SAP Watch

Nov 3 2008   10:31AM GMT

How can SAP get back on track?

JackDanahy Jack Danahy Profile: JackDanahy site expert Axel Angeli isn’t one to pull his punches when it comes to SAP and he makes no exception in discussing SAP’s technological direction in this guest blog.

Has SAP lost its mojo?

In the good old days, SAP was an extremely successful technology company and the darling of many analysts. This was in the last millennium, before SAP lost its belief in its own strengths and virtues.

In 1998 SAP undertook its last breath-taking act of innovation when it introduced the BAPI framework. BAPIs had been designed to transform SAP’s transaction-based system into a component-based service suite where any functionality would be programmatically accessible. Back then, SAP was already an SOA-aware software package! But this successful path was needlessly abandoned in the years to come. The old SAP crew that defined the company’s technological success — in terms of ABAP, RFC and Batch Input — was mugged by the dogma of Java.

Even though it was implemented in the mindset of assembler programming, Java is a language that decorates itself with the feathers of object-oriented programming. It is unstable, unreadable, incomplete and completely redundant, since it did not introduce a single new feature that the world had been waiting for.

How should SAP escape the Java trap? Go back to its old merits. Make a clear and non-negotiable decision in favor of ABAP. This would also mean polishing up ABAP with a more modern syntax. The new ABAP 7.2 kernel has already taken some first steps in this regard. Technically the rudders are in the right direction. What is missing is the clear commitment of the SAP board.

In terms of SOA, there is another problematic area that requires immediate action: Process Infrastructure (PI). SAP doesn’t have the power to make technology changes from the inside and needs to shore up PI by buying a standing technology as an enhancement package for the existing stack. IBM bought Mercator Datastage TX, Progress took IONA, Oracle snagged BEA and Software AG showed mercy to suffering webMethods. The number of possible candidates for purchase is limited. If looking for quality products with an inherent Event-Driven ESB architecture, there are mainly Fiorano, ActiveBPEL or the not-for-sale SAP partner Seeburger. 

In part one of this blog, published last week, Axel Angeli discussed why he isn’t fond of SAP CEO Leo Apotheker, SAP’s service fee increases or elements of NetWeaver. In the next part of this blog, Axel will explain what SAP is, in his opinion, doing right.

4  Comments on this Post

There was an error processing your information. Please try again later.
Thanks. We'll let you know when a new response is added.
Send me notifications when other members comment.
  • throwexception
    "Even though it was implemented in the mindset of assembler programming, Java is a language that decorates itself with the feathers of object-oriented programming. It is unstable, unreadable, incomplete and completely redundant, since it did not introduce a single new feature that the world had been waiting for." Wow... Someone has an Axe to grind with Java. It didn't become one of the main language of server-side development without merit. Even .NET and C# borrows heavily from Java's syntax and interpreted language architecture. Not sure what your issue is with Java, but outside of SAP, it is one of the great things out there.
    0 pointsBadges:
  • Michael Koch
    Alex, I don't share your general criticism with regards to Java. Even though I am an SAP Dev Consultant with strong ABAP background, to my mind there is no doubt that the Java language is one of THE IT success stories of the last 8-10 years. I think one reason why SAP jumped into Java was the fact that there is more Java know-how around than there is ABAP know-how. Therefore Java Developers are cheaper than ABAP resources. Why should SAP at this stage publicly declare that it abandons its Java products in the near future and potentially lose customers? Isn't it much wiser -especially in the current climate- to communicate that you develop ABAP stack further and maybe then offer a migration path in a year or two? On the whole I am not entirely sure how valuable it is to discuss the death and demise of one SAP stack to another. There is a general shift (eSOA, BPX, openness) going on, regardless of programming languages. Kind regards, Michael Koch
    0 pointsBadges:
  • Jan Petersen
    Generally speaking I think you have a good point. Having worked quite a lot with several Java based "packages" from SAP, I have to say that I don't think SAP has made any significant difference in the Java space (mildly put). However this does not mean that the Java language is the problem. Claiming that Java is "unstable, unreadable, incomplete and completely redundant" just reveals that you apparently have no experience with Java other than that of SAP (which is a painful one). Today most companies from banks, insurance to industry are using more and more open source Java based tool. SAP doesn't see and understand these important trends. Another problem, as I see it, is that SAP are failing to realize that the company are simply not able to build everything them selfs - maybe this is German thing (no offence) but the future is neither monolitic nor "closed". Integration plays an increasing role as well. SAP should focus on what they are good at and leave the more infrastucture like work to others. In other words: Don't waste resource building a RDBMS, Java VM, Java EE server etc. Concentrate on aspects where SAP can make a difference. Best regards Jan Petersen
    0 pointsBadges:
  • Panayot Dobrikov
    Jan - you are only partially right. I agree with you that statements of "unstable, unreadable, incomplete and completely redundant Java" are incorrect, and to normal java developer, or company that makes extensive use of Java this statement might look funny. We all know that Java is used for mission critical business, already, and noone can change that with marketing. You, however, should also understand the amount of operational experience that SAP has, and is bringing in the equation, which is partially the reason for such statement, I guess. Some of our solutions involved thousands of Java components running in single JVM - no other application server or java infrastructure on the market can handle that with the neccessary modularity, isolation, lifecycle management, performance, scalability and so on. I also disagree with your statements about core/non-core. If your statement is correct, it would imply that Oracle should do nothing else than DBMS, and Microsoft should do nothing else than OS and office. But they do. And there is good business reasoning in that, and good customer value in doing it. SAP has also proved of being open in both integration open source, where relevant and appropriate, as well as donating and contributing back. Also, recall that SAP was the first certified Java EE 5 vendor, and the first one that promoted Eclipse/WTP as default Java EE toolset. This was in moments when Java EE had little support from major vendors. Every other vendor followed in that later. Finally - on the "german thing" - agh, you know, I still like my Porsche better and won't change it with Ford (no offense implied too, and good luck to all people around the world in these crisis times).
    0 pointsBadges:

Forgot Password

No problem! Submit your e-mail address below. We'll send you an e-mail containing your password.

Your password has been sent to:

Share this item with your network: