- From: <Kevin_Thomas@ovid.com>
- Date: Mon, 16 Jul 2001 13:21:24 -0400
- To: "LeVan,Ralph" <levan@oclc.org>
- Cc: "'Johan Zeeman'" <j_zeeman@hotmail.com>, ZIG <www-zig@w3.org>, www-zig-request@w3.org
Hello Ralph!
One tiny response to one tiny part of your message:
(I've translated keywords into CAPS)
>> I would also like to have the rationale for abandoning RPN
>> explained.
>Because RPN is harder than CCL. All RPN does is make it easier for
>programmers to parse the queries. So, which would you rather type in:
>
<rpn><rpnRpnOp><rpn1><op>dog</op></rpn1><rpn2><op>cat</op></rpn2><op>AND</op></rpn1></rpnRpnOp></rpn>
>or
>dog AND cat
Actually, I like the RPN version better. It's partly because it enables us
to put together queries that are unambiguous.
For instance, AND is a keyword in the search
dog AND cat
but what about
head and neck injuries
OK, so I suppose you could tell people to put search terms in quotes, so
this example would become
"head and neck injuries"
But should the system also accept
head "and" neck injuries
?
Then there's OR . Which gives rise to things like
NY State Health Department
which is fine, but what about
OR State Health Department
Are AND and OR the only keywords in this syntax? What about ADJ and SAME
and NEAR ? What about ROOT and TREE and THES ? I am highly fearful that
the keyword list will grow, thus breaking things.
I'm not saying that these problems can't be solved. Far from it. Ovid's
software is one package of many that takes in end-user queries and handles
queries that have AND and OR and other keywords. Let's just understand
that the standard-setting process will have to pick specific rules for the
various cases.
In case it's not clear, I'm pretty much agnostic on the ZNG issue per se,
so please don't interpret this as taking sides between anybody and anybody
else.
Just a thought,
Kevin
Received on Monday, 16 July 2001 13:23:04 UTC