Its been a little while since SA-REST became a member submission and now
is a good time to discuss the next steps. Several members of the
community have expressed interest in following up on the submission
(for example, during the Frankfurt workshop).
We've listed some of the items we could think of below
(in the order of their importance as we see this now).
1. Addressing W3C team comment
The use of the title attribute is discouraged. Although the HTML
specification is not clear on this issue, the practice has been to use
that for accessibility and other uses of the title attribute
interferes with it. However the suggested class-value pattern (see
http://microformats.org/wiki/value-class-pattern) introduces a number
of extra elements (spans) that seems a bit more intrusive and rigid. A
discussion is needed to figure out a balanced approach where existing
elements are reused as much as possible.
2. Consistency with HTML 5
HTML 5 seems to be doing a thorough job on covering the complete scope
of HTML and there is support for annotations. For example there are a
number of annotation features introduced for Accessible Rich Internet
Applications (ARIA). Our annotations should not interfere with these
annotations or better yet - we should reuse these annotation
3. More use cases
There are some use cases that are listed in the submission although we
are open for more. In fact a collection of use cases and usage
patterns is a good supplement for the specification.
At a later stage a comprehensive (and preferably at least one open source)
toolkit needs to be associated with
the specification. We've seen the hesitation to adopt a specification
due to the lack of tools in many cases. A few tools have prototyped
and some could be easily support the specs (see below).
[Our own group is likely to work on use cases and adapt/extend tools in the contexts
of some collaborations with Biocatalogue and NCBO we have already identified
and hopefully we will share the outcomes soon. Some other tools -- eg SWEET that use
hREST can easily support SA-REST. So there is possibility of improving
exchange and coordination related to #3 and #4.]
W3C team's suggestion is to consider an XG (instead of
a workshop I had in mind)-- as part of an XG,
we can have F2F, perhaps on the side lines of a conference
some of plan to go.
Ideas / comments / suggestions are welcome.
Amit Sheth/Ajith Ranabahu
Ohio Center of Excellence on Knowledge-enabled Computing (Kno.e.sis)
Wright State University, Dayton, OH