <Bjarni> I really wonder when ONE set needs more than 16 units
12:12:48 <frosch123> Bjarni: road vehicles can also be articulated
12:13:08 * Bjarni tries to imagine a sane road vehicle of more than 16 units
12:13:19 <frosch123> ask the australians
12:13:24 <SmatZ> Bjarni: for some "well-known" train consist
12:13:58 <Bjarni> <frosch123> ask the australians <-- the longest I have seen was the truck and 6 wagons, making it 7 units
12:14:21 <Bjarni> since it transported iron ore it was the heaviest road vehicle in the world to go on public roads
12:22:32 <peter1138> Well, UKRS has long articulated maglev...
12:23:48 <Gekz> we have really long road trains
12:23:55 <Gekz> there was one that had many wagons
12:23:58 <Gekz> I can't remember how long
12:24:02 <Gekz> but I'm sure it was more than 20
12:35:48 <CIA-5> OpenTTD: frosch * r13847 /trunk/src/articulated_vehicles.cpp: -Codechange: Add MAX_ARTICULATED_PARTS and set it to 100.
13:07:53 *** dih is now known as Guest1599
13:08:15 *** Guest1599 is now known as dih
13:11:15 <CIA-5> OpenTTD: truebrain * r13848 /branches/noai/ (5 files in 2 dirs): [NoAI] -Add: AIVehicle.GetCurrentValue() (requested by Michiel)
13:19:03 *** grumbel has joined #openttd
13:29:29 *** ChanServ sets mode: +v tokai
13:42:42 *** Noetloj has joined #openttd
13:43:05 <Noetloj> Hey, I've managed to crash OpenTTD. I tried using the original generation for land, not terragenesis.
13:43:13 <Noetloj> I got out of sprite errors :(
13:43:44 <Noetloj> (it all works if I use terragenisis - but I don't want that)
13:43:54 <SmatZ> Noetloj: I tried it 3 times now, no crash :-/
13:44:06 <SmatZ> is it reproducible for you?
13:44:08 <Noetloj> I'll screenshot the settings I use to crash it.
13:44:16 <Noetloj> and it's done it twice in a row.
13:44:27 <Rubidium> Noetloj: what version are you using?
13:44:43 <Rubidium> and are you using 32bpp sprites?
13:45:18 <Noetloj> two ticks, let me update it.
13:46:43 <Noetloj> Same bug, but now it gave me an error report.
13:47:37 <Rubidium> are you using the dos or windows grfs?
13:48:52 <Noetloj> Those are the settings I used for that crash (Out of sprite memory)
13:49:13 <Noetloj> and I've just tested the exact same on Terragenesis, it worked perfectly :)
13:49:21 <Noetloj> (as in it worked, not it worked and also crashed)
13:49:57 <peter1138> Does the progress meter move at all?
13:50:10 <orudge> this is 0.6.1 as well though
13:50:19 <orudge> progress meter doesn't move at all
13:50:22 <orudge> (using that same random seed)
13:50:23 <Noetloj> peter1138: stuck at 0/2
13:50:33 <orudge> this is using what was the latest trunk with YAPP applied
13:51:40 <Noetloj> Is the bit of screen I get before it crashed.
13:51:47 <Noetloj> (on all 4 crash times)
13:51:59 <Noetloj> So no, no meter movement at all :p
13:52:07 <peter1138> Mmm, before it's terraformed.
13:52:22 * orudge is just compiling the latest trunk to see if it still happens there, and shall try to reproduce it with the debugger
13:52:47 <Noetloj> I hope I didn't break your game
13:52:52 <Noetloj> oh wait, orudge has it too.
13:54:44 <peter1138> It'll probably be those 'special' sprites...
13:54:51 <peter1138> The ones that are not used for Terragenesis.
13:55:59 <orudge> with no newgrfs and latest trunk, it seems to be fine
13:56:14 <Noetloj> want me to wipe my grf list and try?
13:56:28 <orudge> as I had the same problem
13:56:39 * Noetloj sits back and lets the professionals work it.
13:58:13 <glx> yeah I can reproduce with 0.6.2-RC2
13:58:24 <Noetloj> me and orudge aren't insane.
13:58:45 *** [1]plakkertjes has joined #openttd
13:59:07 * Noetloj switches back to terragenises and goes off to play.
13:59:57 <orudge> yes, problem seems to be occurring in latest svn. See
14:00:39 <Noetloj> Tis interesting watching you professionals debug, just fyi.
14:01:13 <glx> at least I can get more info now than when using the pdb :)
14:05:07 *** [1]plakkertjes is now known as plakkertjes
14:11:38 <Bjarni> <Noetloj> me and orudge aren't insane. <-- I didn't see any proof of that :P
14:11:51 <Bjarni> in fact I didn't see any proof either way
14:12:09 * orudge is perfectly sane... on full-moon Tuesdays, when the wind is blowing in an easterly direction
14:16:37 <CIA-5> OpenTTD: bjarni * r13849 /trunk/ (config.lib src/unix.cpp): -Fix: [OSX] 10.5 appears to be more picky when setting up compiling of universal binaries
14:22:56 <hylje> go on and make us an awesome town building AI
14:23:12 <hylje> currently, as far as I'm aware, it throws dice to decide what to do
14:23:25 <hylje> without any planning whatsoever
14:24:13 <peter1138> Should be possible with a GRF :o
14:24:32 <Noetloj> if you teach me the language of OpenTTD, I'll do it myself :p
14:24:38 <Noetloj> I need something to do for a few days.
14:24:44 * hylje drops a C++ manual on Noetloj
14:24:54 <Noetloj> You'll need to be more specific than that.
14:25:21 <peter1138> The GRF Specs wiki would be more useful :D
14:25:44 <Noetloj> It can probably not run around naked.
14:26:11 <hylje> nor does it come with a pony :(
14:27:39 <orudge> I like horses, best of all the animals.
14:28:00 <peter1138> Hmm, I think the only restriction on stadiums at the moment is just only one in a town.
14:30:12 <hylje> change the default max to like 30 and see how long it takes to people notice something's amiss
14:40:17 <Noetloj> peter1138: can't you make it restrict them to a town/city of like, 2000 pop? (cause ya love meh ;))
14:40:31 <Noetloj> but seriously, at the moment you get stadiums in the most stupidfest of places :p
14:42:20 <Noetloj> Gah, I can't build a tower for water.
14:42:24 <Noetloj> What does OpenTTD define as a "town
14:42:50 <glx> it's the only way to detect if we are in a town
14:46:29 <Noetloj> What does OpenTTD define as a house..
