Searching your web pages

[Kevin C. Marsh]

The Big Glee Bopper <thom at indiana.edu> wrote:
>
>I think you have to be sorta cautious of Z39.50. In the NISO 
>specifications it explicitly states: This standard is intended 
>particularly for use by systems supporting information retrieval services 
>for organizations such as libraries, information utilities, and union 
>catalogue centers...

Which is why librarians and other collection managers should use it.  I
never said that every web page needed Z39.50, only "serious knowledge
collections".

>Z39.50 came before the web is is not really designed for browsing which 
>is what most folks do with the web. 

I'll agree that Z39.50 specifies a search and not a browsing interface.
Browsing has its place, but did you ever try to browse through a collection
of 1 million photographs?  _Do_ most people browse collections on the Web
rather than using search engines?  Not in my experience.
>
>I think Z39.50 is important but I think IR has significantly changed, 
>browsing being the biggest difference. 

The Web has significantly changed IR by providing a ubiquitous network
infrastructure and set of information access tools (web clients).  People
browse in physical libraries too, but that has little impact on the need for
effective catalogs.  Z39.50 by itself was never user-friendly enough or
widely implemented enough to have a significant impact, but once Web->Z39.50
gateways appeared that changed rapidly.  The Web has made Z39.50 compliance
more important, not less.  Now I can create a Web page in a matter of
minutes that will launch a single search against a selection of Z39.50
databases worldwide and integrate the results in a single hit list displayed
on my Web client.  

>I am not arguing for anyone to 
>support a vendor with a proprietary format or approach.

Again we agree here - open, nonproprietary standards are needed.

>--Thom

Kevin C. Marsh, Executive Director
Information Access Institute
IAI at neosoft.com     http://www.neosoft.com/~iai