Re: Problem report #53 -- real problem

[Richard Stallman]

Can people please try to help with this?

    Is it right to assume that XINT (pos) is always >= BUF_BEGV (XBUFFER 
(buffer)) ?
    If yes, then this problem report is FALSE.

This is a real problem.
Currently it IS possible to call text_property_stickiness
with a POS value that is before BEGV.

text_property_stickiness is called from get_pos_property,
which is called from find_field, which is called from
various user-level functions in editfns.c.

I made sure the other callers can't call it with a POS value before BEGV.
But this one case is hard, and I am not sure what is right to do here.