"ingo" <nomail@nomail> wrote:

> The print factory where I worked (solvent based rotogravure printing) was
> sectioned. Huge safety doors would close of sections in case of emergency.
> People would try to outrun the closing of these doors, the building is 250 m
> long with 5 sections. Not a situation you want.

Jeez.  Sounds like that scene from Moonraker.

> To get proper routes out we added more outside gathering points for
> people(counting). Maybe adding an extra gathering point (green dot?) on the
> other side of the building works?

I was just trying to make the solver take as long and convoluted path as I
could, and the algorithm that I have now only takes a start and target point as
input.
I think I can sequentially solve for different paths and store those in
different path number array, but at the present time, that's about it.

I was just trying to conceptually figure out how to code a 1-way door

After that, I figured some thing to try might be:
  solve a single designated path
  solve multiple single paths
  solve a progressive path with sequential destination points
  solve a single path then repeat using a given obstacle
  find the closest destination of many to a given starting point

But mostly I was just thinking that I might have to use different versions of
the map, or someone might want me to test this on another site, and I wanted to
minimize the amount of work I might need to do to start having the algorithm do
its thing.

I tried loading yesbird's povlab website to see if I could use that to speed up
the parsing, but it's in a "forbidden country"   <eyeroll>

- BE

Post a reply to this message