Re: [MIT-Scheme-devel] a MIME body structure parser for non-IMAP folders in IMAIL

[Taylor Campbell]

   Date: Tue, 29 Nov 2005 01:06:02 -0500
   From: Chris Hanson <address@hidden>

   On 11/28/05, Taylor Campbell <address@hidden> wrote:

   > OK, thanks.  (Is it a bug that string ports disregard line ending
   > normalization settings?)

   I guess it's a bug.  The theory of strings is that they are ISO 8859-1
   encodings with normalized line endings.  In practice, though, they are
   sometimes treated differently (for example as UTF-8 encodings), and
   the "theory" is just the default state when you have no additional
   knowledge about the string's interpretation.  Consequently, it makes
   sense for string ports to support translation when a string is used in
   a non-default way.

   Probably I was thinking only of the theoretical model when I
   implemented string ports.

   I'll make this change.

I see that you've changed the relevant files to make string I/O ports
use the generic I/O layer, but it doesn't seem that they actually
exploit the layer: the GENERIC-NO-I/O-TYPE has no operations that
actually affect the I/O itself, only those that affect the (ignored)
coding & line ending settings.  Am I missing something, or was this a
big think-o?