> >what if the consumer has a legitimate need to access something that just
> >happens to look to the virus filter like it contains a virus?
> >what if the filter doesn't just check for viruses, but also things that
> >look (to it) like porn?
>
> Those examples are problematic as they demonstrate situations where there
> could be a valid reason for having filters on the network border.
define 'network border'. :)
(now try to define it in a world with ubiquitous portable/wireless devices,
where legitimate interaction patterns no longer correlate well with
network topology. imho, topology-based filters are a dead-end. )
> In those cases you can argue that the consumer isn't the user sat in
> front of her browser, but the guy that decides who can use the network
> and what they can use it for.
you could argue that. I'm not at all sure it's either socially valuable,
or that it makes sense from an architectural point of view.
however, without trying to nail down the definition of 'consumer' just
yet, my point is that different users have different needs, and
the filters which one might find useful, will cause harm to another.
if you want the system to work predictably you can't have third parties
who have no responsibility to the consumer or the content-provider
imposing arbitrary constraints on how the system operates.
Keith
This archive was generated by hypermail 2b29 : Thu Nov 18 2004 - 11:21:29 MST