Re: ForTran-Ada + flamette + question

[barmar@mit-eddie.MIT.EDU (Barry Margolin)]

In article <230@sdcjove.UUCP> jonab@phoebe.UUCP (Jonathan P. Biggar) writes:
>Please DON'T confuse Ada with its current implementations.  There is
>no reason that the bodies of a generic quadrature package cannot be
>shared.  Just because current implementations don't do it (YET!!!),
>don't damn Ada.

A language is only as good as its implementations.  If no existing Ada
compilers produce shared code from generic procedures, my first guess is
because it is hard to implement.  If it is going to be several years
before we have smart enough implementations, that means that for several
years Ada programmers will not be able to use this feature as
extensively as they would like.  Theoretical compilers are only useful
for producing theoretical code.  Most customers want real code.

While the language Ada isn't totally to blame, it is one of the
culprits, because it is so complex that it is hard to implement well,
even given three decades of compiler implementation technology.  It's
hard not to confuse a language with its implementations; the only metric
one has is existing generated object code.
-- 
    Barry Margolin
    ARPA: barmar@MIT-Multics
    UUCP: ..!genrad!mit-eddie!barmar