[Freeciv-Dev] Re: Comments on CMA 2.6

[Raahul Kumar <raahul_da_man@xxxxxxxxx>]

--- Christian Knoke <ChrisK@xxxxxxxx> wrote:
> Am Mittwoch, 24. Oktober 2001 12:35 schrieb Raahul Kumar:
> > --- Raimar Falke <hawk@xxxxxxxxxxxxxxxxxxxxxxx> wrote:
> > > > > No why should it? Which one should CMA use if you have two
> > > > > combinations which only differ in food. It chooses the one with
> > > > > the bigger food value.
> > > >
> > > > You won't like the answer: It should choose the one with the
> > > > lower value and leave the better tile for another city.
> >
> > I must have been asleep. Why should the CMA leave the better tile for
> > another city? At this stage we do not know if there is even another
> > city. If I missing some prior details set me straight.
> 
> The player may not want to get the maximum availlable, e.g. food,
> when he doesn't want his city to grow, he decides to grow later
> (by rapture) or because the city size is 8 and there is no aqueduct.
> In these cases he may want to have some good food tiles availlable
> for other cities.
> 

With the current interface I could set the food surplus to 0, and possibly
get the same result. Raimar, am I correct? I suspect that if it comes to 
choice between two tiles, Tile A has 2 food, 1 shield, Tile B has 1 food, 1
shield, the CMA will always pick Tile A.

That is a really hard design decision to code around, because the minor fitness
algorith trys to optimise the secondary attributes after the primary reqs are
met.  I fail to see how Raimar can code around this issue. In fact, do NOT.
Leave the allocation of tiles algorithms as simple as they can be.

I think this problem calls for either the overlapping tiles agent, or b)
The Hack. There is a need for priority ranking of CMA cities. The reason
for that should become clear.

The Hack consists of
does_city_radius_overlap
if other_city who radius overlaps is not a CMA city(What do we do if it is a
CMA city? Some criteria for ranking one CMA city ahead of another is needed)
The user sets want_to_manage_tiles(This is an option in the agents dialog)
While that flag is true, the CMA allocates no overlapping tiles. After
the flag is set to false the CMA can allocate tiles that do overlap that the
other city did not grab.

Basically, the CMA allocates overlapping tiles after the human player has
allocated overlapping tiles. Raimar, if there are overlapping tiles between
two CMA cities, which city wins? I believe this a basically a question of which
one was founded first.


> >
> > > I have never thought about this. A very good argument. I will
> > > probably implement a weight of 0.
> 
> Thinking about it, I'm not sure if this is sufficient: the CMA
> may decide to use the good tiles nevertheless? 
> 
> > >
> > > > A ratio 2:3 will give you the same result as 3:4 or 4:5, if all
> > > > other stats are zero. I think this is more transparent for the
> > > > player.
> > >
> > > I don't think so. In my solution 2:3 would yield the same as 4:6 or
> > > 6:9. Does anybody have a third opinion?
> >
> > Raimar is obviously correct. In fact a ration of 3:4 or 4:5 is 0.75
> > and 0.8 respectively. They're not even the same as each other.
> 
> Well, you missed the point, that there is a logarithmic scale. So this
> is about a comfortable user interface.
> 
> With my suggestion you can have less steps (six) within the slider, and
> a finer adjustment at the same time.
> 

I'll give you the fewer steps. It is obvious that you can remove ratios
like 6:9 8:12 etc if you have a 2:3 ratio. Your idea that
a ratio of 3:4 and 4:5 is equivalent is one I dislike. I like being able
to micromanage.

 

> 


__________________________________________________
Do You Yahoo!?
Make a great connection at Yahoo! Personals.
http://personals.yahoo.com