Re: `insert' deactivates active region

[Clemens Fischer]

* Richard M. Stallman:

      Stefan Monnier wrote:
>     I do think it is a bug.  The problem is that deactivate-mark is global
>     whereas mark-active (which is deactivated through the deactivate-mark
>     variable) is buffer-local.
>
> It is not a bug, but it could be a misfeature.
>
> However, changing it would be risky--other things might break.  I think it
> is safer to keep this unchanged.

excuse me, sir, but i do agree with stefan.  it doesn't make any sense to
interlock editing operations in *different* buffers by way of the global
"deactivate-mark".  yamaoka san had to work around this and solved my original
problem.

in a read-only buffer the mark was lost when anywhere else an unrelated buffer
was modified!

  clemens