DG> Replying to a message of Glen McNabb to Darryl Gregorash:
DG>  GM> Actually, I was quite impressed with the 80186/88
DG>  GM> processors. I really like the pusha and popa instructions
DG> Nice little toy, but I preferred the NEC V20/30 myself.. too bad Intel 
DG> never made their bit manipulations opcode compatible with the NEC chips.
There were several nice things about the V?? chips and several not
so good things. One of the nicer and little known things about the
V series chips was that the earlier ones contained a Z80 micro
inside as well. With a special opcode you could switch processors.
One of the less nicer things was some of the convertion opcodes.
Givin that one of the undocumented parts of the 80XX set could
convert bytes to nibbles but could convert it to smaller values
because the the "pre-byte" in the opcode. The V?? series chip
would ignore this prebyte. Which is one of the reasons it would
operate faster. This difference can be used to detect if that
peticular IC was used in the computer. I don't know if it still
does this today. I don't use a computer that has these IC's
anymore.
DG>  GM> chip for that purpose. The last and best XT I ever had was
DG>  GM> a Hedaka 12mhz 286 XT. It came with all the software to
DG> I'm not familiar with that particular machine, but I do know of a couple 
DG> more XTs that used 286 CPUs. Someone like Acer or some such little clone 
DG> company made one of them, and that started that particular bunch of dudes 
DG> on the road to what they are today.. it was definitely a rinky-dink 
DG> little company that is a big name today.
I did warranty service for Acer once. Most were pretty nice machines.
The thing I liked with the 286 XT was that it would use a 16bit memory
buss while it was happy with my 8bit cards. It could set cycle times
on the IO bus, use the 384k from 1meg IC's for cache, ramdrive, and
printer buffer. Most of these things were not available at the time
this motherboard came out. It could also relocate bios's into ram
memory so that they could run at 16bit and at buss speeds. That was one
of it's larger advantages. However, it didn't like it when I changed
the RLL card to IDE. It choked on that. I would tell AT only programs
that it was an AT while not being one. At the time, I couldn't afford
to replace my entire computer. It was a nice intermediate step between
XT and AT. Ah..... those were the days...
DG>  GM> take advantage of it's chipset. It could cache the hard
DG>  GM> drive and xms without loosing it through a reboot. I would
DG>  GM> also outrun a 286 AT at the same speed.
DG> Well, that particular one sounds like it had a cache that compares 
DG> (relative to the CPU of course) with the caches available today.. any 
DG> kind of cache would blow the sox off an AT that had no caching at all :)
It could. It's advange was in that single chip buss controller. It could
remap the upper memory so that it didn't have to switch back and forth
from protected to real modes. It was the one thing that a 286 was really
slow at. It was faster, in some cases, to swap on the drive than on the
memory...
It's biggest disadvantage was that it didn't have EMS memory. Which, at
the time, was somewhat new and many programs were not so memory hungry
as they are now. I was quite happy with running the software that I was
using. This was during the time when programmers/public were unhappy
with being forced into replacing the computers with ones that did.
You know, when programs didn't require VGA, sound cards, 386sx or
better, etc...etc....
You'd be hard pressed to find anything that don't require Pentium,
1meg+ VGA, speedy CDrom, sound cards, and Win95 nowadays on the
consumer market. I find it odd that educational software requires
this when parents are the least likely to afford these computer
upgrades. It's kinda sad....
Glen..
--- ProBoard v2.16 [Reg]
---------------