Thanks.

> return ''

If we've run out of loop, there must not be a common prefix; our common prefix, therefore, is the empty string. We return that as a literal.

Huh?
I don't understand this...
It seems to me that if you run out of loop, then all the characters of all the non-shortest strings have agreed with those of the shortest string.
So your common prefix is the shortest string, files[0].

It seems unlikely that I would have noticed this while you and your other readers did not... but I can't figure out why the way it's written would be correct.
If you answer this, I'd appreciate an email. I don't normally check this blog.

I'm not sure what you mean by it being hard to tell posts apart from one another. Are the HR and H2 on each post not doing the job? (No sarcasm meant—I'm asking earnestly.)

What's hard to read about my website layout? (Other than the fact that all the articles get shunted down below the sidebar half the time. That's a bug that I still haven't figured out; something to do with the wrapping nature of heading elements, I think.)

(In my defense, your website layout is really difficult to read.)

files = [pair[1] for pair in sorted((len(fi), fi) for fi in files)]

I typically use tuple unpacking in this case:

files = [name for length, name in sorted((len(fi), fi) for fi in files))]

OTOH, Python has a function that does exactly what you want: os.path.commonprefix. It uses min and max in a really neat way I wouldn't have thought of, which is most likely more efficient than sorting the list.

http://svn.python.org/view/python/trunk/Lib/genericpath.py?rev=51719&view=markup

files = ['/foo/bar/xyzzy_one', '/foo/bar/xyzzy_two', '/foo/bar/xyzzy_three/four', '/foo/bar/xyzzy_']

You get no output.