> I must add my now-standard disclaimer, which has evolved over the series of so-called "compatible" reviews I have amassed recently. I have yet to see a compatible that was truly 100% compatible.
And this is why the Mindset (and many other almost-but-not-quite-100% IBM compatible products like it) failed: because they didn't realize that 100% compatibility with the IBM PC actually was possible. Compaq had proven that all the way back in 1982, three years before this article was written, with the release of the Compaq Portable (https://en.wikipedia.org/wiki/Compaq_Portable).
Compaq had realized that the only truly proprietary component in the IBM PC was its BIOS; to speed development and keep costs down, IBM had chosen to use off-the-shelf components for everything else. So Compaq launched a clean-room reverse-engineering project to clone the IBM BIOS. That cost them a pretty penny, but once it was done they had the key to a truly IBM-compatible machine.
By 1984 -- again, a year before this article -- chipmaker Phoenix Technologies had also reverse-engineered the BIOS, and unlike Compaq they were willing to license their design out (see https://en.wikipedia.org/wiki/Phoenix_Technologies#Cloning_t...). So making a "100% compatible" PC clone became something anyone could do, just by cutting a check to Phoenix.
So for companies like Mindset, coming to market in 1985 with a kinda-sorta-IBM-compatible machine meant coming to market too late. With multiple companies already selling 100% compatible machines and lots more on the way, kinda-sorta compatibility was no longer good enough.
Computer manufacturers realized that 100% PC compatibility was possible. But before 1985, 100% PC compatibility was not the sine qua non it would be just a few short years later.
Tandy, in particular, knew they could clone the PC in 1983, and had it on the table as an option for their next big TRS-80 machine. They opted instead to go beyond the PC's capabilities with the x86-based, PC-incompatible Tandy 2000. But they were Tandy -- one of the big players in the USA home computer market, and in their hubris they thought they had the clout to change the standard. IBM was a latecomer to the market where Tandy was already dominant.
Well, you know how that story goes. The Tandy 2000 was favored by a few power users but flopped in the general market. Tandy dusted off its plans to build a PC clone and released it the very next year as the Tandy 1000.
Note that both the Tandy 2000 and the Mindset were both based on the 80186 CPU. It was impossible to build a 100% IBM-compatible 80186 machine because the 80186 was a system-on-chip with timers, a memory controller, and such that were mapped to I/O port addresses that conflicted with the standard PC port addresses.
Also noteworthy on Compaq compatibility with PC... Compaq came out with an 80386 PC-compatible before IBM did. As did at least one other company around the same time, IIRC (ALR?).
I don't know the reason IBM was beat to the punch on 386, and maybe HN knows. My first guess is that they were working on all the changes for launching the PS/2 series (e.g., different bus). There was also IBM's PC/RT (before the RS/6000), and I guess maybe they thought they'd move away from x86 and all the clones.
IBM had made big commitments to enterprise customers to get them to adopt the PC's successor, the PC/AT (https://en.wikipedia.org/wiki/IBM_Personal_Computer/AT), which used the older, transitional 80286 processor. These commitments created a big drag on lots of other IBM projects, as they couldn't do anything that would be perceived by those enterprise customers as devaluing their investment in the AT. So not only was IBM slow to adopt the vastly superior 386, it also insisted on things like OS/2 having to be compatible with the 286, where Microsoft was free to let their competing product, Windows NT, abandon the 286 and all its baggage completely.
This is a great illustration of the idea of the "strategy tax" (see http://scripting.com/davenet/2001/04/30/strategyTax.html), where various parts of a business that are deemed non-strategic have to give up things that would make them more competitive to prop up another part of the business that is deemed strategic. IBM's bet on the 286 was so big that they were still pushing 286 machines well into the PS/2 era, by which time the clone makers had long since flocked to to the much more capable 386. Those commitments to enterprise customers meant that the 286 became an anchor tied around the waist of IBM's PC business.
Thank you, the enterprise commitments explanation makes a lot of sense. Did that enterprise bet make business sense at the time, or were decisions biased by the traditional businesses and leadership (despite IBM having already shaken up industry with the PC)?
Ah, the Mindset -- basically an x86 Amiga released a full year before the actual Amiga -- about which I'd heard some but have never seen (and YouTube has but a single video showing its capabilities). Strange how these old obscure computers get lost to time.
And this is why the Mindset (and many other almost-but-not-quite-100% IBM compatible products like it) failed: because they didn't realize that 100% compatibility with the IBM PC actually was possible. Compaq had proven that all the way back in 1982, three years before this article was written, with the release of the Compaq Portable (https://en.wikipedia.org/wiki/Compaq_Portable).
Compaq had realized that the only truly proprietary component in the IBM PC was its BIOS; to speed development and keep costs down, IBM had chosen to use off-the-shelf components for everything else. So Compaq launched a clean-room reverse-engineering project to clone the IBM BIOS. That cost them a pretty penny, but once it was done they had the key to a truly IBM-compatible machine.
By 1984 -- again, a year before this article -- chipmaker Phoenix Technologies had also reverse-engineered the BIOS, and unlike Compaq they were willing to license their design out (see https://en.wikipedia.org/wiki/Phoenix_Technologies#Cloning_t...). So making a "100% compatible" PC clone became something anyone could do, just by cutting a check to Phoenix.
So for companies like Mindset, coming to market in 1985 with a kinda-sorta-IBM-compatible machine meant coming to market too late. With multiple companies already selling 100% compatible machines and lots more on the way, kinda-sorta compatibility was no longer good enough.