Re: invisible

[Richard Stallman]

    > The feature that Stefan implemented for moving point out of invisible
    > regions doesn't do it in quite the right way.  It needs to move point
    > further in the same direction that it moved in order to enter the
    > invisible region.

    The code I wrote is supposed to behave exactly like you suggest.
    If it does not, please report it as a bug.

I recall reading the code and getting the impression that it
implemented some other behavior.  But that was a long time ago.
I'll take your word for it, and we can fix it if it doesn't
entirely work.

What about the other point: that this behavior should make it easy
for programs to avoid situations where two different point values
correspond to the same location on the screen?