IRC logs for #openttd on OFTC at 2011-08-12
⏴ go to previous day
00:08:47 *** Xrufuian has joined #openttd
02:27:18 *** rhaeder has joined #openttd
02:39:35 *** Blacklite_ has joined #openttd
02:43:42 *** ccfreak2k has joined #openttd
02:43:42 *** SirSquidness has joined #openttd
05:35:08 *** sla_ro|master has joined #openttd
06:04:04 *** Cybertinus has joined #openttd
06:20:11 *** HerzogDeXtEr has joined #openttd
06:24:24 *** Br33z4hSlut5 has joined #openttd
06:29:19 *** TWerkhoven has joined #openttd
06:37:45 *** DayDreamer has joined #openttd
07:40:48 *** Belugas has joined #openttd
07:40:48 *** ChanServ sets mode: +o Belugas
07:42:18 *** Vikthor has joined #openttd
07:50:59 *** Biolunar has joined #openttd
08:30:36 *** Progman has joined #openttd
08:32:40 *** Eddi|zuHause has joined #openttd
08:46:48 *** Alberth has joined #openttd
08:46:49 *** ChanServ sets mode: +o Alberth
08:58:44 *** Biolunar has joined #openttd
10:01:11 *** andythenorth has joined #openttd
10:13:58 *** Adambean has joined #openttd
10:17:21 *** Brianetta has joined #openttd
10:31:57 *** TramOfDeath has joined #openttd
10:33:57 <TramOfDeath> There's a glitch with OTTD where music won't loaded unless located in install/gm
10:34:02 *** andythenorth has left #openttd
10:34:26 <TramOfDeath> Particularly, SJA
10:34:28 <planetmaker> Music has always to be located somewhere in one of the gm folders
10:35:03 <TramOfDeath> It doesn't play if it's in conten~1/gm
10:35:29 <planetmaker> the folder name also would be content_download/gm
10:35:44 <planetmaker> but that's only where ingame download content should go
10:35:54 <TramOfDeath> I know full name, you got my point
10:36:09 <TramOfDeath> SJA is an ingame download after all
10:37:45 <TramOfDeath> I have a "portable omni-install" of OTTD
10:38:53 <Rubidium> might be that those 'portability' wrappers mess things up
10:40:03 <Rubidium> though I wonder why those portability installation things are needed; probably because "readme" doesn't entice to read or so
10:40:09 <planetmaker> whatever this 'portability' thingy is
10:42:19 <TramOfDeath> erm, wrong one, remove #tips at end
10:42:20 <planetmaker> most likely it's an issue then with case of paths
10:43:13 <TramOfDeath> so it's install/content_downloads/gm
10:44:20 <TramOfDeath> not %documents%/openttd/content_downloads/gm
10:44:33 <planetmaker> ~/.openttd/content_download/gm is the default user dir compiled in our binaries for linux
10:44:56 <TramOfDeath> oh, no s... my bad
10:45:16 <planetmaker> in any case: I suggest to try one of our official binaries
10:45:43 <TramOfDeath> The installer USES official binaries
10:45:50 <Alberth> and reading the do-no-readme file may also help as it explains in detail how openttd looks for files
10:46:30 <planetmaker> the readme, though, tells that some/path/install/... is not where OpenTTD ever looks
10:47:22 <TramOfDeath> so it will not look for music in content downloads when portability is enforced by placing the CFG?
10:48:28 <TramOfDeath> in the manual, it is clear that portability is enforced by having openttd.cfg made in the install folder
10:48:53 <TramOfDeath> c:\openttd\openttd.cfg
10:49:51 <Alberth> I don't know, my Linux doesn't play the music, and I didn't bother to fix it
10:50:13 <TramOfDeath> c:\openttd\content_download\gm music doesn't play even though being selectable
10:50:49 <TramOfDeath> but same music pack moved to c:\openttd\gm plays finely.
10:51:25 <planetmaker> and openttd.exe is where?
10:51:40 <TramOfDeath> c:\openttd\openttd.exe
10:52:58 <TramOfDeath> Read the wikihow manual first.
10:55:14 <Rubidium> I wonder where OpenTTD places the music when downloading it. As it extracts it, it might very well be needed that it's in a subdirectory of content_download/gm/
10:56:08 <TramOfDeath> content_download/gm/pack_name
10:56:39 <Rubidium> and/or that it requires some metadata from the content download system to be there as well
10:56:57 <Rubidium> but then... *why* are you putting stuff that is NOT downloaded by OpenTTD in content_download?
10:58:22 <TramOfDeath> SJ anthology IS from content system
10:58:45 <TramOfDeath> Scott Joplin Anthology
10:59:36 <Rubidium> and OpenTTD downloaded it into that directory?
10:59:46 <Rubidium> or did you move the files yourself?
