Re: I-D ACTION:draft-ietf-pkix-certstore-http-01.txt

[Peter Sylvester <Peter.Sylvester@xxxxxxxxxx>]

> >>This is acceptable, but I prefer one that is more straightforward to implement
> >>with web server tools.
> > 
> > That's probably the best argument for choosing MIME multipart/RFC 2585 rather
> > than a SEQUENCE OF, the server shouldn't need to do anything more specialised
> > than "fetch value based on key, via HTTP".  Any special-case processing can be
> > done by the client.
> 
> 
> I don't agree. Complexity should be put in the server, not the client. 

Another argument may be that using a 'simple' standard developement kit
for a server to create unnecessary overhead for a client (and useless
data) instead of having a small and fast logic may result in
having all kinds of possible and undesirable errors and protocol messages 
above HTTP/MIME to escape at some point.