jfm@ruddles.sprl.umich.edu.engin.umich.edu (John F. Mansfield) (11/02/88)
OK, now for a major question, not all of us have time to rewrite all our old code in C. So the question is what are the compilers that will be available for the nExt machine? Will the company support anything or will it be up to some gallant third party hackers to generate the necessary compilers to let everyone else use the beast? Sorry to mention Fortran, but I use it.
dmocsny@uceng.UC.EDU (daniel mocsny) (11/05/88)
In article <3f6ce6e8.59b7@sauron.engin.umich.edu>, jfm@ruddles.sprl.umich.edu.engin.umich.edu (John F. Mansfield) writes: > OK, now for a major question, not all of us have time to rewrite all our > old code in C. So the question is what are the compilers that will be > available for the nExt machine? > > Sorry to mention Fortran, but I use it. Don't apologize for using Fortran. Once I took a week out to pick up Pascal, I never touched Fortran again, but if it works for you, great. You've made an intellectual investment in learning the tools that were available before you were tied up making money. If you spent your time trying to learn all the languages available, you might find one much better suited to your application. However, you wouldn't get any work done in the interim. You can't spend your whole life trying to get ready to work. The most expensive computer component you own is the code you have written. Why must the computer manufacturers constantly attempt to destroy your wealth by not acknowledging this? If they won't provide compilers, then why can't they make a serious effort to provide source-code translators? Instead, this is a fringe market left to third-parties or individuals. If you had a robust Fortran-to-C translator (especially if it could de-spaghetti-fy your code), then you could effectively ``mine'' your Fortran source and move ``up'' to new languages and hardware, without losing your investment. Incidentally, the best way to learn a new language is to study such a translation of code that you know and understand. That way you see how to do things in a new language that you already know how to do in the old language, speeding your comprehension. Then you can start picking up those new features that have no obvious parallel in the old language. Individuals and corporations have $ billions sunk into Fortran and Cobol code. If the computer vendors refuse to acknowledge this, forcing customers to sink additional $ billions into re-writing all of it, they might as well be tossing Molotov cocktails into your office. Dan Mocsny
rodgers@cca.ucsf.edu (Rick Rodgers) (11/05/88)
In article <3f6ce6e8.59b7@sauron.engin.umich.edu>, jfm@ruddles.sprl.umich.edu.engin.umich.edu (John F. Mansfield) writes: > > > OK, now for a major question, not all of us have time to rewrite all our > old code in C. So the question is what are the compilers that will be > available for the nExt machine? > > Sorry to mention Fortran, but I use it. We use both Fortran and ditroff quite unapologetically! Many UNIX-based systems (such as the superb S statistical package from AT&T Bell Labs) rely on the presence of Fortran. The word we have from the local NeXT people is that Green Hills (can anyone out there evaluate this firm for us?) will be supplying a Fortran; they also anticipate a third party vendor for ditroff. Why is the BSD f77 not considered part of the MACH/UNIX system? Given the amount of interest in this newsgroup and NeXT's emphasis on networking, why aren't THEY reading this newsgroup and answering these sort of questions directly??? -- R. P. C. Rodgers, Statistical Mechanics of Biomolecules, Dept. of Pharm. Chem., University of California, San Francisco CA 94118 (415)476-8910 (ARPA: rodgers@cca.ucsf.edu, BITNET: rodgers@ucsfcca, UUCP: ...ucbvax.berkeley.edu!cca.ucsf.edu!rodgers)
bzs@PINOCCHIO.BERKELEY.EDU (Barry Shein) (11/05/88)
From: uccba!uceng!dmocsny@tut.cis.ohio-state.edu (daniel mocsny) >The most expensive computer component you own is the code you have written. >Why must the computer manufacturers constantly attempt to destroy your >wealth by not acknowledging this? If they won't provide compilers, then >why can't they make a serious effort to provide source-code translators? >Instead, this is a fringe market left to third-parties or individuals. Whoa whoa whoa...hold on there a minute... Are the computer vendors *really* constantly attempting to destroy your wealth by not acknowledging this? How many systems do you know of that don't provide (reasonable) Fortran (eg) compilers? It's all market driven, if there's a market the compiler will appear soon enough, it's an obvious product. I also don't know why you seem to feel that third-party compilers are somehow an inferior solution. In many cases they are first rate and probably better than what a vendor would produce, why do you think hardware manufacturers automatically have some magic ability to manage good compiler projects? Companies like Borland, Greenhills and even the Free Software Foundation (not a company) have produced first rate compilers by any standards. Maybe you could put a little more meat in your complaints so we can see if we're misunderstanding you? >Individuals and corporations have $ billions sunk into Fortran and Cobol >code. If the computer vendors refuse to acknowledge this, forcing customers >to sink additional $ billions into re-writing all of it, they might as >well be tossing Molotov cocktails into your office. No, they might as well be tossing Molotov cocktails into *their* offices, I presume you will ask and not purchase systems which don't have the compilers you require? -Barry Shein, ||Encore||
dan@Apple.COM (Dan Allen) (11/06/88)
In article <384@uceng.UC.EDU> dmocsny@uceng.UC.EDU (daniel mocsny) writes: >Individuals and corporations have $ billions sunk into Fortran and Cobol >to sink additional $ billions into re-writing all of it, they might as >well be tossing Molotov cocktails into your office. That may be well and true, but individuals and corporations are also always rewriting theirsoftware, adding features, deleting old features, porting it to different machines, etc. If the dictum "Plan to throw one away" is followed every so often, then often many other benefits will be seen by rewriting old Fortran and Cobol code in newer languages like C, C++, or Modula-2, for example. Just like ANSI C is providing a window to bring old C sources gradually forward to function prototypes, individuals and corporations should see the handwriting on the wall and begin rewriting their systems in a more modern language. Many universities today are no longer teaching Pascal, let alone Fortran. Stanford, for example, is crafting their new compiler to compile C and C++. Not even Pascal. Interesting... Dan Allen Apple Computer