integration use cases

[chris]

Dear IBA-net members,

To get us started here, maybe it is worthwhile to line out some of the
things we want to tackle.

Integration can be and has been understood in quite different ways.
Some think of integration as not much more than links between
corresponding parts of a web site. It seems pretty clear that we want
to go beyond that here, but what is the integration we want to
achieve?

I still see different possibilities:
For one thing, for example a site presenting canonical scriptures in
Chinese might want to integrate other versions of the same text in
Chinese, or even versions in other canonical languages -- or modern
languages. Rather than linking to another website, it would be much
more convenient to pull the corresponding content into the current
webpage to enable the user to compare the different versions
directly. For this, we need to have a way to identify corresponding
texts (and parts of the texts, ideally down to the paragraph level) as
well as a way to request specific parts of a text. There exist some
proposals for doing this (for example the Canonical Text Services
Protocol at http://katoptron.holycross.edu/cocoon/diginc/techpub/cts),
but we would need to adopt this to our needs. Beyond the technical
specifications, we certainly need to do some footwork and develop
tables that map between the texts in our databases. CBETA for example
uses a certain uniform identifier for its texts, like T51n2076 for the
text number 2076 in volume 51 of the Taisho Tripitaka. I assume other
projects have similar identifiers for their texts. So one task would
be to put together such a mapping table that can be used to find
similar texts (this should be flexible enough to not only identify
other versions of the text, but also adaptions, commentaries etc.)

Another, somehow different form of integration would be if, for
example, searches we offer on a website would go out and query
websites of different associated projects and then present them in a
uniform list. To be useful, this would require some kind of search
interface API that is published and can be used to query different
sites.

These examples are of course mostly relevant to canonical texts and
other types of material will require different kinds of integrations
-- a dictionary for example could be queried directly by the web
application showing the text and offer a tight integration between
text and dictionary.

So maybe we should collect some more use cases of kinds of integration
we want to achieve and then see what we can do with that.

Any thoughts or comments welcome,

Christian