How much does Microsoft pay you to market 2005? I realize that honesty is not a big part of some people's jobs but really, it is inconceivable that, unless you are marketing "tart", you could assert 2000 is "unstable" and 2005 is "stable". (Spreading a little FUD, are we? That is the Microsoft/Oracle/Sun way isn't it?) It stretches credulity for you to assert that the feature set in 2005 is in any way superior to 2000. Look at the overwhelming addition of complexity to every single task! Do you REALLY think it is a plus to require interface users to click 3,4,5,6 or more times to discover something that previously took 1 or 2 clicks? Do you like the idea of opening multiple windows to view a log file?If so, are you a)crazy, B)just too eager with the crack pipe or c)do you have nothing else to do than click crappy interface designs some amatuer designer did not think thru before imposing them on us?
The facts are that 2000 is stable, it enabled MS to take the market from Oracle, 2005 is a complete re-write of the technology, MS always screws-up the initial write/re-write of a technology and MS will NOT get it right until they have written the new software package 2 or 3 times. As evidence consider that MS did not succeed with SQL Server until 2000. After it got the Sybase source code it tinkered with making it right in 6.0/6.5 and again in 7.0 then finally got it with 2000. It tinkered in the same ways with Windows, Word, Excel, PowerPoint and Exchange Server.
The ideas MS demonstrated in 2005 will not likely be of any use until (maybe) 2008 or 2011. That is an historically verifiable, highly likely probability! So, spare us the sycophantic marketing, try to be an honest person about SQL Server and then maybe we can take you seriously again.