Why these original FORTRAN quirks?
in [Fortran]
|
From: Gene Wirchenko on 25 Oct 2006 12:52 nospam(a)see.signature (Richard E Maine) wrote: >Rostyslaw J. Lewyckyj <urjlew(a)bellsouth.net> wrote: [snip] >> purely Fortran? But I'd consider that to be rather unrealistic in a >> production environment. Linkers, loaders, libraries etc. are a fact >> of life. > >I think we are not communicating here. The linkers, loaders, etc, are >part of the Fortran environment. That's why the Fortran standard (like They are? I used FORTRAN, back in the capitals days. The linker and loader were part of the OS. Many language systems used them. [snip] Sincerely, Gene Wirchenko Computerese Irregular Verb Conjugation: I have preferences. You have biases. He/She has prejudices.
From: wclodius on 25 Oct 2006 13:18 Gene Wirchenko wrote: > nospam(a)see.signature (Richard E Maine) wrote: > > >Rostyslaw J. Lewyckyj <urjlew(a)bellsouth.net> wrote: > > [snip] > > >> purely Fortran? But I'd consider that to be rather unrealistic in a > >> production environment. Linkers, loaders, libraries etc. are a fact > >> of life. > > > >I think we are not communicating here. The linkers, loaders, etc, are > >part of the Fortran environment. That's why the Fortran standard (like > > They are? I used FORTRAN, back in the capitals days. The linker > and loader were part of the OS. Many language systems used them. Of course they are part of the Fortran environment. They were also part of the environment of any other language system that relied on them. A (viable) (fortran) compiler must (be designed to) deal with their conditions and influences (it is in effect surrounded by them) and the result of using the linkers and loaders must be a code that must be be usable within the system (including the OS) it is used in. <snip>
From: wclodius on 25 Oct 2006 13:23 Rostyslaw J. Lewyckyj wrote: > Richard Maine wrote: > > > Rostyslaw J. Lewyckyj <urjlew(a)bellsouth.net> wrote: > > > > > >>In many cases, systems, external names are more limited in length. > > > > > > When using f77 and earlier long-name extensions, that was an issue I > > recall seeing. F90 requires that the system allow names up to at least > > 31 characters. If the system doesn't naturally allow them, the compiler > > is supposed to compensate somehow. Yes, ways exist and compilers have > > used them. For example, some compilers automatically generate short > > names and have a translation table for the full names to the short ones. > > > > To my knowledge, there are no f90 compilers that fail to correctly > > accomodate 31-character names as required by the standard. There have > > been a few pretty flakey things briefly sold as f90 compilers. I suppose > > some of those might not have gotten it right; they weren't around for > > long enough for the matter to come up. Some of them didn't get a lot of > > things right and had pretty brief lives as a result. > > > > Whoa there! I was refering to the limitation of the surrounding > machines' OS and other components. F90 may require 31 character > namelength support within the Fortran components. But then the > Fortran implementation will require translation contortions > to work within the system.<snip> True but the same applies for any other language system. In particular, I know of no system supporting a Fortran 95 compiler that does not also support a C++ compiler, and the requirements of the C++ compiler far exceed those of Fortran 95.
From: wclodius on 25 Oct 2006 13:33 Rostyslaw J. Lewyckyj wrote: > <snip> > In the extreme case, suppose N=2, which with 8bit chars allows > up to ~32000 unique external names. With the F90 liberal rules > some F90 Fortran program might exceed the system design constraints > and become impossible to do. :) (I know it's an existance argue > ment for illustrative purposes only) The standard explicitly admits the existance of constraints on the processor that may not allow it to compile larger standard conforming programs. The program you are discussing will not be portable to such a system. If enough users of the system are willing to pay for the ability to run such large programs on the system, then the linkers, OS, etc. will be upgraded facilitate that, and a Fortran processor will benefit as a result.
From: Greg Lindahl on 25 Oct 2006 15:44
In article <1161797007.580772.82680(a)h48g2000cwc.googlegroups.com>, <wclodius(a)lanl.gov> wrote: >True but the same applies for any other language system. In particular, >I know of no system supporting a Fortran 95 compiler that does not also >support a C++ compiler, and the requirements of the C++ compiler far >exceed those of Fortran 95. The first time I saw C++ generate a 20 kbyte identifier, I was horrified. Now I'm kind of blase' about it. -- greg |