11:01:06 <TramOfDeath> OTTD downloaded it, it got extracted as content_download\gm\scott_joplin_anthology-2.0.0
11:02:31 <TramOfDeath> It was selectable but didn't play from there
11:04:34 <TramOfDeath> After I moved it to openttd\gm\scott_joplin_anthology-2.0.0, it plays just OK
11:07:25 <planetmaker> it works for me from ~/Documents/OpenTTD/content_download/gm ...
11:07:39 <Alberth> dbg: [grf] Checking ~/openttd/industry_building/bin/content_download/gm/openmsx-0.3.1/openmsx.obm for base music set
11:07:39 <Alberth> dbg: [grf] Adding OpenMSX (96) as base music set
11:07:57 <Alberth> it seems downloaded music ends up at the right place and gets found too
11:08:24 <Alberth> I cannot test whether it gets played
11:09:05 <planetmaker> found is needed as it can be selected
11:10:10 <TramOfDeath> yikes, registering in another place...
11:11:05 <TramOfDeath> just for one silly piece of p00p to be removed...
11:11:47 <TramOfDeath> I forgot my OTTD acc... :@
12:53:43 *** Progman has joined #openttd
13:19:38 *** Maarten has joined #openttd
13:22:52 *** Kurimus has joined #openttd
14:25:21 *** Prof_Frink has joined #openttd
14:54:52 *** Devroush has joined #openttd
15:02:02 *** supermop_ has joined #openttd
15:19:17 *** mib_m4a6jn has joined #openttd
15:20:07 <planetmaker> that was quick ;-)
15:26:03 <Eddi|zuHause> the problem when you have range-fight riders, you don't have any useful units to capture cities with...
16:35:58 <aditsu> hi, there's a multiplayer game that I can't join anymore
16:36:17 <aditsu> it loads and shows the frozen map, then after a long time it says I joined and then immediately disconnects
16:36:30 <aditsu> what can I do about it?
16:37:14 <planetmaker> probably not much. That's the behaviour if the game is too "heavy" for a client to join
16:38:45 <aditsu> hm.. does the game run slower when a LOT of trains got stuck? or it doesn't affect it?
16:39:32 <michi_cc> If stuck means "waiting for free path", then yes, there is an effect.
16:40:37 <aditsu> I noticed that happening in some of the other companies earlier (when I could still join)
16:41:31 <aditsu> and it's a big map too
16:45:48 <Rubidium> I guess it's his computer; it's even too slow to stay connected on IRC ;)
16:45:50 <planetmaker> of course the amount of vehicles and the map size are what makes a game eat much cpu
16:53:03 <aditsu> I'm sure the game can run on my computer, it's just the time when it connects.. there should be something about that, perhaps a timeout setting...
16:57:10 <aditsu> or if the game was multithreaded
16:58:39 <planetmaker> upon connection to a server, you computer has to a) download the map and b) compute the game state from when the map download started to what it is now
16:59:00 <planetmaker> that might be too much in the time it has.
16:59:09 <planetmaker> there's a lag setting for clients which is server-side only
16:59:17 <planetmaker> which allows more or less lag for clients
16:59:30 <planetmaker> or a server can choose to pause until a client finished connecting
16:59:40 *** TheMask96 has joined #openttd
16:59:40 <planetmaker> but... both usually doesn't help much
16:59:52 <aditsu> that means I should contact the admin, but I can't without connecting to the game
17:00:06 <Rubidium> being able to run a game does not mean the game runs as fast as it ought to run
17:01:04 <planetmaker> do you have a rough guestimate on total(!) vehicle number and map size?
17:01:14 <Rubidium> it is't 0.05s slower per tick than normal you won't really notice it, but those seconds add up quickly
17:01:30 <aditsu> map size is 2048*2048, total vehicles... I don't really know
17:01:41 <planetmaker> it's half speed :-P
17:02:01 <glx> 2048*2048 is usually too big for any computer :)
17:02:03 <planetmaker> well, that's already a lot, esp. on such big map
17:02:20 <planetmaker> so it's really your computer being too slow for this game in multiplayer mode
17:02:53 <aditsu> I wonder if any of the other players has a significantly faster computer, otherwise nobody can connect
17:03:07 <planetmaker> might as well be.
17:03:27 <planetmaker> Esp. the multi-core cpus usually perform worse than the lesser-core cpus as then the single cores are faster
17:04:42 <aditsu> I got a Core(TM)2 Duo CPU E8500 @ 3.16GHz
17:06:49 <planetmaker> it should get you a long way, but also that one has its limits
17:07:10 <planetmaker> esp. when there are many boats involved
17:12:20 <aditsu> I don't think there were any boats in that game.. anyway, I'm trying to contact the admin, he seems to have multiple games in parallel
17:25:14 *** frosch123 has joined #openttd
17:30:00 <frosch123> hmm, i guess ottd also plays a loud alert when there is a tornado warning for your region
17:30:43 <Eddi|zuHause> i can't reproduce that
17:45:20 <CIA-2> OpenTTD: translators * r22734 /trunk/src/lang/ (catalan.txt unfinished/persian.txt):
17:45:20 <CIA-2> OpenTTD: -Update from WebTranslator v3.0:
17:45:20 <CIA-2> OpenTTD: catalan - 1 changes by arnau
17:45:20 <CIA-2> OpenTTD: persian - 13 changes by Peymanpn
17:45:42 *** Ammler was kicked by DorpsGek (Wrong channel. Retry in #openttdcoop.)
