🚀 go-pugleaf

|In article <1galsqINNr88@gap.caltech.edu> jack@ccsf.caltech.edu (Jack Stewart) writes:
|>As I understand, the main reason for MIME to use base64 encoding (as
|>opposed to uuencode, 8bit transparent, XDR, or BER ) was that they
|>needed an encoded scheme that would remain uncorrupted accross all
|>platforms and architectures and existing mailers.  Apparently this is
|>a real problem with IBM VM/CMS machines.  base64 encoding was the only
|>thing that would survive these machines.  Hopefully these machines
|>will go away some day and then we can use a more reasonable encoding
|>mechanism.

ag129@cus.cam.ac.uk (Alasdair Grant) writes:
|Yes, but this is a minor problem compared to the presence of machines
|which use the 7-bit American ASCII code, where the top bit is a parity
|bit.

  Since no-one buys GCOS anymore, who cares? (half a (:-)).

  Seriously, though, there are very few machines still in existance which have
a byte size of 6 bits or less.  Given that common communications channels
are 8-bit, it is reasonable to expect that an 8-bit-clean smtp path and
an 8-bit-clean file system will allow the elimination of 7-bit ascii as 
soon as it's worth the vendors' while.  16-bit is the real problem.

--dave 
[ps: Honeywell, IBM 7070s and DEC-10s ran to 36-bit words, and tended
 to use 6-bit almost-ascii codes. Honeybuns actually could store
 strings of 7-bit characters, assuming you really, truly, wanted to.
 Sane honeybuns ran Multics, however, and used 9-bit bytes. --dave]
-- 
David Collier-Brown,  | davecb@CCS.YorkU.CA | lethe!dave
72 Abitibi Ave.,      | 
Willowdale, Ontario,  | York Postmaster and
CANADA. 416-223-8968  | occasional sendfail(8) consultant.