|
![](/i/fill.gif) |
"Greg M. Johnson" <gregj:-)565### [at] aol com> wrote in message
news:3dbd7276$1@news.povray.org...
> "Mark Hanford" <mar### [at] the internet com> wrote in message
> news:3dbd668a$1@news.povray.org...
> > I might have a go at implementing a version of the cubic space,
but
> > this causes the calculation time to increase HUGELY. It would
create
> > some great patterns, though ;)
> >
>
> I tried it in 3d but the cells kept marching in one direction. I
know that
> this can happen in 2d but I couldn't swear that this wasn't a coding
error
> due to mislabeling of all the nearest neighbors...
>
I think I might give it a go, just for the exercise. My current algo
uses the process similar to
CellCount = CellCount+(Cell1=1)
CellCount = CellCount+(Cell2=1)
...
CellCount = CellCount+(Cell8=1)
I shudder to think what my current strategy would look under a 26
neighbour system. Especially as each of the above checks also sees if
it should wrap the edges with
"If LeftEdgeKills then skip Else CellCount=CellCount+(Cell1=1)"
for all the Row=0,Col=0 cells...
I might have to do some real work, and look at other "proper"
implementations of these things. So far I've kind of just made it up
as I went along. Oh, and if anyone's got a new P4-2.8GHz spare, hand
it over...
--
Mark Hanford
Putting POV code on the web? Make it look pretty with POV2HTML...
http://www.hanfordonline.co.uk/povray
Post a reply to this message
|
![](/i/fill.gif) |