14:46:49 <orudge> a building in a town bascically, Noetloj
14:47:00 <Noetloj> Someone living in a caravan, or a favela house would class that as a house.
14:47:05 <Noetloj> OpenTTD obviously does not.
14:48:13 <Noetloj> wow, my OpenTTd screenshot folder is 200MB D:
14:49:10 <Noetloj> That is now a town with houses? :(
14:49:38 <SmatZ> Noetloj: build it over a house (probably)
14:49:44 <orudge> try other locations nearby
14:50:15 <Noetloj> yay, that worked. Thanks SmatZ :D
14:51:33 <Noetloj> I've never played with sub-trop before in length :p
14:59:10 <rortom> where can i find the code for the station gui
14:59:21 <rortom> means the gui that comes when i click on a station
14:59:26 <rortom> not the station list :\
14:59:40 <frosch123> did you try station_gui.cpp :p
14:59:43 <glx> should be in a _gui.cpp file
15:00:43 <frosch123> struct StationViewWindow : public Window
15:03:04 <rortom> why is tha rating and accepts on two different pages?
15:03:15 <rortom> i would put it all into one page
15:03:28 <rortom> "page" = viewable area
15:04:20 <rortom> the question is if that is wished to continue like that
15:06:01 <frosch123> ask the portable players with displays like 5x5 pixels :s
15:17:31 <CIA-5> OpenTTD: frosch * r13850 /trunk/src/ (4 files): -Fix [FS#2146]: Fix resp. implement refitting of articulated vehicles during autoreplace.
15:18:33 *** pm is now known as planetmaker
15:18:33 <peter1138> +static bool EnginesGotCargoInCommon(EngineID engine_a, EngineID engine_b, VehicleType type)
15:18:54 <peter1138> He'll never learn 'have' :D
15:19:17 <peter1138> Nice job frosch123 :D
15:19:23 * peter1138 assumes it works, of course.
15:20:03 <frosch123> I only read the 'Engines' and 'CargoInCommon'. 'Got' did not add anything important :p
15:20:17 <frosch123> Well it works in the cases I tested.
15:20:38 <frosch123> That does not mean there aren't still any corner cases, as I encountered quite a lot :p
15:21:50 * peter1138 wonders what src/ai/default/default.cpp:285-290 are for...
15:22:02 <peter1138> Seems to count the length but not do anything with it.
15:25:54 <frosch123> peter1138: that code is untouched since ottd 0.1
15:26:56 <Noetloj> How long ago is that?
15:27:30 <frosch123> ai.c (modified mar 9 2003)
15:28:06 <Noetloj> wow, I thought OpenTTD was only a few years old.
15:29:15 <ln> there weren't computers 2005 years ago.
15:29:22 <Yorick> Noetloj: ludde started in 2003
15:29:43 <Yorick> you found a bit of original unmodified code :)
15:41:58 <Yorick> what if it goes to contain 37 cargoes?
15:42:40 <Yorick> this would be twice as big...
15:42:51 <Yorick> does that still fit on the screen?
15:43:04 <Yorick> assuming 300x400 screen
15:47:28 <SpComb> hmm... looking at the client code for the receive-map progress bar, it looks like the server just splits the data up into many small packets, and then the client just updates them GUI whenever it gets a chance during pauses between getting these packets?
15:51:50 <hylje> peter1138: probably something absolutely critical to the system
16:04:41 <SpComb> I guess that's the only option that I have for implementing a progress bar for NewGRF downloads
16:05:02 <SpComb> although I'll probably write the server in Python, and not try and integrate it into the OpenTTD codebase yet
16:06:16 * SpComb isn't particularly fond of the OpenTTD network code
16:06:30 * Yorick is implementing it in python
16:07:42 <Yorick> to implementing all packets and know what they mean
16:08:58 <Eddi|zuHause2> <rortom> mod to station gui <- the window should also show the "provides" list [not only when building a station]
16:09:18 <SpComb> I was looking towards creating something that only handled NewGRF stuff, with a slant towards high performance
16:09:26 <SpComb> Yorick: what kind of code do you have?
16:12:22 <SpComb> ugh, threads in irc_lib.py :(
16:12:40 <Yorick> hmm, I don't think it's needed anymore
16:15:43 <Yorick> doesn't he implement irc in a 'proper' way?
16:15:55 <SpComb> well, as a Twisted protocol handler
16:17:49 <Yorick> we recently stepped to a third-party irc-lib, that's what the threads are doing
16:19:01 <hylje> it may be python but it's not written as python. i suppose i'd want to give some patches..
16:21:01 *** DJNekkid has joined #openttd
16:22:12 <Yorick> hylje: if you'd want to give some patches, go to openttd-python :)
16:24:10 <peter1138> There are reasons why it's not done.
16:24:30 <Eddi|zuHause2> YAPF does this by checking signals ahead, roads do not have signals
16:25:07 <peter1138> Havoc with caches, and other vehicles may have moved by the time the vehicle gets there.
16:25:21 <SpComb> hmm... tricking the network code into connecting to the NewGRF server instead of the OpenTTD server will take some work
16:25:34 * SpComb is learning to hate global variables
16:25:46 <peter1138> Learn to embrace them. They are your friend.
16:26:05 <Eddi|zuHause2> keep your friends close, keep your enemies closer
16:27:16 <hylje> maybe i dont dare to change names around despite style
16:27:25 <rortom> i modded the gui a bit further:
16:27:47 <Roujin> peter: i understand the problem with vehicles that have moved by the time the vehicle gets there,, but what would be the problem with caches
16:29:26 <rortom> request for comments? :|
16:30:20 <Eddi|zuHause2> statistics would be cool
16:30:34 <Eddi|zuHause2> not only amount of cargo, but maybe also amount of trains
16:30:56 <Eddi|zuHause2> and still no providing list :(
16:34:01 <peter1138> Roujin: Vehicles move all the time, so caches would be invalid all the time. Or somethign like that.
16:35:08 <rortom> adding stats seems to be very easy in fact...
16:35:54 <Eddi|zuHause2> well, like i said in the thread, it could add dynamic values for the first 10 tiles, like it does with the first 10 signals
16:37:16 <Eddi|zuHause2> rail_look_ahead_max_signals = 10
16:37:39 <Eddi|zuHause2> the functionality is there, just has to be adapted to road
16:38:24 <Roujin> `hmm.. have to look into yapf deeper then..
16:39:09 <Roujin> don't have a clue how yapf does caching yet..
16:39:46 <Roujin> now that's encouraging :D
16:40:38 *** LilDood has joined #openttd
16:42:16 <hylje> if not len(rawcommand) > 0: ..