17:50:41 *** CiprianI has joined #openttd
17:51:11 <CiprianI> How to install 32bit graphic pack?
17:52:42 <Rubidium> nobody has, as of yet, made a reasonably finished 32bits graphics pack that works with standard OpenTTD
17:52:50 *** andythenorth has joined #openttd
17:53:24 <Rubidium> there are some that require that you install some modified version of OpenTTD, but most I hear of that are requests for help to install that (which I can't really help with: see the 32bpp graphics forum for more details)
17:54:22 <Wolf01> and you oddly forgot about my almost finished brickland 32bpp graphics pack
17:54:47 <Wolf01> ok, there are the terrain and tracks only but...
17:55:35 <planetmaker> last time I asked, I was told that the same 32bpp packs work also with trunk
17:56:30 <Rubidium> planetmaker: "finished" is the operative word there
17:56:45 <planetmaker> quite so. sadly so
17:59:09 <Alberth> planetmaker: you don't seem to be able to set hangar positions in newgrf, is that conclusion correct?
18:00:12 <planetmaker> Alberth: in current trunk?
18:00:24 <planetmaker> one cannot, as they must be part of the statemachine knowledge
18:01:00 <Alberth> hmm, how does the game detect a click at a hangar tile then
18:01:20 <Rubidium> there's a list of tiles that's a hangar
18:01:24 <Alberth> the source does have a list of tile positions
18:02:00 <Alberth> Rubidium: exactly, which means a newgrf airport cannot move the hangar currently
18:03:01 <Alberth> (I think so, at least)
18:09:44 <supermop_> hah, tornado warning
18:11:11 <planetmaker> supermop_: xkcd? ;-)
18:14:27 <CIA-2> OpenTTD: rubidium * r22735 /trunk/src/newgrf_text.cpp: -Fix [FS#4718]: triggering NOT_REACHED when playing with a NewGRF that supplies genders/cases for a language that you removed from your installation
18:31:49 *** andythenorth has joined #openttd
18:34:11 <CIA-2> OpenTTD: rubidium * r22736 /trunk/src/saveload/strings_sl.cpp: -Codechange: constify some numbers
18:36:48 <CIA-2> OpenTTD: rubidium * r22737 /trunk/src/saveload/ (cheat_sl.cpp company_sl.cpp strings_sl.cpp): -Fix [FS#4717]: some corrupted savegames could crash OpenTTD instead of showing the "savegame corrupted" message
18:39:18 <supermop_> when I saw the earlier comment about ottd, I just assumed frosh was being insane, or that somehow there was actually a tornado there and he had confused the local warning system with his computer
19:12:19 *** supermop has joined #openttd
19:15:20 *** Hyronymus has joined #openttd
19:20:27 <Alberth> lots of discussion in the dutch train set thread :)
19:26:36 <Hyronymus> yes, discussion is going well
19:34:04 <KittenKoder> Variety is always good.
19:36:34 <Hyronymus> try that on your gf :P
19:38:59 *** Illegal_Alien has joined #openttd
19:49:49 <andythenorth> "omg it's a girl" :P
19:50:24 <KittenKoder> How dare girls play video games!
19:50:37 <KittenKoder> Don't they know it's all boys stuff!
19:50:51 <KittenKoder> andythenorth, don't remind me, it's the reason I hate my mother.
19:51:46 <KittenKoder> If she hadn't been such a wench about my coding, I would have worked at MS.
19:52:46 <Hyronymus> and now you work at?
19:53:08 <KittenKoder> Freelance network person.
19:53:42 <KittenKoder> But that's not the suck part.
19:54:03 <KittenKoder> The suck part is that I didn't get to learn enough early on to be big.
19:56:52 <KittenKoder> I am of the mind they make monorail suck on purpos4e.
19:57:53 <planetmaker> KittenKoder: then draw us a new default monorail (or maglev) ;-)
19:58:05 <planetmaker> Personally I could do with other default maglev tracks...
19:58:11 <KittenKoder> Not talking about the look.
19:58:13 <planetmaker> similar to what you suggest in the forums
19:58:56 <KittenKoder> I'm in Seattle, famous for the toy monorail ... eye candy from some expo.
19:58:59 <Alberth> perhaps they looked at the real world monorail specs?
19:59:06 <KittenKoder> ... and the more recent failed monorial system.
19:59:19 <KittenKoder> Alberth, that's my point. ;)
20:00:20 <KittenKoder> Modern monorail designers must just be designing these things to be fail on purpose.
