The challenges for EIIF

The exchange on the mailing list has been interesting, but, I must 
admit, can also be seen as frightening.

What is by now obvious to all is that there seems to be a huge number of 
"standards" in this area. This is not surprising, as in practice  EM 
touches on many different dimensions of society, and therefore has 
potential relationships to much technology in use somewhere in society. 
But is it feasible to analyze them all and present a coherent 
description of this multitude? Surely not. Some standards will only be 
identified (part of the inventory), others will be described (at 
different levels of detail).

We need a pragmatic approach for our work -- with respect to what we 
want to achieve and how we intend to achieve it.

I feel like making a few statements now, and then see to what extent 
there are other opinions in this group/list on the WHAT and the HOW.

 - there is too much stuff out there in the world that could be relevant 
for EIIG, more than we can handle in practice. We have limited personnel 
resources available within EIIG, and we need to use these resources in 
the most productive way.

 - we need to identify the focus of EIIG. What are the critical areas 
(key concepts) that we argue are (or, must be) at the core of the EM 
area that EIIG tackles. This may be a limited subarea of what could be 
called EM, but it is better to deliver constructive recommendations in a 
subarea than a shallow analysis of a huge domain with fuzzy borders.

 - we should identify some usage scenarios that drive the selection, 
demarcation, and analysis process. These can help us keep on track 
towards a focus area, and also  make it possible to characterize what 
kinds of concepts fall within the current scope of EIIG.

 - we should also have a policy about how existing standards can be 
re-used. It is in practice meaningless to define an all-encompassing 
model; one should  piggy-back on or  in other ways  make use of  
existing models (standards). This  is re-use of existing investments, as 
well as allowing us to explicate what we regard as our domain-of-interest.

 - personally I would like to understand EIIG as an exercise in how to 
achieve *some* value-generating interoperability by having 
interoperation formats and protocols, rather than as the final total 
solution to the use of IT in EM. Most likely you all agree with this, 
but I also want to strongly emphasize the word "some", as this is the 
only way to make progress.

 - then we should be explicit about the evolutionary view on our 
proposals, standards/technology, and the EM-area itself; we do not aim 
to provide the final sufficient solution, but we want to establish a 
foundation that *enables* further development/evolution in EM 
interoperability.

Remarks:
 - yes, this IG has a limited charter, and some of the issues I raised 
may be more relevant for a potential EM-WG. Nevertheless, EIIG should 
have a stance that goes beyond the EIIG charter.
 - an important part of initial work is to clarify how we interpret the 
IG charter, especially in terms of deliverables (in practice, the 
success criterion for a W3C group is its deliverables, as this *is* the 
only trace left after completion of work).
 - much of the mail exchange seen so far can be understood as  creating 
a coarse-grained map of the territory, and in that sense has value to 
our work. The detailed maps we will produce will only cover parts of the 
entire territory.
 - some political issues have been raised (open source, different status 
of standardization initiatives, public vs. commercial stakeholders, 
agency infighting, profession defense, etc). We should avoid going 
head-to-head in such political battlefields, though sometimes an 
awareness of the forces of power is valuable.


I found one remark made by Chamindra early on this list intriguing, 
after mentioning Sahana:
> However in our exploration of standards we found there to be gaps in 
> various needed standards for this domain. Either those gaps can be 
> filled by tailoring existing standards to the domain or it might require 
> a whole new standard. Of course we did not want to re-invent the wheel 
> in our project in creating new standards and it makes no sense for 
> standards to be created in isolation, thus the need for this group and 
> the need for a diverse set of practitioners, academics and solution 
> vendors to contribute to defining this framework.
>   
What were these gaps? Do we here see some specific examples of the need 
to re-engineer the conceptual models that are not well covered by 
existing standards? And does this quote also imply that certain 
standards may be used "as is"?

I suggest that practical experiences -- such as Sahana -- in building 
open standards-based EM systems are valuable input to the initial phase 
of our work, especially about what areas are not well served by standards.


Regards,

/olle

-- 
------------------------------------------------------------------
Olle Olsson   olleo@sics.se   Tel: +46 8 633 15 19  Fax: +46 8 751 72 30
	[Svenska W3C-kontoret: olleo@w3.org]
SICS [Swedish Institute of Computer Science]
Box 1263
SE - 164 29 Kista
Sweden
------------------------------------------------------------------

Received on Wednesday, 2 April 2008 14:10:51 UTC