Probably, the patch is not a good solution, cause it replaces unicode() function with str(). The deal is that the called right before len() returns wrong length of string, so followed styling function (don't remember it's name) mangles string.<br>
Currently I don't completely understand the core of bug.<br>There was similar bug in my experience with unicode symbols in C, when each entered russian letter was counted as two characters.<br>However, it seems very important to fix that, cause me, for example, almost gave up to use Eric due to this bug, although it's a very good IDE for python/Django, probably the best I ever seen.<br>