Did you test it against the test suite? And did you test it for speed changes? The reason why this is being done, in the first place, is purely for speed reasons. Added extra checks or loops significantly slows down the speed of jQuery selectors. The mergeNum technique is the fastest means of checking for unique-ness in a set of elements. If you can find a method that's just as fast, but doesn't have the property, then we'll by all means use it.
Additionally, while your patch does change some things around, it doesn't change the fact that the mergeNum property will still be added to some elements, thus still effecting applications like Wymeditor. It just doesn't happen in this very specific case (which is what your patch is attempting to work around). My suggestion would be to add the following code to Wymeditor to remove the property: $(...).find("*").each(function(){ this.mergeNum = null; }); I mentioned this before, but I still think it's the best solution. It's a very specific problem with a very specific solution. --John On 5/24/07, [EMAIL PROTECTED] <[EMAIL PROTECTED]> wrote:
Hi, please take a look at my patch for ticket http://dev.jquery.com/ticket/1143 I'm not sure why mergeNum was actually used, I think a simple boolean should be enough. Cu, Volker. On 21 Mai, 14:39, Jean-Francois Hovinne <[EMAIL PROTECTED]> wrote: > Thanks for your help. > > Actually, I can remove the attribute using removeAttr, but the extra > DIVs remain (for example if you create inline elements in lists). > > IMHO, as the problem only occurs in MSIE - no mergeNum issue in Gecko > nor in Opera - I'm rather thinking about a browser specific issue. > > Thanks again, > Jean-François