[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
I agree with John (and thus with Neil's concerns) on this point. If we're
going to add information to an existing command, let's add it as early
as possible in the per-message transaction -- specifically, on the
MAIL FROM command.
This has several advantages. First, the information is known early enough
so that per-recipient size limits can be enforced. (I cannot speak for
other implementations here but I know that this will be easy to implement
in PMDF. Each recipient is mapped to an associated channel/transport
as it is accepted, and that's where size limits are also expressed.) Second,
getting the size information away from the DATA command encourages people NOT
to implement any kind of fancy "skip until we're near the end" stuff as a
performance improvement. Third, it doesn't clutter up the DATA command's
parameter space, so that space is available for things related specifically
to the transfer operation (like compression and dot quoting). Message size
is largely orthogonal to transfer issues -- even counted transfer facilities
don't necessarily relate to total message size in any direct way.
More to the point, does anyone see any significant disadvantage to adding
this information to the MAIL FROM command?