Secondly, yes, the interfaces were close enough to the
same that some
BIOSes would actually display an error message saying "mouse inserted
into keyboard port" or /vice versa/. Helpful but not as helpful as
just bloody working with it, whatever socket it was in, would have
been.
Actually, that doesn't imply the signals are simular, only that the
machine can detect what's plugged in to a given socekt.
At one time there were mice that could work on both the PS/2 standard a=
nd
on an RS32 serial port. Typically the mosue cable
ended in a 6 pin
mini-DIN conenctor, there was an adapter packed with it that had a DE9
socket on one sida and a 6 pin mini-DIN on the other. And in most (if n=
ot
all) cases, thei swas a simple adapter linking
pins on the 2 conencotrs=
,
it didn't contain any circuitry.
But the Rs232 interfce is not remotely close to the PS/2 mouse interfac=
e.
It even uses different voltages., The mouse was
designed to work out wh=
at
it was pluggedf into (possibly by detecting said
voltages) and outputte=
d
data accordinging.
Now, as it happens, the PS/2 keyboard and PS/2 mouse protocols are almo=
st
identical. The IBM PS/2 TechRef is not nearly as
good as the ones for
older systems, but it does have some useful information. And it shows
that hte bit-level protocol is the same. The IBM keyboard/mouse interfa=
ce
chip in the PS/2 could talk to both devices,
there were commands to
enaable nad disable them, etc. And it identified which device had sent
the data when you read the incoming data from it.
That is my point. They /are/ similar protocols, so that (eventually)
they were made interchangeable.
Yes, they are similar protocols, but you can't deduce that from the fact
that some machines allow the keyboard and mouse to be swapped round (work
on either socket). You cna only conclude it by reading the IBM TechRef
(or equivalent accurate documentation).
It's entirely possible that the protocols are totally differnt and the
machine detects what's conencted to each port in some way and then uses
the appropriate protocol to talk to it. While this is not the case for
PS/2 keybaords nad mice, I can think of machines that did do thsi.
Serial-stroke-PS/2 mice never were. You couldn't
use a Logitech
serial-to-PS/2 convertor on a Microsoft hybrid-port dual-protocol
mouse, for instance - I tried.
THat's not my point. My point is tha the Async serial (RS232) and PS/2
interfaces are totally different. But there exist mice such that if you
conenct some of the wires int the cable to a PS/2 port it works. If you
connect some of the wires to an RS232 port, it works too (it's not cerain
it's the same wires for both interfaes becasue the only 4 pins of the
mini-DIN socket on a PS/2 machine are wired up, there are 2 not used.
It's possible that these pins are wired up on the mouse, and that the
RS232 conenctor adapter makes use of them).
So while the protocols and even the voltage levels are totally different,
the same mouse can work on either without the user havign to fiddle with
anything. It automaticallt detects what it's plugged into and behaves
appropriately.
Incidentally, PS/2 mice are serial too (What I am saying here is that while
an RS232 poet is an example of a serial interface, it's not the only
one)
-tony