ietf-822
[Top] [All Lists]

Re: Canonical Encoding Model wording suggestion

1995-06-02 14:41:19
< hansen(_at_)pegasus(_dot_)att(_dot_)com writes:
<< I agree with all of this except the last sentence. The MIME systems I
<< deal with that use NL conventions all do it within just the text body
<< parts. Hence, your statement is not true for such systems.

< You'll have to describe in more detail what it is you're doing.  What
< exactly is a "text body part"?

Any body part whose content type says "text":

    Content-Type: text/*

< The systems I've had described to me which do something like this are
< incompatible with MIME.  That is because what is or is not "text" is not
< deterministic and a reader cannot always tell whether or not a 0a octet
< maps to a 0a or to a 0d0a in the canonical MIME form.

If the content type says it's text, then you can make assumptions about it.
If it doesn't say it's text, then you can't, so you have to leave it alone.
(If there is no content type, you can assume it's text and treat it
accordingly.)

                                        Tony Hansen
                          hansen(_at_)pegasus(_dot_)att(_dot_)com, 
tony(_at_)attmail(_dot_)com