use euclidean distance (instead of rectilinear distance) to calculate mouse movement distance about jquery-hoverintent HOT 3 CLOSED

The comparison logic hasn't changed in the last six years. I just found something that worked and stuck with it. You're the first person to have looked at the guts close enough to care/comment.

Would there be any benefits when switching to Euclidean distance? Fewer false positives?!

The only caveat I can think of is that the default sensitivity might need to change, thus breaking anyone's custom override were they to upgrade without reading a disclaimer. I can live with that.

If there's a benefit, it certainly can be changed. If you have a mind for math and care to do it, I'll take a look at the pull request. I found your preliminary patch for ignoreScroll to be well commented. If it trades out one set of false positives for another, then we might leave well enough alone.

What do you think?

from jquery-hoverintent.

There will definitely be fewer false positives, especially for users that use trackpads, or high system-wide mouse sensitivity settings, and those who use very specific sensitivity and interval configuration options with hoverIntent.

As it stands, I can move the mouse perfectly diagonally a certain number of pixels, and that distance will generally be greater than the sum of the component movements in each dimension. This necessarily means that the fix will generate fewer false positives.

In simpler terms: the "bounding box" for the pointer's movement threshold in the current plugin is a diamond, and the fix would make it a circle.

I'm willing to submit a pull request with documentation (and an update to the default value).

from jquery-hoverintent.

:bump:

@briancherne, any interest in merging #14?

from jquery-hoverintent.