James Weinheimer wrote:
> "ISA, Dept. of Social Welfare" wrote:
> > > DC.Publisher = "Eastern Kazoo Web Hosting Service"
> >
> > I think this is too literal or physical an interpretation of publisher. A
> > hosting service provides an area where a published document can be
> > distributed or displayed, like a bookshop.
>
> Yes, or like the items strung around a printer's office. It seems to me that
> Alex's interpretation of "Publisher" deprives it of all meaning. Is it really
> useful to add "GeoCities" to every other record?
A web hosting service provides a lot more than just a location for
displaying web pages.
Apart from designing the layout, producing the graphics and providing
comment on structure, the service hosting a web site is responsible for
censoring its content. If the service receives complaints about a
document it hosts, it's the responsibility of the service to take down
those pages, or face criminal charges brought to bear by Senator
Alston's cronies.
I'm Australian, of course, so you people overseas are free to continue
to assume that a web hosting service is a DIY billboard. In Australia at
least, the content carrier (ISP/web hosting service/comewhatmay) is
responsible for the content it carries.
In fact, the Australian Government is about to accept an Industry Code
of Practice which will force every Australian user of the Internet to
use software like "Net Nanny" to make sure we view stuff that our
fundamentalist-puritan next door neighbour might find offensive (but
that's a different topic, not related to metadata - or is it? PICS comes
to mind). There are also movements in Australia and overseas who intend
to make it compulsory for people publishing content to rate their
content (and for people viewing content to only be able to access
censored material).
Thus, naming the web hosting service in the metadata is essential, in
order to trace back site like "www.bigboobs.com.au" - a hypothetical
site which contains explicit material dealing with breast enlargement
procedures - to the actual service running the site, which in this
instance might be "Eastern Kazoo Web Hosting Service". You wouldn't
include a Library as a publisher, because anyone who complains about the
content of a Library's collection is simply asked to leave ("I don't
like those rude pictures in the Kama Sutra!" "Well, sir, you did open
the book up, didn't you?").
> Yes, or like the items strung around a printer's office. It seems to me that
> Alex's interpretation of "Publisher" deprives it of all meaning. Is it really
> useful to add "GeoCities" to every other record? In this case, what good would a
> search on "publisher: geocities" be?
Most sites on Geocities will have a ".geocities.com" suffix, so in this
instance, it's easy to blacklist the whole Geocities domain, rather than
just sniffing out the "dangerous" pages. So in this one instance,
listing Geocities as a "publisher" makes less economic sense than
filtering based on URL.
Then again, if you're cataloging the Internet, it may be useful to your
clients to be able to separate out stuff that is *not* "published" by
Geocities. Filtering out content based on publisher will be a lot
cheaper than filtering out content on the actual content. Thus the
manufacturers of Net Nanny and similar DIY-tunnel-vision software simply
block out the entire Geocities domain (and the entire Tripod domain, and
every other self-publishing service on the Internet). It's cheaper that way.
Alex
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|