W3C Push Workshop, 8-9 September 1997

[Other Papers] [Briefing Package]

Title:    Application Standards for "Pushing" Content and Streaming Media

Speaker:  Hadi Partovi, Microsoft Corporation


Today's world of "pushing" content and streaming media is filled with a
number of related but distant technology efforts. Each provides a
solution for a focused problem, but few promise flexibility to provide
an application standard for the overall problem of pushing or streaming
content. The requirements for such a standard should be that (1) it
allows an author to author content once and publish the content using
any of the major delivery mechanisms - pull, smart pull, true push, and
streaming. (2) it provides an efficient scalability story for scaling to
different client-device capabilities, different network bandwidth
capabilities, and different overall network sizes. (3) it enables the
next generation of technology in content publishing - searching,
indexing, profiling, filtering, and personalizing content, independent
of the publishing mechanism. (4) it unifies the declarative metadata
format that most of these technologies already use today, providing
commonality to the syntax and schema, along with a procedural interface
for manipulating the metadata.

It is essential to have a standardized meta-data object model that can
be declarative (file format) or procedural (programmatic access), along
with a *vocabulary* for pushing or streaming content. Microsoft believes
that the eXtensible Markup Language (XML) provides the ideal description
language for defining metadata for push purposes. XML has matured a lot
in the past year, and in the past few months many corporations and
organizations have come around to agree on XML as the universal language
for data. The discussion of push should not be centered around the
metadata format, however, as that belongs in the RDF group. Today, the
agreed-upon metadata format is XML, and tomorrow it will be the RDF
model on top of XML. The discussion of push should center around
providing a *vocabulary* on top of the agreed upon metadata format of
XML today, RDF tomorrow. Microsoft's position on push is that the
Channel Definition Format (CDF) proposal provides a great XML-based
vocabulary for push. The CDF vocabulary enables smart-pull or true-push
delivery of content for offline use, and offers bandwidth savings and
push capabilities on different client devices or different networks.
Today the CDF vocabulary is built on XML, and future *extensions* to the
CDF format can and should be based on the proposals and ideas formed in
the RDF discussion group, because of the advanced searching, filtering,
and querying technologies that can be built upon the RDF data model.