This session is taken up to discuss new JES2 requirements that are voted on later in the week. When you submit a new requirement for the JES2 project to consider, it must be discussed here first.
IBM reported their formal responses to requirements turned in last SHARE. These were:
| Requirement | IBM Response |
| 88.254 Translate Table Support for Local Printers | Long range consideration |
| 88.255 Optionally Suppress $HASP190 for SNA Workstations (mine!) | Long range consideration |
| 88.251 PSF SMF Exit #5 Needs Extra Information | Long range consideration |
I submitted a new requirement here as well. Whenever you send a printout to our little laser printer (Xerox 4045), the operator has to issue a $S command at the console. I have been unable to automate this due to a bug in JES2, and I have also been unable to get IBM to admit the bug through normal channels. So I wrote up the requirement, presented it here and Jim Walker from TUCC stood up and said: "This sounds like a bug". I could have kissed him. John Hutchinson (IBM) caught up with me afterward and asked me for additional information concerning the bug. I gave him all my correspondence with IBM concerning it (I had come prepared). Hope this works.
Another requirement was withdrawn when it became clear that it described a feature announced in 3.1.1.
A lengthy letter from a systems programmer at Bar Ilan University in Israel was disscussed and a requirement drafted from it. Briefly, when an application has a LOT of SYSOUT datasets, JES2 allocates a very large number of "protected buffers" which eat up the private area. Eventually the job can go down with a recursive S40D ABEND. The requirement we wrote requested that a method be provided so that the installation can place an address-space-wide limit on protected buffer allocations.
I was amused that the group also adopted a line from my requirement: that IBM apply the fix to ALL VERSIONS OF JES2, not just the latest and greatest. We might turn the JES2 project into a fighting machine yet.