20:01:40 <andythenorth> probably SF didn't want to draw it
20:01:47 <Alberth> oh, you are not talking about the OpenTTD monorail
20:02:22 <Alberth> planetmaker: very few industries do that
20:02:49 <planetmaker> Alberth: yes. I wonder why
20:03:01 <KittenKoder> I may have to make some sci-fi based monorials if and when I make the sci-fi maglevs.
20:03:02 <planetmaker> looks better this way IMHO
20:03:16 * planetmaker considers to just change the sawmill to what I just posted
20:03:25 <Alberth> KittenKoder: too obsessed with high-speed to be much bothered about reliability, I guess
20:03:31 * planetmaker would need to remember where the lumber is from
20:03:41 <andythenorth> planetmaker: some places in FIRS I've used both input + output cargos
20:03:45 <andythenorth> just looks better
20:04:17 <andythenorth> that's ISR lumber
20:04:21 <KittenKoder> Alberth, probably.
20:04:30 <andythenorth> drawn by mart3p IIRC
20:05:04 <Alberth> or perhaps 'speed' sells better than 'reliable' for the people that pay the bill :)
20:05:06 <planetmaker> yeah, then I'll find. It was already used in OpenGFX. I just didn't know anymore where I got it from back then...
20:06:20 <KittenKoder> Anyhow, FYI for those who might care, when I do make the sci-fi trainset, it won't be to include high speed earlier, but so players can have a science fictiony looking set, nor will I make 1,000 mph trains with 50,000 hp ... etc.
20:07:12 <KittenKoder> Woohoo! Just got the 2cc Shinkansen in my current game ... those are so pretty.
20:09:01 <KittenKoder> Personally, I love how the maglev look and the whole "frictionless" concept, but I think they move too fast and there is not enough variety ... and the OGFX ones suck. >.<
20:10:17 <planetmaker> KittenKoder: make better ones. Honestly
20:10:30 <KittenKoder> I forgot the name of my favorite one ... appearance wise I mean, of the 2cc set.
20:10:58 <KittenKoder> planetmaker, I will, just want to play with my new tracks a bit first to find all the problems.
20:12:18 <KittenKoder> I should play a game with breakdowns turned off so I can just pick which trains I like.
20:14:30 <andythenorth> has anyone taught their editor to colour in nml?
20:15:38 <KittenKoder> The c syntax coloring seems to work well enough for me.
20:17:16 <KittenKoder> I like the platforms from NewStations ... the ones that show the little people.
20:18:02 <KittenKoder> The Industrial Renewel set is awesome to, but doesn't have all the cargo for FIRS.
20:18:26 <planetmaker> no station set has
20:18:39 <planetmaker> isr is open-source
20:18:49 <planetmaker> thus... can be picked up and updated
20:18:56 <KittenKoder> That's why I was going to try stations, but if NML doesn't support them yet, I hate clunking with ASM.
20:19:02 * andythenorth tries to figure out nml
20:19:19 <planetmaker> yeah... that's holding me back, too, KittenKoder ;-)
20:19:49 <KittenKoder> Just all the bit counting makes my head hurt. >.<
20:20:00 <planetmaker> the bit counting in NML?
20:20:27 <KittenKoder> Well, ASM is what I use to use.
20:20:31 <planetmaker> it's a matter of getting used to, though
20:20:35 <KittenKoder> A looooong time ago before there was c.
20:20:52 <KittenKoder> Looked at NFO and had flashbacks.
20:21:12 * andythenorth is 20% baffled by NML so far
20:21:18 <KittenKoder> Remembering the days of trying to get animated sprites in Apple IIe.
20:21:48 <KittenKoder> andythenorth, the documentation for NML is lacking, which is what baffled me.
20:21:55 <KittenKoder> They need more contributors.
20:21:56 <Alberth> without assembler? ugh
20:22:06 <andythenorth> mostly I think I just can't read ALL_CAPS
20:22:10 <KittenKoder> Yep, Alberth .... simple op codes.
20:22:19 * planetmaker should change the default config... clearing limit reached :-P
20:22:23 <planetmaker> I set it to 10 tiles
20:22:27 <andythenorth> ALL_CAPS is just rectangles to me
20:22:36 <KittenKoder> ALL_CAPS is standard coding syntax for constants.
20:22:55 <planetmaker> KittenKoder: what is missing?
20:22:57 <KittenKoder> Has been for decades to.
20:23:34 <KittenKoder> planetmaker, mostly examples and more in depth explanations.
20:23:53 <andythenorth> so everything in caps is a constant?
20:24:05 <planetmaker> well... examples are the existing projects ;-)
20:24:13 <planetmaker> andythenorth: should be, yes
20:24:23 <KittenKoder> I saw those. ;) But they don't show everything yet. :p
20:24:32 <andythenorth> but constants that have parameters?
