But still, as evidenced by the many users who said "fine by me" when Microsoft recently revealed that its SQL Server 2005 database release would be delayed, not many database administrators want to go through this hairy process more often than is absolutely necessary.
Because its a serious project, whenever a major new database release comes out, one of the first things we want to know is what the early testers think of it. We want to know its stability, what kind of high-availability metrics we can get out of it, its ease of installation, whether it delivers on the ease of manageability the major database vendors are promising, and whatever else we can glean.
The problem is, getting good, unvarnished feedback from beta testers isnt always easy. But a number of readers took the time to point out to me that getting feedback on IBMs beta of its upcoming DB2 update, code-named "Stinger," from VA Software and putting that in a recent column wasnt much smarter than going straight to IBM sales and expecting something besides a sales pitch.
Thats because VA Software is actually in an alliance with IBM. It derives much of its income from Big Blue. I knew that, so I have no intention of shifting blame elsewhere.
The reason I wanted to get input from them on Stinger was because VA Software is heavily involved with the open-source community (it handles the Web sites for the Open-Source Development Network, which runs sites such as Slashdot and NewsForge, for example).
What with Stingers support for the 2.6 Linux kernel, I wanted to check in with that community to get a reading on how well IBM is doing in staking out a claim as the database of choice for Linux.
What I should have done, though, was to delve deeper into that broad community and find a more unquestionably unbiased source, and for that I apologize. I also should have made clear the relationship between VA Software and IBM. Again, my apologies.