(link dead, sorry)
The Intentional Programmer
(That's a fairly old link; there's plenty of newer stuff out there about this,just Google - but a lot of it is VERY techy).
I know that Charles Simonyi is a very smart man. He's the guy who invented the so-called Hungarian notation so popular with Microsoft Programming (personally, it drives me nuts but I can see the value).
So my feeling that he's doomed to failure here is probably just due to my inability to see what he sees. There is also the matter that other people who report on this aren't as smart as he is either, so the things I read that talk about this are probably not fairly representing his concepts either. For example, I've seen articles saying that "non-programmers will be able to make improvements without the intervention of programmers", but nothing I've read straight from the horses mouth seems to imply that. In fact, it looks to me like it's going to take a pretty smart programmer just to understand this at all. But never mind that: I still think the basic concept is flawed. If the implementation is beyond the grasp of mortals, that's one thing, but shaky foundations are quite another.
Here's where I see the problem: the concept is intent. Charles Simonyi wants a language that can specify intent, and write "correct" programs. The problem with that is the same problem there is with any programming language: garbage in equals garbage out. His is a modeling language, which has been described as the programming equivalent of WYSIWYG. I just don't see how that protects you from logic errors. We are a long way from simply stating that our intent is "keep track of my inventory". You need to define inventory, its inflows and outflows, and write the very complex rule sets that "keep track". Sure, bumping things up an abstraction level might mean less mistakes (Perl is easier than C which is easier than assembler), but it doesn't eliminate mistakes, and it brings in its own layer of overhead and debugging difficulty.
Well, improving things without reaching perfection isn't failure, and I may be reacting more to some of the hype other folks write than to anything Simonyi himself has promised. Anyway, it may be something we'll hear more about in years to come, or it may be just another tool that didn't make the grade.
Got something to add? Send me email.
More Articles by Tony Lawrence © 2012-07-19 Tony Lawrence