20:24:37 <andythenorth> I've never run across that before :O
20:24:42 <andythenorth> it's a new way of coding for me
20:24:46 <KittenKoder> I have been spoiled with Java .... Sun had a VERY comprehensive example set from the start.
20:24:57 <planetmaker> KittenKoder: well... always free to ask. And documentation patches are warmly welcome, too
20:25:19 <planetmaker> or elaborately commented example grfs.
20:25:24 <Alberth> andythenorth: no #define with parameters in a previous FIRS ?
20:25:27 <KittenKoder> Constants are also "pointers" ... to functions, callbacks, etc. .... usually for low level ones when using a high level language like NML.
20:25:39 <andythenorth> all just strings
20:26:07 <planetmaker> andythenorth: you had already many #define THIS_IND_ID F0
20:26:10 <KittenKoder> planetmaker, if it will help, I can submit the source and stuff for my maglev track, that may be a good example for some things.
20:26:35 <KittenKoder> I would just have to archive it.
20:26:37 <andythenorth> planetmaker: that's just defining the constant value though?
20:27:03 <planetmaker> ah, you mean macros...
20:27:09 <planetmaker> well, there might be some. Not sure
20:27:19 <andythenorth> this looks like a function call with params to me
20:27:19 <andythenorth> CHECK_INCOMPATIBLE (brewery, 56, CB_RESULT_IND_DISALLOW_UNSUITABLE, return CB_RESULT_IND_ALLOW_LOCATION)
20:27:23 <KittenKoder> There are some macros in NML.
20:27:43 <planetmaker> but what andy just quoted is a cpp macro which expands to NML ;-)
20:28:09 <planetmaker> KittenKoder: did you look at the (very few) examples in the example folder of NML?
20:28:38 <andythenorth> so macros take params?
20:28:48 <planetmaker> I think it has already a railtype example... but might be very new
20:28:52 <planetmaker> andythenorth: yes, they can
20:29:10 <KittenKoder> planetmaker, when I am talking examples, I'm talking more like helpdoc explanations using some "real world" language, like for parameters. So one doesn't have to learn NFO first. ;)
20:29:13 <andythenorth> and this returns the ID for the arable farm? THIS_ID(arable_farm)
20:29:30 <andythenorth> equivalent to calling arable_farm.id() ?
20:29:38 <planetmaker> well, as said: amendments etc to the docs are welcome :-)
20:29:39 <KittenKoder> It's just a recommendation though ... not that I'm saying "NML help docs suck".
20:29:45 <planetmaker> It's not many people who write on it...
20:30:01 <planetmaker> and things are clear when one has done them a few times...
20:30:11 <planetmaker> so especially this outside view, help and amendments are welcome
20:30:43 <planetmaker> I guess there's at most two people who write in NML longer than myself ;-)
20:30:48 <KittenKoder> Help docs are pain to write.
20:30:55 <andythenorth> but this - THIS_ID(town_distance) - is equivalent to calling self.get_town_distance()
20:31:01 <KittenKoder> I was never good at it, I even suck at commenting my source code.
20:31:30 <planetmaker> andythenorth: it's a shorthand to make the identifier name unique by prepending it with THIS_ID. It's a macro defined in each industry
20:31:32 <Alberth> KittenKoder: the whole trouble with these technical documents is that the people writing them are too deep into the matter to be able to explain it clearly to a new user, they don't see what concepts are new, anymore
20:31:43 <andythenorth> oh so THIS_ID gives the context?
20:31:55 <planetmaker> per-industry. Yes
20:31:59 <KittenKoder> Alberth, that's my problem to, putting it into more "real world" language is hard.
20:32:10 <planetmaker> look at the header of an arbitrary industry file. It's define there
20:32:50 <Alberth> KittenKoder: usually I drop down to concrete examples, like 'here I expect this and this to be explained'
20:33:13 <planetmaker> #define THIS_ID(...) industryname ## __VA_ARGS__
20:33:21 <KittenKoder> I usually learn by the examples more than the docs myself.
20:33:24 <Alberth> but that needs a very careful study of the docs
20:33:48 <planetmaker> which then expands in later usages of THIS_ID(something) to industrynamesomething
20:34:07 <andythenorth> working from the bottom up, last parameter is next check?
20:34:22 <planetmaker> you read the switch chains upwards, yes
20:34:32 <andythenorth> it starts to make sense
20:34:41 <planetmaker> switch = action2 basically
20:34:57 <planetmaker> whatever action2 you want, varaction, randomaction, normal action ;-)
20:34:59 <andythenorth> like nfo, the chance of 'just' getting it is minimal :P
20:35:23 <Rubidium> unless "it" refers to a headache
20:35:24 <planetmaker> any programming language with functions works that way...
20:35:38 <planetmaker> first declare, then use
20:35:57 <planetmaker> except those languages where it doesn't matter
20:36:18 <andythenorth> except those where it doesn't matter :P
20:36:24 <planetmaker> any, every - all the same in my mothertongue ;-)
20:36:32 <KittenKoder> Python (technically scripting but it counts), Java, and PHP (again scripting but meh) you do not have to declare first.
