I have been looking at what Word does about the Unicode ranges. In fact it does more than read the OS2 table. Even if the bits have not been set, nor even the codepage, Word will fill in the appropriate values. Within the Word document, there is a list of font names, each with the details of the signature (codepage & unicode ranges). Even if these bits have not been set in the font, (as you can see from Fontlab or from the Font properties extension) values will be supplied by Word. The values supplied by Word in this list are shown in my Fontlist (http://ourworld.compuserve.com/homepages/RaymondM) which just reads the details from the list.
SO (which I hadn't really noticed before), it seems that in Word there is some scanning of the unicode values in the font, then a sorting into ranges, and assigning to bits. The natural question then is, if the font is used in Word, does it really make any difference whether the bits have set in its OS2 table, or not ? Raymond Mercier
