GoKawiilThe Creation of Windows 2.0
Despite what you may imagine from how history went afterward, Microsoft was fully aware of how MS-DOS was starting to be inadequate for current and future computing, especially if GUIs were truly going to become the only way to go from that moment onward; poor Windows 1.x could barely run on the most common machines of the time (which were running on Intel 8086 and 8088, and not on the newest 80286 and 80386), and it’s not like Windows 2.x could potentially do much better. Thus, Steve Ballmer, negotiated an agreement with IBM on a successor to Windows, a true operating system; ideally, the interfaces of both programs were going to be as similar as possible, so that, one day, Windows 2.x users could switch to OS/2 and encounter a familiar environment, albeit with many new functions and additions. In the meantime, though, they needed to develop another version of Windows, so that they could also fix all the issues caused by the rushed Windows 1.x release [ep 10]; plus, it’s not like OS/2 would be ready right away, as it was much more complex than Windows 2.x, so they needed some kind of new release to appeal to the existing market, and slowly accustom people to the wonders of the new UI.
Remember Tandy Trower, the manager who succeeded in shipping Windows 1.x after everyone else had failed [ep 10]? Well, he was back, and he was tasked with working on this same project, so that Microsoft’s Application Group could work on a new version of Word and Excel, which were becoming very popular on the Macintosh; the problem was that most of the original Windows development team had been tasked with working on OS/2 instead, and even Trower himself had to work alongside that same team in order to maintain consistency between the two UIs. Something that we can really appreciate is how Trower suggested that they created a separate team who also employed real graphics designers, rather than simply software developers, to craft the new UI, hoping to achieve a truly user friendly and consistent interface, rather than whatever the developers found easiest to make; this is in line with the development of the Lisa [ep 3] and the Macintosh [ep 7], where UI was above every other aspect of the software. Plus, if you remember [ep 7], even the Macintosh’s icons were designed by a graphics designer, Susan Kare, who then designed the one found in Windows 3.x. This team worked so well that they ended up influencing the usability of many Microsoft products, and even the full overhaul in Windows 95!
Back to our poor Windows 2.x, even if Trower only had eight months to get it all done, he still managed to do it all; he finally got the chance to add all of the features that he had to skip during the development of 1.x, such as overlapping windows, or a proportional system font. Then, he tried to add in as many UI upgrades, whilst never forgetting to be compliant with OS/2’s UI; some of these things were only possible because of a clever trick that allowed you to use extended memory (part of RAM) in PCs, and not just the base amount that the processor permitted you to use. Fundamentally, Intel processors found on IBM PC compatibles only allowed you to use 640 KiB (kibibytes, not kilobytes) of conventional memory for applications, which meant that anything beyond that amount was reserved to other system functions and was not available to the user; still, there were ways to circumvent that, and that’s how Windows 2.x and other programs such as Lotus 1-2-3 managed to have more RAM available for their needs! It was less difficult to achieve this on a 386, but most people could not afford it, so that’s why they couldn’t assume you’d have one; if you’re curious, here’s a video that shows you how this process worked in the 90s, thanks to some really awesome programs!
Even if not everyone had a 386, there were still some useful improvements that could be added to a version that required it, apart from fully preemptive multitasking; one of them involves the fact that the 386 was able to go into “protected mode”, which was more stable than the “real mode” found on other processors, but made you unable to run code not designed for it, such as DOS applications. After a lot of work, in 1987 not only one, but two flavors of Windows ended up coming out; Windows 286 and Windows 386, which were then renamed to Windows/286 and Windows/386 once Version 2.1 was released.
Beta Build
Windows 2.0 April 1987 build – 1987-04 This is the earliest build you can find, and it was shown in the Microsoft Systems Journal magazine in May 1987.
We can see that the UI has the same color scheme as Windows 1.x, but you can now overlap windows, and you have more ways to control them.
Things start not going so well anymore: a nice lawsuit from Apple
⚠️ A Little Disclaimer ⚠️ Before we start with this section: I’m not a lawyer nor a law student, so if there are any inaccuracies, you know why.
... continue reading