Enterprise Architecture: The Disconnect Between Theory and Reality

Dave Oliver — Thu, 01 May 2008 16:45:16 GMT

I took up Mike Walkers offer in the previous posts comment section to look at his post entitled Making Sense of Architecture Standards and found it an ideal steer on what standards are out there, where they are positioned, how they fit together and who will use them and when to use them.

However, what soon dawned on me was that the disconnect between theory and particular context is huge! Many of the standards are no more than helpful steers that as yet don't offer that distilled level of knowledge that is the mechanics and the moving parts of a functional EA initiative, this you have to make yourself. This perhaps may explain why only larger organisations can take the Enterprise Architecture plunge as the amount of work needed to turn high level theory into a userable practice for individual circumstances will take super-human effects and a deal of time. All of which only the larger of organisation can afford.

Taking Henry T Ford's principle of creating masters and experts of easily manageable domains and string them together into a production line is something that as yet hasn't happened in the Enterprise Architects space, we are still the equivalent of the master craftsman in a cottage industry. When Tim Berners-Lee talks about the immaturity of the IT industry, he need not look any further than the Enterprise Architecture for a shining example of that.

One of the great things about the Open Group is the forums where Enterprise Architects can meet and discuss but it is a club that is not free to join, there is a fee. The fee is large enough to mean a justification case and not every organisation, specially now, can make a case stick, especially if the core industry they are in isn't IT. The upshot is that the Open Group seems to have more than it's fair share of IT companies and weirdly not Microsoft which for all the talking is still the one of the biggest players in the game. Like or loath them, you can't ignore them and no-doubt nor does your business.

So where do Enterprise Architects meet then? I've found a whole bunch of EA's at ITToolbox and would be curious to find some more hiding places of the lesser spotted Enterprise Architect. At least here the questions I would naturally ask have at least been tackled in some way shape or form and afew kind souls even have offered their particular solutions up to the web for inspection by others and relish the scrutiny of their peers. This all adds up to meaningful long running conversation that can be followed and inputted into by all, which seems to be the nature of creation these days on the web, i.e. results by collaboration.

In the past an industry or practice had to rely on thought leadership and some inspired person to drive through change and create a following. John Zachman has tried to do this with the self-titled framework. I think in the connected world, this approach has a limited time left. The technology is here to allow hundreds, if not thousands of individuals to help build solutions, an individual making their little piece and letting others refine it. Open Source has show us that this can be done. I see no reasons why methodologies can't be constructed in a similar fashion?

In this day and age where reputations have to be earnt and transparency is the only sure-fire way to credibility and the price-tag has got to be free, thought leaders and committees have a limited appeal and therefore a limited reach so for a methodologies or frameworks to succeed will depend on whether they can be constructed in the new standard. If not, then we will all be in our cottage's or 'silo' for sometime to come, finding our individual solution to a common problems and interpreting the things like TOGAF's ADM a 1000's times over. Just think about all that wasted 'brain-time', when it could be spent putting the touches to a small part of an industry grown common solution, owned by no-one but everyone. There's a thought.

Enterprise Architecture: TOGAF Certification

Dave Oliver — Fri, 18 Apr 2008 10:38:16 GMT

Recently I attended a TOGAF certification course up in merry old London town run by Architecting the Enterprise (AtE). My certification should show up in the list of other 5000 odd architects that have passed in the next few weeks.

So what is TOGAF in a nutshell? Well it's a very generic architecture framework that helps build, in a non-prescriptive way, how to develop parts of an Enterprise Architecture capability.

Notice the caveats, 'very generic', 'non-prescriptive', 'develop parts'. What am saying is that if you want an idiots guide to implementing Enterprise Architecture in your organisation, this is not it and forget that quest because it doesn't exist! TOGAF sits a few levels above; it is the high-level plan that on specific subjects delves a little deeper with advice, guidance and best practice.

TOGAF is a product of the Open Group that is a vendor neutral IT standards organisation. You do get the impression that it's been designed by committee but try and get over any prejudices you may have on that score. There aren't that many other ways you can get many different and competing parties to agree on subjects.

TOGAF and the course initially left me feeling pretty annoyed, the reason being, I was left with more questions than answers. However, after sitting down in a quiet spot and analysing my thoughts I realised a few home truths,

More questions than answers is actually a good sign! Weighing it up, most of my questions are now smaller in scale, there are still a few big questions but not as many, so this is indeed a sign of 'thinking in progress'. TOGAF does give you a starting point, you have to fill-in not just the blanks but interpret how to make all elements work in your organisation, no two organisations are the same.
TOGAF isn't finished, it's pretty obvious it's a 'work in progress', some parts are more mature than others. My advice would be, have a go with what there is if you don't have a better equivalent, it's better than starting from scratch.
TOGAF is generic that means it is not going to cater for your individual organisation and circumstances neatly, there is a risk of 'if at all'. It is just another demonstration on how immature our industry is as a whole and really how few standards we have, there isn't a great deal of agreed collective thinking, just a lot of very similar problems. So get over the fact that you are going to need to do some thinking and work stuff out. TOGAF isn't an Enterprise Architecture flavoured Pot-Noddle.
Am I really going to make a better job of it than the Open Group? I admit that I can be arrogant buoyed by the confidence that no outsider is going to know my set of EA issues better than me. However many of my problems are similar and therefore are going to have similar solutions. It really does pay to have an open mind.

So am I a fan of TOGAF now? Well, I'm past the stage of 'it's better than nothing'. I've got to a point where I am beginning to respect it.

So is TOGAF better than the other frameworks? Well the answer to that is pretty simple, what does 'better' mean in your context? Microsoft, one could argue is a 'standards bodies neutral' vendor and has a very good article comparing the different Enterprise Architecture frameworks giving you a method to work out what is the most appropriate framework for your organisation.

For my organisation, the 'whether to use TOGAF' debate is still rumbling on but for my immediate colleagues and I have decided to run with it and see how far we get. Consequently, you are going to see TOGAF influencing this blog as I will report back my experiences as the journey continues.