Add regression test for hover tooltip being blocked by BoxAnnotation#14965
Add regression test for hover tooltip being blocked by BoxAnnotation#14965muendlein wants to merge 2 commits into
Conversation
Codecov Report✅ All modified and coverable lines are covered by tests. Additional details and impacted files@@ Coverage Diff @@
## branch-3.10 #14965 +/- ##
==============================================
Coverage ? 91.36%
==============================================
Files ? 291
Lines ? 21066
Branches ? 0
==============================================
Hits ? 19247
Misses ? 1819
Partials ? 0 🚀 New features to boost your workflow:
|
|
|
||
| await view.ready | ||
| }) | ||
| }) |
There was a problem hiding this comment.
This doesn't seem fixed. As I discussed in the issue, the order of renderers matters. Here the order is glyph, annotation, so when hovering over (2.6, 4.0), there should be no tooltip visible, unless propagate_hover is enabled on the glyph renderer.
This also needs more cases, with different order of renderers, multiple points and propagate_hover set on and off, for a complete test. However, it's already a waste to have this one as an integration test, so I would convert this to a unit test with a has_tooltip_visible() query implemented e.g. via callbacks.
There was a problem hiding this comment.
Seems like I misread your comment in the original issue. If I understand it correctly now, you are suggesting that the behavior (no visible tooltip) reported in the original issue is the intended behavior, if the order is first glyph then annotation?
If yes, should the order actually matter or is that also part of the bug (as render level is not affected by the order, right?)?
And maybe a more fundamental question, should a BoxAnnotation in general be placed at the annotation level or rather the underlay level. Personally I feel like many users are applying a BoxAnnotation with the idea in mind that it is placed in the background. This also seems to be the most common approach in UX design, so the current bug acts more like a feature.
What do you think?
For now I will put this as WIP, as there is obviously more work to be done.
Uh oh!
There was an error while loading. Please reload this page.