From mboxrd@z Thu Jan 1 00:00:00 1970 From: drb@msu.edu (Dennis Boone) Date: Wed, 21 Mar 2018 17:15:37 -0400 Subject: [TUHS] FORTRAN In-Reply-To: (Your message of Wed, 21 Mar 2018 19:10:42 +1100.) <20180321081042.GA7611@server.rulingia.com> References: <20180321081042.GA7611@server.rulingia.com> <61051ebbca4809c08b60e92014851069e83a07f8@webmail.yaccman.com> <002f01d3c085$6b5a26d0$420e7470$@ronnatalie.com> Message-ID: <20180321211537.BFE39A58617@yagi.h-net.msu.edu> > >Yes, Fortran is as awful for system programming as C is for numeric > >programming that involves throwing multidimensional arrays around. > Note that Pr1meOS was written in Fortran. I did study it but no > longer recall what extensions it had to make that practical. It's just PRIMOS, no E. And the '1' in place of 'i' thing was just a marketing/logo gimmick. Surprisingly few language extensions. Octal constants (:1234567). A file inclusion facility ($INSERT FILE>PATH>XYZ.INS.FTN). Not much else. Originally, much of PRIMOS was in FORTRAN, with some assembler (PMA). Later, significant rewrites and extensions were done in PL/1 derived systems languages (PLP and SPL), and even later some in Modula. Awfulness is relative. Bill Poduska has said that writing most of the system in a higher level language saved them a lot of time, over using assembler. De