Re: Important: API Change Questions

On Mon, 2004-04-19 at 13:06 -0700, Adam Megacz wrote:
> See the threadwise great-great-grandparent of this message.

Well, um, trapping distanceto does not suffer from the same problems as
PosChange:

  "But PosChange never really worked right (for example, if you move
   to the left ten pixels and your parent moves to the right ten
   pixels, you won't get a PosChange)..."

With trapping distanceto, that's not an issue as you'd trap distanceto
between parent and child in that instance.

  "...and making it work "right" imposes a pretty huge performance hit."

This can be done much better with distanceto where you can simply track
the boxes that are involved in distanceto traps on a box-by-box basis.

  "Plus even with a lot of extra bookkeeping I'm still not really sure
   how to do it without breaking the perfection of the "fake root"
   emulation."

Again, this is not an issue with trapping distanceto, because you can
trap surface.distanceto which is directly supported by fake roots.

> Repeating myself like this is why I'm really, really angry about the
> volume of your postings.

You didn't even consider using distanceto properly enough.  What I've
written above is nothing you couldn't have garnered from a few minutes
of reflection on the issue.

But your time is far more valuable than everybody else's, apparently.
-- 
- Charlie

Charles Goodwin <[email protected]>
Online @ www.charlietech.com