[net.cse] Languages.

karl@dartvax.UUCP (Karl Berry.) (10/18/84)

Alex Colvin [dartvax!alexc] writes that Unix* is hopelessly
inadequate in some areas, and not intended as a general purpose
system. I had thought that one of the basic premises of the Unix
design was that it would be flexible enough so that what the
authors didn't include, someone else could, with little effort.
Although I don't hold it up as the Operating System to end All
Operating Systems either, I am curious as to in what areas it
fails.

It is also strange to me that Alex [apparently] lists APL, Snobol,
and Algol as languages that an educated CS student should be
exposed to. Algol is perhaps the precursor of modern declarative
languages, but it is rather dated now. Modula-2, Ada**, or some
such would seem to be a better choice for that area. APL is such a
language which perhaps has its place, but I don't think that place
is a substantial share of the market, and hence hardly necessary.
An applicative language, like Lisp or Prolog [Is Prolog
applicative?] would seem to cover the other major approach to
language theory better. It's true that Unix doesn't have standard
compilers for Modula-2 or Ada, but it does have Lisp and Pascal,
and so seems to cover most important concepts in languages.

Finally, I have no idea what a modern operating system is -- what
have been the advances since Unix?

karl@dartmouth.csnet %%% dartvax!karl