Re: Changes vs. new element

On Fri, Aug 2, 2013 at 3:19 AM, Alf Eaton <eaton.alf@gmail.com> wrote:
> On 1 August 2013 20:03, Karen Coyle <kcoyle@kcoyle.net> wrote:
>>
>>
>> On 8/1/13 11:05 AM, Dan Scott wrote:
>>
>>>
>>> If I can be permitted to fantasize about a library scenario for a
>>> moment, if the search engine recognized via your location or IP
>>> address that you were in or near a library, it could serve as your
>>> library catalogue and display the additional metadata when it was
>>> actually useful to you (much as it detects when you're looking up
>>> movies, it can show you the local movie listings, including name &
>>> address of the theatre, immediately rather than forcing you to click
>>> through).
>>
>>
>> That was my first fantasy as well. See:
>>
>> http://kcoyle.blogspot.com/2012/09/rich-snippets.html
>
> That's kind of what Google Scholar does
> <https://www.google.com/intl/en/scholar/libraries.html>: IP address
> ranges and library serials holdings => appropriate links to article
> full text through the library resolver, when the library has access.

Right-o Alf, and we discussed something along these lines on-list back
in November too:
http://lists.w3.org/Archives/Public/public-schemabibex/2012Nov/0002.html

> It's particularly annoying that - as far as I know - libraries only
> publish this holdings file to Google, rather than making it available
> for everyone.

I think that's the default behaviour, but at Laurentian University
we've made ours publicly available. We were experimenting with a
commercial discovery layer, and my hope was that rather than sending
them the custom format that to express electronic holdings their way,
I could just point them at our Google Scholar institutional holding
file and they could work with an existing format that we were already
updating on a regular basis... but no. Trying to keep this on-topic, I
gather we're all hoping that schema.org markup will be enough of a
standard that we can avoid similar such painful exercises in the
future :)

> Keeping up-to-date with availabililty of particular items would be too
> much for a crawler, as it changes too quickly, so there would need to
> be a push API, like there is for Google Shopping
> <https://developers.google.com/shopping-content/>, updated with every
> availability change. Alternatively, as long as the library can resolve
> an OpenURL query, tools like <http://www.libraryextension.com/> can
> look up availability of single items on demand.

I was thinking that if you have a sitemap that lists all of your
resources, a persistent URL for each resource, and standardized
structured data on the resulting page that shows you the availability
of items for that resource on-demand, what do you need OpenURL for...
but yes, search engines aren't going to want to add a 1 second delay
for a live response to availability, so a push API seems warranted
there; and plugins like LibX / libraryextension etc need some
standardized search API. I helped teach LibX to speak Evergreen years
ago, but standards are good.

I guess we're getting pretty far off-topic from "Changes vs new
element" in schema bibex land, though.

Received on Friday, 2 August 2013 12:10:15 UTC