Socorro: Jeff, Martin, Juergen, Pam, Lindsay, Susan, Jen, Kumar, Tak
CV: Anand, Andy, Akeem, David, Darrell, Ville
Garching: Sandra, Dirk, Julian, Justo
News / Meetings / Visitors
Summer school this week
Peter Teuben and Lee Mundy in Socorro this week discussing handover of ADMIT.
Kelly Sharp in Socorro next week. Will be working on documentation update project.
Documentation Update Project: We are preparing the new Plone server for contributions from developers; currently adding documentation about the documentation. Information about this project, including minutes from team meetings, is at CASADocumentationUpdatePlanning.
Updating wcslib to make is consistent w/ RHEL 6 and 7.
Working on bamboo CI builds. Has been blocked for licensing issues. The goal for this week is to get builds going so we can replace Jenkins, although we will run bamboo and Jenkins in parallel for a short time. You will probably see emails coming from bamboo when builds fail; contact Ville if these are confusing.
Licensing issue has been resolved. To use an open source JIRA license, proprietary projects will be removed from the open source JIRA server.
Upgrade of tool set is a good idea, but will not be done right now.
Akeem is converting acceptance tests to gradel setup. Going well; still have OS 10.11 left.
Osirus has been split from the rest of the cluster. CASA guides are being run manually on the cluster.
4 out of 5 of the accepted test failures are real.
CASA guide failures reflect the fact that the casa guide data has not been updated.
Light on the ground for testers right now due to summer school and PL testing.
Warnings are stable. Google test is online for OS X. We need a nightly job that does a clean build to monitor warnings. Google tests are ready for developers.
We gave two talks on the HPC status and development in CASA at the ALMA Developer's Workshop in Gothenburg last week. The workshop focused on "what is next for ALMA" and also reviewed previous ALMA development studies.
In preparation for the workshop, we run a series of tests using the ALMA pipeline in parallel. We found failures when running the last step of tclean-target, when creating a cube. We sent information about this failure to many of you.
Pipeline tclean memory problems are not an issue for dedicated machines, but the problem should be solved for the sake of users running on multi-purpose machines.
Will begin adding documentation when new server is ready.
Juergen has been doing parallel testing of tclean clean and has sent around a report.
F2F in Chile. Received a list of reqs for C2R4 release. One month to go.
NAN problem has gone away.
James Robnett has been doing some regression testing of the PL.
Mostly worked to prepare the talk for the ALMA Developer's workshop and run many tests on alma-m100 and the pipeline.
Discussed the issue that mstransform (split) does not propagate ephemeris tables when reindex=False. The discussion led to the fact that maybe the ephemerides table should be converted to a real sub-table. Currently, they are physically stored under the FIELD directory, and they show up in the keywords of the MS and nowhere else.
Discussed the issue of updating the SYSCAL table in mstransform (CAS-8618). Mstransform seems to be doing the correct thing, as there is no way for mstransform to know what UNselected Tsys spws it should pass to its output.
Justo González Villalba
worked on the following tickets:
(CAS-8580) Provide recipe to concatenate ephemeris tables and integrate it into the concat task (resolved)
(CAS-8508) Update default Primary Beam Models for ALMA and VLA
(CAS-8642) Make fits reader more robust w.r.t. invalid binary tables (resolved)
Installed RHEL 7.2 and set up a CASA build following Ville's instructions
Preparing for visit to Michel to work on: (CAS-7861) CASA dies on VLA ephemeris datasets
hoping Bryan Butler can provide test data in time
Boost cleanup in bnmin1
improve linux packaging of Xvfb
setup NFS for serving data from Apple/Fedora host to virtual OSX 10.11 guest
work getting Structure101 to use Understandudb static analysis database
add spectral fixed precision option to the viewer ( cas-8633)
add contour line width as an option to imview(...) ( cas-8366)
setup local el7 regressions, plotms fails due to massively increased memory usage compared to el6
debugged binary test failures since el 6.8 upgrade of build machines, fixed by Darrell
CAS-8618: copy ephemeris table in split if reindex=False