Re: [glob2-devel] flags vs. areas: getting the best of both worlds

[Joe Wells]

Kai Antweiler <address@hidden> writes:

>> Currently, it looks at a (somewhat) circular region.  This would
>> simply change to looking at the clearing cells it governs.
>
> Btw: I think flags should be squares.

And/or diamonds.  Circles suck for clearing flags.  Circles are
probably fine (and maybe even better) for war and exploration flags.

> * This would benefit your proposal, because you will have a better
>   feeling what "radius" means in glob2.  And you want the radius
>   to have an effect on the cell selection. 

I'm not sure this makes a difference.

> * Clearing flags as they are today, always cut out too much resources
>   in the middle and to few at the ends.  A square would be better.

Agreed completely!  Switching clearing flags to be squares or diamonds
would be a big improvement on the current ones.

> * Exploration flags: a 45 degree rotated square would allow explorers
>   to cycle faster.  But maybe circles look better.
> * War flags: I don't know.

It is not clear to me whether or not circles are better for war and
exploration flags.

>> How do the globs find their job right now?
>
> I don't know.
>
>>  I am assuming they are given the destination gradient for the flag.
>> We just change that destination gradient to raise the clearing cells
>> the flag governs that happen to currently have clearable resources
>> on them, and we are done.
>
> Sounds good.  I have encountered something called Building Gradient.
> But it didn't consume a lot of cpu.  So if it does what it suggests
> there must be a trick - like recompute only every 10 seconds.

Hmm.  I was thinking that my proposal could take advantage of the
gradients adjusting semi-rapidly to the clearing of the resources.
But maybe whatever happens now (if not based entirely on gradients)
could be adjusted to the new non-circular areas instead.

-- 
Joe