I thought about this a bit. I considered using C, but decided against it in the end.

C lost out because I wanted to use this project as an opportunity to try out full-on TDD, and C's automated testing tools seemed like a pain.

I also wanted to build something that I could release relatively quickly to validate the idea. Not to mention my eagerness to be able to use it for my own address book needs. I'm not very experienced with C so it probably would have taken me six months to get to this point instead of one.

As it stands, the main downside has been that I've encountered a lot of Rubygems-related woes resulting from the choice. Some earlier versions of Ruby are still very commonly installed on peoples' computers and don't work with ppl because they lack certain methods. I'd rather that than another five months of the app not even existing.

I wouldn't be averse to a C port, if the stars aligned just right so that it was both necessary and possible. For now I'm hoping Ruby will suffice!

It would probably be much easier to port to Perl or Python than to C. I think both Perl and Python are already installed on enough computers that no-one would have commented on the issue, unlike what happened with Ruby.