I think there is a long history of application vendors going to great lengths to complain or litigate against Operating System vendors to compete fairly. Specifically, I am talking about "fairness" with respect to the OS-vendor creating, marketing, and selling end-user applications that were broadly popular in the early days of computing. This would include business applications for accounting or database management, spreadsheets, word processors, and other utilities popular in the day but unlikely to be shipped with the OS.

Some "Retro" examples of the controversy between OS and application vendors, dating back to the 1980's, would be:

  1. Apple famously split off Claris around 1987 to be a separate company engaged in the development and marketing of popular application suites for the Macintosh and Apple ][ family (i.e. AppleWorks). Purportedly, this was done to foster a more healthy relationship between Apple and independent application vendors.
  2. As early as 1984, companies like Lotus and WordPerfect seemed concerned with Microsoft using their dominance of PC-DOS as a springboard to compete with new applications, like Multiplan and Word.
  3. The initial investigations into Microsoft's bundling and tying practices began around 1992 with an investigation by the U.S. Federal Trade Commission. An agreement was reached to end the bundling. In the late 1990's, Microsoft would be sued, lose the suit, and only narrowly escape being split into two companies - one for OS and one for applications.

It's not difficult to fathom how the independent application vendors could feel their business was threatened if engineers working on applications could collaborate with engineers working on the OS within the same company. Were they just being paranoid? Assuming this was not paranoia, what was the earliest instance of a mass-market application vendor challenging an OS vendor over the OS vendor competing "unfairly" in the market for similar applications?

It is conceivable, I think, that this was actually an issue of Microsoft anti-competitive practices alone. I'm specifically looking for concrete instances of these complaints before Microsoft's rise. I would assume the large application markets for CP/M (DRI), PDP, VAX (DEC), and IBM machines would have encountered this problem before Microsoft.

  • 3
    I think this is too broad... There are a number of aspects to explore, including insider knowledge of OS details (which became more important as systems got more complex), bundling practices, anti-trust issues (which are older than Microsoft's, see Ma Bell or even IBM) and probably others. Commented Mar 16, 2017 at 23:30
  • I'm really interested in the genesis of the issue in the software industry only. Nothing to do with broad anti-trust. Nothing to do with how the issue has evolved since it first became part of the industry. Can anyone pinpoint IBM as generating the first controversy of this type?
    – Brian H
    Commented Mar 17, 2017 at 0:06
  • 2
    As a side note (I commented this to another question we had recently): I fail to find any advertisement on software packages earlier than ~1980. I can only find hardware ads (and a lot of them). Apparently, there wasn't a market at all for "standard" software before the 80ies
    – tofro
    Commented Mar 17, 2017 at 17:43
  • @tofro I thought that CP/M had a pretty robust selection of independently developed business apps by the end of the 1970s. This being the reason CP/M compatibility was still an issue for the first part of the 80s, and PC-DOS was made to closely resemble it.
    – Brian H
    Commented Mar 17, 2017 at 18:14
  • @tofro See BYTE for examples of software ads in the late 70s. There probably wasn’t much software to advertise at all before then... DDJ didn’t officially allow adverts in its early years. Commented Mar 19, 2017 at 19:46

2 Answers 2


It could be argued that the “war” started with the creation of the software industry, when IBM unbundled its software and services in 1969. Before then, you’d lease a computer and get the software and services along with it. After 1969, leasing or buying the computer was one thing, buying software and services another. Quite a few IBMers and others started their own software businesses, and right from the start they were competing with the OS manufacturer (at least in IBM’s case). (See the last season of Mad Men for a recent depiction of this.)

IBM was under investigation all through the seventies for monopolising the computer market, and was as a result very careful to behave “correctly”. That didn’t stop them from engaging in a wide variety of predatory business practices, and IBM was the industry’s big bad wolf until Microsoft took its place in the nineties.