16:43:29 <Eddi|zuHause2> you don't have to look deep to find the lookahead penalties
16:52:26 <nicfer> city simulation vs transport company simulation, who will win?
16:52:48 <SpComb> does someone else want to write my OpenTTD network code for me? :/
16:53:23 <hylje> i have more than enough pastime with crunching though the code that is allegedly python
16:53:32 <Bjarni> "I have this great idea and I will start coding it now (one hour later) Will somebody code this for me?"
16:53:41 <Bjarni> like we haven't seen this one before xD
16:53:43 <SpComb> it looks like having the NewGRF download window establish a TCP connection to some third-party server and then doing stuff with it is going to be nasty
16:53:43 <peter1138> Pah, just send it in the same way the map's done
16:54:07 <peter1138> Oh, yes, the map comes from the server...
16:54:13 <SpComb> peter1138: there's so many weird global variables involved that the various functions modify that it's really difficult to figure out what I need to modify
16:54:22 <Bjarni> SpComb: you want to transfer newGRF files on the fly?
16:54:53 <nicfer> best suggestion ever: flying trains
16:55:04 <SpComb> currently it uses libcurl and blocking HTTP requests, now I want to implement it using OpenTTD's network stuff is it doesn't freeze the UI etc
16:55:59 <Bjarni> didn't we agree not to download GRF files inside the game itself?
16:56:16 <SpComb> I guess you haven't been following the discussions on that matter here last week
16:56:52 <Bjarni> and I didn't plan to be offline last week :(
16:57:11 <Roujin> Eddi and peter: could it be that Yapf doesn't really cache values for the road pathfinder? Because I got it running now and it doesn't seem like it's using cached values somewhere instead of the current ones...
16:57:23 <SpComb> well, I think we managed to come up with some rough ideas as to an acceptable implementation
16:57:33 <SpComb> so now I'm just trying to work on a prototype
16:57:40 <Bjarni> is it written online somewhere?
16:58:10 <Bjarni> the result of the discussion
16:58:26 <SpComb> well, we only discussed it here on IRC, so there's the logs of that
16:58:41 <SpComb> I was planning on making a topic in the forums once I have a working prototype to present
17:06:15 <Bjarni> it's as hard as I imagined :s
17:08:49 <SpComb> but in essence, the idea is to have a central db of GRF metadata - GRFID, md5sum - and then have the OpenTTD client check the metadata that it gets from the server against that db, and then download those GRFs that are found
17:09:30 <SpComb> the current implementation does a HTTP POST with <grfid>:<md5sum> tuples as arguments, and gets back a list of <grfid>:<md5sum>:<http_url> tuples
17:10:18 <SpComb> then it dowloads the <http_url> into a "cache/" dir, and loads GRFs for use with network servers from there
17:10:36 <Bjarni> but how is that different from what we discarded? :)
17:11:12 <SpComb> having a central db of metadata, and having the offical client check against that gives NewGRF authors more control over their grfs than what they currently have
17:11:56 <SpComb> and in essence the implementation just combines GRFCrawler and the OpenTTDCoop grfpack together, plus some convenience functionality
17:16:03 <plakkertjes> can i ask something? :)
17:16:20 <Bjarni> but since you already did we will have to punish you
17:16:41 <Bjarni> now let that be a lesson to you
17:16:43 <plakkertjes> i'd still ilke to ask itttt
17:16:51 <Bjarni> you should ask the question instead of asking about asking
17:17:52 <plakkertjes> is the payment you get for transporting a certain type of cargo, affected by how far you go by the train? or just the distance between the two stations
17:18:37 <plakkertjes> would've thought so
17:18:40 <Bjarni> you will not gain a profit from making a 200 tile railroad between two stations that are 5 tiles apart
17:19:14 <glx> you will lose money indeed as the travel time will be too big
17:19:30 <SpComb> what about building a station that's 100 tiles from the industry, and just extended to cover the industry as well?
17:19:50 <plakkertjes> its about the stations right
17:20:03 <Bjarni> that would need a strong computer if you allow such a station spread
17:20:20 <glx> the point used as reference is the station sign
17:20:52 <plakkertjes> why would that require a strong computer
17:21:25 <Bjarni> a high station spread needs more CPU time for the calculations regarding the station
17:21:48 <plakkertjes> 2.4 ghz quad core? o.O
17:22:00 <Bjarni> the game is single threaded
17:22:21 <plakkertjes> is that hard to code?
17:22:29 <glx> well it can use a second core for saving and generating land
17:22:44 <glx> multithread means full rewrite
17:22:53 <Bjarni> I tried to add a thread but it turned out that the semaphor handling used more CPU power than it would give
17:23:00 <SpComb> (which would have other positive aspects, though, as well)
17:23:10 <Bjarni> ideal would be a 20% speed boost but worst case it would slow down 10%
17:23:14 <plakkertjes> but why not make it so that on a quad core, certain things are handled by certain cores
17:23:18 <SpComb> bad code -> bad locking needed to keep it working
17:23:21 <Bjarni> and normally it would slow down the game
17:23:38 <Bjarni> so that threading patch was never finished
17:24:05 <plakkertjes> It never slows down on my computer though
17:24:05 <SpComb> well, the code would need to use less shared state before you could effectively parralelize it
17:24:07 <glx> plakkertjes: each client must do exactly the same things for the same initial conditions
17:24:31 <plakkertjes> wowwww, this goes to show how much I dont know about coding programs o.O
17:25:13 <plakkertjes> god Im already fine with knowing PHP
17:25:17 <nicfer> what's faster, a 3.2ghz single-core processor or a 1.6 dual-core one?
17:25:43 <plakkertjes> and also how much cache it has
17:25:45 <hylje> multiprocessor stuff tends to have overhead when compared to single
17:25:49 <Bjarni> the single core CPU would be faster for OTTD
17:25:59 <plakkertjes> :( so my 2.4 ghz is crapz?
