On Mar 18, 2006, at 5:57 AM, Eric Lease Morgan wrote:
> The searchable interface at the following URL is really a rudimentary
> (read, not quite standards compliant) SRU client:
>
>   http://dewey.library.nd.edu/mailing-lists/code4lib/mail/archive/
>
> Put another way, the SRU root is:
>
>   http://dewey.library.nd.edu/mailing-lists/code4lib/mail/sru/

Nice! I probably could've searched the archives to find this couldn't
I. I did notice that the explain response seems to describe a service
at alert.ockham.org and not the code4lib mailing list archive.

Has anyone in the SRU community talked about using autodiscovery to
indicate an explicit link between a web front end and a SRU service?
Something like what opensearch [1] uses:

   <link type="application/opensearchdescription+xml" rel="search"
href="opensearch-description.xml" />

In the code4lib search interface it could theoretically look like:

   <link type="application/sru+xml" rel="search" href="http://
dewey.library.nd.edu/mailing-lists/code4lib/mail/sru/" />

//Ed

[1] http://opensearch.a9.com/spec/1.1/description/#autodiscovery