Advantages and disadvantages of Windows NT 3.1

# Advantages and disadvantages of Windows NT 3.1 Source: [https://dfarq.homeip.net/advantages-and-disadvantages-of-windows-nt-3-1](https://dfarq.homeip.net/advantages-and-disadvantages-of-windows-nt-3-1) I’ve talked a lot about the advantages and disadvantages of old milestone operating systems\. But what were the advantages and disadvantages of Windows NT 3\.1, first released July 27, 1993? That’s a fair question, and there are valid reasons it proved transformational even though it initially did anything but take the world by storm\. [](https://dfarq.homeip.net/advantages-and-disadvantages-of-windows-nt-3-1/windows-nt-3-1/)Although few people actually used it, Windows NT 3\.1 was a landmark release for Microsoft\.Windows NT 3\.1 was a landmark for Microsoft: its first fully 32\-bit operating system\. It was the weird offspring of[Windows 3\.1](https://dfarq.homeip.net/disadvantages-of-windows-31/),[IBM’s OS/2](https://dfarq.homeip.net/os-2-warp-a-tribute/), and[DEC](https://dfarq.homeip.net/what-happened-to-digital-equipment-corporation/)‘s VMS\. How could it have three parents? Don’t ask\. Just smile and nod\. OK, I’ll explain\. IBM and Microsoft co\-developed OS/2 as a successor to[DOS](https://dfarq.homeip.net/advantages-and-disadvantages-of-ms-dos/), which was popular but really showing its age by the mid 1980s\. But they had a fundamental disagreement over how 16\-bit Windows fit into everything\. Eventually the disagreements led to a very public divorce that left IBM confused\. IBM soldiered alone, developing and eventually releasing OS/2 2\.0 and OS/2 2\.1, both of which worked rather well and gained a cult following\. Microsoft took its nascent code that it intended to form the base of OS/2 3\.0, renamed it Windows New Technology, and eventually released it as Windows NT 3\.1\. So how does DEC VMS fit in? VMS was \(and remains\) an operating system for minicomputers\. Microsoft hired Dave Cutler, the chief architect of VMS, to be the chief architect of Windows NT\. Cutler’s work on VMS and other DEC operating systems influenced the internal workings of NT\. Being 32\-bit, it was more stable than[16\-bit Windows 3\.1](https://dfarq.homeip.net/windows-3-1-released-april-6-1992/)\. It also had the ability to do something resembling real security\. The implementation wasn’t perfect, but the capability was there\. It also did real pre\-emptive multitasking like OS/2 and[Amiga](https://dfarq.homeip.net/amiga-1000-ten-years-ahead-of-its-time/)\. Believe it or not, in 1993 there was some question whether Intel x86 would dominate the future of the CPU field\. Microsoft hedged its bets by porting NT to any 32\-bit chip it could think of\. Its initial target was the Intel i860, then the MIPS R3000\. When it was released, it ran on two architectures besides Intel x86\. One was MIPS, the chip architecture later used many consumer routers, including the venerable[Linksys WRT54G](https://dfarq.homeip.net/is-the-linksys-wrt54g-obsolete/)\. The other was the ill\-fated DEC Alpha, a criminally underrated chip from the 1990s\. PowerPC followed in 1995, and Sun SPARC was planned\. In 1993, this versatility was an advantage\. Notably missing was ARM, but ARM wasn’t getting the attention other RISC chips were in 1993\. But porting to ARM later proved straightforward because Windows NT was designed to be portable\. So was porting to[AMD64](https://dfarq.homeip.net/athlon-64-how-amd-turned-the-tables-on-intel/)\. Aside from the improved stability and security, it had a new filesystem, NTFS, that was more efficient and faster than the ancient FAT filesystem DOS used\. NT 3\.1 was mostly compatible with 16\-bit Windows, and its “Windows on Windows” approach to running 16\-bit apps was more transparent than OS/2’s approach\. Finally, its user interface was more familiar than OS/2\. It was completely different under the hood, but if you could use Windows 3\.1, you could use NT 3\.1 too\. ## Disadvantages of Windows NT 3\.1 In 1993, memory was expensive, and Windows NT wanted 16 megabytes of it\. In 1993, 4 megabytes of RAM was standard and some systems still came with as little as 2\. At the time, 16 MB of RAM cost around $600\. The rest of the computer cost around $1,000\. So you paid a 40% premium to get enough memory to run NT comfortably\. Setting up NT 3\.1 was also difficult, since Plug and Play was still a couple of years