Why did Apple and Microsoft succeed when there were many other companies like the MITS? originally appeared on Quora - the place to gain and share knowledge, empowering people to learn from others and better understand the world.
Answer by Stan Hanks, CTO of Columbia Ventures Corp, on Quora:
I’ve been analyzing this for a company which I advise, trying to steer them in a more productive direction. I recognize the hallmarks, have for a long time, but only recently found the words to describe it.
It’s platform and ecosystem.
So, back at the start, with the Altair 8800 and the Processor Technology SOL-20, and the IMSAI 8080 - there was this huge dearth of software. Hardware abounded - there had to have been at least five or six dozen different designs you could purchase, and it was pretty trivial to cobble together a system of your own design in those days.
But software… that was missing. If you wanted to write a program for your Altair, you had to know what devices were on which I/O ports, how the memory buffers mapped into your memory space, and you had to write all of the code to do that on top of whatever the hell it was you wanted to write - if you wanted to knock off Colossal Cave Adventure, that was cool, but you not only had to write the game, you had to write all the system level code that made it possible to run and load and communicate with the user.
That changed in late 1974, when Gary Kildall released CP/M. CP/M abstracted out the specifics of memory mapping and I/O port mapping and turned calls for I/O and file management (because it was clear that floppy disks were going to be a critical part of the landscape) into jump vectors in a BIOS with all the specifics of how to do any given operation hidden under the covers, and completely opaque to the user-level software.
It was revolutionary for microprocessor based systems, and was the foundation on which the personal computing revolution launched forward. Now, to develop software, you didn’t need to know anything about the underlying hardware, you just needed to know that it ran CP/M - and you might need to specify some minimums like “must have 24K RAM and two floppies”.
So, the vast proliferation of hardware companies started to rationalize: some, such as Cromemcoand North Star Computers, tried to deliver “business ready” computers, that looked like office equipment and had higher price points. Others, not so much - they tried to focus on low cost, with user assembly as a way to further reduce costs. The mid-market pretty much evaporated - anyone trying to make a buck by selling something that someone else was selling much more cheaply died, ignominiously. There was no longer anything compelling about an Altair versus any other of the S-100 bus systems.
In 1977, things changed again. It began to occur to some of the more prescient people in the industry that consumers, in the main, didn’t want science projects - they wanted computers that were more appliance-like. So, in 1977, we saw three new market entrants: the TRS-80, the Commodore PET, and the Apple II.
These systems were very appliance like, they looked right at home sitting on a table in your living room, no weird wiring no build-it-yourself, just un-box, plug in, and shazam! you have a computer!
Not running CP/M, these computers had a problem: they needed software. The different manufacturers set out to create libraries of software, often paying developers to support their platform in addition to CP/M for the popular titles. And they grew, and developed, and turned into fairly respectable markets - the TRS-80 sold over 200k units, the Apple II, taking production another ten years or so beyond the TRS-80, sold over five million.
Note that in this time, Microsoft was just another software house. They had a key product, Microsoft BASIC , which very nearly everyone ran.
The thing that killed all of 8080/Z-80 computers, both CP/M and otherwise, was really the advent of the Intel 8086. It was significantly faster, addressed more memory, and was a 16-bit CPU, which was similar to that used in academic and industrial computing. It was clear that this was the wave of the future.
Once again, there was a seachange. While no one came out with a consumer product immediately, several companies, including Seattle Computer Products developed 8086 based boards for the S-100 bus, breathing life back into that market. Eventually, our old buddy Gary Kildall even dropped the CP/M-86 operating system, in anticipation of a boom in that space. Most people expected a repeat of the earlier 8080 days.
Things didn’t go quite as planned. SCP was losing sales since they had hardware ready to go well before Kildall’s OS was ready, so they wrote their own - QDOS (for Quick and Dirty Operating System). It was shipping in September 1980, and was renamed 86-DOS in December.
