Image Map Image Map
Page 3 of 3 FirstFirst 123
Results 21 to 26 of 26

Thread: Any really good FORYRAN 77 books out there?

  1. #21

    Default

    If there is a strict FORTRAN-77 compiler, it is not going to be very useful for Games that require system-level access. Most of the useful Fortran compilers have extensions for working with HEX and Octal.

    The RM compiler has enough extensions to be mostly compatible with the DEC Fortran compilers. The Microway NDP-Fortran compiler runs under PharLap DOS and is based on the Green Hills compiler. I use it extensively, has no problem using 4GBytes of RAM under DOS. The RM Compiler comes bundled with PLINK86 overlay linker, runs in real-mode DOS. Both have bit-oriented extensions, ISHFT, IAND, IOR, etc. I ran into one compiler decades ago that did not support HEX values and formatting, no Bit-oriented extensions. Dumped it quickly. Even Microsoft FORTRAN-80 for CP/m implemented ENCODE and DECODE statements, the predecessor of internal Reads and Writes.

    I remember the CDC-7600 with it's A10 formatting for 10 characters in a 60 bit variable. That was a PINTA to convert to a 32-bit machine.

  2. #22

    Default

    When you refer to RM do you mean IBM Pro F* 1.0 or something later? I would guess code.compiled with that would at least be as fast as quickbasic executables, no? I also have QuickC to fall back on.

  3. #23
    Join Date
    May 2009
    Location
    Connecticut
    Posts
    3,708
    Blog Entries
    1

    Default

    Quote Originally Posted by 2icebitn View Post
    Krebizfan, what was the title of the book you found online?
    FORTRAN 77 with MTS and the IBM PC by Carnahan and Wilkes. There are a couple of versions of it, some related to the Michigan Terminal System archives and at least one over on Google Books.

    https://babel.hathitrust.org/cgi/pt?...view=1up;seq=1 might be a good starting point.

  4. #24
    Join Date
    Jan 2007
    Location
    Pacific Northwest, USA
    Posts
    27,360
    Blog Entries
    20

    Default

    Quote Originally Posted by BrianS View Post
    If there is a strict FORTRAN-77 compiler, it is not going to be very useful for Games that require system-level access. Most of the useful Fortran compilers have extensions for working with HEX and Octal.
    It's entirely possible to have extensions in a conforming F77 compiler. The proviso is that such extensions could not interfere with the operation of F77-standard-conforming programs. F90 pretty much said that such extensions must be disabled by default; that is, they have to be enabled explicitly.

    The RM compiler has enough extensions to be mostly compatible with the DEC Fortran compilers.
    (RM - Ryan-McFarland) But which DEC FORTRAN compiler? I found the DECSystem 10 compiler to be anything but conforming.

    I remember the CDC-7600 with it's A10 formatting for 10 characters in a 60 bit variable. That was a PINTA to convert to a 32-bit machine.
    That--and the use of arbitrary delimiters led to an interesting "What does this mean?" PSR. That is, H*FOOF* was just as valid a Hollerith constant as 4HFOOF. I've written about the "ambiguous" CDC FTN statement before. (hint: It involved a variable named FORMAT. Before F90, there are no reserved words in FORTRAN. Any pre-F90 compiler that uses them is, by definition, nonstandard.)

    The use of 12 bits to express lower-case and special symbols in NOS led to oddball problems. 10Habcde fills the same word as 10HABCDEABCDE. Then there were "R" and "L" Hollerith constants. F77 with its "CHARACTER" type pretty much made those obsolete.

    But most FORTRANs of the late 60s and early 70s were full of oddball extensions. Even 7090 FMS had oddball extensions, such as "B" in column 1 signified a boolean expression...

    Yet, for all of this, it's entirely possible to write a portable program in pre-F77 FORTRAN. Several of the early microcomputer cross-assemblers were written in FORTRAN. I believe PALASM was also written in FORTRAN.
    Last edited by Chuck(G); May 22nd, 2018 at 04:58 PM.

  5. #25

    Default

    Quote Originally Posted by 2icebitn View Post
    When you refer to RM do you mean IBM Pro F* 1.0 or something later? I would guess code.compiled with that would at least be as fast as quickbasic executables, no? I also have QuickC to fall back on.
    Ryan McFarland sold their compiler under their own name- V2.43 is very stable. It was also marketed as the IBM Professional compiler. I still have the manual and 5.25" 360K floppies. I also had RM version 1.x, had some issues with compatibility on my Leading Edge Model M, maybe others, for the 8087. They fixed it in V2. I disassembled the V1 compiler and library file to fix the problem on my machine, then submitted a bug report. That was when it was new.

    The RM compiler generates inline 8087 code, was very efficient. I still use it, but use the Microway compiler for the increased memory. I do a lot of realtime device control in Fortran/assembly, prefer the control that DOS and PharLap DOS give. PharLap gives most of the memory management that I had under VMS, like mapping a disk file to memory using VMMDRV.EXP.

  6. #26

    Default

    You know what Brian, I never had a hero. But you're my hero dude. Keep typing. I'm loving all of this.

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •