On 2/19/2012 10:12 PM, Fred Cisin wrote:
I am amazed that an XT/370 would be called a
"mainframe".
:-) The XT/370 is not the last word in such units
(
http://en.wikipedia.org/wiki/PC-based_IBM-compatible_mainframes).
And, . . .
would all machines that were once called "mainframes" that do NOT tun 370
code, now NO LONGER be mainframes?
I've heard them referred to as "legacy
environments", which is ambiguous
since people call the 370 architecture "legacy" as well. The only terms
I've heard folks use are the manufacturer or model: I've folks talk
about the "Burroughs", and the "UNIVAC System".
I won't argue that a complete definition needs to include these other
machines, but the term has been largely "redefined" to be 370-based
systems at this point. I will still stand by my statement that the term
refers to 370-compatible systems, but I did find a few notes
(
http://www.eweek.com/c/a/IT-Infrastructure/Unisys-Unveils-New-ClearPath-Mai…)
referring to Unisys' machines as mainframes. I'll concede the point,
but I think one would be hard pressed to find many in IT that even know
about Unisys' big-iron solutions. Most people do know them in relation
to their Tandem machines.
Maybe, the best contemporary definition is "a current computing system
that can natively execute code written in a machine architecture
commercially available before 1970" (or some arbitrary date that covers
the various machines one typically identifies with a mainframe
moniker). Note that this would preclude such Hitachi systems that
actually run on Xeon CPUs and emulate the z Architecture, but I'm sure
some wordsmithing could fix that while keeping things like Hercules
emulator from falling into the definition's space.