Bug Tracker

Changes between Initial Version and Version 1 of Ticket #13182, comment 8


Ignore:
Timestamp:
Jan 15, 2013, 11:23:43 PM (6 years ago)
Author:
gibson042
Comment:

Legend:

Unmodified
Added
Removed
Modified
  • Ticket #13182, comment 8

    initial v1  
    11Well, you sure get my vote for most obscure bug of the month. This one was a doozy, basically coming down to element ancestry cache keys conflicting for sub-runs within a parent selector containing nested `:has` checks... even writing a [https://github.com/jquery/sizzle/blob/5c51caa03824a86ca841aa9613b05f1ad5750228/test/unit/selector.js#L672 unit test] took about 15 minutes!
    22
    3 This will be fixed in 1.9.1, but I '''strongly''' recommend refactoring. Even non-nested `:has` selectors exhibit pretty bad performance, but nesting makes it an order of magnitude worse, and—since the same can be [http://jsfiddle.net/XcNCz/2/ achieved] with descendant selectors—is always unnecessary.
     3This will be fixed in 1.9.1, but I '''strongly''' recommend refactoring. Even non-nested `:has` selectors exhibit pretty bad performance, but nesting makes it an order of magnitude worse, and—since the same can be [http://jsfiddle.net/XcNCz/2/ achieved] with descendant combinators—is always unnecessary.