Lambda the Ultimate

In this clip from an IT Conversations interview Jon Udell tries to explain what dynamic languages are.

I think this clip makes it clear that (a) the situation as regards the terminology used to describe type system issues is beyond hope and (b) the issues Jon tries to deal with are quite real.

As we noted many times in the past, "strong" typing is not the same as "explicit typing" and most of the goals Jon talks about are achievable with statically typed languages.

In fact, dynamic - or scripting - languages encourage a development process in which programs are modified and translated iteratively, so having the compiler check the "morphed" data structures and infer their type shouldn't really be a problem, and in fact can help the developer. Naturally, since not all the code is changed at once a language that would appeal to Jon would allow the programmer to restrict the scope of type checking to specific parts of the program (e.g., only routines that may in fact be invoked).

Whatever your opinions regrading typing, I think it is very clear that we should move the debate away from having Pscal and Java as the only examples of strong typing, and C as the only example of weak typing. I hope LtU would set a good example for the rest of the community...

For some reason many new LtU members choose not reveal their real names. This makes the site less friendly in my opinion, and perhaps even discourages discussion.

I urge people to use their real names, unless they think there's good reason not to, in which case do what you think is best.

Thanks.