RC1 Issues

Owned by Steven Hughes
Last updated: Jun 06, 2017
2 min read

Issue and Proposed Solution

Owner

Solution:Status

Debug written by SPICEOrbitKernelReader.cpp causes tests to time out and appear as crashes.

Fix is trivial.  Comment out the debug.

SPH

Done

STKEphemReader rejects Scenario Epochs with precision more than millisecond.

Modify STKEphemerisFile::UTCGregorianToA1ModJulian to handle more precision for setting epoch.

DJC

Committed and ready for a test run

STKEphemReader has issues when reading files with covariance.

Modify reader to ignore covariance.

DJC

Committed and ready for a test run

STKEphemReader assumes DistanceUnit is required but it is not a required field in the STK spec.

Modify the reader so that we do not require DistanceUnit.

DJC

Committed and ready for a test run

Missing test for TESS ephem case

Create a test case and commit that uses a TESS STK ephem.

DJC

 

Splash screen too big on Mac

Resize the splash screen to be consistent with the size for R2016a.

SPH

Done

Final analysis of Ephem test failures

Verify these are comparator issues.

SPH

All issues resolved and passing on all OS.

Final analysis of navigation failures

DSC/SES

Reverted the SpaceWeather files in the repo and this seems to have corrected the extra failures.

Update ionosphere file.

Put the old file in the appropriate place in the Jazz repo so that it gets used in regression tests.  Then, update the file in mesa-file application with the new version

DSCupdated prior to noon today (6/5/2017). 
Low Thrust 
  • Low Thrust is allowed in R2017A but it is not described anywhere in the help.  Thus, most likely, a user would not know to use it, which is probably good since Low Thrust is not fully working yet.  Should it be disabled, by default, in the gmat_startup file?   The user would then only be able to access the feature by setting Run_Mode=TESTING.
  • DJC comment:  Turning it off in the startup file and enabling it via Run_Mode are 2 different things.  The former does not require a code change.  The latter does not.  For the Nav code, we are enabling/disabling parts of the plugin via RunMode, which is where the TESTING switch comes into play – we could have accomplished the same thing by moving the "TESTING" features into a separate plugin, but that is pretty complex for nav features.  Here we are concerned with the plugin features in their entirety.  So I'd recommend turning off the plugin if we are disabling the feature.  Didn't we have this plugin available in R2016a (as an alpha, so turned off by default)?  We probably should have it behave the same way in R2017a as in that release.
  Â