I'll investigate. My role is admin rather than tech support but I'll see what I can do.
The problem appears to be that the forum software is not "8-bit clean." It is possible to encode English text with only seven bits of data (128 variations), but other languages that use a Latin alphabet require character codes that are eight bits wide (256 variations) to accommodate character-diacritic glyphs and other so-called "special characters." (These, of course, are only "special" if the text is confined to English.) There is
an international standard for these alphabets that all modern software should adhere to. As Andrew Rodney points out earlier in this thread, the forum software accommodated standard 8-bit characters until someone started mucking with it. If whoever is currently responsible for maintaining the software can locating an archived version of it and compare it with the revision that is running now, that would be a good starting point for isolating the cause of the problem.
AddendumIn the event it isn't clear why it is important to use the standard international character-encoding scheme for Latin-based languages in a forum that is predominantly in English: (1) some place names relevant to forum posts may require diacritics; (2) anyone posting in English from a modern computing device (one manufactured in, say, the last 25 years) is likely to be typing eight-bit character codes.