RE: Browsing the cache

> From: David Diaz i Torrico [SMTP:ddiaz@drac.com]
>
> At 13:49 11/08/99 +1000, Marc-Adrian Napoli wrote:
>
> In my localnet won't be such privacy problems, anyway a cache browser
> won't
> violate the privacy more than the logs do.
>
        The only way you will get most portal type sites (i.e. most web
        accesses) to work in offline mode is by very aggressive
        caching. Such caching has been demonstrated to make other
        users' web email become readable instead of ones own, unless
        one is very careful to work out which pages really can be
        safely cached in spite of their apparent non-cacheable status.

        You will also have to do ad-blocking, as many adverts are now
        dynamic.

> Anyway, is there any tool which does that, or will I have to program it?
>
        The only way of force validating with squid is to scan the logs
        (or process swap.state) and issue conditional GETs for
        everything. Such activity will be against the conditions of
        use of some popular sites, and at least one squid user has
        recently added throttling to stop this sort of activity
        by his customers. Forceably caching nominally uncacheable data
        is likely to be outside the terms of use as well, if it results
        in adverts not being updated as frequently. IMDB consider
        ad blocking a violation of their conditions of use.

        If you do revalidate automatically, you will defeat the squid
        expiry mechanisms, as the pages will appear to have been
        accessed recently.
Received on Wed Aug 11 1999 - 05:36:28 MDT