Re: [soa-rm] end plans for SOA-RAF

[Rex Brooks <rexb@starbourne.com>]

I don't think we should necesarily dismiss this idea out of hand, but I 
also don't think that we should break up the current document either.

I think that SOA Governance and Management are not critically 
architectural issues in their nature. They are, in fact aspects of all 
parts of the information architecture that we do not cover in the depth 
that they deserve. In part this is because we lack the expertise, but I 
suspect that we are not alone in this and it is not our unique deficit. 
As business moves more and more into the use of SOA, an understanding is 
evolving that all parts of the SOA Ecosystem require governance and 
management.

In fact, unlike overall business management or IT-system-specific 
management (of which SOA Governance and Management is still just one 
part), this is an area that is not well understood. tI think it demands 
its own careful study. I suggest that what we may be faced with down the 
line is the necessity for reinventing the science and art of governance 
and management for Enterprise Architecture as a whole rather than just 
focusing on the SOA aspect, and, perhaps this may apply for government 
as well. However, be that as it may, I am NOT suggesting we even 
consider tackling that.

I suggest we consider leaving the less well-developed sections in, but 
rework them SLIGHTLY to emphasize that this is an evolving area and that 
we are considering that it may need its own specification, e.g. SOA 
Governance and Management. In fact, it may be that we need to rewind our 
thinking about this area all the way back to generating its own 
Reference Model before a more elaborated model, for which I don't 
actually have an appropriate name right now.

Cheers,
Rex

Thornton, Danny R (IS) wrote:
>
> I would agree that the proposed alternative would require no less or 
> more time to support than moving forward with the current SOA RAF 
> document. I also understand the intent of extracting pieces into 
> separate documents to have concise and even documents to choose when 
> socializing SOA concepts. However, when pulled apart section three and 
> parts of service description are about the only parts of the document 
> that could stand alone. There is much information published similar to 
> what is in the other sections of the SOA RAF. When pulled apart in 
> this way I think the support for and extended meanings for section 
> three and service description are diminished.
>
> This will make for a good discussion.
>
> Danny
>
> *From:* Francis McCabe [mailto:fmccabe@gmail.com]
> *Sent:* Tuesday, April 20, 2010 8:34 PM
> *To:* Laskey, Ken
> *Cc:* soa-rm@lists.oasis-open.org
> *Subject:* Re: [soa-rm] end plans for SOA-RAF
>
> A couple of thoughts:
>
> 1. It would have been good to have had to opportunity to discuss this 
> in the RA forum before bringing it to the RM
>
> 2. It seems v. complicated - a process within a process. We do not 
> have the bandwidth to manage it.
>
> 3. The net effect will be to gut the spec.
>
> 4. I suspect you seriously underestimate the amount of busiwork needed 
> to partition the spec into standalone documents.
>
> 5. The RA was designed as a whole; not a whole collection.
>
> In short, I am not in favor of this proposal.
>
> Frank
>
> On Apr 20, 2010, at 7:49 PM, Laskey, Ken wrote:
>
>
>
> This is one of the agenda items and I thought it might be useful to 
> provide an unusual approach in order to get us thinking.
>
> The underlying assumption is we do not have the cycles or resources to 
> complete all the unfinished sections we believe to be important for 
> the RAF. One end state is a Committee Spec where we have things at 
> various levels of completeness. I consider this to be an unattractive 
> option because I think that some good work will be overshadowed by the 
> portions that are still weak. If everything in the document was at a 
> consistently high level of completion and some areas just weren’t 
> addressed, I think that would be acceptable. An uneven document is 
> less so.
>
> An alternative I’d like to propose, though not fully worked out, may 
> be the following. Let us re-couch the RAF as a collection of 
> expositions on various aspects of SOA. Some of these have been through 
> significant review and can be identified as something like RAF Key 
> Aspects. Other areas that are less developed can be identified as RAF 
> Aspects under Development (RAF-D). Areas where we have said little can 
> be RAF Aspects for Consideration (RAF-C). The process going forward 
> would be to keep the TC together as a forum to keep developing these 
> areas as resources and interest warrant. However, there would not be 
> weekly meetings and much of the work going forward would be done by 
> individuals or small groups. For example, white papers such as the one 
> on Willingness that Dave Ellis and I are supposed to write could begin 
> as a short collection of key thoughts under a RAF-C. As we have time 
> to further develop our thoughts, this could advance to a RAF-D. The 
> RAF-D and RAF-C phases would not require TC review and approval, 
> although comments would always be encouraged. It would, however, take 
> a TBD approval process (and likely a meeting of the whole) to advance 
> to a RAF Key Aspect and there would likely be TC action if there was 
> significant heartburn with something else that had been written.
>
> I believe the advantage of this approach is we keep together a forum 
> and a product that can grow and can encourage others to participate. 
> We can continue to grow material as experience with SOA grows but with 
> a low level of effort from the group as a whole. It also is not an end 
> state where things are left permanently hanging.
>
> This idea is not completely thought out, so there is room for 
> improvement. However, if we do this at the level of Committee Spec, I 
> don’t believe we run afoul of OASIS process and can continue 
> advancement of a product that the organization can point to with a 
> continuing sense of accomplishment.
>
> Again, this is something out of left field but it can hopefully stir 
> some creative options.
>
> Talk with you all in 13 hrs.
>
> Ken
>
> ---------------------------------------------------------------------------
>
> Dr. Kenneth Laskey
>
> MITRE Corporation, M/S H305 phone: 703-983-7934
>
> 7515 Colshire Drive fax: 703-983-1379
>
> McLean VA 22102-7508
>

-- 
Rex Brooks
President, CEO
Starbourne Communications Design
GeoAddress: 1361-A Addison
Berkeley, CA 94702
Tel: 510-898-0670