On 10/14/13 4:06 PM, "Maurice Amsellem" <[email protected]> wrote: >------- >Intuitively, I would say that both fixes are OK, but the latter one has >side effects because of the deferred behavior. > >So should I commit the initial fix, although it's not optimal?
I guess it depends on how much time you have. Obviously, if everyone fixed bugs in a non-optimal way we'd eventually have performance issues that would be hard to resolve, but sometimes a fix does require a performance hit. If you have time, figuring out if the tests just have a timing assumption that needs fixing would be best, but I won't veto your original fix. -Alex
