xemacs 21.4.19 Mule, problem with abbrev (utf8 instead of iso-8859-1)

[Uwe Brauer]

Hello

This is a odd problem:

I write informatica which should be expanded to 
informática, however
it gets expanded to 
informática  


Edit-abbrevs indeed gives me:

"informatica"  17    "informática"

Which is not the right thing.

So I open my .abbrevs_defs file (which looks ok) the coding system is
set to UTF8, so I set the coding back to iso-8859-1, write the file,
restart xemacs, but no again that problem occurs.

    -  why does xemacs 21.4.19 no work with a UTF8 coding for the abbrevs?

    -  why can't I set the iso-8859-1 coding?

Uwe Brauer