Normally new features (here: new event) are not back-ported to fixes. What is the risk that the fix will damage anything?
True, but here we actually had a regression (db grid does not draw custom-drawn check boxes).
The applied patch adds new behaviour AND, as a side effect, solves this.
The solution might be to separate this in two patches - one which solves the problem with drawing grid in check boxes and another one which adds new feature.
Now, I can provide the two patches, but then, Werner, you should revert the changes, apply the first patch,
request backporting, and then apply the other.
Werner, what do you think? If you agree with this scenario, revert the changes, and I will provide the two patches then. Please tell us your opinion.