httpd-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Roy T. Fielding" <field...@kiwi.ics.uci.edu>
Subject Re: the Twun (progress)
Date Mon, 05 Apr 1999 06:59:40 GMT
>Hmmm, I'm not sure about that. If I include a
>  Cache-Control: must-revalidate,max-age=0
>in the response, the proxy-cache is forced to do revalidation. Which
>etag will the proxy use for the revalidation? The one from its cache entry or 
>the one from the new request? I suspect the latter one, but maybe I've missed 
>something in the actual HTTP/1.1-draft. In this case it seems to me that the 
>trick will work.

That depends on the implementation.  Given that very few proxies currently
implement cache-control, let alone etags, the behavior will be random.
In the long term, it will probably be both the etag in the request and
the cache's etag, since you can send both and determine what was active
based on the response.

But this whole idea is totally ridiculous -- you can achieve the same result
far more reliably and easily by adding a user identifier to the URL,
just like quicken.com, and that will work with any browser.

In any case, we sure as hell aren't going to add this thing to the stock
Apache server, so the purpose of this discussion eludes me.

....Roy

Mime
View raw message