Merging EA Frameworks

re: Merging EA Frameworks

Michael Høtoft — Wed, 06 Aug 2008 19:57:15 GMT

Hi Nick,

Dont forget the "secretive" IAF framework from CapGemini.

re: Merging EA Frameworks

NickMalik — Wed, 13 Aug 2008 21:42:04 GMT

Hi Michael,

Microsoft has a framework as well, that we use in our consulting business. I didn't list it. Nor did I list the framework that IBM uses.

Consulting frameworks are not something that a consuming business can leverage without purchasing consulting services. Let a book be written or some very public information made available, and then it may show up in a public comparison like this one.

That is not a slam. It is reflective of the fact that I cannot compare things without information, and these companies are not quick to make their details public, much less expose them to a comparison.

The ONLY reason I list Zachman is because he has published some materials and given sufficient information in the public domain that I can make a comparison.

--- N

re: Merging EA Frameworks

Gaja — Sat, 23 Aug 2008 01:09:01 GMT

It is really possible to merge all the architecture into one? My thought is, there is not going to be one silverbullet that will solve the problem. Just like which project methodology is good, waterfall, iterative, agile... Architecture framework is similar to that. I would pikc the one that fits the problem statement.

re: Merging EA Frameworks

NickMalik — Sat, 23 Aug 2008 04:41:49 GMT

Hi Gaja,

A framework is not a silver bullet. It is a tool. Right now, a set of tools (different frameworks) have been developed to solve similar problems from different viewpoints. Each tool, in its own right, is not complete, so you MUST use many tools. It is difficult to use two or three tools, because every problem has to be solved in different ways. Very disjoint.

But the tools do not have to be disjoint. There can be an integrated tool that helps solve many problems.

The frameworks are all growing to solve new problems, each to fill in their own gaps. They will each grow towards some idea of "complete." In this post, I have provided my idea (today) of what I think complete may look like.

I'm not suggesting that a silver bullet will fix things. But I am suggesting that a better understanding of common problems can lead to a consistent tool that is useful for more situations, and therefore easier for many people to use.

--- N

re: Merging EA Frameworks

Jean-Jacques Dubray — Sun, 24 Aug 2008 01:46:19 GMT

Nick:

you should take a look at Praxeme (http://www.praxeme.org), this is a modern EA framework that was built with SOA in mind.

They have translated most of the documents now (as it was originally written in French).

re: Merging EA Frameworks

NickMalik — Mon, 25 Aug 2008 01:34:31 GMT

Hi JJ,

Thanks for the link to Praxeme. I'll take a look.

--- N

re: Merging EA Frameworks

Richard Veryard — Mon, 01 Sep 2008 11:44:16 GMT

Maybe it's because I'm British, but I think the current version of MODAF is more SOA-friendly than DODAF. (With its emphasis on strategic capabilities, it's also perhaps more Microsoft-Motion friendly.) See my post on MODAF Version 1.2.

http://soaprocess.blogspot.com/2008/07/modaf-version-12.html

re: Merging EA Frameworks

NickMalik — Mon, 01 Sep 2008 22:29:43 GMT

Hi Richard,

You are correct, I should also include MODAF in any good comparison, given their public nature. I do like their SOA work. If anything, I don't think that they went far enough in their SOA support, failing to indicate SOA services at the informational level (As far as I can tell, not being an expert on MODAF).

But their approach is much better than most.

--- N

re: Merging EA Frameworks

BA — Tue, 02 Sep 2008 12:51:56 GMT

I think an ideal EA framework can start out by being generic, but gradually will move towards domain specilization. Eg: Transportation industry will have its own EA framework, that can be used within that vertical. Most of the frameworks mentioned here, suit broadly any kind of industry, and have roots based on a technology framework. Hence, we see lot of system/implementation specific details within such frameworks, but less on governance, enterprise and information aspects. I guess various organizations have to implement these frameworks, in order to come up with a COTS framework which can be plugged and played. Futuristic? :-)

re: Merging EA Frameworks

NickMalik — Tue, 02 Sep 2008 18:01:30 GMT

Hello BA,

I think you are saying: stuff evolves from general to specific... right?

If so, I agree. I'm hopeful that the 'industry specific' frameworks can be variations / adaptations built on top of a fairly solid core framework.

Right now, the core framework is lacking, but it is becoming.

--- N

re: Merging EA Frameworks

Data Modeler — Wed, 24 Sep 2008 20:41:41 GMT

I curious as to why you say the Zachman Framework is highly proprietary given that it is a framework, not a application, not a methodology, and does not specify in any manner what the primative models should look like or which standard they should follow.

re: Merging EA Frameworks

NickMalik — Thu, 25 Sep 2008 03:57:02 GMT

Hello Karen (Data Modeler),

Zachman Framework is owned and tightly controlled by ZIFA, the company run by John Zachman to this day. While John makes presentations on a regular basis, he has not released his framework into the public domain and he earns his income consulting on how to use it, leverage it, and benefit by it.

Nothing wrong with that. We call that capitalism. We also call that 'intellectual property.'

Given that the framework components of Zachman that ARE in the public space (if not the public domain) are without guidance, but there is a great deal of information that is useful, but is only provided when you hire ZIFA, I think it is completely fair to term the Zachman framework as proprietary.

Disclosure: I work for Microsoft. We have some proprietary stuff too. I didn't include that stuff in the comparison at all!

If I had, my adjectives would have been similar. Big difference: our framework stuff didn't change the world. John Zachman's work did.