R2018A-RC2 updates relevant to SOHO:
R2018A-RC2 works differently than the build SOHO has. In particular, RSSStep works differently. Inform SOHO of this. SOHO should use fixed step. If SOHO must use RSSStep, investigate possibility of setting MaxStep to a much smaller value. (I think the SOHO build "effectively" set this to 60 seconds inside the code).
We also should inform SOHO of the OLSEUseRMS parameter. Since this was the prior mode of operation using GTDS, they may want to use that option and reset the sigma multiplier to 3.
Also need to mention setting the nutation update interval to 0.
Also should mention implementation of Marini model, they may want to switch to this since it is consistent with what was used for GTDS.
Also should mention that the ephem propagator now reads big-endian files. SOHO ops currently has to maintain a little-endian file solely for the purpose of running the GMAT compare, but this will not be necessary with R2018a.
We can get a copy of the SOHO operational script, make our proposed changes, and ran a sample of dates to make sure there are no issues.
(We also need to check with SSMO to see if they use relative error control for their scripts)