"DBR" == Denis B Roegel writes: DBR> Among the language attributes, there should IMHO be "default DBR> encodings." These encodings would be either a pair of < input DBR> encoding, output encoding > or a pair of lists of encoding, DBR> which would enable priorities. For instance, English should DBR> have as a default < ASCII, OT1 >. French should have < latin1, DBR> T1 >, Russian could have < koi8-r, X2 >, etc. For Russian, T2A seems to be much more convenient than X2 (the latter encoding does not contain latin chars at all, but they are often used inside text in Russian). As for the default input encoding for Russian, it is not obvious that koi8-r should be the default. There are several really widely used input encodings for Russian. Perhaps, specifying the default input encoding for Russian via language attributes could help? DBR> These encodings would then just be triggered when one writes DBR> \selectlanguage{french}, etc. I think that one should use the \selectlanguage command with caution: it is a "global" command with too many changes (captions, typesetting peculiarities, etc, i.e. such things that typically should have the only one global value per document), so more "local" triggers like \foreignlanguage and \otherlanguage are more appropriate for language switching. I think that it would be good to introduce in BABEL some mechanism which will define for (each or some) languages commands like \russiantext and \textrussian, so that {\russiantext ...} and \textrussian{...} will work as it is common in LaTeX. These commands should then change the font and input encodings. BABEL could provide a unified mechanism for declaring such commands in language definition files. Best regards, -- Vladimir. -- Well, O.K. I'll compromise with my principles because of EXISTENTIAL DESPAIR!