SIGMPE met Wednesday morning in a small room 212 at the Moscone center.
A post-meeting roster count shows 65 registered attendees although I
would wager that we had considerably more actual bodies present.

After some initial housekeeping and pointers to HP3000-L,
comp.sys.hp.mpe, the Interex HP3000-L executive summary, Bruce's Webbot
site, the 3k.com FAQ/resources, and the new 3kworld.com pointers, the
meeting proceeded with reports on existing agenda items from CSY.

Gopi from CSY Bangalore began with last years top 10 improvement ballot
items, indicating that #3 ability to kill a single process has been done
and is in a beta patch for 6.0 (:ABORTPROC), and Posix smoothing issues
are ongoing.

Gopi then followed with some key items in the MPE 6.5 release, including
the following:

* increased system limits
  - up to 16Gb meory
  - disks per system and devices per class up to 511 from 255
  - DSTs from 16K to 64K
  - fixed record files and KSAM files from 4Gb to 128Gb!!
  - VT sessions up to 2600, TCP connections to 20000, UDP to 10000
* multiple run queues and improved internal locking in dispatcher
* AIF interfaces to user-defined job queues
    [started brief discussion of who was using them, only a few
     hands came up; but discovered not many users were on 6.0 yet
     due to Y2K and/or waiting on powerpatch 1 for 6.0]
* SYSDIAG to be replaced by HPUX type diagnostic system, but still
  requires passwords.
    [Stan Sieler noted that he dislikes this facility on HPUX now
     and looks forward to disliking it on MPE as well]  :-)

Gopi's presentation was done by powerpoint with a projector, but due to
the marked absence of an overhead projector the subsequent topics
covered by Jeff Vance, Craig Fairchild, and Scott Maclellan were done
in audio format :-(  We heard some brief highlights of new features in
MPE 6.0 from Jeff Vance, particularly :ABORTPROC and a job info CI
function, which resulted from the Customer Funded Enhancements project.
We then got the short form of the large files enhancement and the new
KSAM64 filetype.

I then reviewed some of the SIG agenda items that were not addressed by
CSY for discussion, with some feedback on the "login banner" item, but
otherwise not much discussion.  Despite the traffic and "verbosity" of
this list, the attendees were unusually quiet :-)

As the number of standing-room-only attendees increased, a conference
volunteer appeared with an offer of a larger room which would be
available at the end of the hour.  We then dismissed for a break to
reconvene in the much larger room 302, complete with overhead projector!

We began with the "new and improved" presentation of the large files
enhancements, complete with details of memory mapped file support and
new AIFs to deal with buffer areas with large files to avoid
complications with PA-RISC 1.x space register issues.

A discussion then followed on NM versus CM command interpreter, as the
lab is looking to "outgrow" the old CM procedures.  The only issues at
hand were some parsing/syntax errors returning different CIERROR values
due to the centralized NM parser versus the command-specific parsing in
the CM version.  This was viewed as "not a big deal" by the majority,
and CSY outlined a possibility of making the "default" version of the
parser being determined by a CI variable.

With no further input from the floor and time running out, SIGMPE was
adjourned.

With that said, I'm off to grab some lunch and listen to Harry
Stirling's presentation which promises to be quite an event.

Jeff Kell <[log in to unmask]>