On Apr 19, 2006, at 11:30 PM, Tom Schmidt wrote:
--------------------------SNIP--------------------------------------
Ed,
Chris' very recent post reminded me of the other issue(s). There's a
fundamental difference between the started task use of static
symbols and
the batch job use of static symbols: STCs bring their own
initiators so
there is usually very little delay between the reference and use,
whereas
the batch jobs could have between a few moments to several years
between
reference and use. (The longest that I remember seeing a batch job
sitting
in an execution queue on a JES2 system was between 1 and 2 years; I
got
bored and stopped looking after a long while.) With a multi-year
lag to
execution there are bound to be unforseen issues.
With the ability of resolving (FSVO 'resolving') symbols via SMF
JCL exits
and other vendor mechanisms is there a real need for IBM to spend
time on
this "problem"?
Tom,
There are more and more installations that are loosing their
Technical support (witness as an example the JES2 exit question that
came up here several weeks ago).
There is less and less people doing such customization as a result.
I believe its in IBMs interest to come up with a(n) built-in solution
(s) as we witness on here a constant need for such a facility. I
think the most needed one (this is a guess on my part) is the ability
for using a date/time for dsn's. Yes there are pseudo solutions out
there but only because someone got tired of waiting for IBM to come
up with a basic solution.
The GDG issue is another (calling the GDG's in order of creation
(0,1,2,3,4 etc) (which is not address by the symbolic issue). The
need has been there for 25+ years I remember needing it in the early
70's and writing a program to do so so I know it can be done.
IBM doesn't (usually) want to hear about needs (I think the whole
GUIDE/SHARE requirements process is a laugh to IBM). It would be
interesting to do a study to see how many rejected requirements have
been really implemented 10 years after). I have seen a simple
requirement that was rejected by IBM a few times only to be satisfied
by an IBM product 4 or 5 years later. When asked IBM (at GUIDE)
pleaded ignorance. We (at GUIDE) liked to pester IBM about
requirements a lot more that the people at SHARE seem to (this is
only a guess as I never got involved at SHARE (due to previous
reasons stated on here ).
I am willing to let IBM slide a little bit on the symbolic issue but
the PARMLIB option (to me anyway) does away with IBM's objections.
Ed
----------------------------------------------------------------------
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
