>>The feature I'd rather see would be to take collating sequence from the OS, so to allow for "arbitrary.nls" as an option in "index ... collate...". IOW instead of asking for new collation sequences to be added to the engine, allow the engine to use what's available on the system. More specifically, to default to the current language set for non-unicode apps.
>
>embolded: as vfp is ISAM, in case of remote dbf (for those not already frightened off by SMB/oplocks trouble) each client has to potentially create/open a collation-named specfic index - if I had already reindex under german collation sequence for adress, you would need a factory to open your specific collation sequence for adress? How would my machine update the index created with "arbitrary.nls" when adding records if my machine does not have "arbitrary,nls" ?

Probably the way a RDP session on a machine in any random country awaits me with serbian keyboard layout: automagically. The needed files appear on the machine where they are needed.