off\. Driver support was also anemic\. Most consumer hardware didn’t have NT drivers at first, so you had to dual boot with DOS and Windows 3\.1 to use a lot of computer hardware we take for granted today\. Power management was non\-existent, so battery life was much better on a laptop if you ran something else\. Almost anything else\. It also couldn’t take advantage of power saving features of the new Energy Star PCs like an[IBM PS/2E](https://dfarq.homeip.net/ibm-ps-2-e-the-first-energy-star-computer/)\. In the early 1990s when I would get into operating system debates, I advocated for OS/2\. People always said, “What about Windows NT?” That was a short argument because I only ever met one person who ran Windows NT 3\.1 on a regular basis\. I was[working at Best Buy](https://dfarq.homeip.net/memories-of-the-land-of-blue-shirts/)and he came in with a pretty long shopping list\. But we couldn’t find a sound card, CD\-ROM drive, or flatbed scanner that would work with NT 3\.1 at the time\. Or at least I couldn’t guarantee any of it would work, since it didn’t say so on the box\. DOS worked\. Windows 3\.1 worked\. Even OS/2 usually worked\. For NT, you were on your own early on\. The best case scenario involved lots of hunting for and downloading drivers from[BBSes](https://dfarq.homeip.net/what-does-bbs-mean-bulletin-board-system/)and[Compuserve](https://dfarq.homeip.net/what-happened-to-compuserve/)\. Windows NT grew up, but 3\.1 was a bit of a rush job and it showed\. It was really NT 3\.51, released in May 1995, that gained a lot of use, and NT 4\.0,[released in July 1996](https://dfarq.homeip.net/windows-nt-4-0-released-to-manufacturing-july-31-1996/), that took Windows NT into something resembling mainstream\. For several years, Windows NT was something that people talked about while they ran Windows 3\.1 instead\. Windows was one of the most important changes of[1990s computing](https://dfarq.homeip.net/computers-in-the-1990s/), and Windows NT had a lot to do with that, but the promise of there someday being one Windows that could be everything to everyone, and delivering it a decade later was probably a once in a lifetime event\. Microsoft has to move faster than that today\. ## Legacy Windows NT 3\.1 is probably the most influential operating system nobody used\. It lived up to its promise, as every Microsoft operating system from Windows XP onward is its direct descendant\. If you ever wondered why you have to hit[CTRL\-ALT\-DEL](https://dfarq.homeip.net/a-non-revisionist-take-on-ctrl-alt-del/)to log in, it was so Windows NT could get C2 Orange Book certification from the U\.S\. Government\. Windows NT 3\.1 was Microsoft’s attempt at a “Unix killer\.” That didn’t exactly happen\. Both Unix and NT are far, far more widespread now than anyone had any right to imagine in 1993\. But NT definitely did put a dent in proprietary Unix\. There’s a lot less AIX, Solaris, and HP\-UX around today\. Much of that is because of Linux, but NT had something to do with it too\. And Windows NT turned out to be far, far more successful than[Microsoft’s ill\-fated Unix product](https://dfarq.homeip.net/sco-v-ibm-winds-toward-resolution/)\. Oddly, even though NT was designed to leave Intel behind if necessary, it[may have saved Intel’s x86 architecture](https://dfarq.homeip.net/why-intel-cant-quit-x86/)\. Since NT could run on cheap PCs, that was what most people bought, and they put up with the limitations of a $1,600 PC\. It wasn’t as good as a $3,000 RISC workstation but it was probably better than half as good\. Sustaining that market share gave Intel the cashflow it needed to eventually make x86 competitive with those other chip architectures, notably starting with the[Pentium Pro](https://dfarq.homeip.net/intel-pentium-pro-how-it-could-succeed-and-fail-simultaneously/)\. You don’t hear[much talk of x86 going away anymore](https://dfarq.homeip.net/random-question-will-arm-overtake-x86/)\.  David Farquhar is a computer security professional, entrepreneur, and author\. He has written professionally about computers since 1991, so he was writing about retro computers when they were still new\. He has been working in IT professionally since 1994 and has specialized in vulnerability management since 2013\. He holds Security\+ and CISSP certifications\. Today he blogs five times a week, mostly about retro computers and retro gaming covering the time period from 1975 to 2000\.