Re: New Internet Draft on protecting children AND free speech

On Tue, 6 Jun 1995, Duncan White wrote:

> In Message-Id <Pine.LNX.3.91.950605094824.4773F-100000@okjunc.junction.net>
> Michael Dillon <michael@junction.net> writes:
> 
> > On Mon, 5 Jun 1995, Nathaniel Borenstein wrote:
> > > Borenstein, entitled, "KidCode: Naming Conventions for Protecting
> > > Children on the World Wide Web and Elsewhere on the Internet Without
> > > Censorship".

> > What I propose is that ratings be served up by special purpose "rating" 
> > servers. These servers can be operated and/or financed by the 

> I disagree with Michael when he suggests a separate ratings server - that
> unnecessarily increases the complexity of Web transactions (already overloading
> the Internet!).  At the moment, one client talks to one server to fetch one
> document and that works (mostly) fine.. 

The conversation between a browser and a rating server will be a very 
light load. Not much more than DNS conversations

> So I suggest that the work should be
> done at the existing server end, within the framework of HTML rather than as
> separate GETs or new http commands.

Using a ratings server requires no changes whatsoever to server software 
or to the documents on a server. WWW publishers don't have to do a thing 
to support a rating server. By its nature, it will work with any Internet 
resource accessible via a URL which includes USENET news.

> But KidCode sounds rather too specific to me - surely all we need in order to
> solve decency problems and many other "indexing" style problems is an agreed
> set of indexing keywords that HTML document authors are encouraged to place in
> the HEAD section of HTML documents?  Creating those indexing categories is hard
> work, especially to do it properly in an International, multi-cultural context,

Hard work is right. Even harder to get everyone to agree on categories 
and classifications. That's why the actual ratings need to be flexible 
and extensible, so that any group can set up a rating server and develop 
classifications that meet their needs.

> > The problem we are trying to solve with rating servers is of larger scope 
> > than the problem you have discussed in the KidCode proposal. Even if we 
> > cannot find a way to merge both proposals into one, we should at least
> > agree on a common system of ratings.
> 
> Of course, the biggest issue here is the whole subjective/objective indexing
> problem, where one culture's "suitable for all" may offend a particular
> cultural or religious belief elsewhere.  I would strongly discourage any
> "moral majority" approach to ratings where a single faction within the US
> tries to impose its moral code on the rest of the world.
> 
> Instead, I would argue that we must restrict ourselves to factual indexing,
> using an agreed set of cataloguing keywords in a consistent way.  As some
> sensible people once said "implement mechanism, not policy" :-)

Ooops, I could have said that better. What I mean is that there is likely 
to be a core set of ratings which people can extend in an orderly way. 
And, yes, the mechanism is the important part here. It should be possible 
for a browser to implement arbitrarily complex ways of weighing one 
rating against another from two different servers using fuzzy logic or 
regular expressions or whatever. While most ordinary individuals will not 
be able to configure such a complex scheme, we should leave the door open 
for it, since it is not entirely clear how browsers will actually be 
designed around the rating scheme.

> Now, how would this link up with Acrobat and PDF which presumably don't have
> HTML headers :-)

Or MPEG and RealAudio?


Michael Dillon                                    Voice: +1-604-549-1036
Network Operations                                  Fax: +1-604-542-4130
Okanagan Internet Junction                     Internet: michael@junction.net
http://www.junction.net  -  The Okanagan's 1st full-service Internet provider

Received on Wednesday, 7 June 1995 15:38:00 UTC