GitLab

This patch includes the previous 2. It also modifies the first screen of whats interesting and adds Messier objects to the list. As a part of that, I also had to add a way to identify that the object was below the horizon.

This will download wikipedia info text box in addition to the description.
It can dynamically display the content in a tabbed pane or right next to each other depending on the width of the page. Both the descriptions and the info text are now saved after loaded the first time for easier loading later.
I also captured the wikipedia image from the info text box and added that to the list of images available in the thumbnail picker.  I also improved some of the layouts in whats interesting

There is a bit of Fuzz, cause astyle was doing it's thing. (Has anyone
run astyle recursively yet? There are still some unformated files.)

To clarify things: I just added a "drawComet" method to the skypainter,
an option to the kcfg and the UI and one tweak to retrieve an instance
of the sun.

To note: the findPA method could be tweaked, by somehow caching the Suns
position, because by now it is re-projected for every comet tail. (Well
that's actually OK, because you hardly see more then two at a time.)

Skip PHD2 and LinGuider when starting INDI Web Manager drivers since these two are just place-holders for the apps

In case of conflict in i18n, keep the version of the branch "ours"
To resolve a particular conflict, "git checkout --ours path/to/file.desktop"

I was working on the Whats Interesting tool because I think it has a lot of potential.  I added photos and descriptions to the list.  I added photos to the detail view.  I also changed some of the layouts.  And finally I added a telescope slew button.

Detect if flat image is already saturated before calculating fitting. Only utilize polynomial fitting if we have 4 or more points since it can quite unreliable at low counts

Initial work for Image Guiding algorithm by Bob Majewski. All guide images must be explicitly converted to floats and then partitioned into NxN regions before sending them to the algorithm

I found the drawing issue that was making the horizontal bar for different pixel heights of the window.  It took me some time to dig up what exactly was causing the black line.  But I finally did.

It was a result of clipping that was taking place in Skymapqdraw.cpp:

    // Set Clipping
    QPainterPath path;
    path.addPolygon(m_SkyMap->projector()->clipPoly());
    psky.setClipPath(path);
    psky.setClipping(true);

Which was based upon some code in Projector.cpp in the method void Projector::updateClipPoly()

The issue was that in creating the clip path, it was stepping around the circle from 0 to 180 degrees and then from 0 to -180 degrees.  As a result, it created not a circular clipping region but two semicircular clipping regions.   I painted the path so that I could clearly see what was going on:

Then I tried reversing the direction for the second half circle from -180 to 0.  This did not work, it produced an unbounded clipping region.

But then I tried doing a full circular path from 0 to 360.  This not only gave a circular path, it worked perfectly for the clipping region and fixed the problem.

Now what I don’t know is whether the reason for two separate half circles was based upon some issue with the sincos method accurately calculating the other half of the circle.  The reason I suspect this is because someone went to a lot of trouble to write two separate loops that are nearly identical.

As far as I can tell, this method works, please check for me.

CCMAIL:kstars-devel@kde.org

This patch should allow the remote solver to print its solutions nicely in the table the same way the online and the local solver already do.  It also adds an autoscale button for the solutions graph.

Read indidrivers.xml as the first file. Silenty ignore replicas. Do not allow device manager to start INDI services for remote drivers

Create INDI driver entries regardless of what is currently install. Drivers that are not locally available are treated as remote-only drivers

Do not process WCS on loading if we plan to create a WCS file. Make loadWCS asynchronous and use signal to proceed in the chain

Do not make thumbnail selector dependent on detail dialog since it can be used also in observing list