20:36:53 <Rubidium> Java uses methods! ;)
20:36:55 <KittenKoder> I'm sure there are a few others.
20:37:22 <KittenKoder> Hmm .... yeah, it is different, but methods are really just isolated functions, if you think about it. :p
20:38:49 <Rubidium> but, lets rephrase ;)
20:38:56 <planetmaker> andythenorth: wrt the callbacks: they can be defined directly from the graphics switch (=action3) of an industry
20:39:08 <KittenKoder> It depends on the compiler and how that works.
20:39:20 <Rubidium> any programming language with functions that is compiled in a single pass
20:39:54 * andythenorth can't remember what Flash does, but that's probably a Good Thing
20:40:10 <KittenKoder> A very good thing.
20:40:40 <andythenorth> it will do same as Flash - I think functions have to be declared first in both
20:40:41 <Alberth> a *language* defining the number of compiler passes? weird
20:41:25 <KittenKoder> Well, the compiler defines the language ... but meh ...
20:41:33 <KittenKoder> Actually ... I'm wrong.
20:42:01 <KittenKoder> The compiler no longer defines the language strictly, many can handle more than one now.
20:42:05 <Alberth> Rubidium: that would work, but it also limits the set of languages to just Pascal, I think :)
20:42:36 <Alberth> KittenKoder: and often there are several different compiler implementations for the same language
20:42:56 <KittenKoder> It's gotten a lot more complex than a=b.
20:43:04 <andythenorth> Pascal is the only language I have been taught
20:43:04 <KittenKoder> Or better ... a == b
20:43:09 <andythenorth> I skipped the lecture :P
20:43:22 <andythenorth> I did the code test
20:43:33 <KittenKoder> I never liked Pascal's syntax.
20:43:35 <andythenorth> I used 'top down programming' apparently that is bad :P
20:44:01 <KittenKoder> Top down is old skool!
20:44:15 <Alberth> andythenorth: with pascal that's good, too bad you have to write your program bottom-up then :)
20:44:38 <andythenorth> or maybe I failed to use top down
20:44:43 <andythenorth> anyway, whatever I did was bad :P
20:44:52 <Alberth> (as in 'main' is at the bottom of the program)
20:45:19 <KittenKoder> Python doers that to .... annoys me.
20:45:48 <KittenKoder> Sounds like a beer name in my head.
20:46:01 <andythenorth> top down is like defining all your interfaces as stubs
20:46:08 <andythenorth> then fleshing the detail code out later
20:46:32 <andythenorth> that's not going to work :P
20:46:40 <KittenKoder> Doers .... like Coors only more active.
20:47:00 <andythenorth> top down design isn't design
20:47:04 <andythenorth> it's implementation
20:47:18 <andythenorth> you'd need a design before you could do it
20:47:57 <KittenKoder> I should check the HEQ thread.
20:48:22 <Alberth> top-down design is also a way to reach a solution, but it is VERY difficult if not impossible for anything non-trivial
20:48:23 <Rubidium> isn't design almost always top-down?
20:49:07 <Alberth> normally we jump at random between top-down, bottom-up, and aspects (and more probably)
20:49:10 <Rubidium> as in: first ou need to know what the goal of the code is, before you can think about details and how it should work algorithmically
20:50:02 <Rubidium> only question is whether you do a breadth of depth first search for the solution
20:50:09 <Alberth> theoretically yes, practically you take details into account in the global goal already
20:50:26 <KittenKoder> I love how HEQ looks in game.
20:50:28 * andythenorth has always mixed them
20:50:38 <KittenKoder> The pixel artist did awesome and the set is just nice to have.
20:50:44 <andythenorth> top down *versus* bottom up seems a false distinction
20:51:08 <KittenKoder> I do spider web coding.
20:51:10 <Alberth> KittenKoder: andythenorth is the pixel painter :)
20:51:14 <andythenorth> too many websites are made top down, the results can suck
20:51:22 <KittenKoder> andythenorth, excellent work on HEWQ.
20:51:25 <andythenorth> I can't think of a game I ever designed top down
20:51:53 <andythenorth> most flash games there's some crux point of interaction that has to be solved
20:51:56 <andythenorth> so you start there
20:52:02 <andythenorth> and worry about the rest later
20:52:17 <Alberth> andythenorth: top-down and bottom-up are just theoretical notions, you never do them in a pure form in reality
20:52:20 <KittenKoder> Spider web coding is the only coding. /sagenod
20:52:37 <andythenorth> Alberth comp sci lecturers seem to think you do them in pure form :P
20:52:42 <KittenKoder> All others are obsolete!
20:52:43 <Rubidium> Alberth: I think that modelling it like a graph search algorithm would be better
20:52:46 <KittenKoder> Obsolete I tells yah!