17:26:19 <glx> but for multithread apps multicore is better
17:26:35 <Bjarni> if you buy a new game in a computer store then odds are that it uses threading and will use all cores
17:26:46 <plakkertjes> yup, assasins' creed uses it
17:27:17 <nicfer> only accelerates 3d games
17:27:30 <Bjarni> I say we might be special by only using one core and no graphical hardware
17:27:43 <plakkertjes> isnt openttd just pure processing
17:27:52 <plakkertjes> (meaning no graphical)
17:28:12 <plakkertjes> for 3d thats not 2 good
17:28:15 <plakkertjes> but for 2d its fine
17:28:43 <plakkertjes> what language is openttd written in? is it hard to learn? o.O
17:28:43 <glx> well there's some 3D stuff in openttd :)
17:28:49 <Bjarni> btw there is one use for more than one core. When the game saves it saves in the memory. Another thread then compresses the savegame and writes it to the disk while the game continues
17:29:01 <Bjarni> meaning you will not really feel a slowdown when saving/autosaving
17:29:10 <Bjarni> like you risk on a single core system
17:29:28 <plakkertjes> whats 3d in openttd o.O
17:29:52 <plakkertjes> it gives you the sense of height
17:29:57 <Bjarni> it's not 3D in terms of a GPU
17:30:17 <plakkertjes> we'd be able to zoom in and angle the camera
17:30:28 <plakkertjes> thank god tahts not the case
17:30:32 <Bjarni> but openGL can use the GPU for 2D
17:30:33 <plakkertjes> rollercoaster tycoon 2 = yay
17:30:42 <plakkertjes> rollercoaster tycoon 3 (3d) = :(
17:31:13 <glx> yes 3D is not always good (look for TTD for PS1)
17:31:17 * nicfer wonders why no-one started a openrct proyect
17:31:35 <Bjarni> that's due to the EULA in rct
17:31:46 <plakkertjes> i even have openttd on my ipod
17:32:05 <plakkertjes> i loved that game
17:32:25 <plakkertjes> and doesnt rct use the same sounds as openttd
17:32:35 <plakkertjes> I dunno how to make programs :(
17:33:01 <plakkertjes> that would be 5 years of hard work, and you'd have like one rollercoaster
17:33:23 <Bjarni> <plakkertjes> I dunno how to make programs :( <-- well... when you are done then you know how to do it
17:33:48 <Bjarni> that's the best way to learn how to code
17:33:51 <plakkertjes> but, whats the language for all those programs
17:34:09 <Bjarni> if you start from scratch then you decide
17:34:30 <plakkertjes> google is your frienddd
17:34:42 <plakkertjes> ASM Auto Recycling Ltd
17:34:57 <plakkertjes> assembly language
17:35:37 <glx> asm is not the easiest thing to start with ;)
17:35:40 <plakkertjes> do i need any programs to do so? or just notepad o.O
17:35:42 <nicfer> also remember that commercial games are just recycled versions of previous games in the series
17:35:50 * planetmaker smells a sadist or masochist out there
17:36:16 <plakkertjes> was a nice rewrite
17:36:33 <plakkertjes> it had almost the same commands
17:36:38 <nicfer> instead open source games are the same game but enhanced every time more
17:37:26 <nicfer> with recycling I mean when they finish a game they start another one and name it [nameofthegame] 2 or something similar
17:37:50 <glx> only if the previous worked ;)
17:38:18 <plakkertjes> The equivalent assembly language representation is easier to remember (example in Intel syntax, more mnemonic):
17:38:27 <nicfer> here openttd 2 (or 2.0.0) will be the same game but with tons and tons of new features ;)
17:38:40 <Bjarni> I can read what you just wrote
17:38:46 <plakkertjes> but the actual meaning
17:39:24 <plakkertjes> Move the value 61h (or 97 decimal; the h-suffix means hexadecimal) into the processor register named "AL".
17:39:27 <plakkertjes> its explained o.O
17:39:31 <plakkertjes> just didnt read on
17:40:09 <Bjarni> MOV is moving one "unit". a "unit" has the same size as the number of bits in the CPU
17:40:37 <Bjarni> coding ASM without a list of commands could be a bit tricky though
17:41:25 <glx> there are many different opcodes for mov (depending on source and destination)
17:41:35 <plakkertjes> The tutorial is completely focused towards the computers that function with processors of the x86 family of Intel, and considering that the language bases its functioning on the internal resources of the processor, the described examples are not compatible with any other architecture.
17:41:38 <plakkertjes> is that my computer? o.O
17:42:30 <plakkertjes> what does it mean :)
17:42:51 <planetmaker> processor family. Same basic instruction sets
17:42:54 <glx> it's a processor family starting with 8086
17:45:40 <plakkertjes> and the groups of 4 bits are called nibbles.
17:46:09 <glx> Bjarni: as already stated it doesn't matter for me ;)
17:46:25 <Bjarni> luckily there are more than two people in this channel :)
17:46:34 <plakkertjes> question: wasnt their something with a computer that worked not only with 0 and 1 but also with 2?
17:47:14 <plakkertjes> i guess it wouldnt be called binary
17:47:25 <plakkertjes> bi = 2, and binary = 0 and 1 right
17:47:45 <Bjarni> Oxford has an experimental quantum computer that can handle 0, 1 and 0 and 1 at the same time (3 stages). However you don't want to use that one
17:47:56 <Bjarni> and it's like 1 Hz or less
17:48:08 *** LilDood has joined #openttd
17:48:09 <Bjarni> not to mention I think it's unstable
17:48:35 <Bjarni> and has dangerously strong magnetic fields
17:48:35 <glx> doesn't use a cat in a box ?
17:48:38 <SpComb> and if you manage to typo your source, it might just trigger armageddon
17:48:52 <blathijs> I think that a quarnary (4-stage) system is used in datacommunications / storage every now and then
17:49:18 <blathijs> Has some nice properties wrt redundancy and ECC IIRC
17:50:17 <Bjarni> <blathijs> I think that a quarnary (4-stage) system is used in datacommunications / storage every now and then <-- there is a 4-stage system used in chip designs because it has a lower power usage
17:50:17 <SpComb> you mean encoding a binary signal using a four-state code?
17:50:33 * peter1138 molycoddles his autoreplace.
17:50:39 <Bjarni> it's not used much due to tradeoffs though ;)
17:50:49 <SpComb> (as a phyiscal layer protocol)
17:50:57 * peter1138 ponders backporting frosch123's fixes.
17:51:06 <Bjarni> SpComb: yeah in the physical layer
17:52:02 <Bjarni> basically: will it compile or not. That's all I need to know
17:53:17 <Bjarni> if I just commit it then people will test it for me without realising it :P
17:54:20 <blathijs> Bjarni: I'll give it a go
17:55:34 <blathijs> Bjarni: Why is it needed exactly? And what does this "max allowed" and "min allowed" mean?
17:56:08 <blathijs> And shouldn't configure just set HAVE_BROKEN_ICONV properly instead of adding conditionals there?
