Is NT really new technology?

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.

A Brief History of NT
NT's history is closely tied to that of David N. Cutler, NT's chief architect. After graduating from Michigan's Olivet College in 1965, Cutler worked for DuPont. Although computers weren't his first interest, he ran simulations on Digital machines as part of his job at DuPont. Before long, Cutler was knowledgeable about software and decided he wanted to develop OSs rather than application software. He joined Digital in 1971 and worked at Digital's famous "Mill" facility in Maynard, Massachusetts, developing OSs for the PDP-11 family. RSX-11M is the first OS in which Cutler incorporated major concepts and design principles that later surfaced in NT. RSX-11M is a PDP-11 OS Digital developed for industrial and manufacturing control.

In 1975, Digital realized that its competitors were developing 32-bit processors and that this technology would lure customers away from PDP's 16-bit architecture. Gordon Bell, a legendary figure in computer history and then vice president of engineering for Digital, drove the development of the 32-bit processor, which Digital eventually named VAX. By this time a star within Digital, Cutler was part of the initial VAX development team. Digital had charged Cutler, along with Dick Hustvedt and Peter Lipman, with designing VAX's OS, VMS. Digital's primary design goals for VAX hardware included backward compatibility with PDP-11 processors and enough flexibility that VAX could be the basis for low-end desktop workstations as well as enterprise-level servers. Digital also made VMS backward compatible with RSX-11M and designed VMS to run on different size machines. Of this development period, Digital states in its company history that it was "betting the business" on VAX and VMS. In an eerie echo of Digital's statement, Bill Gates recently claimed that Microsoft is "betting the business" on NT 5.0.

In 1977, Digital announced VAX-11/780 and VMS 1.0, making the first product shipments in 1978. As the project leader and one of VMS's main architects, Cutler continued work on successive releases of VMS, but he became restless at Digital. In 1981, Cutler threatened to leave Digital. To retain its star developer, Digital gave Cutler about 200 hardware and software engineers. Cutler moved his group to Seattle and started a development center. This elite group's goal was to design a new CPU architecture and OS that would lead Digital into the 1990s. Digital called the Cutler group's hardware project Prism, and its OS Mica.

In 1988, Digital executives cancelled Cutler's project and laid off many of its group members. Cutler decided to leave Digital, but before he could do so, Microsoft executives learned of the development and realized they had an ideal opportunity to hire Cutler. At the time Cutler left Digital, the release of VMS was version 5.0 (today's version is 7.1).

In August 1988, Bill Gates hired Cutler. One of Cutler's conditions for moving to Microsoft was that he could bring around 20 former Digital employees with him, including several Prism hardware engineers. Microsoft readily met this demand­the company knew hiring an OS architect of Cutler's stature was a coup, and few engineers had Cutler's track record. In addition, Gates felt that Microsoft's long-term future depended on the development of a new OS that would rival UNIX.

Microsoft's internal project name for the new OS was OS/2 NT, because Microsoft's intention was for the new OS to succeed OS/2 yet retain the OS/2 API as its primary interface. The success of Windows 3.0 in April 1990 altered Microsoft's thinking and its relationship with IBM. Six weeks after Microsoft released Windows 3.0, Microsoft renamed OS/2 NT as Windows NT, and designated the Win32 API (a 32-bit evolution of Windows 3.0's 16-bit API) NT's official API. Gates decided that compatibility with the 16-bit Windows API and the ability to run Windows 3.x applications unmodified were NT's paramount goals, in addition to support for portions of the DOS, OS/2, and POSIX APIs. From 1990 to NT's public release in August 1993, Cutler's team was in a mad dash to complete NT, and the project grew to involve more than 200 engineers and testers. Figure 1 shows a timeline of the major events in the history of NT.

   Prev. page   [1] 2 3     next page



You must log on before posting a comment.

If you don't have a username & password, please register now.

