...

Is it the sort of thing where it's just work putting a list of all of them (line numbers, &c) up on a web page, maybe crosslinked into cvsweb? Then people can go and pick them off, one at a time.

ok dokey - so, is DC still running a source view web system? If so, you could easily enough run a daily job like: find . -name '*.py' -print | xargs grep -n 'except:' | cut -d: -f1,2 to get the list of them from the source, and put the root of a viewcvs URL on it. Hm, viewcvs doesn't really cut it for this, but lxr would... whatcha think?