On 12/02/2019 12.56, Liam Proven wrote:
On 2/12/19 11:32 AM, Carlos E. R. wrote:
I used one laptop with about 1 MiB, back in the day. It might be more, it might be less, too long ago to actually remember (~1995). It was MsDOS, possibly Win 3. The only laptop on the company.
I have seen 1 MB laptops, sure. Sony did a series of 286 and 386 aluminium-cased tanks that were very popular among people who did adventurous stuff like round-the-world yachting.
But not in 1995, no. These were late-1980s/early-1990s machines.
Here's a 1989 advert for 80286-based laptops:
https://books.google.cz/books?id=B43yj9NtswMC&pg=PA202#v=onepage&q&f=false
By 1995, a normal mid-range PC was a 486 and a cheap laptop was a 386SX with 4MB of RAM. That is _why_ Win95 targeted a 4MB 386: it was a basic spec in 1995.
Yes, my own machine then was a 386-SX, possibly with 5 MiB. I still have it in this room. But I have doubts about the laptop: my boss didn't have much money to spare.
Totally agreed re Win3, though. It would just barely crawl along in Real Mode in 640 kB of RAM, run OK in Standard Mode in 2 MB of RAM on a 286, and 386 Enhanced Mode needed a 386 (obviously) and 4 MB of RAM.
For the ultimate experience, Windows for Workgroups 3.11 on a 4 MB 386 gave you 32-bit Disk Access (also in ordinary Windows 3.1) but also 32-bit *file* access... via a driver called VFAT.VXD, which is why the Linux FAT driver is called VFAT.
AHHHH! I did not know that tidbit. Yes, I do remember about the 32-bit access, and the fight to activate it on machines that wouldn't. For those that were too young to be there, activating "32-bit disk access" also meant bypassing the BIOS native code to access the disk, using instead Windows own code, thus not having to switch CPU mode for every block read or write. The speed improvement was a big difference. -- Cheers / Saludos, Carlos E. R. (from 15.0 x86_64 at Telcontar)