Re: Informix left standing

Steven Hauser wrote: > E-Z Informix Market Forkast-o-rama analysis: > > OK, I confess, I read a book, > "Megamistakes, Forecasting and the Myth of Rapid Technological Change" > 1989, by Steven P. Schnaars, some marketing weasel. So now I am an > armchair expert with 20/20 hindsight. > You read the wrong books.You need to read books on how to sell to the cynical consumer. ;-) > The Basic questions: > Who are the customers? > How large is the market? > Will the new tech offer a real benefit over existing and > subsequent substitutes? > Is the new tech cost-effective relative to the substitutes? > Is the customer benefit worth the price charged? > Are cost efficiencies (of the product) possible? > Are social trends moving toward or away from this market? > Does the innovation require users to do things differently? > Does the innovation go against customs, culture, or established > business practices? > The basic questions are "Who" "What" "Where" and "Why"Add to that "How much". What you are first trying to do is *market research*. First Informix has to ask: "Who are my current customers." Which is actually: " Type of organization " " Size " " Location " " Type of customer." Then "Why did they chose Informix " Which breaks down to : " (1-5) Rate the importance of Cost" " (1-5) Rate the importance of Performance" " (1-5) Rate the importance of Support " " (1-5) Rate the importance of the Salesperson. " " (1-5) Rate the importance of additional support from PS group." Then "What types of applications are you using Informix." Then "Where does Informix fit within your IT organizations strategy" [ I think you get the idea. ] Now ask the same questions only with respect to the UDO. > From the above questions Informix 7.x answered everything asked in > [SNIP on 7.x] OWS/ODS has nothing to do with SMP. From an applications standpoint,an SMP box looks like a regular box. Disregarding processor affinity, (Which is a bad idea from a SMP perspective), the application doesn't know or doesn't care which CPU it runs on. You can create more virtual CPUs than real CPUs. ODS/OWS's great advantage was that it was multi-threaded and allowed for parallel operations to occur internally. This technology was written up in the ACM transactions on databases roughly in 1989 - 1990 time frame. So you can see the 5 year lag time from research to practical applications. (Note: The article mentioned specifically Informix was the only corp. to take an active role in this research.) The product was a hit because it addressed a specific issue which all RDBMSs faced. It was an improvement of the internals. So that current applications did not have to change too much, if at all. Faster transactions, Faster queries. The *ONLY* major drawback is that it is much harder to tune. > > > Informix 8.x has problems with several of the questions, like who are > [SNIP] The problem with XPS is that its not a complete solution, and it isharder to manage. Now you have to consider the location of the data with regards to the CPU. Depending on your queries, you could hit all CPUs or only one. (CPU = node in this case) Again, the problems that XPS don't solve are in the datamanagement. (HPL and ONBAR came later and there are still problems.) But the one thing that is important is that it was created in response to an existing problem. > So the new strategy of 'Data warehouse, no low end'. How does this > stack up? Yes there are customers, trends and business culture, > [SNIP] This is the confusing part.Datawarehouse? That market, for all its hype does exist. But there are already niche players, like RedBrick. Oracle and Sybase are also chasing after the same market. I think what Bob is trying to say "Focus on our strengths not our weaknesses." Meaning, that Informix can compete against Red Brick, Oracle, Sybase and Microsoft, within the Data Warehouse market place. Therefore we compete there. However, the message comes accross as: "We will compete in the niches where we currently have a technical advantage. We will concede the comodity market because we don't see the profit margins there and we can't compete against Microsoft." The problem is also that the competition is also chasing those markets. Microsoft is even talking about putting NT on the Big Iron. (Sick thought I know.) Also as time goes by those advantages enjoyed by Informix are going to be lost. > > > Well, how the hell am I supposed to know? Well, stay in school. You might learn something useful.(Or you might end up like me ;-) The issue with Informix is never a technical one. Its one of mindset of the organization. Most of the posters here have been around Informix far too long. Some have centralized their skillsets solely on Informix. Why? Simple. Because the product is good. (Well for the most part. ;-) The frustrating part is with the people side of Informix. (Read: Engineers Good Guys. Sales, Marketing, Upper Management Bad Guys.) [Oooh! Now I've gone and done it. I slammed upper management. The Informix Gods will now throw lighting bolts at me!] Business/Marketing decisions are being made base on assumptions and poorly defined research and input. This brings us to UDO/Universal Server. While I do not question the *vision* of Informix, I question the research that it has done in *developing* this technology. I also question its ability to go from today's corporate needs and to *sell* this technology so that it may achieve its vision. (Going from today Point A, to the future (UDO) Point B). Bob made some hard and fast cuts. While this saved the bottom line in the short term, it has effected its credibility in the long run. Informix has to realize that its customers for the most part are cynical. If your salespeople can't explain the technology and understand it, how can they sell it? Look, its so easy sitting back and be critical of Informix's management. While we don't know all that they are thinking, and planning, we can see what their actions are after the fact. We are confused. Why? Bob is saying one thing, and the Informix pundits are saying another. -A shakesperean fool