Reader Comments

I enjoyed Mark Russinovich’s excellent “Windows NT and VMS: The Rest of the Story” (December 1998), which compared VMS and NT internals. I was hatched under VMS, and I’ve often found familiar widgets under the NT hood. Like many manufacturing systems administrators, I was sucked into the vortex of Microsoft Office systems several years ago. Nevertheless, I must continue to provide PC access to VMS manu- facturing files. Do you remember which network operating system (NOS) you need to share VMS files with PC-land? (Hint: During this software’s golden era—–one Tuesday afternoon in 1992—–the software’s market share made Banyan VINES look like a sound investment.) OK, I’ll whisper it: “VMS Pathworks.” Guess what? Thousands of VMS boxes that run manufacturing applications day and night still need Pathworks to share VMS files with a PC network. Does Compaq care? Don’t bet Grandpa’s bell-bottoms. Sure, we’ve seen Pathworks upgrades—–ones that work on only Alpha boxes running this week’s version of OpenAXP. What about all those VAX stations that Digital Equipment built to run forever? What do administrators like me do now—–toss 10 development-years’ worth of production software in the bin and throw out our most reliable hardware? Or, dream wistfully that Compaq will develop a conscience before curtailing support for our old systems?<br> --Alan Hodge

Alan Hodge- August 06, 1999

As a <i>Windows NT Magazine</i> subscriber for the past 2 years, I thank you for the accurate, useful, and timely articles. Mark Russinovich’s December cover story “Windows NT and VMS: The Rest of the Story” has historical value, but this history is well known. This article seems either Microsoft-sponsored (trying to give credibility to NT by tracing its roots to DEC’s VMS and David Cutler) or akin to an academic research paper. History cannot be changed. David and VMS are the origins of NT. But, who removed VMS’s stability from NT? What about NT’s equivalent of VMS’s Mount Verification (MV)? I am a systems administrator, and we have a VAX VMS and five NT 4.0 servers. The VAX VMS has run without a crash or problem for the past 4 years. Meanwhile, the NT servers have crashed several times in the first few months of use. Plenty of tweaking and adding extra hardware have decreased the number of crashes, but NT will never be like VMS. Will NT make it to its 20th anniversary, as VMS did in October 1998?<br> --Kersi Marolia<br><br>

<i>Microsoft was unaware of the article until after the magazine went into print. In addition, the idea that Microsoft wants the general public to know that NT’s core is closely based on VMS is unlikely. Have you ever heard Microsoft advertise this link? I’m not sure what you mean by “academic.” If you mean “of little practical value,” you’re correct. However, many people find the untold history of the OS interesting. In the article, I try to clearly describe what NT takes from VMS: the core architecture, philosophy, and in some cases, even the algorithms. NT is still very young compared to VMS, and NT’s developers wrote a lot of the OS’s layered management and application support (e.g., the Win32 environment, the administrative tools, the device drivers) from scratch. The kernel has undergone a lot of tweaking and enhancement. I’m not going to make excuses for NT, but I’m also not going to make general statements about NT’s stability based on only personal experience. Without an academic (scientific) study, that conclusion is meaningless.<br>--Mark Russinovich</i>

Kersi Marolia- August 06, 1999

At Last the truth is comming out about what sort of theft goes on inside the M$ Corp corridors... I am not going to say any more except that WindBlows will never regain it's strangle hold over the computing world The arrival of Linux was just the start of the onslaught against the M$ Corp

Pete

Pete Nikolic- May 21, 2004

What was I thinking? The arrival of Linux has had little impact. VMS has a greater chance than Linux. :)

Pete Nikolic- June 10, 2004

Fascinating. I've just started out learning about the WinNT setup, using the "Inside Windows NT" book. (I have to - I'm hoping to start developing stuff for the MS WinNT platform as well as the Linux/*BSD, and every little bit helps. (Mono and PNet) ;)

Now it looks as if it mightn't be such a bad idea to hunt down the VAX/VMS Internals and Data Structures book as well. Thanks for the info.

Wesley Parish- July 04, 2004

This is very amusing and new to me. Linux was written from scratch yet Microsoft claims it might face crippling patent lawsuits. We now know this is laughable as the only company with a distant chance in hell of winning a lawsuit for damages against Linux was SCO, and we all know the laughing stalk they've become.

