Re: Explicit Linking is Impossible

Tim Berners-Lee (timbl@www3.cern.ch)
Thu, 6 May 93 18:29:33 +0100


> Forwarded from comp.text.sgml...

> From: drmacro@vnet.IBM.COM
> Date: Tue, 4 May 93 11:10:35 EDT
>

> At relatively small scales, it is possible to create hypertext
> where authors hand craft the links between various places, say
> via typical IDREF mechanisms

> [...] This is impossible once the number of cross-library links
goes
> above about 50 because of maintenance problems--there are simply
> not enough hours in the day or people on the job to create,
maintain,
> and test these links. The next order of magnitude up
(cross-enterprise
> linking) is so clearly impossible as to be not worth considering.

You have a goood point of the scalability of the web. This has been
made by some people here who have the same wories. A solution which
works in software systems is to have, in any work, a set of anchors
which are explicitly declared "public" just like a global
declaration of a routine. This declation means that the author
will maintain that (named) entry point. It may be that the server
will want to keep a table of entry points to map them onto real
places, and indeed in the HTTP2 protocol there is a specified but
unimlplemented by anyone yet as far as I know (SBUAYAFAIK,
vapourware for short) facility for an HTTP server to return a new
URL as a name server, instead of the document.

Other anchors whch are not declared public would not be linkable to
outside the document. This would be a convention which you could
enforce locally.

This scheme scales quite well, as for any size entity (article,
chapter, book, shelf, department, library...) there would be,
independent of its internal size, a fixed number of global entry
points, so the problem scales as log N ie better than the
organisation which wrote it. (Not a dig at IBM but a comment
on hierarchical management perhaps)

Ed's system of referiing to searches in the index
"Look up the section on environment
variables in the O'Reilley MH book"
is similar, though heuristic if you don't contract with
ORA that they there will always be a section on envirnonment
variables in the book.

Tim Berners-Lee
CERN