This thread has gone on and on in the usual way, with posters talking at cross
purposes.
I do not wish to trivialize Peter Relson's comment, but one interpretation of
what he said that seems to me to be at once reasonable, helpful, and consistent
with his words is just that
1) an amode(24) program that is unaware that amode(31) and amode(64) exist will
not itself be impacted by the existence of amode(31) and amode(64), and
2) an amode(31) program that is unaware that amode(64) exists will not itself
be impacted by the existence of amode(64).
THis formulation, which is also consistent with the rules of English syntax, is
notable for what it does not say or imply. It does not
1) say or imply that an amode(31) routine that invokes an amode(24) routine
will not be impacted adversely by that amode(24) routine's ignorance of its
requirements, or
2) say or imply that an amode(64) routine that invokes an amode(31) routine
will not be impacted adversely by that amode(31) routine's ignorance of its
requirements.
An invoking routine higher in the sequence
amode(64) > amode(31) > amode(24)
than a routine it invokes must take its own environment-preserving precautions.
It cannot expect a routine that is a fortiori ignorant of its more stringent
requirements to take these precautions for it.
To repeat myself now, since it would appear to be necessary, Peter Relson's
formulation is notable for its asymmetry. It does not, for example, say that
an amode(64) routine cannot be adversely impacted by the blissfully ignorant
functioning of an amode(31) or amode(24) routine that it invokes. It says only
that that this blissful ignorance will not be intruded upon in its private
garden.
John Gilmore Ashland, MA 01721-1817 USA
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
