[occi-wg] Weekly conf call: consensus on the formats issue

[Sam Johnston]

Afternoon all,

Though the main call was shorter than usual yesterday afternoon, Gary,
Shlomo, Andre and myself pressed on for 3 hours to see if consensus could be
found (again) on the formats issue. This is clearly something people feel
strongly about so it's unsurprising that it's taken us this long to get this
far, but I think we've finally found a happy medium (touch wood).

Here's my take (3 hours condensed into 3 lines):

   - Our number one priority is machines (rather than users) so we want
   native representations defined by existing standards (e.g. OVF) over
   home-grown (POX) and/or human-friendly ([X]HTML) representations.
   - Resources are "enhanced" by metadata which consists of links to other
   resources, categories, attributes, etc. and HTTP provides a perfectly good
   metadata channel in the form of HTTP headers so we'll use these to expose
   this information.
   - We'll also define a single, XHTML5 based representation that is usable
   by both users and machines. All of the information that appears in the
   headers will also appear within the XHTML5 rendering such that "enterprise
   applications" can extract it with XPath/XSLT/etc. without having to look at
   the headers.

I'm very happy with this compromise, which basically involved elevating both
existing documents to the "must" requirement level. The load for
implementors is minimised and both simple and complex clients are supported.
Furthermore it does not restrict the roadmap and enables us to trivially add
support for semantic web technologies including microformats and RDFa which
I know will make some people on this list very happy.

Thanks everyone for your perseverance with this issue. Please speak now if
there's any comments, questions or concerns about this (or forever hold your
peace :)

Sam
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.ogf.org/pipermail/occi-wg/attachments/20091029/d4ba6c30/attachment.html