Developers meeting

[Abe Kornelis]

All,

today we had another developers' meeting.

We discussed the status of scripting.
Anthony has two open pull requests for scripts.
We discussed how to test these pending changes.
With the changes not actually being in the z390 repo yet,
it appears that not everyone has a clear vision on how
these pending changes might be tested.
Anthony agreed to write a short how-to.

Then we discussed the status of the doco conversion process.
Anthony indicated he's still making progress, but also
wondering whether the structure is okay.
He has an open pull request for doco changes and invites
comments from anyone ....
Melvyn expressed his concerns that the zVSAM D&L
(Design and Logic) might be hard to convert.
We agreed Melvyn will send a fresh copy to all participants,
so we can all check the updates and Anthony can assess
whether or not conversion will be problematic.

Next we covered the changes by Dave Kienow.
This covers two broad changes: one to encapsulate
and improve java I/O during assembly to sequential files,
and one to issue an appropriate abend when a privileged
operation is executed at run-time.
Current testing reveals some problems with the I/O-related
changes. We've decided not to spend time on figuring out
how or when the bug was introduced - we'd rather move
forward by getting the issue resolved.

Don remarked he's refraining from applying maintenance to
java code while these changes are still in the works.
We discussed that git helps to auto-merge versions as long
as changes to a source file do not overlap. I.e. two persons
can change different sections of source file without creating
a version conflict. Don agreed that it's mainly lack of familiarity
with this phenomenon that makes him cautious.
We agreed to try with a relatively small change to get
more familiar and confident with the process.
After all: the whole purpose of adopting git was to enable
us to do more work in parallel ...

While discussing zVSAM Melvyn remarked he's run into
two issues with recent changes in the java layer.
He reported the issues to Don, who agreed to look into it.

Anthony then brought up the issue of the zCics build process
not working correctly any more. Melvyn assumed that this is
caused by the restructuring operation that split out the source
files into separate subfolders. The bat files that perform the build
have not been adjusted accordingly yet.
Anthony and Melvyn agreed to work together to get it fixed.

Finally we decided to have our next meeting on August 23d.
Don will send out a proper invitation the weekend before that.

Kind regards & Happy programming to you all!
Abe
===