It hit the OS space with Linux, it changed the world of encyclopedias with Wiki and it capsized the media space with blogs. ‘Open’ is the word to watch for and get ready to see it in the esoteric nook of SOA. On another of those memorable trips to India, Rob Levy, Chief Technology Officer and Executive VP of BEA Systems does some crystal ball gazing for us. Excerpts:
Are you content with the market response so far to BEA 360?
It’s too early to judge anything from initial response. It’s not a market campaign but a formalisation of what we, at BEA, have been doing and intending to deliver.
What are the contours ahead for the realms of SOA (Service Oriented Architecture)? Do you see application market here moving beyond the enterprise level?
We see SOA as a mechanism for changing the paradigm of application development in future. I see two trends in the SOA space. One is the enterprise level development and another is the action on the social level. Till now application development is purely enterprise in a top-down approach, where a business problem leads to a product that moves from project owner to developer and finally pushed on to the customers. In the future, SOA will embrace the ‘outside-in’ wave. Look at the way Wikipedia changed the landscape of information and encyclopedias by letting the people outside trigger a topic, create something in multiplication, improve upon it as and where via participation of a mass of users. Same way, what if the customer wants to do something about a new problem, he can enter this futuristic application development space, find the open tools he can use, do something and let it stop there in a bucket in a disintegrated form. From there on, either someone else will pick up the intermediate application and do something more on it. Or people may collaborate together and thus applications are constantly created and sustained in a fluid space. The world of applications and SOA will change from ‘inside-out’ to ‘outside-in’.
Does that mean you endorse the open standards philosophy?
Yes. In fact SOA relies on loose coupling of components while they are properly integrated for business applications. The knowledge about an application doesn’t rely elsewhere, but in the system itself, in the components of a SOA. It helps someone who might want to know more, modify or try some permutations and combinations over these components and in turn create a new application altogether or address a new business problem. It’s like two strangers who want to interact for a given problem or idea. One’s native language is French and the other’s Hindi. They are happy and successful in these languages on an independent basis but when the issue of a temporary interaction comes, if each of them knows a common language, say English, the problem is solved. It’s all about a common platform.
Would this open world of enterprise application or SOA lead to security issues?
You need to understand this from a fraction anatomy. For instance, cheque processing is a small part of credit management that in turn may be a part of private banking. Now if we zoom on the smallest part here, it is as complex as its parent part. So security increases even at the minutest fraction level. It will move from identity checks to entitlement screening as SOA takes futuristic shapes. A person would be given access to a component not on the basis of who he is but on the basis of what his activity and role area encompasses.
How will this translate into commercial implications and business models for companies like yours?
Business will become more dynamic and thus more open. An application will be built under the premise that tools inherent in it would be used for a new application which is not there till now. Businesses will have to turn more agile. Companies will not make a product perfect initially in this future scenario. They will release it fast, let the users re-use, rebuild or give feedback after which, the product can be improved upon and released again into the market. What’s happening now is that companies do double tests, strive for perfection and that leads to delays in product rollouts. In the outside-in approach, we will do it fast and fix it later after customer feedback that will bring more and sustainable value. It will also allow for greater participation of smaller players besides the users.
Compared to the current scenario which is more like a buffet where one has to pay for a seat and time in the restaurant no matter what and how much one eats, the scene in open-component set up will be need-based. Options will range from pay-per-use to open to packaged software depending on a user’s costs and needs. Instead of the subscription model, the next level will see more of web-servers and pay-per-use models.
Does that mean you admit the impact of companies like Jboss in the application space who are pitching themselves on the open-wave?
Sure these companies do exist. They started late and learnt from us and are more significant from the question of ‘Do we need a third player in a market reigned by BEA and IBM’ than a threat? It’s more of a development approach than a business threat. And in general, as a CTO, I love innovation. Open source offerings are a valid development approach. We like the concept because it encourages innovation. I know that the power of 5000 people applying their minds on a development outstrips the power of 500 people. But managing 5000 people is not a cakewalk, it’s a control issue.
Can you elaborate on that?
An upside in the open source case may at the same time be a downside too. A developer in an open community for instance, does need to work on a problem he doesn’t feel like working on if it’s not appealing enough. But for my company, it’s an imperative because my client needs it. Testing on platforms is not glamorous but a customer might require it on a priority basis. I think hybrid models make more sense here as they take care of the open community needs (for developers) and closed system needs (for customers). This is something BEA has entered into with the acquisition of Kudos.
But have the open-source companies caused any dent to your market shares?
They would never dent it, on the contrary, they will help us expand it as newer applications and markets might be thrown open.
Your comments on the low-cost proposition by these players vis a vis the high-priced packaged software?
In life, there are no free lunches. One gets what one deserves. With factors of time, maintenance, total cost of ownership and returns, a CIO sees the bottom line and not what he is shelling out at the moment. Open or packaged, you eventually pay somewhere, nothing comes free.
© CyberMedia News