Open akhilman opened 4 years ago
When I was linting non-keyed implementation by benchmark's linter, it said that Seed behavior isn't strictly non-keyed. So the results aren't 100% comparable because each framework uses different algorithms. So.. if you think VDOM modification will make it better and helps to improve user apps - I'm ok with it. If we want to change VDOM to satisfy this benchmark, we shouldn't do it.
Definitely agree do what's best for the user not an artificial benchmark.
The key benchmark cannot be fully implemented (see https://github.com/krausest/js-framework-benchmark/issues/694). Our keyed VDOM implementation recreates one of the swapped element. The reason for this choice is that the browser still recalculates the element when it is moved inside the DOM.
Possible solutions: