Richard Stallman of course was an advocate of free software, and VMS was proprietary, so he would've disapproved of it on that basis alone. (To be clear, I am not discussing here whether he was right or wrong, only the historical facts.)

But I came across this fascinatingly elliptical remark on https://www.tech-insider.org/vms/research/1982/0111.html suggesting that he also had some technical criticisms:

From: Richard M. Stallman <RMS at MIT-AI>
To: info-vax at SANDIA

The deficiencies of VMS I described two years ago,
which are in the file RMS;VMS > on MIT-AI,
as far as I know have not been fixed.
They add up to a lot of problems with trying to make
a multi-process command environment (such as is used
on ITS and Twenex) to run on VMS, plus lots of assorted
uglinesses and things that can't be done.

A Google search turned up no sign of that file. I would be interested in just what his criticisms were.

  • 4
    No idea what Stallman was complaining about, but VMS processes were fairly heavyweight, a CLI did not run as a separate process, and as far as I can recall, subprocess creation was fairly obscure until the advent of the library routine LIB$SPAWN to make it easy. Not sure when that appeared.
    – dave
    Commented Dec 31, 2021 at 23:12
  • 6
    Stallman has well-known public contact info [email protected]; you could always write to him and ask. Commented Jan 1, 2022 at 19:34
  • 1
    Not really a multi-process issue, but as I recall, you'd have to work really hard to get a ^Y into your user-mode program, since the command interpreter, running in supervisor mode, would have already enabled it as the 'interrupt' key. And ^Y is 'yank' in emacs.
    – dave
    Commented Jan 2, 2022 at 1:06

2 Answers 2


I have access to the file RMS; VMS 7 dated 1980-07-12. I can't make it public, but I can read it and I believe it's ok for me to paraphrase and summarize it.

RMS argues VMS lacks several important features, and suggest he fix it by adding them himself, working together with a VMS expert he names. The feature list is:

  • Access to multiple processes for users.
  • User-space device drivers, like ITS.
  • Improved terminal control.
  • Adding pipes.
  • 7
    Any particular reason why you cannot make it public? I would imagine Stallman himself wouldn’t mind his views republished, plus quoting is generally considered fair use anyway. Commented Jan 3, 2022 at 11:39
  • @user3840170, the files are inside MIT's archives and they decide what can go public an not. Commented Jan 8, 2022 at 13:44

A Google search turned up no sign of that file.

I'd be surprised if a search would turn up anything as RMS;VMS indicates a local file on his account on a machine called MIT-AI. Likewise a reply to this file was located at RWK;VMS on the same machine, so it's been an internal exchange at the MIT AI lab. This is very early Usenet.

I would be interested in just what his criticisms were.

Hard to say, but the discussion this got inserted to was raging on usenet newsgroup FA.info-vax with arguments about VMS being closed source, not easy to change and behaving strange on terminal input, while Unix was considered less capable, more responsive to change but lacking important tools. So it came down to spending big bucks on VMS to get a turnkey system for less IT savvy scientists vs. Unix being cheap but needing a well-paid system hacker to keep it running the way an department needs.

Of course, RMS' arguments may have been different from what was mentioned here.

Some of the posts can be found on this gopher.

A few tidbits may as well be found in Lars Brinkhoff's ITS-Archives covering the same time frame.

Historical Remarks:

  • These messages are so old, they are still in A-News format.

  • At first Usenet had no hierarchical structure. All groups starting with FA. were gated from Arpanet. The 'Great Renaming' in 1985 turned FA.info-vax into comp.os.vms. With the decline of Usenet hard core comp.os.vms users moved in December 2008 to a mailing list called ... well .. 'info-vax' which is still quite active today (2021).

  • 10
    Perhaps rms was merely annoyed at having to use RMS ?
    – dave
    Commented Jan 1, 2022 at 2:51
  • @another-dave Uh, looks like you got a point there :))
    – Raffzahn
    Commented Jan 1, 2022 at 2:57
  • 3
    Personal files will rarely make it out to my public repositories. MIT does have RMS; VMS 7 and RWK; VMS 1, 2, and 3 on file, but there's some amount of red tape to get them out. Commented Jan 2, 2022 at 11:52
  • 2
    RMS is also Bob Spunik! Commented Jan 2, 2022 at 11:53
  • @LarsBrinkhoff Great that they are still around. Thanks.
    – Raffzahn
    Commented Jan 2, 2022 at 13:59

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .