I'm struggling on this challenge too (for the same reasons I think)
Havn't figured yet the solution but the last solved date is 2015/01/15 (as of today) so looks like it's still doable.
Looks like the core of the challenge is to trick the regex BUT, looking for strlen documentation have revealed that chars like ä count twice so....
and strlen do count the NUL byte but the regex doesn't accept it (at least, the way I sent it)
After that, does a-zA-Z differs from A-Za-z ? (don't think so but meh...)
Keep trying and you may be the 269th solver
PS : I think your #1 solution won't work and the 2nd has to be prefered (especially if Ruby)
But keep in mind that I havn't solved it yet!