Closed stockiNail closed 2 years ago
Coverage decreased (-1.08%) to 95.274%
Looks ok, but I don't really like that we have to be one pixel inside the element before the tests work. Also should test different borderWidths of the line (1 for the other edge case, or should it actually be 0.1?). Also without label.
Without 1 inRange return false. I need to go in deep and debug it. I have the feeling it's related to intersect function and the epsilon. Let me debug it and I'll come back with the result.
Coverage decreased (-1.08%) to 95.274%
It's correct because the events part has been removed but it's present in the other PR.
Looks ok, but I don't really like that we have to be one pixel inside the element before the tests work. Also should test different borderWidths of the line (1 for the other edge case, or should it actually be 0.1?). Also without label.
Without 1 inRange return false. I need to go in deep and debug it. I have the feeling it's related to intersect function and the epsilon. Let me debug it and I'll come back with the result.
I think it would be fine if the test passes with borderWidth: 1
. If it does not, then I think its worth taking a closer look.
Looks ok, but I don't really like that we have to be one pixel inside the element before the tests work. Also should test different borderWidths of the line (1 for the other edge case, or should it actually be 0.1?). Also without label.
Without 1 inRange return false. I need to go in deep and debug it. I have the feeling it's related to intersect function and the epsilon. Let me debug it and I'll come back with the result.
I think it would be fine if the test passes with
borderWidth: 1
. If it does not, then I think its worth taking a closer look.
I found the bug in intersect... Let me test again (with borderWidth: 1 and others)
Ok the bug was here:
It must be:
return (sqr(x - xx) + sqr(y - yy)) <= epsilon;
I'm committing the changes and then I'll rebase it.
I think after this we are good to go with 1.3.0?
So the version bump would be next. The we could give it a week or 2 before starting to merge v2 things. The week or 2 so there is room to fix any new bugs found in 1.3 without creating any additional branches and stuff.
I think after this we are good to go with 1.3.0?
So the version bump would be next. The we could give it a week or 2 before starting to merge v2 things. The week or 2 so there is room to fix any new bugs found in 1.3 without creating any additional branches and stuff.
Completely agree with you!
Coveralls - ubuntu-latest-firefox — Coverage decreased (-0.0%) to 95.947%
This PR remove events test cases with inRange method tests.