The problem with Informix
Little Orphan Annie
Opinion IBM has a problem with Informix. It is one of those problems that is actually a good problem, but that doesn't mean that it doesn't cause headaches.
The problem is twofold: first, IBM's main database product is DB2 - but actually Informix is a very good database and always has been. In appropriate circumstances, you would expect Informix to outperform DB2 and it certainly has a lower management overhead than DB2, even when you take into account all the autonomics (which are impressive) that IBM has introduced in the new release of DB2.
So, there are circumstances when it makes sense for IBM to promote the use of Informix as opposed to DB2. But, and here is the nub, it cannot be seen to be promoting anything other than DB2, since the company has put all of its database eggs into the DB2 basket.
Actually, it's worse than this. The second problem with Informix is that IBM has branded its entire data management story under the name of DB2. Funnily enough, existing users of Informix in particular (but also UniVerse, Unidata, Redbrick, et al) tend not to be too happy when IBM tells them that they are part of the DB2 umbrella.
In other words, there is some seriously bad branding and marketing going on here.
Funnily enough, IBM is actually doing a lot with Informix, it has recently introduced the new CopperEye Datablade and it has some major enhancements planned for Informix in future releases. Moreover, IBM is keeping to its promise about enhancing both products from the other: the latest release of DB2 reuses a number of important bits of Informix technology and I understand that the next release of Informix will see some significant traffic in the other direction.
So, in a sense, all is rosy: IBM is keeping its promise to existing Informix users, it is selling the product to new users, it is enhancing the product, and DB2 and Informix are moving closer (albeit slowly) together.
However, IBM's exclusive focus on DB2 as a brand obscures whatever good news there is about Informix. It means that the company cannot publicly identify markets and opportunities that are more suitable for Informix, as opposed to DB2.
This would not necessarily be hard to do. For example, DB2 has never been a major player in the embedded database market. That is, the market for ISVs and VARs to develop applications with the database embedded within it. Conversely, this has always been an area of strength for Informix. Of course, there are other areas of strength where making such a clear delineation may be more difficult, but that does not mean that the attempt should not be made.
IBM's argument is that one day, maybe decades from now, DB2 and Informix will merge into a single entity. If that is true, and if it is also true that each product has its own strengths (as they clearly do) then a parallel database strategy would seem to make much more sense than the single focus that the current structure imposes on the company today.
Copyright © 2004, IT-Analysis.com
Sponsored: RAID: End of an era?