Notes of z390 developer meeting

[Abe Kornelis]

All,

Earlier this week we had a z390 developers meeting.

We were happy to have Bob Schulien and Larry Simmons joining us.
Welcome to both of you - and sure hope to see you again!

Meeting cadence: we decided not to change the timing and
to resume/continue meeting monthly.

We discussed zCobol support. The NIST test suite that Don used
when he developed zCobol has been found to be incomplete.
Don to look for a complete copy on his machine.
Anthony is looking into a port of the Cobol compiler
that came with MVS 3.8J. The biggest challenge seems to be
the I/O module which uses access method functions that are
not supported in z390.

Optables: Various instructions had been included with the
wrong optable. Almost all optables are now fixed.

We discussed the issues we have with the z390 java sources.
They are large and difficult to maintain, using many shared
variables. We would rather see cleaner interfaces and more
use of standard object-oriented programming techniques.
No conclusions were drawn and no decisions were made.

There are still regression tests in the repository that never
made it into our standard regression test set.
Abe to look into this.

We discussed whether we should create a new distribution.
The last one had issues with its version number.
Anthony to look into defining a version file that we
can maintain manually.

Finally we brought up the topic of the Java compiler version.
We concluded that the current approach - forcing version 8
compatibility mode - is okay and is expected to remain supported
for a very long time to come.

Next meeting: Tuesday, June third (Europe, Americas)
which is Wednesday, June fourth in Australia.

Kind regards & happy programming,
Abe
===

[Larry Simmons]

Abe,

Thanks for the follow-up. I plan to attend meeting on Tuesday, June 3.

Larry,