Skip to Content
In this thread here: Preparing for SMLT (“Was dem einen recht ist, ist dem anderen billig”, Part II) Wolfgang (Valtin) used the phrase “reliable data models for SAP applications”, and I am very grateful to him for doing so. The reason I’m grateful is that he correctly “read between the lines” of my post about SMLT. That is, the reason I brought up the topic of preparing for SMLT was because SMLT-preparation would be a lot easier if there were “reliable data models for SAP applications”. And as I have said before, the absence of such data models will prove to be a real hindrance in SAP’s attempt to convince its customers to come along on the long march toward Enterprise SOA. To provide a concrete example of why this is so, let me share a “hypothetical” experience which I had during a “hypothetical” OTC initiative. A very senior and experienced FI/CO consultant (and I mean VERY) asked me to see if it would be possible to intercept F-28 processing between this selection screen: Figure 1: image and this screen: Figure 2: image Not being familiar with F-28, it took me a “little” while to find out that: a) yes, there is an intercept in the “right place”: open_fi_perform_00000900_e b) but this intercept will not allow a customer to change what displays on the screen in Figure 2 (the two fields in red) because a copy of postab is all that is passed to the exit, not the “real” postab. And don’t even ask how much longer it took me to find the code that is filling postab from the underlying tables. So – suppose someone wants to write a service that does what currently can’t be done in F-28. Then to be cost-effective, whoever writes this service will have to know all about the code and data model underlying F-28 already. Otherwise, the learning curve will be too expensive. But does the SAP community really think that it can realize the Enterprise SOA vision using only consultants who already know all about the code and data model underlying any particular transaction, such as F-28 in the case above? And if not, what does the SAP community think is more dangerous: c) divulging the data models relating structures to tables or d) delaying the time-frame for realization of SAP’s Enterprise SOA vision (because currently, there aren’t enough consultants with reasonable rates who know enough about SAP internals)?

To report this post you need to login first.

2 Comments

You must be Logged on to comment or reply to a post.

  1. Dirk Herzog
    I think your idea is right, we need a consistent data model to do the myriads of web services we will be building in the next years. But your example is a bit faulty. I don’t think there should be two web services for this topic but SAP should fix the user exit here. SOA doesn’t eliminate the need for user exits and I would just ask for a way to manipulate the postab in the exit.

    Best regards
       Dirk

    (0) 
    1. David Halitsky Post author
      Glad you agree on the need for consistent data models. 

      Please note that out of politeness, I didn’t ask whether such consistent models actually exist in Walldorf.  I made the assumption that in fact they do – otherwise the Repository Information System couldn’t possible work.  So I was concentrating on the issue of “divulging” the data model, because as of 2000, SAP was arguing that this was “too dangerous” from a “trade secret” point of view as well as a “stupid customer” point of view.

      With respect to the example – yes – you’re correct that an “active” exit instead of a “passive” exit would fit the problem.

      But if I were on SAP’s legal team, I would have to protest an “active” exit in this case and say that a customer voids its support agreement if it chooses to use the “active” version of the exit.  This is the way it used to be done: the customers can do anythign they want as long as they’re willing to live with the results … and pay expert rates to fix their mistakes afterwards.

      Again, thanks very much for taking the time to reply.

      Best regards
      djh

      (0) 

Leave a Reply