Friday, January 30, 2009

The Two Sides of SOA ... take a look at the flip side

A lot of folks still think SAP is falling behind in the SOA space, especially after reading media that paint a limited view. For example, Informationweek online in Germany recently only referenced one Gartner Report, in which SAP was shown as lagging behind IBM, Oracle and Microsoft.

»Magic Quadrant for Application Infrastructure for New Systematic SOA Application Projects«.

In this study, SAP only falls into the "niche" quadrant, which lead to a lot of commentary in the past several weeks.

However, if you look more carefully, you'll see that this report only refers to situations where companies are building up their SOA infrastructure "from scratch". Honestly, this is neither the area SAP is targetting, nor does it sound like a good idea to proceed with this course of action when you have an existing application landscape in place that is already SOA-enabled.

This is where SAP is focused. Delivering a SOA-enabled set of end to end industry processes, that can act as your business process platform, to allow rapid development of composite applications. In this space, the situation is quite different, according to this report, also from Gartner.
»Magic Quadrant for Application Infrastructure for SOA Composite Application Projects«.

SOA has two sides, and this year the difference will become more clear.
- The "build from scratch side". (Have fun programming business logic)
- The "leverage a SOA-enabled application platform for rapid composition side"

The choice should be easy to make based on your requirements. But I think we also still have a steep hill to climb to explain this story to all of our customers, especially since it the SOA-hype has subsided and there is very little "innovation buzz" around concepts like "use what you have".

These days, however, my reaction would be "Sounds good to me" ...

What do you think?

Thursday, January 29, 2009

SOA has arrived ... now what?

These days, it has become a bit more quiet around the topic of SOA.

This is very good news. It means, that it has arrived, and people are now taking advantage of service oriented architectures, as opposed to experimenting with it, or debating it. This is a very normal process and the same happened to other hype topics in the past.

Now what? Should we just move on to the next topic "de jour", the next 3- or 4-letter acronym to hype up, and forget about SOA?

I think not.

I think in the next 2 years, especially with the brutal pragmatism imposed all of us in IT by the global economic crisis, it is more important than ever to speak about service orientation. Forget for a minute about the architecture, and focus on where the whole concept came from in the first place.

Service oriented.

For me that implies looking at the business first, at the value your business provides, and how the applications you are running are supporting that value. Once you have clarity around that, you can reap the benefits of service oriented architectures.

If you just talk about SOA, business people will just roll up their eyes, as opposed to their sleeves, and you will not get funding for a project.

I think Software as a Service, On-Demand and cloud computing are the killer application for SOA, and will truly help us along on our joint journey to Software AT your service.

But we all need to talk less about SOA (technology) and more about business services and business value to succeed in making technology more relevant, and significant to any business.

Thats my call to action and new years resolution, what is your's?

Wednesday, January 28, 2009

2009 is the year to get into the cloud .. but not head first

If you listen to the pundits, and I really try to do that as many times as I can, 2009 will be a year of transformation for the enterprise software market. The large incumbents like IBM, Oracle, MicroSoft and SAP are slated to gain share of wallet, there is going to be a consolidation of the "stack" (StackWars), and the emerging concepts around Software-as-a-Service, Cloud Computing and Virtualization are predicted to gain traction.

So far so good. Ho Hum. Any real news here?

I think 2009 is the year to get into the cloud ... but, like I mention in the header, not head first.
Cloud computing to many sounds like a renaissance of mainframes and time share, especially when you combine it with the benefits of virtualization and Software-as-a-Service consumption model.

The real breakthrough benefit, however, is not delivered by where the software is running. Or by how it is being consumed. I believe the transformational power of this concept is rooted in where it has its source.

The source of the powerful appeal of cloud computing is not the concept of taking existing enterprise software and "sticking it" into the cloud. This will not work, especially in the case of highly customized, industry specific enterprise software, where there are just too many issues around integrity, elasticity and availability to be a real alternative to the current on premise model.

The source of the powerful appeal of cloud computing is the way software will be developed for the cloud ... by combining it with the "wisdom of the crowds" approach of social media. I believe "Cloud-Ready" software will evolve more organically than the software produced in the classic prescriptive programming models of the last century. There will be trial and error, and many many feedback loops. There will be collaborative innovation, and incremental break throughs. There will be customization on the fly and there will be mashups at the personal level, to really make the software fit your pesonal style. And there will be powerful integration with the existing software, already running in the enterprise. And, as I postulated in my last post, there will be a continuation of Middleware, in the guise of Service-Ware, or Process-Ware, that will not run in the "middle" of anything, but "around" everything to deliver governance, integrity, consistency and transformation.

That, in my mind, is the appeal of cloud computing.

The Cloud, like so many concepts in the past, will not replace anything in computing, but add a new, exiting, much more fun dimension to enterprise software.

That's my 10 cents for today (P.S. 10 cents vs. 5 cents because of inflation)


Tuesday, January 27, 2009

Is Middleware a concept of the last millennium?

I would like to kick off my blog with a question.

As software evolves to an online consumption model, is Middleware a concept of the last millennium and will be replaced by an "integration as a service" model? And if yes, by when?

To this point, this does not seem the case, especially in enterprise software. In the enterprise software space, it becomes ever more important to ensure process integrity not just inside the walls of the enterprise, but across the entire value chain. And that "chain" evolves more and more into a true collaborative network, with rapidly changing endpoints and governance models.

But maybe the concepts of Middleware have to be re-thought?

Maybe its no longer about something that sits in the "middle", but something that "surrounds" the processes and provides services to deliver integrity & governance, transport & translation and more.

Maybe Middleware is dead. Maybe this is the age of ServiceWare? ProcessWare?

I look forward to your thoughts and comments on this