As programmers age, they face age discrimination — true or false?
On the surface this appears to be true, but when you accept superficial explanations you pay the price. In my experience, both forward and reverse age discrimination are prevalent, except what’s really going on is a mismatch between expectations and you. Managers expect employees to progress. If you’re 55 years old and still want to write code, that’s out of line with expectations. If you’re 23 years old and want to succeed at executive-suite strategic consulting, good luck — gray hair and exposed scalp is an unstated, but vital qualification.
When it comes to evaluating Web services, the latest next-big-thing in information technology, gray hair and exposed scalp helps. Assuming, of course, that history serves as guide to the future.
The vision of Web services is that you’ll assemble any application you like by stringing together publicly available components from multiple vendors. The use of standard interfaces and ubiquitous Internet protocols will make this fast, cheap and easy.
Don’t bet the farm. It could happen, I suppose, assuming:
- Software vendors make a 180 degree shift in support policies. Right now, if you have just two vendors and one problem you solve it yourself, because your vendors expend all of their efforts pointing fingers at each other while telling you they don’t support your configuration.
- Vendor and product selection becomes quick and simple. Right now the process of selecting a software package from a single vendor requires serious effort. With Web services you’re assembling your application from the work of dozens of vendors. In my experience, comparing multi-vendor solutions is more complicated than single-vendor solutions, not less.
- Trustworthiness is omnipresent. At InfoWorld’s recent Next Generation Web Services conference, a security panelist made a telling point: Just because a component advertises its behavior through UDDI (Universal Description, Discovery and Integration) doesn’t mean it’s telling the truth. Even ignoring the likelihood of malfeasance, we’ve all encountered DLLs that don’t behave as advertised. Why should publicly available components be any different? Hmmm?
Web services is a promising architecture. Many of us expect it to be important in the not-too-distant future. The tricky thing about the future, though, is that it hasn’t happened yet.
The best guide we have is history. So far, the Web services hype factory has ignored its lessons.