Was MySQL 5.1 Ready For Release? Its Users Will DecideWas MySQL 5.1 Ready For Release? Its Users Will Decide

The author of MySQL, Monty Widenius, has taken a shot at Sun's release of MySQL 5.1, saying it wasn't ready yet, and pointed an accusatory finger at Marten Mickos, senior VP of Sun databases. He's also not bothered to deny rumors that's he's resigned from Sun. I don't know about the rumors, but it's clear to me that Elvis has left the building.When I phoned in for a briefing on MySQL 5.1, I was told by Sun public relations representatives that my interviewee, Robin Schumacher, director of product management, "wasn't prepared to comment on the blog" by Monty Widenius.

This statement I took to be part of the new, transparent Sun. Mention something that I knew nothing about but refuse to say anything about it. Schumacher gave me a technical briefing on the changes in MySQL 5.1, or more precisely, 5.1.30. The new feature set made for a good a MySQL update story Monday. Then I went and looked up Widenius' comments.

In a Nov. 29 posting to his "Monty Says" personal blog, Widenius advised MySQL users: "Don't expect that all critical bugs that you may have encountered in 5.0 to be fixed in 5.1. Even if we have fixed a big majority of the bugs from 5.0, some really critical ones still haven't been addressed." Regard the news features of 5.1 as being only "of beta quality." Needless to say, Schumacher didn't comment on this charge.

To round out the picture, Widenius added: "The reason I am asking you to be very cautious about MySQL 5.1 is that there are still many known and unknown fatal bugs in the new features that are still not addressed."

If there was any doubt that he opposed this release, he clarified that with the following: "Quality is not regarded to be that important. To quote Marten Mickos: 'MySQL 5.1 will be released as generally available in or before December because I say so.'" Needless to say, Schumacher made no reference to such a statement.

Nor was Mickos available to offer a countervailing view when I asked Sun for comment yesterday. But what's going on here is apparent in two ways.

First, Mickos is answering to those to whom he is responsible at Sun because he's realistic about where things stand, and Widenius by continuing as an employee of Sun (confirmed by Sun) but making these comments is less so. Widenius sees the change in who Mickos is responding to and is publicly protesting that things are not being done the way they used to be. Another way of phrasing that is that Widenius is not being listened to the way he would like to be listened to, and he doesn't like it. One might ask, did he not realize something like this might occur when he sold MySQL AB, which was basically his company, to Sun and pocketed his share of the $1 billion? Judging by his posting, my answer is, no. He believed it when they told him things would still be the same. They're not.

Meanwhile, I believe Widenius is disengaged from MySQL under Sun's ownership and is investing his share of the $1 billion in Finnish start-ups, such as IT Mill, which lists him as an investor. That's OK, but like I said, he may still be on the payroll but he's left the building.

So is Mickos doing the MySQL user community a disservice by releasing 5.1 as "generally available," meaning no longer beta software and ready for production purposes? This is a much trickier question, given the knowledgeable detail that Widenius issued in his broadside.

I think and hope the answer is no. Software can always be made more bug free with more time and more testing, but time is not a luxury that Sun can afford these days as losses mount. Realistically, there's always a debate over whether software is ready to release. Sometimes it makes sense to get a product into the hands of the customers and fix it further from there; sometimes that's a recipe for disaster.

But Widenius may yet be proven right if Sun has rushed the release of 5.1 to get more software into more hands and glean the revenue. I don't doubt that Sun would prefer to gamble a little bit with customer's tolerance than watch its open source strategy and company fail with its virtue pristinely intact.

But what I think matters less than what we are about to learn from the marketplace. This is a test of what happens to successful open source code once its purchased by a large, publicly held, profit-hungry company. If 5.1 was mostly ready and not seriously deficient, the response from users will be muted. If it contained serious bugs that should have been addressed before GA, the answer will be voluminous, vociferous, and a biting statement about the fate of acquired open source code. MySQL users, beware, and be forthright in your verdict.