<hanco...@bbs.cpcn.com> wrote:
> On Aug 27, 9:28 pm, glen herrmannsfeldt <g...@ugcs.caltech.edu> wrote:

>> But the architecture includes the privileged instructions.

> That's a tricky question.  The vast, vast majority of users of the S/
> 360 series do not use privileged instructions, nor of course touch the
> microcode.

> Do utility programs like IDCAMS or SYNCSORT (an external purchased
> package) use priviledged instructions?

> If we count priviledged instructions, should we then start worrying
> about hadware design?  Undoubedtly Z series is completely different
> physically than a S/360 was, with all sorts of "pipes" and look-aheads
> previously not available on the low-end machines.

> But that's all transparent to the application programmer, even one
> programming in Assembler, isn't it?  If someone has a 25 year old
> program, it still runs, and that's what counts.

>>> Rather, I would say the 'children' have extensions to the original
>>> architecture.

>> For problem state, I agree. Even so, it seems that enough features
>> were added to S/370 that it deserves to be a new architecture,
>> though I couldn't say how many new features you need, or how
>> significant those features should be.

>> Changing the address space, to 31 and then 64 bits, is definitely
>> a big change. First order, I follow IBM when they give it a new name.
>> It might be that XA/370 and ESA/370 aren't different enough.

>> -- glen