We have also looked at "size" in the IMS project. We are including it
in our metadata dictionary. Generally, the thought was that it is not
size per se that is relevant, but delivery lag to a user in an on-line
situation that is frequently of interest. This is sometimes expressed
as "quality of service". I understand that generally, people don't do
well with delays of more than 6 seconds. And 3 seconds delays are
considered noticable.
Some of the things we considered were biggest block, which would be the
largest block in a object that would have to come down to a user to
produce the next screen or experience. This can be translated into time
delay at an effective bandwidth. We also considered the number of
TCP/IP interchanges as a metric, this provides some means of factoring
in the network delays. This left out the problem of working with
streaming media.
In the end, we have adopted "size" as just the number of bytes. Crude,
but we consider it a place holder with only minimal utility. Basically,
we felt it important to recognize the need and allow the field itself to
evolve, rather than wait until we had the "perfect" solution.
The decision to leave it out of the Dublin Core does not remove its
potential use as an extension. The unofficial quality of such an
extension fosters the evolution of the field through use experience.
Probably a good thing. I like the fact that the Dublin Core is minimal
but stable. I trust the field taxonomies are approached in the same
spirit.
--Tom Wason
[log in to unmask] wrote:
>
> Dear all
>
> I have recently joined the mail list, and have been following the debates
> with interest and admiration. It's late in the process, and I understand
> that various documents have to be put to bed, so I hesitate to raise such a
> fundamental question (but maybe it has a very straightforward answer): why
> isn't "Size" one of the basic 15 metadata elements?
>
> It seems to me that size indications are given in regular cataloguing
> information (such as the number of pages in CiP data). Size (in the sense
> of quantity) of components is often given too (e.g. number of
> illustrations). For electronic files, we find from our user feedback that
> size of document or object is a significant piece of metadata - at least
> as important as coverage, say. I would expect publishers these days to be
> including size indicators as part of their metadata.
>
> Maybe I've missed something and it will be dealt with as subelement (e.g. I
> could see it being a subelement of "Format" for electronic files, and maybe
> a subelement of "Description" for non-electronic?).
>
> Sorry if this is an issue that's already been discussed a thousand times!
>
> Cliff Morgan
> Publishing Technologies Director
> John Wiley & Sons Ltd
> Chichester PO19 1UD
> UK
--
--------------------------------------------
Thomas D. Wason, Ph.D.
Director of Research and Evaluation
Institute for Academic Technology
University of North Carolina at Chapel Hill
730 Airport Rd., Suite 100
Chapel Hill, North Carolina 27599
919.962.9286
919.962.4321 FAX
|