Ok, on the "!" marks.

The Segmentation fault exits python and isn't catchable as far as I know. It's happens in the compiled tokenize.c file. The python side error detection doesn't get a chance to catch it.

The problem is present without the patch applied and when using the console help so it wasn't anything I did.  I could use a hack to bypass that particular file, but I don't think that is the correct answer to this particular problem.  Better to get that issue resolved first.