Competing with the OS vendor is a fact of life in some parts of the software industry, notably in gaming, starting back in the Atari days and continuing today with Nintendo etc. (the latter also being their systems’ gatekeepers). In the application world there were fewer opportunities for this kind of strife: there weren’t many companies which were interested both in developing operating systems and applications running on those operating systems. Applications as we understand them nowadays weren’t a thing before the advent of micro-computers: before that, non-batch applications were stand-alone devices (typically, word processors), so there was no competition between the operating system vendor and the application vendor. The first software application is perhaps Electric Pencil, released for the MITS Altair in 1976 (and MITS was in no position to compete with application vendors). I don’t think any of the companies you mention ever had a serious application development arm: Digital Research developed operating systems, language tools and GEM; DEC built hardware and operating systems, but no applications; IBM did have software such as ATS but it never dominated any application market after the unbundling.

An early example of an OS vendor interested in producing applications and trying to control the market is Atari; they developed the operating systems for their range of 8-bit computers in the late seventies, and also developed a range of applications in the early to mid-eighties (AtariWriter for example). At first, they only provided technical documentation for their platform to companies willing to sign an NDA, and it was only with the publication of De Re Atari in 1982 that documentation became widely available (coincidentally, Atari released their own official documentation in 1982; De Re Atari itself was written by Atari employees and published by APX). Atari weren’t the only OS vendor publishing applications — Apple had their Apple Writer — but as far as I’m aware they’re the only vendor which didn’t play nice with their platform’s (potential) ISVs.

In the business world it’s often said that complaints about companies are only posturing until they become a lawsuit. Looking at the software industry from that angle, I can’t think of any lawsuit by ISVs against an OS vendor competing with them in the application space until the Microsoft anti-trust lawsuits in the early nineties. There were lawsuits in the eighties, but primarily between OS vendors (Apple v. Microsoft) or between application vendors (Lotus v. Borland etc.). The PC software market wasn’t dominated by an OS vendor until Microsoft’s dominance in the late nineties: before that, the dominant applications were WordPerfect in the word processor space, VisiCalc then 1-2-3 in the spreadsheet space, and dBase in the database space. The story is slightly different on the Macintosh, where Claris had a few market-leaders (FileMaker in particular), but Microsoft quickly dominated the Macintosh software market.

Things really came to a head with the sudden rise of Windows with the release of version 3.0. At the time, Microsoft was one of the only vendors actively developing software for Windows (Word and Excel); the other incumbents were either developing for OS/2 (Lotus) or largely resting on their DOS laurels (WordPerfect; see W. E. Pete Peterson’s very interesting Almost Perfect — Microsoft were actively courting ISVs so that their platform would be a success). At the time, the future of the PC was still very much up in the air, and many industry pundits still thought that DOS was dead and people should be looking at NeXTSTEP, Unix, OS/2, or even MacOS, and many ISVs were spending time porting their software to other platforms (Lotus with Improv on NeXT and 1-2-3 on OS/2 and MacOS, WordPerfect on everything from the Atari ST to Unix...). Microsoft were in the unique position of having applications ready for Windows, and suddenly having a decent platform (Windows 3.0 meant that Word and Excel could run much better than anything in DOS, without any changes). At this point, the lack of separation between the OS and application divisions became a real issue: while Microsoft had no doubt been able to take advantage of undocumented features in DOS, that didn’t seem to have much effect in the application market; in Windows it became much more obvious that Microsoft’s applications performed better thanks to the use of undocumented features. Microsoft were also in the unique position of being able to bundle applications along with the operating system, and that’s where the main complaints came from (and the lawsuits, albeit somewhat artificial — you could argue that Novell’s nineties buying spree was more about building a case against Microsoft than building a really competitive application suite).

In the utility space the situation is different again, with many vendors being driven out of business when equivalent tools were integrated into the operating systems. The vendors that survived were those that either licensed their technology to the OS vendor (Central Point, Norton) or those that uped the ante and produced even better tools (Quarterdeck, at least with its QEMM). Lawsuits did occur too; the most famous of these are the suit brought by Stac Electronics against Microsoft after the release of MS-DOS 6.0 with Doublespace, and of course United States v. Microsoft which centred on the bundling of Internet Explorer with Windows 95 (before which there was a booming Internet browser industry, not just Netscape Navigator).

