Gartner EA Summit – Day 1


Mike Walker's Blog: Gartner EA Summit London 2008 So I figured since I already started a thread on the Gartner keynote (Gartner EA Summit - Emergence Enterprise Architecture Keynote) I should finish what I started, huh?

Just so everyone knows, this is Gartner's London EA Summit event and not the North American event. It does seem a lot smaller than the other EA events. I am not sure if it is the case in attendance or just the facilities but it does seem a lot smaller.

I will do a recap of my key thoughts so far on the conference. I hoped around a bit between different tracks based on my interests. Keep in mind I am not going to divulge too much on Gartner's materials here. I'll mostly offer up my opinions on the sessions.

Understanding the Market and selecting an EA Tool

This was the first session I went to after the keynote. I think Robert Handler did a good job of presenting this topic but it is pretty clear that this is for folks that are new to the EA Tools space. For the most part it was very much 101. Keep in mind this isn't a negative comment, just the context.

Even though it was some what beginner in nature there was some fresh and good nuggets in there that are worth pointing out explicitly.

  • The notion that there is a trade-off between the Usability (Simple/Clean/"Keep it Simple Stupid") and Functionality (Rocket Scientist, "You want it, we got it")
  • EA Tooling Status
  • Recommended using Office or what is referred to as the EVP Suite if your organization is at a certain level of maturity (Gartner's Level 1 - 2). This makes a lot of sense and this is the value proposition that we have proposed to organizations with the EATK.

What I did learn in the session was a status on the EA Tool vendors. If you follow my blog you may of seen my simple assessment of this space based on my experience. I talk a lot about the ability to execute, fragmentation of the EA Tooling market and so on. There were very similar themes in the status of tooling. One thing that I didn't factor in was the consolidation of tooling and the impact on existing players. I want name names, but there were some well established names in the presentation that were mentioned as declining due to the impact of consolidation in the market.

One other observation here is that when Robert talked about EA Tools he evaluated them on features. While this has worked in the past, I think we should move toward a capability type of approach when evaluating tooling. I think we will start to realize that what we traditionally thought of a EA capability is really something we already had, such as workflow, security, ECM, repo, etc.

What's Next: Context Delivery Architecture

This was an interesting one for me. When I looked at the title, I think I was expecting something much different. There were some things I liked about this topic such as the notion that this is another Architecture Style. I also, thought there were some interesting nuggets littered throughout.

However, I found there wasn't much meat to the presentation. Maybe it's too early, but I would argue that the Web 2.0 space is all over this and has a ton of great material. Microsoft as an example, is all over this. My colleagues back in Redmond such as Karthik Ravindran, Paul Stubs and Masashi Narumoto (see: Contextual Service, Contextual Service Example, Architecture for Consumer Oriented Services Cont'd) are doing some really cutting edge stuff around this very topic. I really didn't see the holistic pitch around the space mentioning who and what are the key technologies to enable this space. There was a lot of redundancy around Facebook and Nokia. For example these were not mentioned:

  • Myspace - Social Networking
  • Twitter - Microblogging
  • Microsoft - Live Services
  • Apple - iPhone and GPS location aware apps and games, music recommendations, etc.
  • Zune - music recommendations, subscriptions, social music aspects, etc
  • XBox 360 - Gamer Tags, Persona aware, Messenger Integration
  • Zooomr - Photo geo-tagging
  • Plazer - Location Services with integration into blogs and micro-blog engines
  • etc.

There were also a lot of non-qualitative numbers being thrown around. There is probably more data points behind the scenes in research papers but would of really liked to see that information. Information like, where did this data come from that drove these predictions (enterprises, consumer application development, hobbyists, micro-ISVs, SMBs), probability of the prediction and the impacts on current designs and decisions.

I do think that this is going to be part of the future of IT especially as the next generation IT folks join the ranks. It is going to be expected that context services are enabled. The way in which we do things will seem somewhat antiquated to future generations and they will expect these automation services to be available.

Adapting TOGAF Globally

I both had the pleasure of attending Joel Goodling's presentation and also having conversation over dinner as well. He did a great job articulating the value of the using a general purpose EA framework such as TOGAF.

He walked through how he configured TOGAF to fit inside his corporate culture. By being mindful of that terminology and existing organizational structures will play a large role in the success of TOGAF he did some very smart things like renaming the ADM phases. He also removed some of the ADM steps (mitigated that with baking those missing bubbles into other processes).

All in all a great session.

Communicating the Value of EA

I may of become a new fan of Brian Burke. He really hits the nail on the head. There were a ton of tips in this presentation that should really be taken by all architects not just EAs. I would of actually renamed the session to something like "Communication is key to an Architects Success". Needs a little work, but you know where I am going with it...

Even though this is still a bit of 101 as well Brian highlighted some good situational leadership ideas. He reinforced the need for clear communication plans and the notion of communicating frequently even if the message is going to change.

 

Skip to main content