[Date Prev][Date Next][Thread Prev][Thread Next][Author Index][Date Index][Thread Index]
Upgrading the news URL
- To: xanadu@xxxxxxxxxxxxx
- Subject: Upgrading the news URL
- From: Dan Fabulich <dfab@xxxxxxxxxxx>
- Date: Wed, 12 Mar 1997 09:31:06 +0100 (MET)
- Reply-to: xanadu@xxxxxxxxxxxxxxxxx
- Sender: xanni@xxxxxxxxxxxxxxxxx
I've recently become interested in hypertext systems like Xanadu and
Hyperwave, and I've been considering what would be the most efficient
means of implementing a fully integrated (backlinking and fine-grained)
hypertext system. I've come up with a solution which you might find
intriguing.
Hypertext could run off of existing Usenet architecture. With VERY
little effort.
Usenet ALREADY meets most of the requirements laid down in the Xanadu
FAQ: bdghjlnoq are already part of the system; ki could be easily
implemented with programs like PGP which use public key encryption
algorithms. ap will be implemented, slowly, across the Internet, as
it becomes harder and harder to sniff and spoof. c is probably on the
way as well, though there are many who would wish otherwise.
That leaves ef, the important missing features. But to our advantage,
the most popular newsreaders ALREADY support URL linking within Usenet
posts. Some, like Netscape and IE, can view these posts as full HTML
documents. So with minimal effort we could upgrade the "news:" URL
to allow the user to specify a particular byte range in the body of a
unique message-id: news:<message-id>@host[/<startbyte>[-<endbyte>]]
So a URL to reference the first paragraph in the body of the Xanadu FAQ
would be: news:5ferud$241vg@xxxxxxxxxxxxxxxxxxxxx/000-512
Simply by making this miniscule change to the "news:" URL, Usenet could
instantly become an embryotic hypertext system. It would even conserve
bandwidth, as people would link to the previous posted sections in a
debate, rather than reposting them at length. Articles would be
backlinked and referenced to any MIME media type via URLs and the
References: header field, which could link to any other article on
Usenet. We could forget the difficult task of implementing hyperg
servers across the Internet and get the whole thing done quickly, easily
and efficiently.
Thoughts on this?
-He who laughs last thinks slowest-
dAN