Submitted by This is an article written 20 years ago by Mark Russinovich, which compares VMS and Windows NT.
When Microsoft released the first version of Windows NT in April 1993, the company’s marketing and public relations campaign heavily emphasized the NT (i.e., New Technology) in the operating system’s (OS’s) name. Microsoft promoted NT as a cutting-edge OS that included all the features users expected in an OS for workstations and small to midsized servers. Although NT was a new OS in 1993, with a new API (i.e., Win32) and new user and systems-management tools, the roots of NT’s core architecture and implementation extend back to the mid-1970s.
And now… The rest of the story: I’ll take you on a short tour of NT’s lineage, which leads back to Digital and its VMS OS. Most of NT’s lead developers, including VMS’s chief architect, came from Digital, and their background heavily influenced NT’s development. After I talk about NT’s roots, I’ll discuss the more-than-coincidental similarities between NT and VMS, and how Digital reacted to NT’s release.
Great read.
Pascal Zachary’s book “Showstopper! The Breakneck Race to Create Windows NT and the Next Generation at Microsoft” is also great reading for anyone interested in how NT came to be.
Especially the information around the software environments (I forget their technical term for it) that it would support. NT could run OS/2 programs, win16 programs, win32 programs, POSIX programs, and one other one, I think.
And the creation of the HAL lead to it being usable on x86, MIPS, PowerPC, amd64, Itanium, and probably others I’m forgetting about.
Edited 2018-06-22 21:49 UTC
Subsystems? Well, it’s what was used to create the Linux support.
I think Personality is what they used to describe the software environment applications ran in/on top of.
I was under the impression that the NT kernel had “subsystems” and NTFS had “personalities”.
Yeah, personalities. From the book (as found via Google Books):
OS/2 PM was the primary personality for most of NT’s development (it was originally the next version of OS/2, after all), but it was switched shortly before release to Win32.
Edited 2018-06-23 03:36 UTC
Some things in the article don’t seem to me to really mark any resemblance of VMS to Windows NT, like how the kernel and current user program share the memory map (the kernel basically exists in the memory maps of all programs) has been almost universal to VM OSes, and the idea of pools of paged and non-paged memory are basically universal to OSes running on x86 (and most common CPU architectures now). Linux, BSD, Haiku, and basically all modern OSes share these traits.
But that said, I don’t really know if these things were common to non-DOS/Mac OS OSes in the very early 1990’s.
Edited 2018-06-22 22:47 UTC
David Solomon’s book “Inside Windows NT” treats this topic with lots of detail.
Cutler is one of the greatest thinkers in operating system design. MSFT was fortunate to snag him from DEC while that formerly great company imploded.
Yes, today’s most popular desktop OS and second most popular server OS is based on concepts formed at Digital Equipment Corporation when Cutler et al. were developing VAX/VMS for minicomputer clusters developed to replace the PDP-11 in the 1970s.
Edited 2018-06-23 05:20 UTC
Agreed, Mr Cutler has had a huge amount of influence on computing and I’ll credit MS the NT project. That’s one chunk of software (the kernel side) they did pretty damn well. It’s not perfect but no software is. Windows 2000 was a revelation. NT 3.5 had shown how solid NT could be and 2000 put, IMNSHO, a pretty good GUI on top AND stay very solid indeed.
To the general public he’s probably less well known than Linus Torvalds (and not many know him!) but that’s often the way.
It’s funny how that works. I think of these people as the singles that never made it to number one in the “hit parade”. 20 years later everyone’s forgotten about most of those number 1 singles but some songs, even though not chart popular, end up hanging around history and influencing untold numbers of other people and their music. It isn’t that their song wasn’t great, there’s often something glitzier or flashier (or sillier) or easier that got the popular vote at the time.
It was sad to see the demise of DEC. If for no other reason than Alpha is one of the best CPUs I’ve ever had the pleasure to hack on.
Something I found interesting was not only were there similarities between VMS and Windows NT, but if you line up their initials you see Microsoft’s product is just one letter further along in each position: VMS -> WNT.
Sort of like HAL from “2001” is just one letter removed in each place from IBM.
I never noticed the VMS » WNT one-letter displacement probably because most abbreviate Windows NT as WinNT.
Was the one-letter displacement on the abbreviation a guided coincidence?
Dec gave us more than just the seeds of windows nt,
Storageworks gave us most modern hard drive tech between 1 and 2.1 gigabytes.
Alpha gave us the amd athlon
I’m wondering what the change in management was that stopped the innovation in the 1990s. Before compaq bought them.
>I’m wondering what the change in
>management was that stopped the
>innovation in the 1990s
It wasn’t so much that innovation stopped. DEC kept doing what it was doing. Kept inventing cool stuff. But the world passed Digital by. The stuff they were inventing wasn’t what the world needed anymore, even if it was cool.
The PC landscape changed, and the entire industry was changed. Digital was locked-on to their world view: Expensive, vendor-specific, workstations (and all hardware, really)… proprietary networking (DECnet anyone?)… dumb terminals connected to big machines in the basement. The world changed. To use a hackneyed but accurate phrase (that I first learned at DEC), “the paradigm shifted” and DEC was stuck on the old paradigm.
People discovered that they didn’t need to pay $30K+ (1990) dollars (minimum) for a VAXStation in their office so they could answer email and create documents. The VAXStation in my office in 1990 retailed for more than $65K, IIRC.
People discovered that they didn’t need to pay $20K-$50K+ for a network router, either. Yes, you could (literally) drop a DEC router out the window of a building and have it survive… guess what? People decided that level of robustness probably wasn’t needed in their network equipment. Oh, yeah… And there was this little thing called TCP/IP that became, ah, “rather popular”…
This kind of “paradigm paralysis” (sorry, again) is what gets most big companies eventually. Look at Ford. Look at GE. It’s what threatening MSFT right now.
I wonder if David Cutler and Linus Torvalds ever met and if they would hit it off together.
One of Cutler’s quotes: ‘Unix is akin to a religion to many. Sorry, I do not believe in that religion.’
A very cute article, and quite true.
What’s really MISSING from the article are the contributions that the Prism OS architecture (that Cutler and team were working on at DEC in Bellevue when the project got cancelled) brought to NT. This is understandable, because the Prism design details were a closely-held secret, even within Digital (design documents were numbered and individually accounted for). NT is a direct descendant of Prism, which itself adopted many VMS and RSX-11M concepts. Probably the most striking difference between VMS and NT is the concept of OS Emulation Subsystems… and this concept was *very* well developed on Prism.
Cutler is a very fine engineer, and what really stands out about his engineering ability (aside from the fact that his Macro-11 coding was pure poetry to read… truly unbelievable) is how very pragmatic it is. Never a theoretician, and not formally trained in CS, Cutler designs stuff that *works* and meets practical needs.
Somebody needs to write a really good history book about NT. Showstopper is just terrible, and suffers from a lack of technical insight and appreciation of what’s important on behalf of the author.