Carl Bunje led this meeting. He began by noting that although this is a "Customer
Council" meeting, Supplier members are also users of IT and therefore the
contributions of Supplier members to this meeting will be just as welcome as those from
the Customer members.
Carl explained (see slides - available to
members-only) that this topic - business agility and the role an IT standards framework
could play in enabling it - emerged from ideas he raised at the January 2004 (San Diego)
Board Meeting. A paper on this was discussed at the April 2004 (Brussels) Board Meeting,
and the paper was published in the May 2004 Members Newsletter (available on The Open
Group web site). We can't get by without our IT systems any more - business today depends
on having effective IT systems. The paper reviewed competitiveness and business agility,
and the enabling role of information systems and visions of future IS - virtualized,
adaptive, autonomic, service-oriented, heterogeneous - seeing the enabling role that
standards has, and consequent opportunities for The Open Group in developing and promoting
standards.
How can we be sure that we are working on the standards that will fit together to
enable us to achieve the business agility that our businesses need for the future? A
standards framework would provide common ground in a common vision among suppliers and
with customers. A standards roadmap would show paths towards convergence, warn of
conflicts, and identify gaps where new standards are needed, so pointing towards the areas
where enterprises need to encourage development of new standards-based interoperable
solutions. We can encourage this direction by establishing the common ground - among
Suppliers’ visions, and with Customers’ needs. Related items and
activities we would need to address include good awareness of current standards activities
(e.g., OGSA, etc.), the DMTF management standards, The Open Group Standards Information
Base (SIB), and our TOGAF and Architecture Development Methodology (ADM). We have a lot of
good standards and tools that we can start out with.
Carl invited member discussion, and suggested we could structure our thinking by asking
ourselves the following questions:
- What would be the expected benefits and usages of such a framework to Customer
enterprises? Roadmaps, development guidance, other ...
- What would need to be included? What level of detail and specificity? What breadth and
comprehensiveness?
- What are the main Challenges? Supplier IP and competitive strategies, customer
engagement and validation, other ...
- Is this worth pursuing? If so, what are the next steps?
Claudia Boldman: The Commonwealth of Massachusetts has issued a policy for procurements
based on open standards. They need guidance on which standards are ready for prime time,
and which are proven as effective. They want profiles too. They need guidance on this.
Doug Barton: NIST are sponsoring convergence of e-business standards and part of this
is creating a catalog of standards needed for e-business to enable them to interoperate
seamlessly. This is a good starting point for information.
Bill Estrem: In the Architecture Forum later this week he will be introducing a
meta-SIB which will identify what sets of standards work together.
Chris Greenslade: Within the TOGAF architecture methodology one of the first things
when going into a new architecture practice is to establish an architecture continuum, and
part of this is a set of standards.When we point them to the SIB, two problems arise: one
is they may not want to use all the available standards so they need a way to decide which
subset of standards they want (our existing SIB is an Open Standards SIB), and the other
is that many organizations use vendor-specific standards. What is needed here is a way to
filter out the standards they don't want and merge in the standards they do want. To be
really useful our SIB needs to provide guidance on both sets of standards, and the fitness
for purpose of each.
Elaine Babcock: Using a roads construction analogy, where standards don't exist or are
under construction but not yet ready or not yet accepted, we could have pointers to them
to provide visibility that they are under development and will become available in time.
Carl asked for comments on what we would need to include, and on the level of detail
and specificity we should aim for. He emphasized that we are trying to capture as much
feedback as possible - we are not aiming to produce a resolution out of this discussion
today - more we want members' feedback on support for this proposal and priorities for how
to address it.
Allen Brown: He sees that all the Forums in The Open Group are doing something on
architecture - Security Design Patterns, Security Architectures for Boundaryless
Information Flow, Architecture Guide for Identity Management, Secure Email, Secure Mobile
Architectures - yet the SIB is not well-maintained at present. It needs to be updated, and
it would not be a big move to do so.
Steve Whitlock: This would be a good idea - it has been a while since he looked in the
SIB.
Bill Estrem: We need to bring the SIB up-to-date (to include WS-I, OASIS, etc.), and
how to maintain it. Pointers to standards is sufficient.
Tom Bishop: The absolute and most basic thing we should do is have a standards
repository for the hardware and software that is available in an IT environment - a place
where you can go to say what's available. It is an indictment of all of us in the IT
business that the most critical piece of information we don't have in a consistent
accessible place is what is available. Some place where we can record what is installed
and when we install something that information is stored in a standard way. Until we solve
this problem we should not expect to make progress.
