Ending several months of negotiations, Sun Microsystems has decided not to join the Eclipse open-source tools effort backed by rival IBM.
[…]
“Microsoft is still setting the pace from a usability and productivity perspective, but the Java vendors are continuing to support two competing platforms/communities for plug-in development,” [RedMonk’s Stephen] O’Grady said. “That just seems counterproductive, if the real goal is to threaten Microsoft.”
This, folks, is how Microsoft wins.
One of the hallmarks of Microsoft’s success has been its unique ability to integrate technologies into a complete solution. In competition with the mishmash of solutions that is UNIX and, more recently, Java, this complete, integrated solution has won every time.
Don’t get me wrong. The number and diversity of components and implementations is one of the greatest strengths of the UNIX, Java, and Linux worlds–this diversity creates a dynamic environment that fosters an unparalleled level of flexibility, competition, and choice.
However, this great strength has also been a great weakness. Too much choice can be bewildering, breed complexity, and lead to incompatibilities around the edges that limit the technology’s overall potential and drive its user base right into Microsoft’s arms.
We can have the best of both worlds, if we choose to; and we can choose to by embracing open standards that allow the flexible, competing choice of components and implementations to interoperate with each other.
The first step is to abandon the “not invented here” mentality that has, for too long now, plagued the computer industry. This, to my eye, appears to be yet another example of that.
Just to show this isn’t another case of the pot calling the kettle black, here’s another example of NIH from my own backyard: Debian’s refusal to adopt the de facto standard RPM package format. Note: I said package format, not package system. I will have more to say about this in the coming days..