is there a way to insert a button into a form so i could trigger a different form? i feel like it should work somehow but i can't figure out the magic words. my goal is to make a "CRUD app" to edit a dataset/databin, or a generator thereof

@ThomasLackner What is a form? Something like TeXForm? Can you describe a bit more what you have in mind?

@halirutan sorry! I meant like FormAction/FormFunction/FormObject. reference.wolfram.com/language/ref/FormObject.html

@ThomasLackner So what you want to have is a button that when pressed changes the arguments the FormFunction reads in?

@halirutan no, i want the ability to send people between forms. i'd like to also add other arbitrary UI elements. for instance, a TabView might be nice in some cases. A button would serve a similar purpose - pop up a different screen/function. I know you can make custom UIs with buttons, but I wanted something I could deploy to the web too, if possible :)

@ThomasLackner This is exactly what I'm concerned about. I'm not sure this will be deployable to the web if you hack around the default behavior.

On the other hand, I rarely use the web features.

@halirutan maybe the best approach would be to make the "container" flexible - that would control navigation so to speak, and be specific to desktop/notebook vs web and take different steps in terms of how you go between screens. then, the form itself could be general. a lot more work sadly

@ThomasLackner But you have seen the multipage forms, yes?

@halirutan yeah i have seen that, thanks. problem is this is more like an "Edit pages" or "Edit users" kinda split, where the screens don't make sense presented linearly. but it is a cool feature.

@ThomasLackner So basically, you want a form that can take it's input and decide to which other form it leads the user?

@halirutan in a sense, yes.. but more tangibly, i want a form in which i can have other UI elements. imagine a search box for instance. or a drop-down to filter the data shown in the editable form below. or a button that takes you to a different screen to view a report, or edit a different type of data.

@ThomasLackner Well, one thing that is possible with multipage forms is that you have access to the values entered in the last step. You probably can build a conditional form that takes the selection in the former step into account.

But in general, I haven't done enough work with form to give a more satisfying answer.

@LegionMammal978 Yes, I see. If I would have to tackle this problem, I would probably first look what algorithms exist in libraries. Polygons on spheres should have a large number of applications in things like earth navigation. There must be a theoretical fundament available.

What I am very curious about is, if it is possible to transform this problem into a simpler one. One idea is a spherical projection that maps all points of a sphere with south pole at {0,0,0} and north pole at {0,0,1} to the plane.

If the distance between two points of your polygon is reasonably small, you can probably work with the points itself. Otherwise, you need to interpolate them by a geodesic. Then, you rotate your sphere (and the polygon) so that the point you want to test is at {0,0,0}. Now, you make the spherical projection:

The black line on the sphere is a large, non-convex polygon. The gray line is the projection ray and what you see on the bottom is the projected polygon. If this works, you can test your point on the plane (which is projected to {0,0}) with the projected polygon with one of the simple 2d algorithms.

There are quite some things that can go wrong: You polygon line crosses the north pole (which is projected to (complex) infinity) and I'm not sure if this holds when you have a very tangled polygon that goes across the sphere. Just an idea.

@halirutan Yeah, was actually thinking of using GeoWithinQ for this purpose

The problem is, all computation on that is geodesic, and I need mine to be exactly spherical

@halirutan Luckily, my polygons actually correspond to constellations in my use case, and none pass through the celestial poles

Wait, no, geodesic lines are spherical

idk, it's 6:27 where I'm at

But Polygons with GeoPositions still don't correctly cross over the east/west borders of the projection, so that's moot

Ima try your idea

(Just seems like some raycasting)

There, figured out how to calculate the surface points from the right angle and declination

I don't think I actually have to rotate the sphere

Just have to figure out the exact mapping for geodesics

And then actually test if the points are in this (curved, possibly inverted) polygon

@LegionMammal978 No, that should not be a requirement. It just made my example a bit easier to understand.

Now Ima go to Math's chat to figure out these projected geodesics

@yode Do you know if there are silent updating with Matlab ?

@andre First time to know that..

@yode I do agree with you that serious bugs going unfixed for a long time are extremely frustrating. And there is certainly room for complaints. But I do not think that SE is a good place for those.

However, when a complaint is reasonably and convincingly phrased, I think it may have an effect on Wolfram Community.