The story of Java is an interesting look at monopolies, market share, and lock in. When you have different platforms, such as Win32 vs. Mac vs. Unix flavors vs. OS/2, etc..., then it is clearly in most developers' interest to go with the largest one-- Win32 in this case. Since Win32 runs on many PC's and most of the new ones, you go with Win32.
However, the case of MSFT Java vs. 'pure' Java is a different calculation. It's clear that Microsoft will get a lot of new PCs to have it's Java, as well as all the users of IE 4.0. However, here it's a case not of who has the majority, but of who has ubiquity. PCs can have multiple versions of Java. And because of this, Sun is likely to win. If you are a developer and write for Microsoft's version of Java, then you are writing to "most" PC's that will run Java. But you can write for *all* PC that run Java by either writing to Sun's Java, or writing to the intersection of Sun's and Microsoft's Java-- which seems to be the actual case for the most part. So it's not about 95% vs. 5% as in the OS wars, it's more like 90% vs. 100%.
Microsoft loses if they don't get people to write Java that doesn't work on Sun's version. Microsoft seems to be having a hard time convincing Java developers to use platform specific API's. |
