Re: [dita] How much flexibility do specializers have to make exceptionsto behaviors that are outlined in the DITA standard?

[Eliot Kimber <ekimber@reallysi.com>]

Ogden, Jeff wrote:
> This is a note that Don Day asked me to send out during yesterday's DITA
> TC call.
> 
> About a week ago I restarted the discussion about what the DITA Standard
> REQUIRES, strongly RECOMMENDS, and what is OPTIONAL.  So far the
> restarted discussion has been between just me, Michael, and Eliot.  What
> I'd like to find out from others on the TC is if these issues are
> important enough to continue to spend time on them or not?
> 
> So please let me know by sending a short reply either to me or to the
> DITA TC list.  I'll summarize the responses before next Tuesday's DITA
> TC call.

I think this is very important.

The DITA specification needs a conformance statement, which it currently 
does not have.

Without a conformance statement we don't really have a standard because 
there's no formal basis on which to judge any given use or 
implementation of DITA for correctness against the dictates of the 
specification, in particular, in terms of what things are optional and 
what are required. The closest there is are some statements in the 
architecture spec about requirements for specializations, but nothing 
that relates to processors.

In order to make a conformance statement the spec must be clear about 
what is invariant and what is suggested. It currently does not do that 
with sufficient clarity (obviously, or we wouldn't be having this 
discussion).

There is also the question of whether there should be different levels 
of conformance based on support for different set of optional features 
or if all features of DITA are required. For example, we have to decide 
if a processor that does not implement support for non-standard 
specializations is or is not conforming. Of course, the easy answer (and 
probably the correct answer for DITA 1.1) is that there are no optional 
features.

And of course this all ties into the fuzzier question of what it means 
to "support DITA" in certain types of tools and whether or not that's 
something specification should address normatively or informatively or 
remain silent on. For example, it might make sense for the TC to issue a 
separate technical report on support for DITA in content management 
systems that offers guidance by defining some testable categories of 
support without trying to define formal conformance levels or anything.

Cheers,

Eliot
-- 
Eliot Kimber
Senior Solutions Architect
"Bringing Strategy, Content, and Technology Together"
Main: 610.631.6770
www.reallysi.com
www.rsuitecms.com