2008-10-31 #1

Re-Invention

Evolution or Revolution?

I wonder if the problem is really that the industry keeps repeating old ideas, as you say, or if it’s that the market doesn’t adopt good ideas in the first place, thus forcing the industry to have to keep repeating itself. If Lisp, NLS/Augment, Xerox Star, Smalltalk had all been widely adopted in the first place, then they wouldn’t need to have been reinvented later, in the forms of Language Oriented Programming, WWW, Macintosh/Windows, Ruby.

I’ve made a living from recognizing good ideas from old Lisp and Smalltalk systems and re-implementing them in Java as something new. If these old systems had been widely adopted, and continued to remain viable, then the opportunity for "reinvention" wouldn’t exist. (Actually, I can think of some counter examples to this. The popularity of WebServices and AJAX is a bit mystifying given that we already had better solutions with CORBA and Java. Perhaps the heuristic of "newer being better" sometimes works against the industry by helping the adoption of some technologies despite merit.)

Playing a bit of the devil’s advocate: why risk trying to invent anything new when it is so much easier and safer to just "reinvent" something old and pass it off as new? It’s the difference between giving people something that didn’t exist before vs. something that didn’t exist "for them". The end result is the same: users end up with something new (to them at least). Microsoft was built by "evolving" other companies' inventions. With their introduction of C# (from Java), they’re left without any other big-ticket inventions to "evolve". Now they’ll have to start inventing things themselves (WinFS), stagnate (Vista), or else wait for Apple to invent something worth copying.

In the end, I actually agree with you. What I’m really saying is that "yes, invention is important" but that it isn’t enough. You still need to push to have inventions adopted. Maybe "invention" (coming up with new ideas) is really the easy part, and "innovation" (successfully bringing them to the market), is actually the much more difficult part.