render@m.cs.uiuc.edu (05/01/89)
Written 9:04 am Apr 30, 1989 by dana@gmu90x.UUCP: >In article <39400015@m.cs.uiuc.edu> render@m.cs.uiuc.edu writes: >>Why? This would be like comparing a pterodactyl and a red-tailed hawk. They >>both fly, but they're at different evolutionary stages. > >If you are going to use such an analogy, please be careful. It is not quite >correct to say that each creature is at a "different evolutionary stage". A >better (although still imperfect) description is that each creature fills a >different ecological nitch (or at least fills it better than other creatures >of that time). We can disagree about the accuracy of the analogy, but the intent should be clear. Each organism/language developed in a specific environment for which it was suited, but the differences between the environments make a comparison of the organisms/languages problematic. By "evolutionary stage", I meant a specific span of time and its accompanying contextual attributes. It wasn't intended to be an exact analogy, I just wanted to use something a little more precise and imaginative than "apples and oranges" as a standard of comparison. If we're going to split any more hairs, give me some time to sharpen my verbal hatchet. >> A better study would >>be C++ vs. Ada, since they're both touted as being on the leading edge of PLs >>for software engineering. > >Given a more careful wording of the analogy makes it more appropriate to the >question of programming language comparison. To yield a useful comparison, >the programming languages being compared should (must?) fill the same >"programming nitch". That is they should belong to the same programming >paradigm. I've always thought that, with respect to computation, there should be some standard means by which different programming languages can be compared. In the old days people did this by lines of source code or object code, while today they use formal models into which both languages can be translated (Turing machines, temporal logic, denotational semantics). I've never really seen a good analysis of the problems inherent in such comparisons, though I might have if I read more about formal logic or cognitive science. I find this stuff interesting, so I wouldn't mind hearing from people who know more about the subject than I do. > Of course, everything in this posting could be wrong... What he said. > J Dana Eckart UUCP: ...!(gatech | pyrdc)!gmu90x!dana Hal Render render@cs.uiuc.edu