Do stacks solve the lock-in problem?

Lock-in occurs when you discover it would be more expensive to dump your IT garbage than keep throwing good money after bad. Lock-in happens whenever a platform, hardware or software, becomes a less-than optimal solution yet customers, knowing that, keep supporting it.

As unpopular as lock-in is with customers, it's just that popular with vendors. It assures them cash flow. It pleases Wall Street. It gives them "cash cows" whose profits they can use to develop something really nifty.

Notice where the incentives lie in the above paragraph. They don't lie in solving the problem. They lie in creating it, and maintaining it.

An open source stack, like LAMP (Linux, Apache, mySQL, PHP) is supposed to solve the lock-in problem. You can control the code used in the entire stack.

One of the big trends of 2006 has been for major vendors (the folks who benefit from lock-in) supporting a version of the stack. There's now an Oracle stack, an H-P stack, an IBM stack, even a Microsoft (endorsed) stack, from Novell.

Installing a stack doesn't solve the problem. Not if you're depending on one of these vendors to keep your stack current and relevant. Not if you're depending on anyone else to create the software that sits on top of the stack.

In the end you have to depend on someone, unless you're rolling your own solution, in which case you're in the development business, and not whatever business you thought you were in. So is a stack any better than a proprietary solution?

I think it's a step in the right direction, but that's all.