I've read in Unicode FAQ about problems of adding digraphs to Unicode and, as I am a speaker of a language which uses them and know about problems which might arise from not representing them (even though that particular digraphs are defined ;) I would like to propose a solution for other languages which, as far as I can see, doesn't cause any of the problems presented in FAQ.
My proposal is that new control code should be added to Unicode, just as there is control code for e. g. directionality of writing, which would mean that characters surrounding it are actually a digraph, and another which would mean that characters surrounding it are not a digraph. This would make very small compatibility problems. New programs would, of course work well with both new and old data as. Old programs could display and print new data perfectly, as the character would be non - displayable. About the only thing that old programs would not be able to do well is comparing new data to old data, and data inputed using old input method. They could even compare new data with data inputed using new input method (which would probably require one keypress for both characters and digraph code to be inputed). However, whatever compatibility problems occur, transition needs to happen only once and in future it would work for all encodings in all languages, even for trigraphs etc. if some languages have them, and even for all future additions to Unicode.
