1k1mgm@kuhub.cc.ukans.edu (Christopher Gunn) (05/04/91)
Was sitting in a bar maybe 20 years ago when a companion pointed out and asked the name of a guy across the room, who was a particularly sleazy drug dealer and ripoff artist. 'Oh,' said another person at our table, 'that's George ------, known as "Slime" to his friends.' A get a funny feeling reading the thread on Fortran 90 status. What's going to happen with a language when its *supporters* hate it? When those potential vendors who didn't vote AGAINST the standard had to hold their noses in order to vote for it (or felt they had to, anyway)? I've been writing programs, mostly in Fortran, since 1967. I actually LIKE Fortran (although I usually use Ratfor for projects of more than a few dozen lines). I kinda like Fortran 90, though lord knows it has some problems. It seems to me that while F90 as a whole is perhaps too large and awkward, it is certainly possible that IF SOME COMPILERS TURNED UP a clean and straightforward usage vernacular might emerge, allowing some enthusiasm and momentum to develop. But that isn't going to happen if compilers arrive--what was that one guy saying?--about 1998 (!!). Or if they cost $35,000. Frankly, if comments of the last few days represent informed insider opinion, then I'd guess that Fortran 90 isn't really going to happen at all. We might wind up with vendor-specific subset compilers on some of the big vector machines, but where else? Where is the outside push [like Ada] or the user pull [like C] that will induce vendors to invest big bucks in producing the compilers and tools, then heavily subsidize their distribution in order to get lots of copies out into the world? I suppose F90 might fail as a comprehensive entity, but still provide some cohesion for vendor-driven extensions in the 1990s, to be consolidated in a Fortran 2000, as happened with F66 and F77. Maybe this is already taking place; I think I've seen ads for hot-rod PC Fortrans that advertise 'Fortran 90 features.' It's late, I'm tired and I ought to shut up. I hope people will prove me wrong. I hope the industry will prove me wrong. But I'm sure not optimistic. Christopher Gunn Molecular Graphics and Modeling Lab SPAN--KUPHSX::GUNN Department of Medicinal Chemistry, Malott Hall 913-864-4428 or -4495 University of Kansas, Lawrence, KS 66045