> Talking about cache and replication, it would be nice also to have more
> and more server admins doing "the right thing" about Cache-Control
> headers, it would avoid dirty hacks to see if a resource can be cached or
> not...
OK, this is getting even *more* off-topic (sorry!), but I've been working on
a study of web server capabilities for a little while, and it may be of
interest in this regard. It does things from the other end; instead of doing
trace analysis, etc, it breaks down popular servers into types of _content
generators_ that they use, and tests them for protocol compliance.
It's still quite rough, and not the most rigerous study you'll see, but have
a look at:
http://www.mnot.net/papers/capabilities.html
(and I apologise in advance for the PDF - I know the type is really small!)
Where is this leading? I'm starting to put together an (even more embryonic)
I-D that deliniates server-side roles in regard to responsibility for HTTP
compliance. The argument is that people aren't just using static files to
manage content any more, and the protocol will hit a wall if it isn't
accounted for in servers.
Would this be wrec's area, http-wg's, or somewhere else? Any interest out
there?
-- Mark Nottingham Melbourne, Australia http://www.mnot.net/
This archive was generated by hypermail 2b29 : Thu Nov 18 2004 - 11:21:26 MST