On Fri, 16 May 2003, Oskar Batuner wrote:
1. Better chance for acceptance. As you told yourself
"... ICAP and HTTP folks would feel very comfortable".
And the protocol acceptance depend on those folks, so
we should help them, not parsers :).
I agree, though the situation is not black-and-white. We may be making
other concessions to keep ICAP/HTTP folks happy. Adding a number in
front of a header does not sound like a major change.
3. Optimized MIME is not radical enough to justify it's
That could be true. Let's postpone the final judgment until the first
draft that will make all optimizations known. Also, this sounds like
an objection to NetStrings-like approach for headers, not other MIME
optimizations. Perhaps we can use some optimizations but not the
As for optimization - short aliases you've mentioned in another
message looks like a very good idea. It helps with all headers (not
just those one wants to ignore) and may result in much bigger
overall saving. And the best thing in it - it is optional!
That is also the worst thing about them -- you cannot force the peer
to use them and, hence, cannot guarantee any savings for yourself
(other than having to write shorter strings). It would be interesting
to know SIP folks experience with that optimization.