Re: Comments on Byte range draft
Jeffrey Mogul (mogul@pa.dec.com)
Tue, 14 Nov 95 11:13:49 PST
I agree that there should be some *opaque* string used to select if the
object is the same or not (string which could be for instance a last
modified date, an MD5 digest,... whatever the server wants)
And the client could for instance blindy use what the server sent as
Content-Digest: (for instance, but we could use a different name if that
one is 'burned' ;-) )
I would recommend against using the Content-Digest as the
cache-authenticator. This means that if a server wants to control
caching, it is forced to generate a Content-Digest that is also
semantically correct for whatever other uses a Content-Digest
is useful for.
In short, Content-Digest is NOT opaque to the client.
It also means that the server must either re-digestify the object
to do a validity check, or keep a database of Digest values.
If the server wants to use a simpler scheme (such as file modification
time) to generate a validator, it should be able to do so.
-Jeff