Re: meta information

Chris Lilley, Computer Graphics Unit (
Mon, 6 Jun 1994 11:31:11 GMT

I thought this statement was important enough that I would repeat it.

Chris Neuss said:
> Seriously, this is important. The only way to actually get people
> to add such information (which is necessary for all kinds of local
> and global indexes) is have editors provide a special field for
> filling it in. But this requires that a standard has to be agreed
> on by the time real WYSIWYG editing tools arrive - and this is
> gonna be pretty soon... :-)))

I have listened to the recent discussion about meta information with interest.
It reminds me of the discussions about meta information that happened in the
scientific datasets community some years ago. As looks like happening here, the
proponents of 'we don't need no meta information' won. So, apart from some
notable exceptions, such as the NASA work on PDS, scientific datasets don't have
proper meta information, there are no real standards worth a damn, and
visualisation systems throw any meta information away. Ho hum.

But discussions of the best, truest and most holy way to alter the DTD to
express this or that have seemed a little closed and high flying. Yes some
people know enough to suggest an addition to the DTD when they want to add a
meta information element - but the vast and increasing majority don't. Many
information providers will not even know the DTD exists; just that they use a
special Web editor to write Web pages. The implementation details are as
irrelevant as they are to the user of a wordprocessor, or a paint package.

Similarly, I have little sympathy for the idea that all meta elements are left
open ended and un-specified, as if each information provider should be capable
of designing their own, well specified, incompatible, system. That may be true
of some people - perhaps many of those on this list - but we are not a
representative selection. Once good authoring tools come into play, we will be a
massively un-representative selection!

I would much rather see a suitable selection of generally usefull elements which
will, inevitably, be adopted as a defacto standard in the absence of anything

Then they will get used, acquire critical mass, and become useful.

At WWW94 I suggested to Dave Raggett that the meta elements be firmed up a
little so that there was a usable set for commonly required ones, and a simple
procedure for registering new ones. That way, the system is extensible and
useable. The WYSIWYG editors just incorporate the latest crop of registered
extensions each time a new minor version comes out.

If new elements are registered this avoids duplication, which is I think self
evidently a good thing. If some elements are standardised they will get used,
which is also clearly of benefit, and if a significant proportion of the web
documents out there have some, usable, meta information - all using the same set
of tags, not all different ones - we have a big win.

Chris Lilley
| Technical Author, ITTI Computer Graphics and Visualisation Training Project |
| Computer Graphics Unit,        |  Internet:            |
| Manchester Computing Centre,   |     Janet:            |
| Oxford Road,                   |     Voice: +44 61 275 6045                 |
| Manchester, UK.  M13 9PL       |       Fax: +44 61 275 6040                 |
| X400:  /I=c/S=lilley/O=manchester-computing-centre/PRMD=UK.AC/ADMD= /C=GB/  |
|  <A HREF="">my page</A>   |