All this to say that as far as I can tell, although the ingredients were there earlier, and some companies did attempt to take advantage of their position as OS vendor to help their applications, Microsoft was the first company to dominate the markets with their OS and their applications, and as such was the first company (and so far only...) to attract serious complaints and lawsuits from ISVs.

  • Mainframe software applications were mainly bespoke. I'm inclined to think that a commercial software industry selling non-custom apps didn't exist before mini-computers took off. Also languages and "system software" aren't apps.
    – Brian H
    Commented Mar 17, 2017 at 13:02
  • Just because it's bespoke doesn't mean there aren't ISVs providing it and feeling the squeeze from the manufacturer. Where did I mention languages and system software? Commented Mar 17, 2017 at 13:35
  • Competition over contract programming between ISV's and a manufacturer isn't part of any ecosystem/market for general-purpose apps - like accounting, database, wp, etc. ISV products in the early days were languages and system software, since these had a broad market appeal to custom and in-house application developers. Application markets came later. I'm hoping to pinpoint the year when controversy first erupted specifically over competition for general-purpose apps.
    – Brian H
    Commented Mar 17, 2017 at 14:17
  • OK — could you update your question to clarify all that? Commented Mar 17, 2017 at 17:04
  • I amended the question to specifically ask "...what was the earliest instance of a mass-market application vendor challenging an OS vendor...?"
    – Brian H
    Commented Mar 17, 2017 at 18:39

That "war" is/was not a general one and not in all brands of computers. Also, it was maybe more present in areas that targeted business customers (which is where the money is) than in typical home computer markets.

Sinclair and Amstrad as notable vendors of home computers and, as such, "OS vendors", for example, didn't even develop their own application software.

Both did re-selling of third-party applications under their own label, in the case of Amstrad only very reluctantly through Amsoft, the unloved Amstrad software arm that was only seen as a means of initial market push and very early ceased away.

Reason for that is that in the early days, the "OS" in this market was considered a non-divisable part of the computer (in most cases, it even physically was, as it was delivered in ROMs). Once the computer was sold, there was no other way to sell a new version of the OS than to sell a new computer.

Also, none of the two mentioned companies hadn't much of in-house software development capacity - Large parts of the Sinclair SW development was done by a company named Nine Tiles, and Amstrad had most of their software contracted out to Locomotive Software. Apparently, they believed the money was in hardware, not software.

  • 1
    I’m not sure about the reason you give — the fact that you can’t sell OS upgrades doesn’t make it any less interesting to use your insider knowledge of the OS to make better applications to sell to users. In many cases in the early years I think the vendors were simply too small to consider developing anything beyond the OS or BASIC (when they developed that), or they just didn’t see the point (or in Amstrad’s case, Alan Sugar didn’t see the point). Atari as a counter-example did develop or repackage their own application software (Atari Writer etc.). Commented Mar 17, 2017 at 17:08
  • @StevenKitt The main reason was something I probably have considered as a given and shouldn't have: Both weren't in any way "Software Companies". To my knowledge, Sinclair contracted out the Spectrum and ZX81 Basic, and the first (and last) "OS" they actually developed in-house was the QL operating system (where the main developer was made to leave before he considered his job done). Same for Amstrad - apparently, all of the Amstrad's ROMs were contracted out to Locomotive Software.
    – tofro
    Commented Mar 17, 2017 at 17:39
  • 1
    Yes, agreed. The relationship between Amstrad and Locomotive is an interesting one; a couple of videos were released recently detailing the early history of Amstrad computers (the CPC range), I'll try to find them again. All that "not software companies" business was where Microsoft's rise started — they realised that turning their BASIC into the de-facto 8-bit standard was worth pursuing. Commented Mar 17, 2017 at 18:04
  • @StephenKitt There's an online copy of Alan Sugar's (3rd-party-written, but apparently authorised) memoirs somewhere. Very interesting read.
    – tofro
    Commented Mar 17, 2017 at 20:08

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .