Re: EDITS for Section 10.16 (Content-Location)
Koen Holtman (koen@win.tue.nl)
Tue, 30 Apr 1996 12:58:21 +0200 (MET DST)
Balint Nagy Endre:
>Koen Holtman:
>> Here is new text for the Content-Location section. The change-bars
[...]
>This may cause the false sense of Content-Location applicable only to
>varying resources, however there are good reasons to use it for
>simple URIs:
[lots of URI's which all serve the same thing deleted]
>identify the same resource. (Its worth mentioning that
>internet-drafts have a lot of URIs on different servers, but that
>problem will be addressed by URNs hopely in the near future.) I
>guess the official name of the http server is 'src.doc.ic.ac.uk', and
>clients (and caches) should use that name, but we cant be sure,
>unless the server says that (using Content-Location). Storing under
>12 keys the same resource isn't economic, we need a possibility to
>spare that disk space!
While this is a big problem, I do not see how it can be solved by
Content-Location headers. Until we have URNs, we can only hope that
HTML authors will generally all use the same link to only one of the
URIs above, so that caches won't end up having 12 identical copies.
Using Content-Location to collapse the 12 identical copies into one in
an un-spoofable way seems like a far-fetched idea for me.
If the Content-Location text makes one sense that Content-Location
cannot be applied to solving the 12 identical copies problem, I
consider that to be a feature, not a bug, of the text.
>Andrew. (Endre Balint Nagy) <bne@bne.ind.eunet.hu>
Koen.