TTS , transport tycoon clone
Yeah the TTD AI is pretty shocking, but at least it goes some way to preventing compete world domination. It would be nice to have some challenge in single player games, otherwise after 100 years you have pretty much everything on the map covered and what's left to do? It would be good to have some more intelligent AI which could provide some (semi) serious opposition to a player's company.
An AI that's not too obvious in it's decisions and that would maybe even consider going RORO after some time would be a nice change
But probably quite hard to program
But probably quite hard to program
GoneWacko. Making [url=irc://irc.oftc.net/tycoon]#tycoon[/url] sexy and exciting since 1784.
That's the problem. It's very easy to program Artificial Incompetence. TT stands as perfect evidence of this ("I know! I'll avoid the sensible option and zig-zag across the landscape instead!"). There's a reason they don't teach us AI until the third year of the degree.
Bugzilla available for use - PM for details.
Then you're being robbed of the chance of dropping it
(Then again, our AI course allegedly has too strong a focus on PROLOG. In fact, from what I'm told by previous students, the "AI" module was more-or-less a course in programming in PROLOG.)
On a further note, do you have plans and designs, or are you working in the XP (extreme programming) manner of "sit down in pairs and get coding"? Do you have your strategies for such things as i18n and l10n yet (e.g. how they will be achieved within the code - gettext, includes, etc.)? Then there are a11y issues - that authentic font is pretty nice, but at minimum zoom the text gets a little small (of course, part of that was down to fitting everything into 640x480px).
It would also be interesting to see how this turns out with the initial demos, not to mention how many holes people can pick in it. I certainly hope it won't be "bug-compatible" with TT
Just as a side note - ever set the opposition construction speed to "Very Fast"? Give them a year or four and they're all over the map, with dodgy airports a plenty, and a plethora of retarded railways.
(Then again, our AI course allegedly has too strong a focus on PROLOG. In fact, from what I'm told by previous students, the "AI" module was more-or-less a course in programming in PROLOG.)
On a further note, do you have plans and designs, or are you working in the XP (extreme programming) manner of "sit down in pairs and get coding"? Do you have your strategies for such things as i18n and l10n yet (e.g. how they will be achieved within the code - gettext, includes, etc.)? Then there are a11y issues - that authentic font is pretty nice, but at minimum zoom the text gets a little small (of course, part of that was down to fitting everything into 640x480px).
It would also be interesting to see how this turns out with the initial demos, not to mention how many holes people can pick in it. I certainly hope it won't be "bug-compatible" with TT
Just as a side note - ever set the opposition construction speed to "Very Fast"? Give them a year or four and they're all over the map, with dodgy airports a plenty, and a plethora of retarded railways.
Bugzilla available for use - PM for details.
My father is (was) an expert in artificial intelligence and expert systems, not as applied to games, per se, but still relevant. I strongly suspect that the "clone" AI will take enormously longer to develop than the game proper, if it ever does appear. It seems to me that a "good" AI that we would all accept, is beyond the capabilities of the available resources, if not of the available art, considering the enormous number of variables involved in running a business.
Development Projects Site:
http://www.as-st.com/ttd
Japan, American Transition, Planeset, and Project Generic Stations available there
http://www.as-st.com/ttd
Japan, American Transition, Planeset, and Project Generic Stations available there
Have you noted that we're programming an engine here atm, so none of the internationalization, localization and accesibility (I translate the abbrevations for those who don't know what they are) are applicable yet.ChrisCF wrote:On a further note, do you have plans and designs, or are you working in the XP (extreme programming) manner of "sit down in pairs and get coding"? Do you have your strategies for such things as i18n and l10n yet (e.g. how they will be achieved within the code - gettext, includes, etc.)? Then there are a11y issues - that authentic font is pretty nice, but at minimum zoom the text gets a little small (of course, part of that was down to fitting everything into 640x480px).
Since our move to C++ we're doing much more strategy than before (the project began in the XP way, as you call it). We don't live anywhere near eachother so we can't just go and visit eachother, most communication goes over the internet.
The move to object orientation also made it possible to use a programming strategy called 'design by contract' (you might have heard of it), allowing us to use none-finished pieces of code already (In fact, we could write contracts for all classes we intend to use, and everyone could already begin extending the engine, even though it isn't implemented).
At least all known bugs in TT(D) will be fixed.It would also be interesting to see how this turns out with the initial demos, not to mention how many holes people can pick in it. I certainly hope it won't be "bug-compatible" with TT
I did, but they usually end up with huge loans and not making very much profit...Just as a side note - ever set the opposition construction speed to "Very Fast"? Give them a year or four and they're all over the map, with dodgy airports a plenty, and a plethora of retarded railways.
"Peace cannot be kept by force. It can only be achieved by understanding." - Albert Einstein
Not yet, but you will at some point have to consider it. It's that much easier to consider it before you reach a point where you need to incorporate it Anyway, I have registered at your forums, so I will probably continue this particular matter there, as time progresses.TBOT wrote:Have you noted that we're programming an engine here atm, so none of the internationalization, localization and accesibility (I translate the abbrevations for those who don't know what they are) are applicable yet.
