Woke up this morning and found that one of my Twitter searches are returning quite a few “MySQL 6 is dead, it’s Oracle’s fault and the world will end tomorrow as a result” (more or less) related tweets flying across the wire (I may be watching a trending topic ascend..?) A number of the tweets point to the “Newsflash: No more MySQL?” article at wings-of-wind.com.
Somehow, a change in the development cycle/release model (which, factually, appears to be pulling features from the previously planned version 6.0 & 6.1 of MySQL into version 5.4 as well as into a mainline trunk from which future releases will spin with greater frequency) has people thinking that MySQL has been cancelled by – at this point in time – Sun Microsystems. Of course, in the best traditions of the “I see a conspiracy here” netizens, the blame is being placed upon the ‘anti-MySQL devil itself’1 Oracle.
The rumored “cancellation of MySQL” is, of course (most likely), complete hogwash as I believe Oracle/Sun cannot announce any acquisition related product plan alterations until after they legally own the company (I believe the SEC would be very cross with them if they did). And until the European Union (EU) approves the deal, Oracle does not own any aspect of the Sun Microsystems assets (or liabilities). Some people would argue that Oracle could ‘indirectly’ influence the plans, but considering what some of the apparent EU sticking points for the Sun/Oracle deal are, I’d be extremely surprised to see Sun or Oracle messing with MySQL in any negative way right now.
So where does that leave “us” (the consumers of MySQL)? Well, it means we’re looking at – factually speaking again – a release model that provides us with more frequent MySQL releases and a development model that, in my view, pretty closely reflects that of the Linux kernel itself (no new major release until it’s truly warranted, and the definition of “truly warranted” is fairly fluffy considering the number of significant changes that’s been happening to the Linux kernel thus far in the 2.6.x release stream). Basically, if you measure the number of releases as an indicator of health, MySQL is going to appear to be a much more vibrant and ‘exciting’ product, I think.
The problem could be if you, like me, measure health by how stable and compatible a product is…
Well, we’ll see what happens. It’s entirely possible that the similarities to the Linux kernel release model aren’t accidental and that, as a result, the MySQL trunk will be considered “upstream” and any commercial product(s) will follow the “distro model” used for Linux whereby a given release of the upstream trunk represents a baseline (starting point) for the commercial products. The commercial versions the get hardened and tested before they’re released “into the wild”. As a result, there will be somewhat of a lag between what’s deployed into production and what’s available in the trunk. Obviously, if you’re working in the Enterprise space, stability is pretty much the only measure you’d use when considering something a successful deployment of technology, so considering how some of you want no less than 7+ years of support for every release of software you deploy and that lag probably will not be much of an issue, at all, for you.
PS: I do not share the perspective that Oracle is anti-MySQL. As a matter of fact, having a (minimally that it’s perceived as such) vibrant competitor in your midst and helping it succeed only means good things for your business and your business’s ability to avoid “anti competitive scrutiny” from the authorities…)
