#2098 das2server ephemeris inconsistency

[Jeremy Faden]

Masafumi pointed out a case where his new ephemeris source doesn't read, appearently because of units, but an older one works when it has different units. Try:

vap+das2server:http://jupiter.physics.uiowa.edu/das/server?dataset=Juno/Ephemeris/Jovicentric&interval=60&item=2&start_time=2017-02-02T12:58:00.000Z&end_time=2017-02-02T12:59:00.000Z

which works, but

vap+das2server:http://jupiter.physics.uiowa.edu/das/server?dataset=Juno/Ephemeris/Jupiter_Footprint&interval=60&item=footlat&start_time=2017-02-02T12:58:00.000Z&end_time=2017-02-02T12:59:00.000Z

does not.

[Jeremy Faden]

See nudnik.physics.uiowa.edu:/home/jbf/project/autoplot/bugs/2098 which explores new code which would read directly into QDataSet.

[Jeremy Faden]

I'm revisiting this, knowing that it was almost done, it's a quiet week on campus, and Chris pointed out the X,Y,Z stream which was never supported.

[Jeremy Faden]

The X,Y,Z is not as trivial as I was hoping. A new object type is needed to represent, and I'll have to do that in the morning. Chris also pointed out that there can be multiple Zs in this form as well.

[Jeremy Faden]

I refactored the org.das2.client package a bit, mostly to clean up names "multiy" is now just "y", added "z". The new xyz mode is supported, and I've integrated the new das2stream parser which parses directly into a QDataSet into Autoplot.

[Jeremy Faden]

I know there's probably more work to be done with append vs join. When you have two blocks of data which are the same form, they need to be appended, not joined.

[Jeremy Faden]

These give consistent results now.