[Simh] Crowther's Adventure game

[Clem Cole]

Again speaking for myself ... not Intel...

On Mon, Feb 5, 2018 at 4:27 PM, Kenneth Seefried <kjseefried at gmail.com>
wrote:

>
> Have you looked at the Intel Fortran compiler (https://software.intel.com/
> en-us/fortran-compilers)?  Intel/AMD-only, of course.  Supports
> Fortran-IV/77 through at least FTN2003 with 2008 bits, plus a bunch of SIMD
> & multicore stuff (SEE, AVX, Phi esp).  Free-ish download for Linux
> (non-commercial use license, mostly).  No idea how good it would do on this
> particular application, but it did a pretty good job on some crufty old
> code when I tried it.
>
>
​Indeed and it has the added advantage of 'having the DEC compiler DNA
ground up and injected back in' to quote a retired Intel Fellow (Rich Grove
- who had lead the Gem compiler at DEC).  A number of the same people are
still developing it (*i.e*. those that have not yet retired).

It is the most popular of the Fortran compilers for the HPC community and
the commercial ISV's that still have code in Fortran.  There is great care
to try to ensure old code from old systems 'just work' as well as bring in
modern Fortran features.  i.e. FORTRAN IV (such as Adventure), FORTRAN77
and VAX Fortran code should pretty much just work (I can say the last time
I tried to recompile adven.f with it - which was FORTRAN IV,  a few years
ago it recompiled and worked with out any changes).  I have personally put
VAX Fortran code through it with much success before I worked for Intel
(when I was an independent consultant).  I have been told by people that I
trust about a number of customers/ISVs that continue to do this regularly.

I do not know how much success FORTRAN-II code has had, but I did help an
old friend from IBM (one of the ASC and 360/90 designers) a few years back,
so at one point, something worked; although I have not idea how much he had
to hack his code to make ti work.

It also is modern compiler conforming to Fortran 2003, 2008, and
the initial draft Fortran 2015 standard - including features such a Cray's
Co-Arrays, parallel code generation et al. * I have not looked* at the
'SPD' nor personally tied it, but I believe it will also fully conform to
the FORTRAN90 and FORTRAN95.   Modern Fortran BTW, is from a syntax stand
point, nothing like what I was taught in the late 60s/early 70s.

Kenneth points out for noncommercial Linux use, it is available at no
charge (and it is the same bits that people pay large sums).   That
download is the same compiler as that for MacOS and Windows and uses a
common back-end/optimizer as that if Intel C/C++ (i.e. nothing has been
disabled).

I have personally brought up a plead for years  to make it available for
non-commercial use on any platform (actually, I >>personally<< believe it
should be free no matter what but that's a different story -- I understand
the reasons why it is not, I personally do not agree with them, but I'm an
OS guy not a compiler person - I used to kid Dr. Gove with the line:  "Rich
are their more developers of your compiler or developers that use your
compiler" - which was not fair, but a fun thing to do).

The point is that Intel takes Fortran seriously, since the codes that come
out of the compiler sells real silicon,  which of course is what Intel
fundamentally does - sell expensive, hot rocks to run real codes.

That all said, there is also gfortran (the GNU compiler) and translators
such as NetLIB's f2c.   Although, I do not know of anyone that uses truly
use  either for real work.  I know people that test with gfortran and check
things out; but all the serious Fortran types in the HPC world that I
am familar use either Intel, Cray or PGC compilers or some mix.   Many
sites have more than one compiler.
ᐧ
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.trailing-edge.com/pipermail/simh/attachments/20180205/4097b74a/attachment-0001.html>