Primary and secondary cell counts are not provided on IARCP64 simply 
because the value of providing such a thing is far less (if there was any 
at all) than the cost of providing it. It makes sense for CPOOL because 
GETMAIN (STORAGE OBTAIN) is in 8-byte increments. But storage above-2G is 
in 1M increments even if you only want 8 bytes. Once you have to get 1M 
anyway, do you really care about primary vs secondary? There is danger in 
presuming that "old needs" apply to "new situations".

Whether IBM's documentation is deficient our not, the fact is that the 
official documentation -- namely things that you can rely upon and we will 
take an APAR to correct (whether that be a correction in code or a 
correction in the documentation) -- is what is in the books, not what 
might or might not be in a macro prolog.

It is not infrequent that additional data that is in a macro prolog and 
not in a book is not part of the intended programming interface.

And it is all to frequent that useful and necessary data is in neither a 
macro prolog nor a book.

The thing I happen to prefer about macro prologs is that the syntax 
diagrams of many of the newer macros are hierarchical and (to me) easier 
to understand than the "brackets and braces" style of most of the IBM 
books, when it comes to what is required with what. New major functional 
areas may have their own books and use railroad-track syntax diagrams 
which, while cumbersome, at least do let you see the complete syntax too. 
It is quite unlikely that the large amount of resources to change from 
brackets and braces to either other style will be allotted.

FWIW, IARCP64 might not have been a good one to complain about. Before 
complaining about IARCP64, did anyone actually compare the book to the 
macro prolog before making sweeping generalizations? I ask this because 
this macro happens to be one for which the book chapter was, to a large 
extent, created from the same source that produced the macro prolog. The 
book chapter for this macro might even be more correct than the macro 
prolog. Still, there's always the possibility of errors, and always plenty 
of opportunity for improving documentation.  Taking IARCP64 as an example, 
it might be worth having a discussion of things that you the users prefer 
in the macro prolog "style" to the book chapter "style".  And I'm also 
curious about omissions you might find. For example, it is fully 
intentional that the modify form is not part of the book chapter - if 
anything, the macro is in error in not preventing MF=M from being 
specified. Feel free to send me notes offline (or, if you think the 
discussion will be helpful to the group, post there).

Peter Relson
z/OS Core Technology Design

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to