[Halld-offline] migrating hdds updates to ccdb

[Richard Jones]

Dear Sean,

After Maria pointed out that an overlap scan of the GlueX geometry in
hdgeant4 generates a lot of warnings, I have gone back through the hdds
geometry and done a massive cleanup. This involves shifting/squeezing a lot
of volumes to eliminate overlaps here and there. Nearly all of the
subsystems are affected in some way. These overlaps have crept into the
geometry over time, as many of us have been involved in updating the
geometry, while no one was doing strict validation.

The good news is that almost none of them are significant at the level of
impacting the physics. The one possible exception to this was the overlap
that existed between the BCAL and the FDC in the forward region, in which
the FDC outer volume was shaving off the innermost few mm of the BCAL seen
by incoming tracks and photons. I discussed how this came about at the
offline meeting today, and we agreed that my fix to the problem is correct.

I have submitted all of these fixes as a single pull request to the
JeffersonLab/hdds repo on github. Once that pull request is accepted, the
updates will immediately begin to be seen in hdgeant simulations using the
master branch hdgeant. However, in order for the updates to be seen by
hdgeant4 simulations, they need to be propagated into the /GEOMETRY tables
in ccdb. I would like to see these updates applied retroactively to the
geometry description, going back all the way to the beginning of GlueX. I
don't believe that any of these updates I have made are run-sensitive from
the beginning of GlueX through the present time. Going forward, there will
be changes of course.

Do you agree? If so, what should be the plan for pushing these into the
ccdb?

-Richard Jones
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20190108/674fa428/attachment.html>