 |
|
||
| Date: | 1993-11-02 | ||
| From: | Paul | ||
| Subject: | RE: Not So Smooth Sailing |
I'm communicado in Las Vegas. I'll be checking my mail at DEC every other day, more or less.
In partial reply to your questions:
The sparklist can get pretty big, but not immense. The 68000 architecture limits segment size to 32K bytes (each spark is currently two bytes - a X and a Y value), but this is probably not absolute, as I can allocate memory from the heap at runtime. In fact, this is the way C++ prefers to operate. So go wild! While I don't have the code in front of me, I believe the answer on shoreline is similar.
As for dealing with "discontinuous shoreline space" - whatever that means - I'd figured on a combination of macro and micro coordinates, allocating space for micro coordinates only as the need arises. Pure land and Pure Sea cells would not need micro coordinates; shoreline cells would. We discussed the math on this some months back, and I've done no refinements since then. I *may* have preliminary data structures in my hotel room; if so, I'll get back to you with what I've got.
And *nothing* is cast in concrete. Suggest a "better" way to deal with smoothing and I'll happily adopt it.
Duk