Thanks! I agree with you that the gettext approach relies on the assumption that everyone speaks the same language. I would add that it also relies on the assu

Thanks! I agree with you that the gettext approach relies on the assumption that everyone speaks the same language. I would add that it also relies on the assumption that every programmer has good writing skills, which is not always the case. In my teams, some programmers are really good at writing code, but not so good at writing a correct English or French sentence. In such a situation, string IDs are better.