[OAI-implementers] RE: Resumption tokens

Tim Brody tdb198@soton.ac.uk
Wed, 26 Sep 2001 12:10:20 +0100

From a general point of view, it would be a "Good Idea" to implement flow
control (resumptionTokens), especially nearer the beginning of a project
while the way your system and OAI works is fresh in your memory!

Ignoring the initial harvest, the majority of harvesting will be based on
changes since two days ago. If you are likely to have a large number of
changes over two days (say, would create a larger than 1mb download),  as a
harvester I would like that broken up.

(In my limited experience XML and the tools to handle it don't scale well)

All the best,
Tim Brody

----- Original Message -----
From: "Carl Lagoze" <lagoze@cs.cornell.edu>
To: "Jeff Good" <jcgood@socrates.berkeley.edu>
Cc: <oai-implementers@oaisrv.nsdl.cornell.edu>
Sent: Wednesday, September 26, 2001 10:37 AM
Subject: [OAI-implementers] RE: Resumption tokens

> Hi Jeff,
> I am bouncing your question back to the implementers list where there
> are plenty of people with experience to answer your question.
> Carl
> -----Original Message-----
> From: Jeff Good [mailto:jcgood@socrates.berkeley.edu]
> Sent: Tuesday, September 25, 2001 10:22 PM
> To: openarchives@openarchives.org
> Subject: Resumption tokens
> Hello,
> I currently run a small data provider (the Comparative Bantu Online
> Dictionary) which contains less than a hundred metadata records.
> However,
> I may soon add about 2000 more records to our metadata inventory. I
> haven't worried about resumption tokens to this point. I am wondering
> if there are any guidelines on when to implement the use of resumption
> tokens--I'm not concerned for the server on my end, but didn't know if I
> should be worried about the harvesters.
> Thanks,
> Jeff Good
> _______________________________________________
> OAI-implementers mailing list
> OAI-implementers@oaisrv.nsdl.cornell.edu
> http://oaisrv.nsdl.cornell.edu/mailman/listinfo/oai-implementers