In his OSCON keynote last Thursday (26 July 2007), Bill Hilf of Microsoft described his perspective of the "two steps forward, one step backward" relationship between Microsoft and the world of Open Source. Amongst the steps forward, I count the establishment of a web site microsoft.com/opensource, releasing some Microsoft software under an OSI compliant license, and -- specifically to MySQL -- the mentioning of the MySQL Connector for Visual Studio in his OSCON presentation.
Microsoft seems to be acknowledging that there is a heterogeneous world out there. It's not "all proprietary" versus "all Open Source". The community mixes and matches at will.
Many might think there are few commonalities between MySQL and Microsoft. But there are more than you'd think, and they're not limited to the common first letter of the alphabet.
The biggest commonality is WAMP. A simple real-world observation is that plenty of the organisations deploying LAMP applications, have developed them in whole or in part under WAMP.
MySQL doesn't have a platform agenda, so we want to make WAMP easy to use. If WAMP turns out to be good enough not just for development, but also for deployment, great! And there MySQL and Microsoft have a common interest: Making WAMP attractive enough for production deployments, as opposed to being a mere playing ground for LAMP as The Real Thing.
So while you can expect MySQL to continue to want to excel as the M in LAMP, and while I suspect it's a safe bet to expect Microsoft to continue to have a platform agenda, I don't think anybody should expect a dogmatic, black and white, all-or-nothing world. In reality, customers and users mix and match, and it's about time we all acknowledge that.
Bill Hilf made two points in his OSCON keynote that I'd like to comment upon as for how they apply to MySQL. In both cases, I'll conclude with a plea to contact me, if the thinking applies to you.
First, he pointed out the order-of-magnitude technical improvements that could very simply be made to the interface between IIS and PHP, once key IIS developers were connecting with key PHP developers. Such opportunities for improvements exist only if the products have been living in complete isolation from each other. And that's not the case with MySQL and Windows, which have peacefully coexisted since last century. This means that the technical fruits to be picked when making WAMP more attractive won't hang quite that low. But we do know that there are technical hurdles to overcome, and documentation to be improved, specifically when it relates to deployment of WAMP applications. Should you be one of the casualties thereof, I'd love to hear from you about how we should make WAMP more attractive. Email me at email@example.com!
Second, he pointed to the fear of the unknown, to the extent Linux is unknown to the Microsoft user base. "Can I go visit a Linux website?" was one of the questions he had been asked, not tongue-in-cheek, but for real. My take on this is that there is a perceived divide "Microsoft vs. Open Source" bigger than the divide out in the real world of IT. And I don't think it's in the interest of the user base to be lead to believe that solutions have to be fundamentalist. Paradoxically, I don't even think it's in Microsoft's interest. I think the world's perception of MySQL's user base may be skewed towards LAMP rather than WAMP, not just because L, A, M and P are all Open Source as opposed to W, but also because of a relative lack of acknowledgement for WAMP as a development (and deployment) stack. Should you have deployed large WAMP apps, I'd love to hear from you about WAMP specific success stories. Email me at firstname.lastname@example.org!
I can see history repeating itself in strange ways. In the early days, many picked MySQL "for development and testing only", fully convinced to deploy on other databases (and that was definitely the case with myself in the 1990s). But the "testing" phase dragged on and on, and in many cases MySQL turned out to be "good enough" for deployment.
That's how MySQL grew.
With proper attention from MySQL, Microsoft and our respective communities, perhaps WAMP in a growing number of scenarios can enjoy the same benefit of being "good enough for deployment"?