ietf-822
[Top] [All Lists]

Re: RFC 2046 message/partial (sect. 5.2.2) clarification

2003-06-10 10:20:12

Message fragmentation seems pretty hairy.  Does anything actually use
it?

According to my logs, I've received just 100 fragmented messages in the
past year.

My software silently reassembles such messages. I only become aware that
messages were fragmented at some point when there are reassembly problems. I
don't recall the last time I've had a problem with message reassembly on my
personal setup.

The number of customer questions we get about message reassembly indicates to
me this feature of our software is still in active use at some sites. Probably
not many, but some.

While I would welcome clarifying text about message/partial for a future MIME
update, I do note that there is a big difference between issues that are
largely pedantic in nature and issues that actually cause real world failures.
The operational reality is that message/partial has achieved a significant
measure of operational interoperability with the present specification, warts
and all. And while clarifications are fine, I would be extremely reluctant
to add text that invalidates existing implementations in any way.

                                Ned