Ray,
I am puzzled by your logic of deducing from a discussion about whether the
client should send the result set name in a query or as a separate parameter
to a conclusion that we don't need the server to return a result set.
As for why we need the server to return the result set, this was discussed
at the meeting and again via e-mail last week, so I don't understand why you
are making such heavy weather of it.
Anyway to try to drum this point home. Let us consider a server with ten
records by the author Smith. Let us also assume that this particular server
always returns results sorted by title and currently has no titles beginning
with A.
We send a search request
Query: au=Smith
StartRecord: 1
MaxRecords: 5
And get back records 1 to 5 by Smith with the result set name RES1
However at this point unknown to us someone adds a new record by Smith with
a title beginning with A
Now if we issue the search request
Query: au=Smith
StartRecord: 6
MaxRecords: 5
The first record we get in the response will be the same as the last record
we received from the previous search request
If however we send the search request
Query: result set=RES1
StartRecord: 6
MaxRecords: 5
We will get back the next 5 records following directly from the first search
request.
Does that clarify why we need server generated result set names?
Matthew
> -----Original Message-----
> From: Ray Denenberg [mailto:[log in to unmask]]
> Sent: 11 July 2001 22:01
> To: [log in to unmask]
> Subject: why do we even want the result set name?
>
> I'm trying to get the ZNG report finalized (by about noon tomorrow), and
> as I'm
> writing this message I have to leave in 5 minutes so please pardon the
> haste and
> brevity of this message; normally I'd wait till tomorrow and think this
> through
> before posting, but I'm hoping some of you may think about this
> tonight....
>
> I'm having a real tough time trying to articulate why we even want the
> server
> to supply a result set name, given the discussion of the last couple days.
> Any
> ideas? Should we drop the result set name altogether; if we do, what do
> we
> lose?
>
> Thanks.
>
> --Ray
>
> --
> Ray Denenberg
> Library of Congress
> [log in to unmask]
> 202-707-5795
|