Nice to see that there is some thinking going into it. It's annoying when everyone looks forward except the developers (I've made that mistake myself - it hurts )
... and a long-ish wait before the land is freed up again. I managed to get an AI company making £x00,000 per year profit, so I had to sabotage them, and wait. Then wait some more. SighI did, but they usually end up with huge loans and not making very much profit...
Bugzilla available for use - PM for details.
Well so I guess AI is not really likely for TTS. No worries on my behalf, I guess with a 65k x 65k map it is going to take a good long while before players run out of things to do - no real need for any competition to slow us down!
Actually that brings me to another point.......what is the distribution of industries/towns going to be like? Similar density to TTD (that will mean 1000s of them!) or more spread out?
Any opinion on this one tycooners?
Actually that brings me to another point.......what is the distribution of industries/towns going to be like? Similar density to TTD (that will mean 1000s of them!) or more spread out?
Any opinion on this one tycooners?
- spaceman-spiff
- Retired Moderator
- Posts: 20634
- Joined: 28 Jul 2002 07:08
- Location: Belgium
- Contact:
Let's hope it doesn't put a factory two blocks away from a farm and gives subsidies every 2 years
I never use them because it's too close
Perhaps we should have a parameter that holds the min. number of tiles their have to be between tied industries !!!
For me, I have no problem with two similar industries close together
I never use them because it's too close
Perhaps we should have a parameter that holds the min. number of tiles their have to be between tied industries !!!
For me, I have no problem with two similar industries close together
Well, of course it would be nice if the spread is a bit better. But it might create unbalance at the start of the game, since you would have to build enormous tracks (Especially in a mountainous (or extreme, which might be a TTS option ) terrain it can be very expensive).
"Peace cannot be kept by force. It can only be achieved by understanding." - Albert Einstein
The spacing-of-similar-industries idea would work. I'd think it would be feasible to define two or three different figures of the sort for each of the number of industries offered in the difficulty settings, as the harder difficulties are probably meant to have the linked industries spaced out a lot.
Grunt
(aka Stephan Grunt, CEO of Grunt Transport Inc. since 1994.)
(aka Stephan Grunt, CEO of Grunt Transport Inc. since 1994.)
Suggestion
One thing that I think could be a good idea, for when the project is in a state for it to be opened up, and ready for code from other people, is having a "plugin" style approach to the AI players.
I'd be quite keen to have an attempt at writing an AI player implementation without having to be completely in-control/in-charge of that part of the project. If the program was structured in a way to allow an API to be created for controlling AI players, then no one person would have to be.
It would also have the potential of making the game take on another aspect for people who are interested in the programming side of things. Instead of just playing the game, people could be writing programs that play the game, A LOT like c-robots style games work. If it was possible to load/run more than one implementation of a AI player at once, then you could play different people's implementation against each other in competition.
Just a suggestion to think about. I really hope you succeed in this project, one of the things I like most about TT and keep me coming back to it rather than running other train games is the graphics, an well structured open source version of the game would be a awesome thing.
I'd be quite keen to have an attempt at writing an AI player implementation without having to be completely in-control/in-charge of that part of the project. If the program was structured in a way to allow an API to be created for controlling AI players, then no one person would have to be.
It would also have the potential of making the game take on another aspect for people who are interested in the programming side of things. Instead of just playing the game, people could be writing programs that play the game, A LOT like c-robots style games work. If it was possible to load/run more than one implementation of a AI player at once, then you could play different people's implementation against each other in competition.
Just a suggestion to think about. I really hope you succeed in this project, one of the things I like most about TT and keep me coming back to it rather than running other train games is the graphics, an well structured open source version of the game would be a awesome thing.
Yes, AI plugins would be cool. Instead of having competitions between ourselves, we could have competitions between our personal AIs...
Development Projects Site:
http://www.as-st.com/ttd
Japan, American Transition, Planeset, and Project Generic Stations available there
http://www.as-st.com/ttd
Japan, American Transition, Planeset, and Project Generic Stations available there
-
- Tycoon
- Posts: 3849
- Joined: 20 Jan 2003 14:51
- Location: Broadstone, Dorset
- Contact:
Personally I am most grateful to Josef for the wonderful Patch, and without him none of this would be here. But, I think once TTSD is finished and running, then there really isn't any reason to continue development on the Patch or anything else connected with it - all the efforts can go into extending TTSD, which will be vastly easier since it was designed to make that possible.
Development Projects Site:
http://www.as-st.com/ttd
Japan, American Transition, Planeset, and Project Generic Stations available there
http://www.as-st.com/ttd
Japan, American Transition, Planeset, and Project Generic Stations available there
There will still be those people out there that don't like TTSD for whatever reason and will continue to use TTDPatch probably long after support for it has been discontinued. Therefore someone will, at some point, pick up development for TTDPatch and add a bunch of new features into it that weren't around at the time it stopped being made...
But I'm talking as if I'm about ten years in the future, and that's not important right now.
But I'm talking as if I'm about ten years in the future, and that's not important right now.
Grunt
(aka Stephan Grunt, CEO of Grunt Transport Inc. since 1994.)
(aka Stephan Grunt, CEO of Grunt Transport Inc. since 1994.)
Who is online
Users browsing this forum: No registered users and 4 guests