Re: Explain-lite (Explain discussion)

Robert Waldstein wrote:

> I think we are talking about 4 separate things under the Explain-lite heading
> (maybe more).

I think we're talking about these things, plus Explain-lite, under the "*Explain*
heading".

>
>   1. sending EXPLAIN information in the init.
>   2. THe EXPLAIN model is too complex. THe search and retrieval is a
>      complexity that is slowing implementation.
>   3. EXPLAIN is too complex at the information level.  THat is, what are
>      all these things being asked for and why.  By the way, I point out
>      that it was the ONE people who required that I populate the EXPLAIN
>      element
>           AccessInfo -> unitSystems
>      Always wondered what a client did with that...
>   4. Should the records be carried in something like XML.

>
>  Do people agree - cause if yes I would like to separate these somewhat, and
> discover where we can't separate the issues.  And I will resist commenting on
> these - am interested whether there is agreement that these are separate.

These are four important topics, but no, I don't think they can be completely
separated. XML can be separated from the other three but these first three are
closely related.

I would like to suggest (and it's not the first time I've mentioned this, nor am
I the first person to suggest this) that we begin by asking the two questions
"what problem should Explain solve?"  and "what information does a client need?"
-- for the second question: rather than looking at the existing information that
Explain supports and asking if it's necessary/useful, reverse the process and
instead ask what information is useful/necessary and then see how well it matches
what Explain supports.


--Ray


--
Ray Denenberg
Library of Congress
rden@loc.gov
202-707-5795

Received on Monday, 20 November 2000 10:05:22 UTC