Re: Rmail and the raw-text coding system

[Mark Lillibridge]

Stefan wrote:
>  I (Mark) wrote:
>  >     It then decodes the BABYL message part:
>  
>  > ...
>  
>  > This process leaves the buffer as a unibyte buffer.  
>  
>  The question for me is why did it choose raw-text here (which results
>  indeed in a unibyte buffer)?  It should have been emacs-mule.

    I assume it did so because the buffer contained "invalid" code
points.  Remember that loading raw-text then converting to multibyte can
(I believe) produce a buffer with essentially arbitrary bytes modulo no
unaccompanied continuation bytes.  Presumably, the result can be
considered invalid by emacs-mule.

- Mark