[Kstars-devel] [ask]Little features, with ISS and transits...

[jb]

Hi all :), Hi Jason, 

> However, I was always reluctant because I feared the CPU burden of having
> Earth satellites zooming by overhead in real-time.

Surely :)

> However, on reading your suggestion, I realize that we wouldn't need to do
> it that way.  As you say, we can simply draw the curve of the satellites
> track, which would be static on the sky.  You could even add tickmarks and
> labels showing the time at which the satellite passed.  I think that would
> be great!

Yep, It was such things I guess. Compute passes on a night (ISS can make 6 
passes on a night) and draw them statically.

I'm not sure we have to include Lots off satellites : 
TLE files are format as : 
ISS (ZARYA)
1 25544U 98067A   05161.65811343  .00015791  00000-0  11912-3 0  6610
2 25544 051.6456 294.3676 0004311 229.4752 339.8263 15.72674954374655

So, steps : 
1) dowload TLE on space-track.org. (could not be done automatically due to 
security reasons..see space-track.org)
But we could include a interface to log-in space-track and get TLE's with our 
space-track login.
(I have a shell (or perl.. don't remember) that can do this...and give lots of 
satelites ;) )


2) Choose which satellites we want to draw...

3) compute 

> We don't really have a plugin architecture; I suggest we add satellite
> tracks as a new class of objects drawn on the skymap.  How many satellites
> are you thinking of including?  Presumably, we will add a page to the
> configuration window that allows the user to select which satellite tracks
> are shown.

Yep... depends on what space-track give. But you can have hundred of 
satellites... but more inrested are ISS, Hubble and envisat...

> So, to get started, I guess we need a SatelliteTrack class that contains
> the satellite's TLE data, and a QValueList<SkyPoint> containing the track
> across the sky.  It should also have a QValueList<QTime> containing the
> time at each position along the track.  It will also need methods to
> determine if it is visible on a given night, at a given location

This Code is in "predict". And, I surely (not so sure :() can have code to do 
this accurately.

> compute the track for a given night, at a given location.  Once this is in
> place, we can start integrating it into the rest of the code (Skymap,
> Options dialog, KStarsData for reading in data files and updating the
> tracks, Whats Up Tonight for predicting satellite passages, etc).

> As for the accuracy of our Sun and Moon positions, they are pretty good.
> KStars does a good job of reproducing solar eclipses.

Cool :)

Clear skies, 

					JB