What do you think about adding a bool property to control the behaviour? Just in case if someone prefer seeing a caret. It will also reduce count of breaking change.

It should definitely not be caret + filled placeholder as it was before. That duplicates information.

Filled placeholder with no caret (which this PR implements) is the best option IMO and an unambiguous improvement. This is what Mathematica and the MS Office equation editor do.

Caret + blank placeholder could also be an option. This is what MathQuill does by default. I think this is inferior as it doesn't show the structure as well and is also a larger change to this library. So I don't think this should be an option, and so I don't think we need any options.

@Happypig375 do you agree?

@charlesroddie
Agreed

@FoggyFinder
The major component of CSharpMath's version is still 0. This means that we can still take breaking changes.

Just that the PR that changes the target to .NET Standard 2.0 is a bigger breaking change compared to others.

Also: thinking about appropriate test.

Should we introduce graphical tests? i.e. Comparison of rendered images, which can be done with CSharpMath.SkiaSharp.

Great then.

Should we introduce graphical tests? i.e. Comparison of rendered images, which can be done with CSharpMath.SkiaSharp.

It would be awesome. But I have never written graphical tests.
All UiTests that I wrote was for WPF apps.

You don't mind, if i try to make blinking cursor there and placeholder with dashed border ?

You don't mind, if i try to make blinking cursor there and placeholder with dashed border ?

I would suggest creating a separate issue and PR.
Small parts are really easier to review.