RE: Positioning Z39.50

On Tue, 21 Nov 2000, Matthew Dovey wrote:
> > Tearing Z39.50
> > into little
> > pieces will *never* make it more appealing to the W3C types
> > who disqualify
> > it for reasons of their own.
>
> I don't want to destroy Z39.50 (not yet anyway ;-)) and am unfortunately
> only too aware of the politics of the W3C.

Awww, c'mon guys. The politics you're aware of are the politics of the
computing industry, where Z39.50 has barely been noticed. That world (with
relatively little prompting from W3C) has lately gone apeshit over XML.
There's no W3C anti-Z39.50 conspiracy at work here, just the natural
consequences of Z39.50 failing to have set the world alight.

If you want my personal view: Z39.50 could do with being available in
bite-size chunks. If someone is writing an XML Schema for bibliographic
content, or writing a spec (eg. UDDI) to characterise network accessible
machine services, they currently have an all or nothing interaction with
the Z world. There's a lack of simple recipies for being "Z friendly".
Perhaps "bit sized chunks" is a metaphor for being chopped up and eaten?
My preferred metadata system (RDF) has certainly suffered in the
past for not being more obviously XML-friendly; I fear Z39.50 is in a far
more vulnerable situation. Casting this as W3C XML zealots against the
plucky Z39.50 outsider isn't going to help any of us understand what best
to do. To put my own view totally on the table: Z39.50 will die unless it
gets a bit more mainstream. Rightly or wrongly, mainstream now seems to
mean XML, and that's something that goes beyond W3C to the computing
industry at large. The challenge is to avoid throwing baby out with
bathwater, and to identify the core ideas, architectures and vocabularies
behind Z39.50 so the lessons of the past 20 years get deployed in the
search systems of the next 20.

IMHO etc.,

Dan

ps. a couple of folks have asked about the possibility of having a daily
digest version of the www-zig@w3.org list available. Right now it seems
(http://www.w3.org/Mail/Request) that this isn't a feature we have
configured for more than a couple of lists, though I'll pass on the
request to the systems team. Full web-accessible archives of the www-zig
list are available at http://lists.w3.org/Archives/Public/www-zig/ though.

Received on Tuesday, 21 November 2000 14:04:09 UTC