Steve et al.

I'm not exactly sure how it fits into this discussion, but you ought to take a 
look at the Lockheed MIS system, which is now an IEEE milestone at:
https://ethw.org/Milestones: Pioneering_Medical_Information_System,_1965-1974 

Its prototype began in 1966 and was fully operational by 1968, running at the 
Mayo Clinic on an SDS-92 and operating about five terminals and several 
printers.  It was then ported to an IBM/360 M40 and was in production on an IBM 
S/370 M155 about 1972.  

Essentially, MIS software was a transaction processing supervisor that took 
over control of the hardware and turned it into a multi-processing transaction 
processor which by the mid-70s was in production servicing four different 
hospitals with hundreds of terminals.  A block diagram of the system 
architecture is on sheet 52 of this document: 
https://drive.google.com/file/d/1GogbWimvoNvh2n1550Qxa0wmHVDtZWVM/view?usp=sharing.
  As it says, "This software is organized in three basic groups—each of which 
operates relatively independently from the others on a time-shared basis—and 
communicates with the others by passing messages back and forth."   In the IBM 
DOS era, DOS was just one of the three "basic groups" with the others being a 
terminal handling group and a transaction processor.  How they managed to run 
it on later IBM OSs is a bit of a mystery, but it ran on IBM mainframes until 
2020!

It died this century, mainly because management never extended its capability 
into the physician's office.

>From a UX perspective, it was renowned for its response time and ease of use, 
>and most current users familiar with MIS say its performance still isn't 
>matched by today's medical informatics systems.

tom

 


On Sun, Aug 23, 2026 at 3:47 AM Steve Lewis via cctalk <[email protected]> 
wrote:
>
> Hey gang, a few months ago I presented about timeshare and some 
> aspects about operating systems since the 1950s.
>
> To buzz through a PDF or PPT version of the notes:
> https://github.com/voidstar78/VCF2026
>
> VCFSW has the talk archived here:
> https://www.youtube.com/watch?v=9hE235ti9XQ
>
> Sorry being a bit dull as a speaker, but a few highlights:
>
> - overview of Douglass Ross "Director Tape" for Whirlwind (only "what 
> it looked like" evidence that I was ever able to find, would be 
> curious if others knew more about it)
>
> - 1965 Honeywell ADMIRAL OS: while I couldn't find much info about 
> that specific OS, what I did find what Honeywell went to bat for us 
> all vs RAND corp, putting an end some concerns about ENIAC hardware 
> patents (that's a very short version of it)
>
> - today we fight about \ vs / slashes, when MULTICS split the 
> difference and used > (greater than symbol), at least as far as I 
> could tell from the manuals
>
> - some verification that UNIX was (sort of) originally called UNICS 
> (see footnotes in the PPT notes), never formally written down but 
> "verified" in the past year
>
> - I made a claim that Datapoint CTOS (1972) was the only OS ever on a 
> cassette tape (not to be confused with the late 1970s CTOS) - that 
> said, I suppose the MCM/70 did as well; really admires Mattis Lind's 
> work on the
> DP2200
>
>
> I never found too much info about "tymshare" (apparently a late 1960s 
> kind of "corporate or business-to-business digital exchange" system, 
> maybe for inventory or check-processing?).
>
>
> Tons of other OS's, obviously couldn't cover them all in a VCF talk.  
> But there are a bunch of addendum notes after the end of the slides:
>
> - Mike Muuss  ENIAC archive (at a .mil ftp if anyone wants to archive 
> it more properly?)
> - I still claim that the word "operating system" was never used before 
> 1960; the "SOS" or "share OS" was a name applied later (1962ish) to 
> those pre-1960s concepts [ afaik ]
> - I couldn't fit in a bit about squozed :) really wanted to, just 
> couldn't make it fit
> - AOSP,  EXEC II,  IBM 1800 derivatives (anyone ever see one of these 
> in the wild?),
>
>
> Oh, not officially launched yet, but I'm now self hosting   oldsayings.info
> (with a telnet front-end at port 42525).  Kicking around some ideas, 
> but it's reliably staying up now with Starlink as the ISP.  I got 
> tired of waiting for fiber - but the county workers are getting 
> closer, I see them running it ranch by ranch.
>
>
>
> Cheers, wish could have made it to VCF West this year.
> -Steve



--
--Carey

Reply via email to