0rrodrig@TEKNOWLEDGE-VAXC.ARPA (Russ Rodriguez) (11/16/88)
Hello:
I'm a student at Sonoma State University and have gotten extremely
sick and tired of the version that came with the machine. Has a more recent
version (18.52?) been ported to this machine? If so, can you lead me to it so
that I can ftp it? Many thanks.
Russell Rodriguez
------------------------------------------------------------------------------
Disclaimer: The opinions of Russ Rodriguez are his own and couldn't possibly
reflect those of his employer: Teknowledge, Inc., of Palo Alto, CA
------------------------------------------------------------------------------
internet: 0rrodrig@teknowledge-vaxc.arpa
usenet: {uunet|sun|ucbvax|decwrl|uw-beaver}!0rrodrig%teknowledge-vaxc.arpa
home: 2420 Slater Street, Santa Rosa, CA., 95404-2143 Phone: 707-571-8172hecker@fed1.prime.com (Frank Hecker) (11/17/88)
I work for Prime Federal Operations in Rockville, Maryland. I'm a long-time user of Prime EMACS for the Prime 50 Series under PRIMOS, and have also recently ported GNU Emacs to the Prime EXL Series (80386-based UNIX System V systems). In response to the query by Russ Rodriguez regarding GNU Emacs for PRIMOS, I must confess that its wording somewhat confused me. Did the phrase "the version that came with the machine" refer to the EMACS editor sold and supported by Prime? If so, be aware that Prime EMACS was not derived directly from the present GNU Emacs but rather was one of the numerous Emacs clones created in imitation of the original TECO-based Emacs. (I believe Prime EMACS was introduced as a product in 1982 or 1983.) Prime EMACS is written in SPL (a Prime-specific dialect of PL/I) rather than C, and as I recall the original version was acquired by Prime from Software Arts, Inc. (remember Visicalc?). I am not aware of any version of GNU Emacs (18.52 or otherwise) that has been ported to the Prime 50 Series under PRIMOS. The potential problems in doing so are as follows: 1) Changes would be required in the area of OS-dependent routines similar in nature and extent to the changes required for VMS support. 2) Adequate performance would require using the -32ix option to the Prime C compiler to generate instructions using the new IX mode addressing (new since 1983, that is). (IX mode for C basically gives you adequate byte addressing; the earlier V mode used 16-bit word addressing by default, with byte addressing requiring a 48-bit pointer.) The resulting code would only run on IX-mode systems (all four-digit models except the 2250). It would _not_ run on earlier systems such as the 750. 3) Even after using IX mode, you would probably have to do extensive performance tuning to deal with some of the peculiarities of the 50 Series virtual memory architecture and the PRIMOS shared library system. I used the original beta version of Prime EMACS and it was quite slow, especially in terms of startup time. The current version is pretty well-tuned, but requires the use of some techniques that are still not documented and supported for customer use. If anyone is still interested in doing a port of GNU Emacs to PRIMOS I will be glad to advise to the best of my ability, and am also willing to supply a standard 1600-bpi tape in PRIMOS MAGSAV format containing the current GNU Emacs version 18.52 distribution. Please respond via email to the above address or feel free to call me at (301) 948-7010 x308. Frank Hecker