Prevent software from running on new computers or virtual machines.

That is a myth I have already mentioned. Routine spread sheets use much mo re time accessing the data and organizing the calculations than they do in the actual calculation itself. So unless you were using a bunch of complex math routines a coprocessor did not speed up the calculations noticeably.

Yes, it would help complex engineering math, but that is not by any means a mainstream application and illustrates my point. A few customers of the 4

Since you can't discuss anything in a civil manner, I'll leave you to your profanity and rude insults.

Bye.

Rick C wrote in news: snipped-for-privacy@googlegroups.com:

You have a thing for your 'myth' word.

NOW. Because now they are huge data arrays that replaced databases in some areas.

Back then they had a smaller set of data and there was a lot of math if math is what the engineer wanted done on the cells. And it was.

Lotus 123 'routine' was MATH on a smaller array for engineers.

Maybe for accountants it was shuffling data.

But you are silly. Spreadsheets were FPU enabled. From early on. And they used it.

Rick C wrote in news: snipped-for-privacy@googlegroups.com:

Sure it was mainstream. Right. For our engineering group it was all it was used for. The accounting group used the mainframe.

I suspect the same for other design groups.

You 'illustrate' nothing and made no point. I was there from the XT on.

You are like the guy who said that dual CPUs were of no benefit. Strange though that my Dual CPU unit was faster at both Windows and Linux, depsite a bunch of dopes claiming zero benefit, zero gain.

Now, it is difficult to buy a single 'core' machine or device. Microcontrollers are still single core. There are exceptions, of course.

There is likely still an 80186 line out there being made somewhere. That was the most used chip in the industry for a long time.

Never made it into a PC, but for a couple protos. The 286 hit before it did as far as PCs go.

Rick C wrote in news: snipped-for-privacy@googlegroups.com:

You have mouthy replies which invoke them. Nice try though.

A good example is that you seem to like jumping on the retarded rename bandwagon, and call me by names other than any I ever used.

Yeah... you are spotless, chump.

No, YOU are silly, you silly goose. Yeah, I've already said engineers and others like them who use more than adding up columns and the occasional com pound interest calculation will gain from a math coprocessor. But accounta nts!? Don't make me laugh... you silly goose.

As I've said more than once, the vast majority of users had no need for a m ath coprocessor and so the 486SX as a viable lower cost option. But you ar en't very good at learning from your mistakes... like your twin Whoey Louie .

Rick C wrote in news: snipped-for-privacy@googlegroups.com:

Yeah, dipshit. What do you think things like 'compound interest' and such get calculated by? 'Data entry girls'? You trapped yourself with that one.

Vast majority? The DX was out for quite a while before the SX was even offered. It was 2 years later, in fact.

And you having any clue what the vast majority wants is a laugh.

I had a friend in Denver and helped him set up his Novell server and

You have no clue what the industry was buying. The SX came out for lame, low end upncoming home users that wanted a savings, not the geek level enthusiast types like me been around since the XT.

After two years of fabbing with a higher than desired rate of bad chips right off the die forcing 100% test, they noticed the main failure was the FPU. I can see the engineer right now... "Hey! I know how we can CAPITALIZE on these failed units!"

Silly indeed.

DOSREF shareware manual for the BIOS is still around at least in parts. The most recent version is from about a decade after the introduction of the XT but essentially covers all the magic addresses of antiquity.

It used to reside on the Simtel archive. ISTR 3.3 was last version.

Most people used the 80188/6 for embedded PC.

What exactly is an "embedded PC"? PC stands for Personal Computer which is not an embedded computer at all. What are you trying to say?

Am Sun, 05 Jan 2020 09:48:29 -0800 schrieb Rick C:

A classical example would be a computer in PC/104 <

Cheers, Peter

They don't use the 80186. The 80186 could not be used for a PC because it had a fixed memory map that did not match the PC. It would have required a wholly new operating system than any running on the PC when the 80186 was a viable chip.

It is not the memory map, but the use of interrupt vector slots.

In the initial 80x86 architecture descriptions, Intel reserved the 32 first (0x0 - 0x1f) interrupt slots for hardware use. The initial 8086/8088 processors did not use all of them, so IBM chose to ignore the reservation and put the BIOS call vectors into the same area. This caused a collision with hardware use in 80168 and newer, and it makes it pretty difficult to have PC software on an 80186/80188.

PC compatible embedded in some other device (eg kiosk, ATM, video slot machine, etc.)

PC/104 possibly also qualifies as embedded pc.

Not an 80186.

Rick C wrote in news: snipped-for-privacy@googlegroups.com:

It was for industrial embedded machine control. Not a PC ever AFAIR. It gained huge decades long success that way too. Then other, less consumptive, more robust microcontrollers entered the market. ATOM and ARM to name a couple.

Rick C wrote in news: snipped-for-privacy@googlegroups.com:

As far as I know kiosks and such NEVER had an 80186 in them. It is

Am Sun, 05 Jan 2020 11:32:52 -0800 schrieb Rick C:

You're right -- I stand corrected.

Hey! You can't talk like that! Don't you know what group you are in???

Lol, not very often you find someone here rational enough to admit they made a mistake.

I'll give my take on it...

I'm thinking of something that might be incomputable.

Maybe square of 2 or the halting problem.

At least with square of 2 it can be computed to a certain precision.

Maybe emulators won't bother to compute up to a certain precision because maybe it would then become infeasible/too slow for practical usage ;)

Bye, Skybuck.

snipped-for-privacy@hotmail.com wrote in news: snipped-for-privacy@googlegroups.com:

Try NetHack. It was one of the first.

That game... IF you play it... will keep you in suspense long enough to keep you off Usenet, and then we will not have our newsgroup invaded by your inane "looky what I am doing" posts.