Re: [Rails-devel] Automatic route calculation

[alexti <al...@sh...>]

Hi Stefan,

I like your idea. It makes sense that artificially complicated track  
layout with lack of station markers in a simpler game may create more  
difficult problem than the realistic scenario in a more complex game will.

If you are volunteering to create such setup for 1870 I can try to build  
similar scenario for 1856 (with couple of diesels, for example). In what  
form can you export such setup? Something simple (for example  
comma-separated tile ids) will do. I won't have time to do it until the  
weekend though...

Alex.

On Tue, 23 Mar 2010 16:35:52 -0600, Stefan Frey <ste...@we...> wrote:

> Alex:
> Some more comments on a potential test for automatic route calculation:
>
> I think there is no need for one of the more "exotic" types (18US, 18C2C,
> 1844) to create a (reasonable) difficult scenario.
>
> It is easy to create a quite involved track layout with maps and tiles  
> of one
> of titles already implemented in Rails (for this even the partially
> implemented 1835 or 1870 would work).
>
> Then rails could supply:
>
> * A map (here I mean the collection) with map hexes and tiles. From my  
> point
> of view it is still open, how it is converted to the graph you need to  
> run
> your algorithm.
>
> * The trains available to the one company could be easily changed. And  
> we can
> make things more difficult by requesting scenarios, which are not  
> possible in
> 1870: For example running a  8, 10 and 12 at once.
>
> * I think, that there is no need for tokens so far (either simply assume  
> that
> there are none (which would allow even more connections than usual) or  
> create
> a network that is usually available to one company after considering the
> effects of tokening, if we want to built a more realistic test case).
>
> If no other wants to jump in, I would volunteer to create such a network  
> on
> the 1870 map.
>
> Stefan
>
>
>>
>> >> Perhaps it can be tested? There is no need to have complete support  
>> of
>> >> such games in Rails to make experiment. We could create a graph
>> >> representing "difficult case" and run algorithm on it.
>> >
>> > Yes, though it could take time to manually build the graph the way  
>> your
>> > algorithm wants it.
>> >
>> >> > Some I would look at would be 18US, 18C2C (for sheer size and the
>> >>
>> >> ability
>> >>
>> >> > for a company to run lots of large trains), 1844 (which adds  
>> tunnels
>> >>
>> >> and
>> >>
>> >> > mountains that affect the route score), and 1860.
>> >>
>> >> Unfortunately, I don't own any of them :(
>> >
>> > The complete rules are available for at least 18US and 1844, and  
>> ps18xx
>> > includes the map/tiles for all of them.
>>
>> Are there some cases you would consider examples of "difficult" ones  
>> that
>> are available in some kind of format? From some PBEM games perhaps? It
>> might be relatively easy to convert them into the graph I need. And it's
>> difficult to get an impression of what realistic end game layout would  
>> be
>> without playing the game.
>
> ------------------------------------------------------------------------------
> Download Intel&#174; Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Rails-devel mailing list
> Rai...@li...
> https://lists.sourceforge.net/lists/listinfo/rails-devel


-- 
Using Opera's revolutionary e-mail client: http://www.opera.com/mail/