Windows NT on the other hand is a wholesale re-write of VMS, steals parts of OS2, and was written by the former employees of the people who wrote VMS in the first place! If anyone is ripe for a lawsuit then it is Microsoft.

I'm sure they've got their backs covered though, and if they don't then they could buy any threatening company out with their 50 Billion in cash. Except for IBM perhaps, that would be a clash of titans.

And to Pete Nikolic, yeah I guess you're right, Linux isn't a threat. That's probably why it continues marginalizing IIS into a niche and dominating the server market ;-)

Anonymous User- February 08, 2005

http://www.euronet.nl/users/frankvw/index.html

Anonymous User- February 22, 2005

The windows source code must be released to the public. Everyone than can laugh at the shoddy programming and stolen stuff in Winows NoTail. Linux is better, bugs are fix faster, but still new hardware support is sometimes a bit weird. Still more stable than winows.

Anonymous User- March 06, 2005

Article Rating 5 out of 5

Dear Anonymous User -March 06, 2005,

It already has been, look up "Windows 2000 Source" in Emule. Some Russian hackers stole the sourcecode and released it - stealing Microsoft's source is like stealing candy from a baby. It's filled with profanities and complaints about how terrible code is, quite amusing really.

Anonymous User- March 08, 2005

Article Rating 5 out of 5

Fantastic article. I have used both VMS and Win/NT intimately for the last 20 years, and Mark's article is right on the money.

Anonymous User- April 13, 2005

Article Rating 5 out of 5

You missed perhaps the most interersting part - the Digtal lawsuit that followed. MS was going to lose so they settled out of court. But the joke was on DEC because although they settled for $50M, it was all in forms that ultimately benefited MS. One part of the deal included MS subsiding the creation of Digital's MS Services practice and the training of DEC's personnel. But this was a move MS was going to make anyway in order to create a global enterprise-class support org such that MS could claim as many MCSE's were certified on NT and there were Unix support professionals in the market. Thus MS could be positioned as enterprise-ready and as supportable as Unix. Secondly, MS guaranteed they'd outsource a large % of their helpdesk calls to Digital call centers. Again, this was MS's model anyway. In that time period when you made a call to the MS helpdesk, and the person on the other end answerered "Hello. This is Microsoft, can I help you?" you were actaully talking to a Digital, NCR, HP, or Vanstar employee in their respective call centers. And even this arrangement was a joke because MS paid so little to DEC and other support partners on a per-incident basis (I recall the math was $25 per call which meant over 10-15 min's in length and you lost money on that particular call), and MS required so much reporting infrastructure and annual training hours, that the support vendors were left with no margin. In the end most of the big vendors finally got out of the MS desktop support business because it was a money loser.

All of this came out of the theft of VMS! Even when they get caught they win.

Anonymous User- June 29, 2005

Article Rating 4 out of 5

Everything comes from somewhere. Linux was not written from scratch. The kernel was coded from scratch, but the ideas and philosophy behind Linux can be traced back to the original AT&T UNIX system. Likewise, much of the software for Linux (C compiler, X Windows, filesystem utilities, etc.) was available on the various UNIX systems long before Linux was even a thought.

The difference, of course, is that Linux *legally* used the ideas and designs already available. Microsoft cannot claim the same thing; by using VMS code in NT, they actually committed copyright infringement. If DEC patented any parts of their system, then MS would also have committed patent infringement.

All in all, this was an interesting and humorous article. I never knew the history of NT either, but at least MS took the design from a stable and robust OS. The instability is not NT's fault; the stability problems come from third-party drivers and poorly written applications, which Microsoft has no control over. I can write a device driver for Linux that will crash the system (guess how I know :) ) too.

Anonymous User- July 15, 2005

Article Rating 5 out of 5

Linux sucks big time....comparing to NT...

Anonymous User- July 22, 2005

Article Rating 5 out of 5

We enjoyed reading it a lot and we really appreciate the work done by Cutler.

rishinitin- September 25, 2006

Article Rating 4 out of 5

 
 

ADS BY GOOGLE