20:53:21 <Alberth> Rubidium: with nested graphs, could work
20:54:08 <Rubidium> either you think a lot about stuff when going down, i.e. do something breadth first searchy, or you dive into a problem, fix it and fix the next closely related problem
20:54:35 <Alberth> andythenorth: everybody uses models to simplify the world, and obviously, if they know how to code, they would not be a teacher ;)
20:55:03 <Rubidium> hmm, maybe it's more like pre-order and in-order traversal of the graph
20:55:27 <Alberth> Rubidium: and even incremental design/implementation can be done in different ways :)
20:55:28 <Rubidium> so the complete solution with all the steps would be the graph (hopefully somewhat tree-ish)
20:55:31 <andythenorth> is there any significance to the nodes? (or what do they represent?)
20:55:53 <andythenorth> do you start at the node with most links?
20:55:57 <andythenorth> start at the edge of the graph?
20:56:11 <andythenorth> do you even know the graph until you've done the design
20:56:33 <Rubidium> andythenorth: I'd see the nodes as kind of actions/decisions, and no you don't know the graph till you're done
20:56:34 <Alberth> probably not until you have finished the implementation :)
20:56:44 <andythenorth> the graph model is as good as any other
20:57:01 * andythenorth has just spent three months prototyping a web app so real developers can code it
20:57:24 <andythenorth> it's basically a graph search, and the skein or plane it's on is an ontology
20:57:33 <Rubidium> so the top node is the "what are you going to make"; the closer to the top node the more abstract it is
20:57:50 <Rubidium> and further away gets more to the actual coding
20:57:57 <andythenorth> in my case it was more like "what are you going to make" is 100% of the graph
20:58:09 <andythenorth> the entities and relationships were the graph
20:58:21 <andythenorth> and throwing away lots of the graph is how we got a workable design
20:58:39 <andythenorth> but the other way of thinking is interesting
20:58:58 <andythenorth> KittenKoder: any specific HEQS requests?
20:59:00 <andythenorth> more is planned - later
20:59:33 <KittenKoder> Some earlier ones would be nice.
20:59:39 <andythenorth> earlier trucks? trams?
20:59:46 <KittenKoder> I noticed most don't appear prior to 1970's.
20:59:59 <KittenKoder> I like the detail in them. :p
21:00:17 <KittenKoder> I really like the logging truck to.
21:00:18 <andythenorth> Rubidium: elephants don't help with design :P
21:00:26 <KittenKoder> The way it actually folds up when empty.
21:00:41 <Rubidium> for pulling logs and the likes through the woods in some century long ago
21:00:54 <KittenKoder> I set up forest road routes just to use that one.
21:02:09 <KittenKoder> You know, those car semis .....
21:02:18 <KittenKoder> That would be cool for engineering and farming supplies.
21:03:10 <KittenKoder> There are some with flatbeds for tractors, I have seen, that "fold" like the logging trucks.
21:03:49 <KittenKoder> I have an elevated freeway and main strip just out my window ... so I see all sorts of stuff through here.
21:05:25 <andythenorth> Rubidium: the graph search model is really appealing
21:05:33 <andythenorth> you should write a paper :P
21:05:46 <andythenorth> query: could you start at any node, in theory?
21:07:10 <Rubidium> you'd have to start at the top node: the idea for the thing you're making
21:07:13 <andythenorth> you'd have to start at one which is known to be true
21:07:15 <Rubidium> without that, there's nothing
21:07:40 <Rubidium> but with pre-order traversal the next step would be actually doing something real
21:08:49 <andythenorth> what if the idea is, I have a cool IO algorithm, what can we do with it?
21:13:41 <Rubidium> say that is the root? ;)
21:13:58 <andythenorth> somehow we get code written :P
21:14:03 <andythenorth> sometimes we even make money from it
21:14:11 <andythenorth> so there must be some process happening
21:14:27 <andythenorth> and the top-down and bottom-down models don't look realistic
21:14:31 <andythenorth> omg - realism :P
21:14:38 <andythenorth> models != reality
21:14:56 <Rubidium> should we make 'realism' the godwin event of this thread?
21:15:46 * andythenorth -> bed in that case
21:16:01 <andythenorth> KittenKoder: I need to do a truck set separate to HEQS
21:16:11 <andythenorth> but that is blocked by a few things
21:16:12 <KittenKoder> That would be cool to.
21:16:45 <andythenorth> adding trailers to road vehicles would be useful in game
21:16:51 <andythenorth> I need a collaborator or two
21:17:00 <andythenorth> and I need inclination
21:17:15 <andythenorth> I have 4 sets already
21:17:19 <KittenKoder> Meh, just make them like you did the logging ones.
21:17:30 <KittenKoder> Start off small, get longer as you go. :p
21:17:54 <planetmaker> FIRS is already much bigger :-P
21:17:57 <andythenorth> mostly I have enough sets to do though :)
21:18:04 <KittenKoder> Seems to be how everyone does it.
