you may be correct about the unwillingness of implementors to change.
however, any new encoding scheme would bear the same burden, and would
further increase complexity due to the need to support both 2047 and the
new scheme.
Jean-Marc Desperrier said:
Keith Moore a écrit:
In my experience the problem with RFC 2047 implementations is usually
that implementations try to "decode" RFC 2047 for the purpose of
internal representation and then to "encode" again for retransmission,
rather than to "decode" RFC 2047 only for the purpose of display.
Unfortunately, this is the way most work and not likely to change.
If I were to write the ideal specification for the developpment of a new
news/mail reader, yes, it would have exactly the behaviour of keeping
the subject intact by not decoding/recoding it.
Unfortunately most developper do not see the advantage of it, and do not
want to make the extra effort this means.
Being able to do it requires to test if the subject of the reply has or
has not been edited by user, and reusing the older subject if there was
no change.
This is extra code to circumvent the problems with RFC2047, and does not
even solve everything, even if it would reduce the worst cases.