More on the TOGAF conference

Okay, back ‘home’ in England after the TOGAF conference in San Diego. Time to reflect a bit.

First: a real sense that I’m not as on my own in my approach to enterprise-architecture as I thought and felt I’d been: there are a lot more folks out there now who recognise the inadequacies of standard IT-centric TOGAF, it’s just that in many cases it was still at the level of a feeling of discomfort rather than explicit articulation.

(In that, I owe an apology to Len Fehskens and Walter Stahlecker, who did indeed articulate that discomfort at the TOGAF Munich conference last October. After I first saw their presentations at Munich on ‘the future of EA’, it did feel that a fair bit of had been all but lifted from the conversations I’d had with them at previous conferences; but I now acknowledge I’d done an Isaac Newton, claiming exclusive ‘possession’ of ideas that were more out there in the general aether. The simple fact is that they’d arrived at much the same conclusions as I’d done, but each from an entirely different direction: I should have celebrated that fact rather than being annoyed about it! :wrygrin: )

Anyway, for me, a lot of very good conversations: the mood seemed far more receptive than before to ideas about the need to get out of the IT-centric rut and move to a more explicit whole-of-enterprise perspective. Having the books definitely helped in that: in street-value terms, I must have given away something like $3,000-worth of books, plus probably much the same in e-books, but it meant that I had something concrete and (literally) tangible to back up my thesis about the need for a broader EA scope, and it certainly helped in terms of establishing credibility. It was really noticeable, though, that the people who picked up on the ideas quickest were almost all outside of ‘mainstream’ EA – either in non-information-centric industries and contexts (such as one of the US federal government departments, or again a large logistics operation), or from countries outside of the US/British ‘axis of IT-centrism’ (such as Norway, Malaysia, Japan, China, Switzerland, and, of course, the Netherlands).

Some parts of the conference were excellent – particularly the business architecture sessions led by Bob Weisman – but some were appalling, bluntly. The lead keynote speaker said almost nothing useful beyond sales-pitch, and even somewhat sarcastically that EA was irrelevant to his own work – which was not a good start…  And at least two of the plenary sessions on cloud-computing were blatant sales-hype, with nothing of substance behind them at all: a bit disappointing, to say the least (which to my mind was true of the entire cloud-computing hype, to be honest – I’m seeing all too many memories of the ‘Business Process Re-engineering’ farrago a few years back, such that it’s clear that no lessons have been learned from that debacle at all). But there were some definite highlights, too, such as Bob Weisman‘s presentation on “Enterprise architecture: the strategic tool for innovation in tough times”, and Chris Armstrong‘s presentation on “Agile enterprise architecture”: in that sense, it was worth going there, regardless of the TOGAF 9 launch.

And TOGAF 9 itself? Well, I’ve had more of a chance to look at it in depth (i.e. something to do in the long long waits at airports, and on the flights themselves…), but I’m still disappointed at the lost opportunity that it represents. To be fair, The Open Group is focussed on “boundaryless information flow”, so the over-emphasis on IT should hardly be a surprise; and the history of TOGAF itself, certainly from version 7 onwards, represents a slow climb up from the IT-centric depths. But although the Open Group may need to emphasise information above everything else, that isn’t true of the enterprise-architecture discipline as a whole: and since TOGAF is the leading framework here, that imposes some really frustrating and unnecessarily arbitrary limitations on where and how we can use it. Hence the disappointment.

There’s no doubt, though, that from an IT-architecture perspective, TOGAF 9 is a huge improvement on the previous version. There’s been a lot of clean-up, it’s far better structured, the Content Framework (adapted from CapGemini’s IAF, apparently) and Capability Framework (from Bob Weisman) look like a good basis for future standards for interoperability and architecture governance. And there’s some explicit guidance on how to link across to SOA and security-architecture – though, like me, some of those practitioners are a bit disappointed that the links don’t go far enough into their respective spaces.

Yet despite all that good effort, it still doesn’t work properly for iterative architecture, or for anything outside of an IT-centric scope. And the reason is exactly the same as before: the absurd assertion that all enterprise architecture can be crammed into a fixed scope of ‘anything not-IT that impacts on IT’ (the proper meaning of what they term ‘business architecture’), ‘information systems architecture’ (IT-only) and ‘technology architecture’ (again, IT-only). It does sort-of work for low- to mid-level EA maturity; but it acts as a rigid block against moving any further in maturity-levels – and that move is what business is demanding now.

The good part, I suppose, is that the critiques and solutions I developed in Bridging the Silos and Service-Oriented Enterprise apply to and work just as well with the new version as they did with the old. I’ve now set myself the target of doing a new ‘TOGAF 9 edition’ of Silos in time for the next Open Group conference in London, in April: on that, Watch This Space, as usual?