Microsoft gets back to its UNIX roots with Interix
2000-02-14
Given the current polarisation of the UNIX/Linux and the Microsoft/Windows camps, you could be forgiven for disbelieving Microsoft’s UNIX past. Today, IT history has chosen to quietly sweep under the carpet the fact that Microsoft played an instrumental part in bringing UNIX to the PC, a heritage which subesquently bequeathed Minix, then Linux. Quietly forgotten, but true. Today, with Microsoft’s announcement to integrate Interix with Windows 2000, it looks like the software giant has decided to take a leaf out of its own history book.
It was back in 1979 that Microsoft and The Santa Cruz Operation (SCO) collaborated to develop Xenix2, which was the first UNIX implementation for the Intel 8086 chip. Development was later handed over to SCO so that Microsoft could concentrate on its single-user operating systems in the form of DOS and Windows 2. Xenix2 gave rise to SCO UNIX, which became OpenServer, merged after a fashion with UnixWare and is now being integrated with AIX to spawn Monterey.
More recently, in September 1999 Microsoft acquired Softway Inc., developers of Interix. This product started life as OpenNT to replace the inadequate POSIX system that was bundled with Windows NT. From the outset the intention was to give companies the ability to more easily port applications from UNIX to NT, by providing a UNIX layer that runs on top of the NT kernel. The trouble is, Software did such a good job of developing Interix that the OS layer has passed compatibility tests of the Open Group, keepers of the keys to UNIX.
As it is, Microsoft are keen to distance themselves from their heritage. On its “Linux Myths” web page, for example, the company states disparagingly that “Linux fundamentally relies on 30-year-old operating system technology and architecture.” The line is (to quote Garfield) that Windows 2000 is new and improved, UNIX is old and inferior. So – why, oh why, in the shape of Interix, is Microsoft investing so heavily in UNIX?
The answer, according to Microsoft, is that by providing mechanisms which bridge the gap between the UNIX and Windows NT platforms, companies will migrate from the former to the latter. This is a tactic which has worked in the past, for example with the provision of facilities in Microsoft Office to read and write competitors files. Companies such as Lotus and WordPerfect provided read-only mechanisms and found to their cost that users were turning to Microsoft as a result. Interix, too, is a two-way bridge and Microsoft’s tactic, though proven, is not without risk. For a start by supporting a UNIX layer the company is, in some way, validating the existence of UNIX. Thirty-year old technology or no, says Microsoft, it runs perfectly well on top of our kernel. So it should, for that matter, and so will it run perfectly well in other guises such as Linux, Solaris, HP-UX or AIX. Secondly, Windows 2000 is already under pressure from a number of fronts. Linux is one example. Also, commentators are recommending that the new OS goes through a purgatory period prior to its more widespread adoption. Finally there remains the looming shadow of antitrust case, which is having knock-on effects in Europe and elsewhere.
This is not the first time that Microsoft has played the UNIX-for-Windows card. Last year, for example, the company teamed with Mortice Kern Systems to provide a set of Unix tools for NT. Like it or no, and despite Microsoft’s best efforts to shake it off, the reality is that UNIX will remain part of the landscape for the foreseeable future. A product like Interix may prove to be an asset, not to lead the company to world domination but to keep it in the game.
(First published 14 February 2000)