17:57:19 <Bjarni> because HAVE_BROKEN_ICONV setup appeared not to work with the flags in question
17:57:55 <Bjarni> the min flag is adding something that used to be implicit for universal binaries (but isn't anymore)
17:58:52 <Bjarni> max allowed is the version of the system it's build on except if it's set elsewhere (which it isn't for us)
17:59:02 <glx> I guess they are not macros nor define (like some OSX stuff)
17:59:37 <Bjarni> I have them as defines in /usr/include/AvailabilityMacros.h
18:00:12 <glx> but you could use a little exe to detect the value to use for HAVE_BROKEN_ICONV
18:01:51 <glx> or update tmp.iconv.cpp stuff (config.lib:2128)
18:01:55 <plakkertjes> pfft it says I need a compiler, where can I find it?
18:04:02 <Bjarni> that depends on your OS
18:04:06 <blathijs> Bjarni: Why is the !APPLE there anyway? AFAICS, the HAVE_BROKEN_ICONV is only set when the having a const char* breaks in configure, so why would you still use a const char* on MacOS < 10.5?
18:04:42 <Bjarni> <blathijs> Bjarni: Why is the !APPLE there anyway? <-- to avoid non OSX systems from looking at the flags I presume they haven't defined
18:04:51 <Bjarni> plakkertjes: I'm sorry
18:05:37 <Eddi|zuHause2> <nicfer> only accelerates 3d games <- modern graphics cards (drivers) can also run arbitrary code on the GPU. some games use that for physics engine
18:05:49 <blathijs> Bjarni: Hun? What flags?
18:06:09 <blathijs> Bjarni: I mean the !APPLE in unix.cpp: #if defined(HAVE_BROKEN_ICONV) && !defined(__APPLE__)
18:06:17 <Eddi|zuHause2> <plakkertjes> what language is openttd written in? is it hard to learn? o.O <- if you have to ask that, then it is.
18:06:21 <blathijs> Bjarni: (Looking at the current version, without your patch)
18:06:49 *** Brianetta has joined #openttd
18:07:12 <Bjarni> I working on a fix for my fix since I realised that this whole issue is due to differences between different versions of OSX
18:07:20 <Bjarni> so the trunk is broken right now
18:07:39 <Bjarni> before I fixed this it was broken for some other people
18:12:11 <Bjarni> now I think I know why HAVE_BROKEN_ICONV failed to be setup correctly
18:12:44 <Bjarni> if I'm right then the solution is way more simple than all the stuff I have been trying so far >_<
18:16:57 <blathijs> Bjarni: I think the && !APPLE there is really a workaround for a broken configure, so adding more defines to that is not a proper solution
18:17:11 <blathijs> Bjarni: Fixing configure to set HAVE_BROKEN_ICONV when needed, is
18:18:32 <Bjarni> now I know what's wrong
18:19:06 <Bjarni> when compiling a test app to get a result to assign to HAVE_BROKEN_ICONV then it skips the CFLAGS
18:19:11 <blathijs> The configure testcase is wrong?
18:19:16 <Bjarni> or rather they haven't been set yet
18:19:41 <Bjarni> so when I build a universal binary the test app isn't based on the correct SDK
18:19:57 <blathijs> That is indeed the major flaw in this configure approach
18:20:14 <Bjarni> it's all so simple now that I detected it
18:20:15 <blathijs> If a testprogram fails, you can only assume that it is because the reason you think :-)
18:22:32 <Bjarni> the question is how to fix this...
18:27:27 <blathijs> Bjarni: What CFLAGS do you need? Can't you set them earlier in the configure process?
18:27:51 <blathijs> In general it seems like a decent idea to first set up the compile env and then do tests
18:28:04 <blathijs> Though perhaps the things you need shouldn't be in CFLAGS
18:29:08 <Bjarni> it's set in make_sed()
18:32:16 <Bjarni> it = the code to set up the cflags
18:33:21 <blathijs> But what kind of CFLAGS do you need, then? Or rather, what CFLAGS?
18:34:25 <Bjarni> T_CFLAGS="$T_CFLAGS -isysroot /Developer/SDKs/MacOSX$osx_sysroot_version.sdk -mmacosx-version-min=10.4"
18:36:09 <Eddi|zuHause2> hm... i should separate my long distance grain and cattle trains
18:36:24 <CIA-5> OpenTTD: frosch * r13851 /trunk/src/newgrf_station.cpp: -Fix (r9393): GetTownByTile() is only valid for houses and roads.
18:36:28 <Eddi|zuHause2> to be able to counter shifts in the production amount better
18:39:26 <blathijs> Bjarni: It seems the -isysroot is also set in make_cflags_and_ldflags
18:39:36 <blathijs> Bjarni: Are they both needed?
18:41:17 <blathijs> Bjarni: Also, a few lines lower in make_cflags_and_ldflags, the -isysroot is converted to something else for the makedepend cflags. I feel that that might be handled better when setting the -isysroot in the first place, I think
18:41:33 <peter1138> frosch123: ... something actually uses parent scope for stations!?
18:42:08 *** Tech-Frog has joined #openttd
18:42:31 <Tech-Frog> Is this OTTD as in Open Transport Tycoon?
18:42:42 <frosch123> Is there a nicer way to detect whether we are resolving industries or industrytiles
18:42:51 <SpComb> there's no "Transport Tycoon" in "OpenTTD"
18:43:08 <blathijs> Bjarni: Also, I feel that the -isysroot might be better placed in cxx_host instead of CFLAGS. Not sure, though
18:44:00 <peter1138> frosch123, uh, isn't it a totally separate resolver?
18:44:01 <Tech-Frog> Well, this channel is about the game right?
18:44:29 *** Tech-Frog has left #openttd
18:44:57 <frosch123> peter1138: industrytile's getvariable calls industry's getvariable for parent scope, but it shall not be passed further to the town's getvariable :)
19:16:18 *** Shorttail has joined #openttd
19:17:38 <Shorttail> Mine doesn't quite work
19:17:54 <bleepy> can get pills for that
19:18:53 <Shorttail> When the buses come in, and the money show up above upon unloading passengers, the money isn't added to my fund
19:19:53 <ln> those damn passengers who don't pay for their trips.
19:20:32 <glx> don't use transfert orders :)
19:20:47 <glx> unless you fully understand how they work
19:21:50 <Shorttail> Well what should I do then?
19:22:00 <Roujin> just give them normal orders
19:22:03 <SpComb> don't add weird modifiers to the vehicle orders
19:22:03 <Eddi|zuHause2> Shorttail: just give goto orders, not transfer
19:22:07 <Shorttail> I played this like 10 years ago and managed to farm properly then
19:22:20 <SpComb> game's different now than it was 10 years ago :)
19:22:22 <Brianetta> Transfer didnt exist 10 years ago
19:22:23 <Eddi|zuHause2> 10 years ago it did not have transfers ;)
19:22:34 <SpComb> (10 years ago OpenTTD didn't exist...)
