The pivotal importance of IBM's PS/2 range on computers, even today • Hacker News Board

Re: The pivotal importance of IBM's PS/2 range on computers, even today

#81

Post by lproven » Sat, Oct 01, 2022, 8:48 PM UTC

Fatnino wrote: post #73

lproven wrote: post #39

Earlier quoted context omitted.

Depends which keyboard I use, but sometimes, same here. My main work PC has a 1991 Model M on it. Best PC keyboard ever made, and I have about 5 of them, carefully kept since people were throwing them away in the 1990s. I remap Caps Lock to Super, and I'm good to go. I really don't need media-control keys or anything, or a right Super key. I am trialling an early model of Das Keyboard on my Mac. The volume-control di…

On windows ctrl+esc is the same as the super key.

No, it's really not.

I mean, yes, Ctrl+Esc opens the Start menu, which the Windows key also opens.

But the Windows key -- before MS renamed it, the Super key -- is a single modifier key in its own right.

Win+R = Run

Win+D = hide all windows and show Desktop

Win+E = open Explorer

Win+SysRq = open System Properties.

There are dozens of them; these are just a few of the ones I personally use.

Other OSes and desktops use it for other things. In GNOME it opens the Overview. In Unity it opens the search function, and Super+A opens the app search function, and Super+S opens the virtual-desktop switcher, etc.

It has lots and lots of uses, and opening the Start menu is just one. Ctrl+Esc doesn't do any of these other functions, and it cannot do them because it's not a modifier key so you can't use it plus another key at the same time.

People often say that they can't use an old, pre-Win95 keyboard because they need the Windows key. What I am pointing out is that it's easy to work around that, by sacrificing another key that's not very useful: Caps Lock.

This also sits in the same location as Ctrl on the old Sun keyboards, which is a convenient position for a modifier key -- it's easy to press CapsLock with your little finger while using the rest of the hand to press the other key of the combination.

Re: The pivotal importance of IBM's PS/2 range on computers, even today

#82

Post by lproven » Sat, Oct 01, 2022, 8:55 PM UTC

cardiffspaceman wrote: post #77

lproven wrote: post #45

Earlier quoted context omitted.

I wrote the comment and the blog post. What wording are you referring to? Yes, Windows supported the 286 too. There was a special Windows 2 version for 286s, and a different edition for 386s; what was novel in Windows 3.0 is that one edition and one install on disk supported all three: 8086, 80286, and 80386, all in one. Windows 3.0 had 3 modes: - if it had - if it had 2MB of RAM and a 286, it started in Standard Mod…

I read about the multitasking of DOS programs in some round-up of DOS multitasking systems (DeskView et al.), and convinced my boss to fund experimentally adding 4 MiB of RAM and a Windows 3.0 license to my 386-based clone. I got a build (cross-assemble) of my code going, and then entered the Brief editor, and pondered what exactly I had enabled. Then an in-house utility was invoked from the makefile to download my c…

Anything that hit the metal directly could cause problems, yes. You had to be pretty careful and test stuff first.

TBH I didn't use these multitasking tools much. I was a support guy then; I worked on other people's PCs so I needed to know how to use the built-in tools to best advantage. I could not rely on additional apps, utilities, etc. because the PC I was fixing almost certainly wouldn't have them.

Late in that era I carried around a wallet with about 20 floppies full of freeware tools and utilities for this reason.

But for me it was occasionally very handy to be able to go and make a directory, or format a diskette or something, while in the middle of something in an app.

But I wouldn't try something that wrote to the hardware, do file transfers or something. Serial ports if you were very careful to set them up so they didn't share IRQs, which of course on the PC was the normal standard config -- COM1: shared IRQ 4 with COM3:, and COM2: shared IRQ 3 with COM4:

While LPT1: was on IRQ 7 and why on Earth can I still remember this 35 years later? >_<

Re: The pivotal importance of IBM's PS/2 range on computers, even today

#83

Post by hilbert42 » Sat, Oct 01, 2022, 10:31 PM UTC

tengwar2 wrote: post #78

hilbert42 wrote: post #72

Earlier quoted context omitted.

Excellent concise summary, glad you and ForOldHack beat me to it. I would have confined my reply to the ring issue having forgotten much of the controversy with other OSes—Netware for instance (even though I'd used it many times). You're right about segmentation being hard work, I can recall cursing Intel at the time and hoping it would bring out a more eloquent processor that was easier to use such as Moto's 68000.…

I find it difficult to think of a context in which Gates would say that 640kB was enough, given that the limitation came from the IBM PC. MS/DOS could quite happily address 768kB on an RM Nimbus, for instance. The more fundamental 1MB limit came from Intel. Gates didn't have to justify either.

There's no doubt that that quote was the talk of the IT world back then. It did the rounds in magazines and news outlets for ages and I cannot recall any of those articles doubting that Gates actually said it.

It was also wheeled out later in computer mags often as mocking comment when the '386 came out as by then PCs had 4MB RAM.

Doubt about the quote's origins only arose about a decade later after Gates denied saying it. I'm now racking my brain trying to remember which magazine cemented the idea that Gates was responsible. I already knew about the quote but this was a big splash and it had a photo of Gates and cohorts on a podium (product launch) at or near the top of the article. Could have been PC Magazine or possibly InfoWorld but I really can't remember.

I'd not really given the matter much thought until now but it does seem strange that doubt still persists (or that it does at all) given the widespread attention news reports paid to it at the time.

Re: The pivotal importance of IBM's PS/2 range on computers, even today

#84

Post by ForOldHack » Sun, Oct 09, 2022, 6:07 PM UTC

lproven wrote: post #48

ForOldHack wrote: post #36

Earlier quoted context omitted.

This is a bit of history, I had to search a while for: "The designers of the 80286 based it on the Multics segmentation model, but in reality segmentation makes it really hard to implement, because of internal memory fragmentation. A while back I wrote blog post about how an alternative 80286 with segment registers pointing to 256 byte paragraphs giving a true 24-bit virtual address which was then paged into a 24-bit…

Segmentation is harder to work with, but it has benefits too. At the time, when it was all new, everyone just complained about it: they wanted the ease of use of flat, unsegmented memory, with just one memory address. More recently, the NX bit has become desirable and promoted. The CPU can flag areas of RAM as data, and mark them as N on e X ecutable. That is all: one bit, so either it is code and you can run it, or…

Excellent. This is either extremely well researched, or you where there, like I was. Thanks.

Re: The pivotal importance of IBM's PS/2 range on computers, even today

#85

Post by lproven » Mon, Oct 10, 2022, 12:35 PM UTC

ForOldHack wrote: post #84

lproven wrote: post #48

Earlier quoted context omitted.

Segmentation is harder to work with, but it has benefits too. At the time, when it was all new, everyone just complained about it: they wanted the ease of use of flat, unsegmented memory, with just one memory address. More recently, the NX bit has become desirable and promoted. The CPU can flag areas of RAM as data, and mark them as N on e X ecutable. That is all: one bit, so either it is code and you can run it, or…

Excellent. This is either extremely well researched, or you where there, like I was. Thanks.

Thank you!

Yes, I was there. Got into the biz in 1988, but was a hobbyist before then.