Re: undo bug?

[martin rudalics]

> I'm wondering why we have to use last_point_position.
> It seems to be asking for problems because it is delimited by command
> processing, whereas the undo records are delimited by undo-boundary:
> there is a correlation between the two, but no equivalence.
>
> I.e. I suggest we introduce last_undo_boundary_pos and use it in place
> of last_point_position in undo.c.
> It'd be set to PT in Fundo_boundary (which could/should also set
> last_undo_buffer).

Couldn't we do away with undo-boundaries and always undo until the next
recorded position of `point'?