19:23:34 <Shorttail> The order shold just be clear?
19:23:55 <glx> you can have load/unload (but it's not recommended for buses)
19:23:57 <Shorttail> In the old one, the buses wouldn't pick up passengers with that XD
19:25:02 <Shorttail> Not the one I played
19:25:15 <Shorttail> I paid to have the buses running, but they never did anything
19:25:50 <Eddi|zuHause2> that could have all sorts of reasons...
19:26:17 <Eddi|zuHause2> anyway, your money should now be green instead of yellow
19:26:28 <Eddi|zuHause2> that means it's actually added
19:27:50 <Shorttail> The one I had then started in a mode where the water was lava
19:31:02 <Bjarni> <ln> those damn passengers who don't pay for their trips. <--- I experienced something like that once. Somebody came up to me and asked if there were crew in the wagons to handle tickets. When I said yes they ran
19:32:07 <Bjarni> took me a moment to realise that they had asked train crew if there were train crew to handle tickets or if they could travel without paying without being caught
19:36:50 *** SaiBot is now known as Joshuaxiong1
19:42:31 <Eddi|zuHause2> 's a pretty genious question to ask :p
19:46:53 <Bjarni> it took me completely by surprise
19:47:10 <Bjarni> so I didn't realised what happened before they were gone
19:47:50 <planetmaker> hehe. So 'yes' was the right answer :)
19:48:06 <Bjarni> the railroad realised that certain people figured out to leave the trains when they spot uniforms
19:48:28 <Bjarni> solution: ticket control from people in civilian outfits
19:49:26 <planetmaker> he. They run around nearly like those ragged alcohol junkies :P
19:49:56 <Bjarni> funny enough the railroad will not tell what kind of outfit they will be using
19:50:08 <Bjarni> I think they plan on not using a standard at all
19:50:18 <hylje> else it'd be an uniform, duh
19:50:48 <planetmaker> probably just "come to work in your usual clothes"...
19:51:19 <Eddi|zuHause2> civilian ticket controllers i know from trams around here...
19:51:33 <Eddi|zuHause2> the conductors in trains i have always seen in uniform...
19:52:01 <planetmaker> well. A conductor doesn't need to be undercover. It's clear that he's on bord :)
19:55:20 <ln> i've seen a DB conductor who spoke finnish.
19:56:34 *** Joshuaxiong1 has joined #openttd
19:57:56 <SpComb> ln: did he look weird?
19:58:56 <Joshuaxiong1> /join #Joshuaxiong1
19:59:36 <Eddi|zuHause2> /kick Joshuaxiong1
19:59:48 <Joshuaxiong1> /join #Joshuaxiong1 - The botnet Owner
20:00:23 *** ChanServ sets mode: +v Bjarni
20:00:23 *** ChanServ sets mode: +v Belugas_Gone
20:01:02 <ln> SpComb: he was a german married to a finnish woman.
20:02:35 <Wolf01> uhm, I reduced the size of the trains to deliver cargo quickly, and now the industry lowered the production 3 times in 3 months by 40% each time
20:03:49 <ln> Wolf01: why don't you work as a conductor at Trenitalia?
20:04:27 <Wolf01> maybe I could make it work better
20:05:51 <nicfer> would be good a game like 'opensimcityttd'?
20:06:09 <SpComb> with eight levels of micromanagement
20:06:34 <SpComb> you could combine elements of The Sims as well
20:06:48 <Eddi|zuHause2> Master of Orion!
20:07:26 <Eddi|zuHause2> and while we're at it, Age of Empires!
20:07:43 <Rubidium> SpComb: you mean a train, ship, flight and car simulation for optimal micromanagement of the movement of your vehicles?
20:07:46 <nicfer> it would be multiplayer, where some players build cities and other ones joins them with rails and roads
20:09:16 <nicfer> well, two-in-one games tend to suck
20:09:31 <Rubidium> but it'll be like 10 games in one
20:11:21 <Eddi|zuHause2> you could even play tetrisolitaire
20:12:28 <Bjarni> <Eddi|zuHause2> Master of Orion! <-- great idea
20:12:54 <Eddi|zuHause2> i think somebody actually started that...
20:13:00 <Bjarni> "what are you doing?". "I'm playing OMOO... do you want to join me?"
20:13:11 <Eddi|zuHause2> freeorion or something
20:13:21 <nicfer> or else, create a way to syncronize a multiplayer city simulator with openttd
20:14:02 <nicfer> so, when a city is built in [opensimcity], that city is created in a openttd multiplayer game
20:14:33 <Eddi|zuHause2> yes, and let drive trains between different multiplayer servers!
20:15:41 <Bjarni> looks like FreeOrion is alive
20:15:51 <Bjarni> 0.3.10 was released a month ago
20:16:34 * peter1138 releases some beer.
20:17:33 <Bjarni> it's open source yet they released an exe and a linux binary only
20:17:45 <Bjarni> where is the source tarball? :)
20:19:10 <Eddi|zuHause2> "The orange icon represents trade. Trade does absolutely nothing in v0.3"
20:20:14 <Eddi|zuHause2> Bjarni: source is in the svn, i presume
20:21:03 <Bjarni> but isn't it common to release source tarballs with stable releases?
20:21:26 <Celestar> peter1138: you there?
20:21:39 <plakkertjes> does owning a country for a certain percentage have any benefits?
20:21:43 <plakkertjes> country = company
20:21:56 <Eddi|zuHause2> Bjarni: tell that to the maintainer ;)
20:22:24 <plakkertjes> why not? why buy shares? to sell for profit later on?
20:22:27 <Celestar> peter1138: do you have a list of comments/criticism? :D
20:22:44 <Celestar> peter1138: does it ... work?
20:22:56 <peter1138> Well I see routes being added and removed, so I think so ;p
20:23:02 <Eddi|zuHause2> i had an idea how to handle the non-nonstop stations properly: have an "autofill" button on the order list, similar to the one on the timetable list
20:23:07 <peter1138> I haven't yet made it draw on the minimap, though.
20:23:25 <peter1138> Eddi|zuHause2, train stops at station, adds it to the order list?
20:23:32 <Celestar> peter1138: I'll make it work during this week
20:23:38 <peter1138> Mmm, I had been thinking something similar.
