When is a PC not a PC? The PC-98

https://youtu.be/EJ_k_QdlNYE

The PC-98 re-entered my life while I was working for Be. NEC had us develop a version of BeOS for what was possibly the last of the line of PC-98s. Once again, I thought the industrial design was pretty decent. The flat-panel LCD monitor matched the style of the case, mouse and keyboard. The design was reminiscent of a SparcStation or maybe a Radius Macintosh clone.

There were lots of these back in the day -- systems with x86 processors that attempted to fill the PC's niche and even ran DOS but were subtly incompatible. In the early 80s the IBM PC was not the clear winner yet, and the BIOS hadn't been reverse-engineered, so there were attempts to come up with similar, but not quite compatible, designs to compete in the same market. By the release of the PC/AT, the IBM architecture was pretty much unstoppable in its dominance.

The Japanese PC-incompatibles, like the NEC PC-88 and PC-98 series, hung on for a bit longer for a couple of reasons:

1) as explained in the article, displaying Japanese hiragana, katakana, and kanji had high resolution requirements that none of the standard IBM graphics adapters could really match until about the time VGA came out, and there was poor support in MS-DOS until the late 80s or so

2) the Japanese liked their special, idiosyncratic, very Japanese computer designs.

To the latter, there was an episode of Pretty Samy (a Tenchi Muyo spin off and magical girl parody) in which the villain was Biff Standard, a glasses-wearing software magnate who wanted to standardize all the world's software. Obviously he was supposed to be Bill Gates, and there was a rising tide of anti-Microsoft sentiment at the time in the West as well... but the episode came out shortly after the release of Windows 95, the first consumer version of Windows to natively support Unicode and international encodings without a special build. After Windows 95 came out in Japan, there was much less of a need for products like the PC-98, and Japanese software could standardize on a plain PC architecture running Windows. Computer enthusiasts at the time felt the malaise of the closing of an era.

Even to this day, when you go PC shopping in Japan, you will find there is a hierarchy of preference among brands. Japanese retailers will promote the Japanese brands (e.g., Sony (now Vaio), Fujitsu, Panasonic, NEC) most prominently, followed by the non-Japanese Asian brands (Lenovo, Samsung) and finally the Western brands (HP, Dell). Apple gets a special pass because it's pretty much the Louis Vuitton of electronics.

After much frustration we discovered that we had PC-DOS but needed MS-DOS (or vice versa).

As far as I know no one got much further than installing the OS; the system seemed dramatically inferior at the time.

One other (even bigger) way in which Japanese computer technology influenced Australia, was that Fujitsu had a lot of success selling their clone IBM mainframes here in the 1980s/1990s. The university I went to – and later became staff member at – had used one. By the time I joined the staff, it had already been thrown out, but I remember discovering all this evidence of its past existence left behind – masses of tangled bus and tag cables under the data centre floor tiles, a storeroom full of 9 track backup tapes, etc.

Fujitsu's mainframes mostly didn't run MVS (although they could), they ran OSIV/MSP, Fujitsu's "clone" of MVS with significantly reduced licensing costs. However, the reason why it was cheaper, is it wasn't actually a clone – Fujitsu had illegally stolen the source code to MVS from IBM. They tried to cover it up by doing things like changing the copyright notices (apparently they forgot a few though), and they also renamed a lot of the OS modules and commands – the modules of MVS' timesharing component, TSO, mostly start with the prefix IKJ, Fujitsu renamed them all to start with PDE instead; they renamed the IBM catalog management utility, IDCAMS, to KQCAMS; etc–which meant most MVS software could work, but only after a bunch of search-and-replace changes to its code. IBM saw right past this feeble cover-up attempt, and Fujitsu (and Hitachi too, which did the same independently) ended up paying IBM many hundreds of millions of dollars to settle the resulting lawsuits. That was likely one of the factors resulting in Fujitsu pulling out of the mainframe market in Australia, and most of its customers either went to non-mainframe platforms (UNIX/Windows/Linux), or else to IBM's mainframe operating systems (MVS / OS/390 / z/OS). Fujitsu's mainframe OS and hardware still survive in Japan, but are stuck on 31-bit – they never made the investment which IBM did to add 64-bit support (with the release of z/OS in 2001).

http://internat.freebsd.org/platforms/pc98/

On PC-98 your boot drive was A: followed by your other drives of the same type and then your other drives of different types.

Let's say you had 2 floppies and 1 hard drive, and booted from the hard drive. The hard drive would be A: and the floppies would be B: and C:. If you booted from floppy the boot floppy drive would be A: and the other floppy B: and the hard drive would be C:.

A lot of DOS software that needed to find a hard drive would start its search at C:.

So looks like, NEC had agreements with IBM and/or other pc brands, and have access to all need information, and invested much into professional products (create chipset was expensive), but for some reason, play games with Nintendo and other local vendors.

Anyone else remember DOS/V Magazine? :)

I recall that one the zip compression utilities could run on both PC-98 and on IBM compatibles.

https://en.wikipedia.org/wiki/PC_System_Design_Guide

This includes the Amiga, Mac, most game consoles, advanced calculators, etc. Do you disagree? Too sad.