On Tue, 11 Feb 1992 09:55:13 -0800 (PST) you said:
The example you gave isn't a problem if you interpret newlines correctly.
Remember that the *only* Internet standard for newlines is CR LF.
Host B made a mistake by interpreting the newline in quoted-printable as a
local newline. It should not do this unless the data in question has been
definitely identified as textual data (and thus subject to newline
considerations).
To avoid a layering violation, you should keep everything in the CR LF format
and convert newline conventions only at the upper level when you are dealing
with binary vs. text objects. You'd probably only do this for content-type
TEXT.
Believe me, I had to deal with this problem. It is possible to do the right
thing.
This sort of confusion is an example of why quoted-printable is probably
unsuitable for binary data in the general case; not that it won't work but
rather that it is hard to do right.
-- Mark --
Do I understand correctly that the right way to do the transport encoding
depends on the type of the object transmitted ? This is contrary to the
credo I heard before : transport encoding can be done/undone without knowing
anything about the type of the transported object.
Anyway, if doing the right thing is possible, but 'hard', most implementations
will do it wrong (there are enough simple things done wrong to prove this
point). A specification that leads to such a conclusion should be considered
broken in the first place.
/AF