20:23:41 <Eddi|zuHause2> that allows the user to remove them manually
20:23:51 <Eddi|zuHause2> but it saves the micromanagement to add them
20:24:05 <Celestar> peter1138: I'll also have the pathfinder working this week
20:24:18 <Celestar> Eddi|zuHause2: a "reverse route" button would help as well :P
20:24:36 <Celestar> Eddi|zuHause2: like you add A->B->C->D->E and it adds ->D->C->B then
20:25:05 <Eddi|zuHause2> maybe... but i think that is less general than the "autofill"
20:25:12 <peter1138> Well I shall have a go at doing the minimap, as it means I may understand some of it more ;)
20:25:26 <Eddi|zuHause2> yes, minimap would be great ;)
20:25:56 <Celestar> peter1138: I'll be on for question from tomorrow 0700 (your time)
20:26:09 <Celestar> peter1138: the question is still whether and how to proceed (branch or trunk)
20:26:51 <peter1138> Personally I wouldn't branch... all that syncing 'n stuff :o
20:27:02 <peter1138> So far it seems fairly contained.
20:27:02 <Eddi|zuHause2> branch is only needed if you expect heavy intrusion into existing code and rather long development time...
20:27:31 <glx> you can also use mercurial
20:27:45 <peter1138> Yeah, that may be an useful.
20:27:52 <Celestar> I dunno mercurial, what is it?
20:28:12 <Celestar> peter1138: I might be less contained when the actual cargopacket handling comes into play
20:28:20 <Celestar> peter1138: but as it will (always) be optional, ...
20:29:19 <Eddi|zuHause2> i'd vote for trunk, that way you'll have the most testers
20:29:21 <peter1138> I just struggled to turn my fan around... with its 25cm fan and 120W motor it's quite gyroscopic...
20:29:48 <Celestar> Eddi|zuHause2: good point
20:29:55 <Celestar> what's Rubidium's opinion? :D
20:30:18 <Celestar> peter1138: the question is whether to make it optional at compile-time (dependency on boost)
20:30:34 <peter1138> No, because that'll cause havoc for network play.
20:30:57 <glx> it should be network safe before inclusion
20:31:29 <peter1138> Well that's why it's not being based on the existing patch ;p
20:31:34 <Celestar> glx: currently, it is TOTALLY network safe. Might be related to the fact that it doesn't do anything yet :P
20:31:43 <Celestar> peter1138: well, that too ;)
20:31:59 <peter1138> Boost should be okay as it does not require extra libraries.
20:32:01 <Celestar> glx: I'll promise to do my best to keep it network safe (=
20:32:04 <Eddi|zuHause2> you can disable network if you disable boost ;)
20:32:27 <Celestar> peter1138: we _could_ put the boost stuff as svn:externals ..
20:32:39 <Celestar> peter1138: if we find a stable revision that we could use
20:34:02 <glx> something like squirrel for noai
20:34:14 <peter1138> Oh, there are some codestyle issues at the moment, but nothing insurmountable.
20:34:30 <Celestar> peter1138: having YAPP and NoAI and paxdest in 0.7.0 would be awesome, wouldn't it?
20:35:13 <Eddi|zuHause2> then people have nothing left to request for 0.8 :p
20:35:23 <Celestar> Eddi|zuHause2: OpenGFX?
20:35:36 <Eddi|zuHause2> so you'll have to start insane stuff like shunting :p
20:36:03 <peter1138> Hmm, what have we got since 0.6?
20:36:40 <Eddi|zuHause2> conditional orders
20:37:16 <Celestar> the new autoreplace (=
20:37:40 <Eddi|zuHause2> svn log | grep -i feature
20:38:07 <rortom> when i write {CARGO} in a language file
20:38:15 <rortom> to what is it resolved?
20:38:30 *** Prof_Frink has joined #openttd
20:40:43 <rortom> i mean i want something like that for vehicles, so i try to understand the system
20:40:44 <Celestar> peter1138: good night. I ya got question, I'll be on in a bit more than 9 hours
20:40:46 <peter1138> How? Deep within the depths of the string system.
20:40:58 <Celestar> peter1138: boost's a b!tch in the beginning :D
20:41:09 <rortom> so is there a {VEHICLE} ?
20:41:39 <Eddi|zuHause2> also new is the separation of the load/unload flags for orders
20:41:40 <peter1138> {VEHICLE} gets the name of the vehicle
20:41:51 <peter1138> New order (hah) gui
20:42:34 <peter1138> name can be either the user's custom name for it, or "Train n" if it has no custom name
20:42:45 <peter1138> (If it's a train, heh)
20:42:46 <Celestar> the new order gui rocks
20:43:00 <peter1138> Celestar, it takes a bit of getting used to, but yeah...
20:43:35 <CIA-5> OpenTTD: bjarni * r13852 /trunk/ (config.lib src/unix.cpp): -Fix (r13849): [OSX] fixed universal binary building without breaking anything this time
20:43:52 * peter1138 tests Bjarni's patch ;)
20:44:16 <Eddi|zuHause2> i'm still not convinced of the idea to not allow both "no loading" and "no unloading"... "go via" is not the same, because it does not allow turning around
20:44:17 <Bjarni> it's always easier to get people to test AFTER a commit
20:44:53 <Eddi|zuHause2> Bjarni: you should really have commited it 5 minutes before the nightly :p
20:45:19 <peter1138> Maybe 'via' should have an option for allow/disallow turning around?
20:45:21 <rortom> i get an error NOT_REACHED at line 923 of strings.cpp :(
20:45:36 <glx> then you did something wrong
20:45:38 <Bjarni> Eddi|zuHause2: surprisingly enough this didn't affect the nightly build
20:45:53 <peter1138> Then you included {VEHICLE} in a string that wasn't expecting it, probably.
20:46:05 <peter1138> Bjarni, famous last words...
20:46:29 <Eddi|zuHause2> problem is that "via" does not allow stopping at all
20:46:48 <peter1138> Yes, because it was designed for waypoints
20:47:06 <peter1138> And turning around at a waypoint was deemed unfortunate
20:47:39 <peter1138> I once had two paths and put a waypoint in to make trains choose the less congested route
20:47:55 <peter1138> But the trains we're getting to the waypoint and then turning around.
