Re: [glob2-devel] Water needed

[Leo Wandersleb]

Kai Antweiler wrote:
> We also could try wells, as a mix of resource and building.
> You must build it.  After that it is only a resource.
> I wanted to do something like that any way for resource transplantation.

Sounds funny. but how about a well to raise fertility of the wheat/wood? now it 
needs water in the proximity and fertility depends on the water found in the 
15-wide surrounding of the resource so a single well interpreted as water would 
barely influence fertility. but if we go for a change in fertility-calculus it 
would be an easy task to better take wells into account:

now to check if a res reproduces, a random field in the 3x3 box with the field 
in question  in the center is chosen. if this field has a res, a second sample 
is taken in the 31x31-box. if this is water, the res grows. this is not a good 
algorithm especially not for wheat.

better procedure: calculate a water-gradient if none already exists (water=255).
now growth takes place if we hit a res in the 3x3 box as before and if 
(syncRand()%1000 < water-gradient*15/255) [numbers chosen to get about the same 
behavior as now]
now it would be easy to simply set water-gradient to 255 in a certain area 
surrounding the well. Also the default falloff from one level to the next could 
be 17 so after 15 steps the gradient is 0 making it easier to regenerate it 
globally if a change takes place.

Next idea for the well: the well is 2x2 and needs workers. they walk in, come 
out 5s after that and splash some water in the near surrounding (walk 5 steps 
randomly and aply their water (blue ball)). Easiest would be to define the 
water gradient by the number of workers in the well both in size as in value 
and not calculate every splashing into the gradient.

Greetings, Leo Wandersleb

P.S.: I'd be interested to help on that or on teleporters.