Shibboleth Developers

[Peter Murray <>]

Task Group 1 (TG1) of the NISO Metasearch Initiative (NISO-MSI) is 
seeking to define the problem of access management to service providers 
and resources within the context of the newly evolving metasearch 
engines.  As you may be aware:

   In a meta-search engine, you submit keywords in its search box, and
   it transmits your search simultaneously to several individual
   search engines and their databases of web pages. Within a few
   seconds, you get back results from all the search engines queried.
   Meta-search engines do not own a database of Web pages; they send
   your search terms to the databases maintained by search engine
   companies.

<http://www.lib.berkeley.edu/TeachingLib/Guides/Internet/MetaSearch.htm
l#WhatAre>

There are general metasearch engines out there that look at the public 
web space (Vivisimo, Metacrawler, and Dogpile), but these are beyond 
the scope of the NISO-MSI.  What we are focusing on are specialized 
metasearch engines that target the "dark web" of library catalogs, 
indexing/abstracting services, full-text publishers, institutional 
repositories, and the like.  These are using tools such as MuseGlobal, 
WebFeat and others with established protocols like Z39.50, 
custom/proprietary XML gateways, and raw HTML-scraping to perform the 
metasearch function.

I suspect the access management issues are leaping to mind now.  The 
metasearch engine is a 'resource' itself for which the end user will 
have to go through some sort of authN/authZ step in order to access, 
and then the metasearch engine will act as a proxy on behalf of the 
user to search service providers for appropriate resources.  And, in 
the end, the user may decide to leap out of the metasearch engine 
environment and operate directly with the service provider.  Through 
all of this, how are the credentials effectively passed around to 
ensure the correct access to the resources.

This type of interaction starts to look like the P2P interaction 
proposed in Steven Carmody's LionShare Shibboleth use case document in 
that we have applications talking with each other rather than 
interaction through a browser.  In my read of the use case, though, it 
goes well beyond the LionShare specifics -- there may be needs to pass 
user attributes from the metasearch engine to the resource provider, 
for instance.

The output of MSI-TG1 as defined now will likely be an inventory of 
methods and "good enough" practices that exist now along with 
recommendations for further development.  Shibboleth will be among that 
inventory of methods, but until a fully Shibbolized world is out there 
other methods will be needed.  (The actual charge of each of the task 
groups will be established at an all-MSI participant meeting in 
Research Triangle Park on April 22-23.)  That said, however, there does 
seem to be a convergence of interest surrounding the creation and 
adoption of metasearch standards and the application of Shibboleth to 
access management issues.  If this is a point in time in which further 
work for the next generation of a Shibboleth architecture is under 
development, it seems wise that our groups work together to ensure both 
of our needs are met.

I now return you to your regularly scheduled discussion of the Shib 1.2 
roll-out...


Peter
--
Peter Murray                      http://www.pandc.org/peter/work/
Assistant to the Director for Technology Initiatives  860-486-6771
University of Connecticut Libraries            Storrs, Connecticut