21:18:16 <andythenorth> FIRS was always going to be enormous
21:18:22 <andythenorth> it was predicted to take 18 mths
21:18:57 <Hyronymus> That's rather quick tbh
21:19:06 <andythenorth> I'm quite obsessive :P
21:19:08 <andythenorth> and I've had help
21:19:10 <KittenKoder> Gives me more reason to make complex train lines.
21:19:25 <KittenKoder> ... as well as more realistic and complex stations.
21:19:26 <andythenorth> FIRS 0.7 will rock compared to previous versions
21:19:31 <andythenorth> it's more polished
21:19:37 <andythenorth> lots of small changes
21:19:43 <andythenorth> well maybe 0.7.1 will rock
21:19:54 <Hyronymus> added some buttons from our original page
21:19:56 <andythenorth> maybe 0.7 is the 'oops' edition with lots of bugs
21:20:18 <KittenKoder> Need variable economic system though, but I don't know it that's possible with a NewGRF.
21:20:29 <KittenKoder> Such as one year something being in higher demand than another.
21:21:03 <andythenorth> that's a whole other thing
21:21:15 *** douknoukem has joined #openttd
21:21:20 <andythenorth> you'd need to be able to manipulate cargo payment rates, can't do that
21:21:31 <andythenorth> well...you can within the spec, but it's way buggy
21:21:49 <KittenKoder> Yeah, I thought it would be a OTTD thing, that's why I submitted in the OTTD recommendation section of the forum.
21:22:26 <KittenKoder> Well, if there's a way to do that, it would be fun to play.
21:22:29 <andythenorth> custom profit calculation callback at stations
21:22:43 <andythenorth> but it requires persistent storage to be any use
21:22:49 <andythenorth> and there isn't persistent storage
21:23:33 <KittenKoder> If they implemented it into OTTD itself, with an advanced setting, it would probably make more sense anyway.
21:24:03 <andythenorth> YACD would be capable of doing it
21:24:06 <KittenKoder> Then it could be done with any vector without the NewGRF being coded differently.
21:24:26 <andythenorth> extend YACD to take account of longitudinal supply / demand
21:24:41 <planetmaker> vectors are not FIRS ;-)
21:24:59 <KittenKoder> I'm still not great with words. :p
21:25:07 <KittenKoder> Industry sets ....
21:25:21 <KittenKoder> What ARE vectors then? In OTTD ...
21:25:31 <andythenorth> they're not an entity
21:25:37 <andythenorth> it's just the name of a set of newgrfs
21:25:56 <KittenKoder> Okay, I newbed ....
21:26:04 * andythenorth ponders how well YACD might work with supply demand
21:26:25 <andythenorth> the problem is that changing supply / demand disrupts routes
21:26:36 <andythenorth> otoh...I've been trying to play with big mixed trains
21:26:59 <KittenKoder> Thus why I would love to play with that feature.
21:27:34 <andythenorth> it's painful without some kind of automatic dispatcher / consist control
21:27:56 <andythenorth> supply/demand worked in railroad tycoon, but only with automatic consists
21:27:58 <KittenKoder> I just think it would be fun, so I recommended it .... :p
21:28:12 <KittenKoder> I never liked Railroad Tycoon.
21:28:25 <KittenKoder> Mostly for the track stye.
21:28:35 <andythenorth> YACD models most of a supply / demand economy, it just misses spot prices for cargo
21:29:43 <planetmaker> g'night andythenorth
21:42:43 <KittenKoder> I hate making third lanes on my rails. >.<
21:47:44 <Rubidium> I'd almost say that your map is too big if you need three tracks next to eachother, but then I remember my old 128x128 savegame
21:48:46 <Rubidium> which has some 4 track bits to cope with the traffic
21:49:23 <KittenKoder> I just wind up putting too many industrial routes through one section.
21:50:49 <planetmaker> though... it needs updating with our other record games.
21:51:18 <planetmaker> (at least records for us - in MP that is sometimes more difficult than in MP)
21:51:19 <KittenKoder> My stations are rarely bigger than four lanes.
21:52:47 <KittenKoder> Even when I was doing 500 train games ... but then, 128x128 with 500 trains would need a track to each train.
21:54:21 <KittenKoder> In this instance I have a machine shop and lime kiln right next to each other.
21:54:36 <KittenKoder> So I put the recieving station between them.
21:55:25 <KittenKoder> ... and both their sources are from the same general area.
21:55:35 <planetmaker> with quite a bit of houses and trains ;-)
21:55:51 <KittenKoder> Now I play using only about 50 trains.
21:56:56 <KittenKoder> Aaah, so city expansion does drain CPU.
21:58:44 <planetmaker> depends a bit on the houses, whether they're animated or not
22:20:28 <KittenKoder> The CG1 is cute, but the horn is grating.
23:12:04 *** Brianetta has joined #openttd
23:40:27 *** Xaroth_ has joined #openttd
23:51:39 *** osaka is now known as woldemar
continue to next day ⏵
