I believe that the PoOP should or could stay like it is, because the normal use
is not as a tutorial for learning ASSEMBLER or machine architecture, but to make
clear some difficulties or peculiarities of seldom used instructions, at least
in my case.
For learning ASSEMBLER, I would use the following approach (and I do ASSEMBLER
training and tried this approach out some times in the past few years):
first start with the "classical" 360 instructions and teach how to use them and
then tell the people about the different hardware extensions (XA etc.) and how
these extensions left their traces in the instruction set. It is IMHO
sufficient, if the student gets an idea, how the large or Y instructions are
constructed out of the classical ones. To look up the PoOP and see, what
variants of instructions exist for LH - for example - , can be left as an
exercise for the student.
We don't need lots of people who are capable of programming in ASSEMBLER
language, so we can rely IMHO on those who are willing to do some more work by
themselves to learn the language. So we don't need a student's text book which
covers all the new instructions. Instead, like I suggested above, a book with
two parts would be sufficient:
- the classical instructions
- the extensions of the architecture, and examples for the new instructions, as
provided by the SHARE presentations on these topics
Am 23.08.2011 13:34, schrieb D.C.:
> Here is a typical PoOP paragaph:
>> For LOAD HALFWORD (LH, LHR, and LHY), LOAD HALFWORD IMMEDIATE (LHI),
>> and LOAD HALFWORD RELATIVE LONG (LHRL), the first operand is treated
>> as a 32-bit signed binary integer. For LOAD HALFWORD (LGH and LGHR),
>> LOAD HALFWORD IMMEDIATE (LGHI), and LOAD HALFWORD RELATIVE LONG
>> (LGHRL), the first operand is treated as a 64-bit signed binary integer.
> I think that the PoOP could be shortened by at least 25% if they
> would just get rid of (and by "rid of", I only intend "reduce the use
> of") such arduous nomenclature...