Billy Mitchell: His background is telecoms, where it was always a specific business
issue that drove standards. We could architect things to the nth degree, but it is easy to
get into an analysis-paralysis - rather if we know the business application that we're
shooting for, it becomes much easier to architect the architectural layer. Then we can
construct building blocks - e.g., for voice over IP, or a broadband modem service - where
developers were able to see where they were heading and aim for and hit certain benchmarks
along the way, while operators could deploy staged developments knowing what is coming and
so roll out the required new technology solutions. That way, the Supply-side could be
assured they were developing what the operators want and would buy.
Carl: Agree there is a market dynamic to this issue. Suppliers are not opposed to
standards unless they intrude on areas of product differentiation between competitors.
However, it is important that the Customer-side tells the Supply-side what they want as
part of the standard offering and what can be kept as the product/feature differentiators.
Walter Stahlecker: The SIB is a linear thing that is not easy to navigate; it would be
a significant help for Customer-members to tell The Open Group how they structure
standards in their organization, so we know what categories they use - we could pick that
up and maybe create from it a common structure for the SIB that will help members find it
most usable.
Chris Greenslade: Heartily agreed with Walter's suggestion, but clarified that the SIB
was originally structured to be part of the TOGAF system and its service categories, but
as other less-tutored SIB entries have been made the structure has become less obvious -
however, for TOGAF-enlightened members it's child's play to use it.
Bill Estrem: Instead of using the old Technical reference model, we could use the
Integrated Information Infrastructure reference model which is probably more appropriate
for people trying to use Web Services, etc. However, the content for these things also
needs to be added.
Doug Barton: Noted that many standards bodies and industries are being represented -
the NIST initiative he alluded to earlier does have a structure for presenting its
standards, which we may benefit from following.
Joe Veghelyi: Assuming we come up with an appropriate taxonomy for categorizing the
standards, we should recognize that each standards-creating body will produce standards
relevant to their own business drivers, so rather than just capturing the titles of these
standards, the value-add we should provide is to explain the relevance of a particular
standard, the relationship it has with other standards, the motivation to create it, etc.
- it is the context for each standard that is especially valuable in a SIB. From this we
will see how different industries take up particular standards and show both the
Supply-side and Customer-side which standards are being adopted so must be supported in
products.
Carl: Looking at the challenges listed in the introductory slides - we would like
feedback on issues like supplier intellectual property, competitive strategies, and
customer engagement and validation, because it tends to be an add-on.
Billy Mitchell: Agreed that customer engagement and validation is the hardest part. In
the telecoms space there were central places where vendors could go to show their product
and its added values - and be validated as having product that is approved as proven.
Maybe The Open Group could create a testing facility that is not a proper test and
certification laboratory, but is more a demonstration lab to show whether a product does
perform the required business. Such a facility is run by the telecoms service providers,
and it has the added advantage that it is open to all, so the small vendor can demonstrate
their products in the same playground as the large vendors.
Carl Bunje: Agree that perhaps a more visible validation and demonstration facility for
products would encourage greater market interest from both Supplier and Customer sides in
the value of testing and certification, as a way to let the Supply-side know what
Customers do want. This issue here of course is who pays for the facility.
Claudia Boldman: Perhaps the question we are considering here should be put
differently. One of the things a Customer will easily tell vendors is what they don't
want. The S/MIME Gateways certification program is a case in point - they pointed out that
interoperability here was a real pain-point, and they took interest in The Open Group
initially because the problem was important to their business and was important enough for
them to chase.
Dave McCaskill: There are a zillion standards out there; this effort needs to be
bounded if it is to make credible progress, otherwise it is an endless task. Where do we
want to concentrate our effort to kick off the framework, and what should be our focus?
Dave believed security, and information management, are the big issues for him and
therefore what he would like to include within the initial effort. The NIST initiative is
clearly a good place to start.
Doug Barton: Within the business that is dependent on IT, the cost of doing business
within your supply chain is a significant proportion of your overall costs - up to 70%.
You can reduce this cost by eliminating costs in your supply chain. When major competitors
in one industry community (e.g., Lockheed Martin and Boeing) do not agree on what the
common set of standards should be, you're adding costs to your end-products. CIOs in
customers notice this. It is therefore important for industry communities to consider this
in their standards strategies.
Regrettably, time ran out for further discussion. However, we will welcome further
feedback from members, so please send comments to i.dobson@opengroup.org,
or input them to your Forum Director who will know who to send them to.
