JQuery :: 1.4 Slower Than 1.3x When Using Append Or Html With Large Strings?

I've been using jquery for some time, and was very excited about the improvements in jquery 1.4 I have found however, that in one particular instance, jquery 1.4 seems to be performing slower than 1.3x, to the point that it has forced me to downgrade the script. I use jquery heavily throughout my web application; in this particular case, when a very large block of html consisting of a bunch of <tr>s is .appended (or .htmled, tried both) to a table, jquery 1.4 in firefox will give the "unresponsive script" error, prompting the user to stop, debug, or continue. The same block of html works perfectly fine in jquery 1.32 (and quite fast too). I haven't had time to do too much experimentation, as this is in a production environment, and thus downgrading was necessary as it was breaking the page, but I would love to figure out why this is happening so that I may optimize the code sometime in the near future. Have the improvements to .html resulted in code that causes higher cpu usage, or that would have a much higher overhead on longer strings? I've commented much of the code around the call, so that it pretty much consists of an ajax call that returns a chunk of html, which is inserted to a table (that I first empty). I thought it might be something with event bindings that occur after the insertion, but removal of these event bindings does not resolve the unresponsive script error. It seems the .html or .append is doing it. Note that this code, as is, works perfectly in jquery 1.3, even with event bindings, etc.

Does anyone have any ideas? My next step was going to be to try returning the entire table including the <table> tags, and doing a replace instead of an append, but if anyone has other suggestions to try, please let me know. Also, just as an aside, what do you guys consider the 'best practice' to be when returning dynamic data for a table (server side sorting, filtering etc from a db) ? Do you return just the data in json, and repeatedly clone a row element, replacing the values in each row (thus decreasing the size of the ajax call, but increasing the client side processing), or return the full html, and replace the innerHTML of the table?