[Halld-offline] De-commissioning the Commissioning Branch of sim-recon

[Mark Ito]

*Executive Summary*: check-in any final changes to the commissioning 
branch of sim-recon by C.O.B., Friday, March 6.

Folks,

After the offline meeting today a few of us were discussing how to 
restore the sim-recon trunk (trunk/sim-recon) as the main line of 
development. Recall at the collaboration meeting there was consensus 
around copying the commission branch (branches/sim-recon-commissioning) 
onto the trunk at some point. Now that Justin has discovered the source 
of our previous problems with the branch (simulated data not 
reconstructing properly due to inappropriate calibration constants) it 
is beyond time that we do the copy, branch to trunk.

That having been said, one of us (David) pointed out rightly that some 
people probably have un-checked-in changes to their commissioning-branch 
code. As such the copy mentioned above, if done immediately, would leave 
their changes off the trunk. To avoid this we will delay the copy until 
5:00 pm on Friday. In the meantime, please check-in any changes to the 
commissioning branch that represent real improvements. Those changes 
will then get swept to the trunk when we do the copy on Friday.

Note that if for some reason you are not able to make this deadline, 
your code changes are not lost forever. There are various ways to put 
them back on the trunk (description a bit too technical for this 
message). See me for assistance if needed.

After the copy is made, then the commissioning branch will be considered 
to be at end-of-life (apart from exceptions alluded to above, in 
reference to missing the deadline). All development after that should 
then begin with a fresh check-out of the trunk. We will make an 
announcement on this list when the copy is complete.

After we move off the branch, there will still be a need to analyze data 
from the Fall run obviously. To do that you need to build sim-recon 
against the commissioning branch of HDDS (i. e., you must still use the 
commissioning geometry). Alternately, you could continue to use the 
commissioning branch of sim-recon, just know that further improvements 
to that code base will not be coming in. For the upcoming run, since the 
trunk of HDDS (geometry) has the hydrogen target and the default beam 
line components, it should be used.

Finally, if  you have objections to this plan, please let me know before 
Friday, 5 pm.

   -- Mark

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.jlab.org/pipermail/halld-offline/attachments/20150304/b1bcb9c1/attachment-0002.html>