On Dec 19, 2007, at 1:29, Aaron Stone wrote:
I have read all of the responses to my earlier post about UDP / byte
ranges, and have done my best to incorporate those responses into this
first pass at documenting an extension for UDP and byte ranges:
http://code.sixapart.com/svn/memcached/branches/binary/server/doc/binary-protocol-plan-udp.txt
I'm going to hack some more on this tomorrow, and will try to reply
more
thoughtfully to all of the responses I received today. Thank you all
very much!
I don't think it makes sense to define any values that belong within
the binary payload in the UDP document (specifically response status
and command codes). This can lead to collisions in other documents.
That would be bad.
I think it's best for us to think of UDP as just a transport and only
define the minimum required for it to work. With that perspective, I
don't see what's missing from the UDP transport that's already in use.
--
Dustin Sallings
