On Wed, 25 Apr 2001, Jim Choate wrote:
>In some ways, in others it is. Your singular reference to TCP/IP for
>example represents a 'central authority', technical but it still
>constrains and coerces.
Not really. The IETF has no coercive power whatsoever. It cannot force
people to use what it designs. For instance, IETFers would probably want
every network capable box in the world to support IPv6, PIM, whatever, but
people go where they want. That is still legacy IPv4 without multicast. If
you look at the entire RFC repository, you'll see that IETF's history is
riddled with unadopted protocols.
>>I'm guessing this is about layering TCP/IP based VPNs on top of the
>>current Net infra, only this time mandating participation in a given one
>>in exchange for your liberty. Keeping the VPNs separate would be
>>where the crypto comes in. It is also where the idea breaks, since
>>without plenty of smut to go around, the non-smut part of the network
>>would probably wither and die. That's where the money is, the killer
>>app.
>
>It's not about smut or money, it's about 'community',
Really? So who would pay for the non-smut part of the Net? Apparently
nobody, since the technology is here and few people are segregating
themselves. Besides, communities can be created without splintering the
bigger Net. Communities in their traditional, geographic sense are
meaningless in net.economy. At most, we're seeing a collision between the
US principle of 'local standard of indencency' and the international,
borderless Net culture.
>'expectations',
People should treat Net-the-medium as the Net-the-medium, not as
Poughkeepsie Daily. If they don't, and start legislating based on that, they
will not have the Net anymore. Just my point.
>'normalization',
This would be relevant if we were having some sort of moral argument. We
aren't.
>'shared resources',
Well, the resources are shared, but contractually. They are private. Please
do not tell that you view the Internet as a public commodity. That would be
foolish.
>It's not about isolation but rather filtration. It's also about
>decentralization of the technical standards, the real long range power of
>Open Source efforts.
Now this I can understand, now that we've seen Crows' version of the story.
He's proposing that we make way for voluntarily built up VPN sorta islands
online, in the expectation that people can then choose who to hang out with
and be happy. Fine by me. Only it does not work. (I'm going to equate the
Net with the Web, since that is what most non-techies do, anyway.)
I thought about this for a while yesterday, and I don't see any real
difference between Splinternet's and the kinds of online communities that
can be built by current censorware set to operate purely on site lists and
block out everything else. They have the same problems with scalability
(who's willing to pay for the enormous resources it takes to keep an upto
date database of sites belonging to your peer group?), limited functionality
(a lot of stuff, i.e. the reason people go online, is turned into some
stuff, which people can get in their local library), governance (how do you
make sure that 'illegitimate' stuff doesn't show up? how do you decide what
is illegitimate? let your net.inhabitants vote on each site to join? and
that would leave you with a Splinternet of size what?) and technicalities
(performance, reliability, scalability).
Splinternets would also welcome new trouble, like keeping the VPNs
absolutely separate, ensuring email connectivity without sacrificing that
compartmentalization (or do you assume people would likely give up global
SMTP connectivity?), hardware/software support which would probably be far
more comprehensive than the one needed for censorware, inconvenience in
choosing what Splinternet to be on, probable inflexibility in cruising
between Splinternets (not so bad with censorware because it can be
reconfigured and standards for the transmission of blocking data exist in
the form of PICS) and so on, ad nauseam.
We've already seen that people do not buy the limitations and cost inherent
in censorware which only works on lists of approved sites. We've seen that
the compartmentalization afforded by software that works on heuristics (like
forbidden word lists) isn't enough to satisfy the splinterers. We've seen
that these limitations cause politicians to pass laws anyway, regardless of
the availability of censorware.
In fact, the most favorable mechanism to implement Splinternets I can come
up with is PICS capable software set to reject unrated content, and
voluntary self-rating with a central metadata certification authority to
please the Real Bigots.
If you view the history of PICS, you'll see how likely this is.
In fact, we might view Crows' suggestion as a reformulation of what the PICS
framework wanted to do, only with the difference that the boundaries of each
Splinternet are pretty much set in stone, and that the resulting system
bares more resemblance to the idea of community, domain or place (which
people for some weird reason apply to online fora) than automatic selection
and filtering based on formalized criteria (like subject categories, site
lists and other metdata).
If one were to mandate such Splinternets in exchange for online anarchy
inside them, the only reason this would not be perceived as compelled
speech, like mandatory self-rating would likely have been, is that people do
not understand how protocols work, but seem to grok the
place/Net/community/agora/forum metaphor.
Sampo Syreeni, aka decoy, mailto:[EMAIL PROTECTED], gsm: +358-50-5756111
student/math+cs/helsinki university, http://www.iki.fi/~decoy/front
