Re: Sybase limitations

Hi Mark. I would just like to confirm that what you and Graeme are stating are accurate representations of the current SMP and MPP architectures of Oracle, Informix, and Sybase. These are important issues for customers to understand since they do affect the way RDBMSs will behave in different environments and applications. The major weakness in Informix's strategy - which continues to hurt it in non-MPP application markets - is its concentration (some might call it an obsession) with the highest end of the RDBMS market. While Informix concentrates in this area, Oracle, MS, and Sybase are expanding very rapidy in the low and mid-tiers. Because of this there is much better application and end-user tool support for these RDBMSs (tool vendors will normally create direct drivers for Oracle, Sybase, and MS SQL Server while lumping Informix in the "ODBC" pile). Customer interest for C/S applications seems to be presently focused on Oracle, Sybase, and MS SQL Server (most surveys as well as my own surveys bear this out) because of its much higer visibility and penetration into the C/S market. High customer demand (for example for Sybase PowerBuilder, Oracle SQL*Forms, MS Visual Basic/SQL Server/Access) creates a large pool of trained developers and consultants who in turn continue to recommend the solutions that they have been trained in. Essentially a snowballing effect is created which has help propel Oracle into dominance and has helped push Sybase from a situation a few years ago where they had negligble share of the Unix market (Informix, Oracle, and Ingres had approximately the same share of the market) to a position of second place (a situation which suggests that Informix has not done a very good job in addressing market demand). My guess is that if Informix continues to pursue its current strategy (i.e., faster and faster and faster RDBMS) it will continue to lose market share. Just like technical weaknesses can catch up to vendors so can marketing weaknesses. Informix's lack of pentration will surely catch up to it as it continues to be niched out at the very high end where they will probably be competing against IBM initially. Ultimately the other vendors will also go after this market but only after they have secured their positions in the high volume (I am speaking abouy the number of seats and licenses) low and mid-tier situations. It is a shame because Informix has a good engine but doesn't seem to know how to compete in any other market other than the very top end. Parallel Query is a good marketing message but it is getting old (Informix has been using this marketing message for over three years now) and has no meaning to the tens of thousands of customers that cannot or do not need to use it. For these customers Oracle Workgroup 2000, Microsoft's BackOffice Suite, and Sybase's System 10/PowerBuilder solution has far more resonance. Ultimately, Informix's excessive concentration on just one niche of the application market will cause Informix lots of problems. Regards, Rich Richard Finkelstein, Performance Computing, Inc.