20:48:03 <Eddi|zuHause2> so "go to" with "no loading" and "no unloading" could be a replacement for that. problem at that point is it does not allow passing through when the train is already late and does not have to wait or turn around at that intermediate station
20:49:02 <Eddi|zuHause2> yes, it'd be a good idea to have "turn around" as a specific order
20:49:31 <Eddi|zuHause2> i think that should be part of future "shunting orders"
20:50:19 *** stillunknown has joined #openttd
20:50:35 <Eddi|zuHause2> that list will get longer very quickly ;)
20:52:59 <Eddi|zuHause2> with "shunting orders" i could imagine stuff like "shunting at {station}, leave wagons behind [with condition (e.g. cargo=0)], take wagons [with condition (e.g. cargo>0)]"
20:54:14 <Eddi|zuHause2> initially, wagons without enginge could be stored in "the void" [at a station], or on a free platform
20:54:32 <Eddi|zuHause2> they need not be actually moved by a shunting engine
21:02:25 <Eddi|zuHause2> what i meant, that could also include "shunting at {station}, move engine to other end"
21:03:11 <Eddi|zuHause2> or "shunting at {station}, go backwards" [for trains with steering wagons]
21:16:52 <rortom> where is the vehicle actually loaded when inside a station?
21:17:57 <Eddi|zuHause2> src/economy.cpp:static void LoadUnloadVehicle(Vehicle *v, int *cargo_left)
21:17:57 <Eddi|zuHause2> src/economy.cpp:void LoadUnloadStation(Station *st)
21:18:02 <Eddi|zuHause2> one of those, i think
21:19:35 <peter1138> Hidden in the depths of FIFO queues...
21:25:54 <Eddi|zuHause2> looks a little spread out ;)
21:26:35 <rortom> but its a nice addon i think
21:26:50 <Eddi|zuHause2> could you do something like "3 Trains" instead of "Train: 3"?
21:27:14 <rortom> i still have to fine tune it
21:27:21 <Eddi|zuHause2> less blank lines, and maybe write stuff after the : (with line break)
21:27:56 <rortom> like the rating for cargo
21:28:04 <rortom> so each cargo gets its own line
21:28:18 <rortom> and each vehicle type as well
21:28:30 <Eddi|zuHause2> maybe, but i was more thinking like the accepts line...
21:28:43 <Eddi|zuHause2> and it probably needs a second scrollbar
21:29:44 <Eddi|zuHause2> i like the general idea of the patch ;)
21:31:18 <rortom> i just feel that its already getting too big to be worth the effort of trying to get into the game
21:32:09 <Eddi|zuHause2> then clean it up and shrink it ;)
21:35:57 <Eddi|zuHause2> especially split the "backend" part of calculating the stats from the "frontend" part of adding the GUI elements to display the stats
21:36:11 <Eddi|zuHause2> splitting is always a good step for inclusion ;)
21:37:06 <rortom> i also added a new vehicle_type string replace in the strings system
21:39:34 <Eddi|zuHause2> needs plural handling ;)
21:40:16 <Eddi|zuHause2> Road vehicle{p "" s}
21:40:16 <rortom> you want to see the patch?
21:40:29 <Eddi|zuHause2> i don't think there's much to see there ;)
21:40:41 <rortom> i think i made somthing wrong
21:41:46 <Eddi|zuHause2> i'm not someone to give you codestyle comments or something
21:43:07 <Eddi|zuHause2> a question: assuming "transferred" means both "delivered" and "picked up", what happens with actual "transfer"? [by the "stupid" algorithm that cargo would be counted twice]
21:44:24 <rortom> the thing in question is SCC_VEHICLE_NAME_BY_TYPE
21:44:37 <rortom> also it takes everything that is put into the station
21:44:59 <rortom> via UpdateStationWaiting(...)
21:45:24 <Eddi|zuHause2> yes, but "transfer" means it is put into the station once, and picked up from the station later, so the counter would be increased twice
21:45:49 <Eddi|zuHause2> so upon pickup, you would have to check if that cargo originated from the station
21:46:06 <Eddi|zuHause2> if not, do not count the packet again
21:46:32 *** de_ghosty was kicked by DorpsGek (Wrong channel. Retry in #openttdcoop.)
21:46:33 *** de_ghosty has joined #openttd
21:46:57 *** welshdragon has joined #openttd
21:50:04 *** extspotter has joined #openttd
21:50:13 <Eddi|zuHause2> maybe it would help your case if you could formulate whole sentences, welshdragon.
21:50:33 <welshdragon> i would like the latest nightly
21:51:08 <Eddi|zuHause2> then substitute "nigtly" for the "*" in the URL in the topic...
21:51:22 <Eddi|zuHause2> my h key is somewhat broken...
21:51:49 <Eddi|zuHause2> rortom: that's not necessarily new...
21:52:28 <rortom> Eddi|zuHause2: yes, but refreshing ;)
21:52:35 <rortom> also i will leave for bed
21:54:46 <Eddi|zuHause2> what's going on here? it's not even midnight and everyone is already in bed...
21:54:59 <Eddi|zuHause2> what happened to the "nightshift"?
21:55:14 <ln> though it is past midnight.
21:55:46 <Eddi|zuHause2> it's not... it's "fünf vor zwölf"
21:56:29 *** extspotter has left #openttd
21:58:46 <Eddi|zuHause2> ["fünf vor zwölf" is often a fixed phrase meaning "the last chance to prevent an imminent desaster" [for example global warming]]
22:00:32 <Eddi|zuHause2> you'll find that quite often in political speeches ;)
22:02:06 <Eddi|zuHause2> i have no idea if other languages have similar phrases
22:06:35 *** planetmaker has joined #openttd
22:09:28 <Eddi|zuHause2> the english counterpart of the phrase appears to be "minutes to midnight"
22:29:07 <Vikthor> (00:02:06) Eddi|zuHause2: i have no idea if other languages have similar phrases -- In Czech that would be "Za pět minut dvanáct" - Five minutes to twelve
22:35:43 *** GoneWacko has joined #openttd
22:44:37 *** ChanServ sets mode: +o Bjarni
22:51:47 <Bjarni> it's not like I'm using IRC right now anyway
22:52:13 <ln> or at least a contradiction
22:59:49 *** Dred_furst has joined #openttd
23:11:39 <CIA-5> OpenTTD: truebrain * r13853 /branches/noai/src/ai/api/ai_road.cpp: [NoAI] -Fix r13833: he, at least he has the guts to say he made a mistake ;) (Yexo)
23:13:26 <CIA-5> OpenTTD: truebrain * r13854 /branches/noai/src/ai/api/ai_controller.cpp: [NoAI] -Add: added 'null' and 'bool' support to AI SaveLoad code (Yexo)
23:24:56 <Eddi|zuHause2> that's worse than that christmas decoration house :p
23:39:12 <Prof_Frink> Well yes, it's not TSO
23:50:41 *** DJNekkid has joined #openttd
continue to next day ⏵