Microsoft, having been contacted by IBM about their not-yet-shipping IBM Personal Computer, had also been casting around for an OS for the 8086. Kildall wasn’t ready, and wasn’t super cooperative when approached by both IBM and Microsoft, and Microsoft bought the rights to 86-DOS just before the IBM launch in October 1981.
And the world as we know it changed, fundamentally.
The IBM PC, with it’s much more powerful 16 but 8088 processor and integrated graphics (a lesson learned from Apple and Commodore and the TRS-80) made it clear that to do serious business computing, you were going to need to step up to more memory, and a faster CPU, and that you just couldn’t get there on the older platforms. And the new platform? MS-DOS, the renamed and spiffed up version of SCP’s QDOS.
But that’s not where it got solidified. That took the work of Rod Canion and his band of merry men in Houston, the founders of Compaq. Just as had happened in the 8080 world, there was a proliferation of 808x hardware companies, but all of them faced the same issue: they might run CP/M-86, but that wasn’t the same as the IBM PC. And everyone feared that it wouldn’t be enough, that IBM would win the market based on, well, being IBM.
So Canion et al developed a new plan. They set out to make a computer that was 100% compatible with the IBM PC. All of the ROM BIOS calls were to be identical, all of the weird locations in memory that you PEEK’d or POKE’d would work the same, your software for the IBM PC would 100% run on the new machine from Compaq.
When they pulled it off - and no one really thought they could - it set the market standard. The new platform became “IBM PC” which meant MS-DOS, and then Windows, and then… wait, it’s still Windows.
Apple, in the mean time, is still selling Apple II’s but looking to the future. And the future they saw was based on the Xerox Dorado, and it was called Macintosh.
Apple had been paying attention during all of this, and knew that ecosystem was key. While the Mac came with MacWrite and MacDraw, they knew that it was “starter kit” stuff and they wanted to foster massive third party development. However, the third party developers knew that there was a chance that Apple would continue to develop and enhance the basic software, and there was indeed clamor from the users for just that. So, Apple spun out the internal software team into Claris, removing Apple as a competitive threat to the third party developers while at the same time providing software that delighted their existing customers.
Apple has continued to work this premise. They sell powerful high-margin hardware with some very basic programs, but count on heavy involvement from a deep and robust developer community to really fill user needs. They have a fully integrated ecosystem. There is a market (the AppStore for both Mac and i-Devices) for software, and users who buy from it know that the software has been vetted to some degree, isn’t malicious, and probably runs. There’s a review system from verified users only so it’s possible to know with some surety what it does. But to put your software in the store, you have be a registered Apple Developer - for which privilege you pay an annual fee. And to ship your software, you need to build it in an Apple developed programming language (Objective C or Swift), on an Apple supplied IDE (Xcode), on an Apple computing device (Mac).
It’s all full circle. Fully closed. And the focus is to deliver things that amaze and delight the customers, at a fair price.
Microsoft on the other hand is all about the software. In the world of Windows, the specifics of the hardware is rendered nearly meaningless by the various hardware and software abstraction layers. (Not that it doesn’t matter at all, such as for gaming or whatnot, but it’s less relevant). No one buys a modern X86 computer “because it runs Windows”’; that’s a ticket to play. More to the point, vast numbers of people at Microsoft work very hard to make sure that there are in point of fact no x86 computers which CAN’T run Windows.
All your third party applications, beholden to Windows. All your manufacturers, need to have Windows OEM deals. And on top of that, Microsoft is also a massive software publisher, delivering Office and it’s component parts, the key business software on which the modern world works. They don’t worry at all about alienating the third party developers; they assume the third party guys know better than to butt heads with them in their core competencies.
Both Apple and Microsoft are ecosystem companies. They got there by very different paths, and have very different ideologies as to why they’re there and how they make money from being there.
But they’re very much both about ecosystem, and being the platform.
This question originally appeared on Quora - the place to gain and share knowledge, empowering people to learn from others and better understand the world. You can follow Quora on Twitter, Facebook, and Google+. More questions: