David Starner wrote as follows. quote
What good would a private use character do here? The private use area is good for Tengwar, Cirth and Shavian (all of which have multiple fonts using the same private use area encoding.) But there's no huge demand to interchange data with these characters, and the few users are probably going to use something less complex then the private use area. Assuming I scan this book in for Project Gutenberg, I'll probably use something like [3], [4], [4,] and [4,h] for the characters, at least in the ASCII version (and there'd be no reason to post a Unicode version if these characters aren't in Unicode.) It's simple, readable and precise, something your solution only has one of. end quote The size of the "demand" does not affect my suggestion. It is a matter of research in encoding scripts. Using the Private Use Area is not at all complex. If one has a font containing some Private Use Area characters the font can be used quite straightforwardly from within a program such as Microsoft Word using the Insert | Symbol facility which that program provides. Characters can also be accessed fairly straightforwardly from a program such as some issues of Microsoft WordPad using an Alt sequence, by holding down the Alt key and entering a decimal code point value using the number keys at the right of a PC keyboard, then releasing the Alt key. Project Gutenberg is a very valuable project. I have recently started reading the notebooks of Leonardo da Vinci which is available at Project Gutenberg. For a file using ASCII text, using [3], [4], [4,] and [4,h] for the characters is probably the only type of method available for the work and probably quite suitable as it gets the job done within the limits of the available technology. However, consider that that file could be processed using a short Pascal program using a eutocode typography file. The Pascal program would read in ASCII text and output a Unicode text file, converting certain character sequences or individual characters. The particular eutocode typography file for the conversion of the format which you suggest would only need five lines of a few characters each, provided that the figures 3 and 4 were only used within the file for those symbols and not as digits, unless you intend using the [ and ] characters as well as the digits. However, a Private Use Area encoding of the special symbols would be needed and a font to display them. If a Private Use Area encoding is produced for the special symbols used by missionaries in the fifteenth and sixteenth centuries, then a few fonts, from various fontmakers, might include the special symbols. Thus your suggested ASCII file for Project Gutenberg could be used to produce print outs using the correct special symbols if desired. The eutocode typography file format is described in the following document. http://www.users.globalnet.co.uk/~ngo/ast03300.htm You suggest that using a Private Use Area encoding has only one of the three attributes of simplicity, readability and precision. I feel that a Private Use Area encoding could be reasonably simple. Care could be taken to make it as logically structured as possible within the limits of an on-going, a bit done now and then type activity, of adding in code point allocations as symbols are found in the literature. The experience gained could be useful when promotion to regular Unicode is considered formally, when the order of encoding used in the Private Use Area character set could be changed around as desired so as to produce a formal encoding. Certainly, without a suitable font readability would be a great problem. Yet once a Private Use Area encoding is published, font support may follow. A Private Use Area encoding can be precise provided that both the originator of a document using the encoding and the user of that document both know what is the encoding and that both have suitable facilities for applying the file which contains the document. In such circumstances the Private Use Area can be of great precision and very useful. For example, readers might like to have a look at the font COURTCOL.TTF which is described in, and downloadable from, the following web page. http://www.users.globalnet.co.uk/~ngo/font7001.htm Readers might like to have a look at the way in which I have expressed the colours in monochrome then perhaps search at http://www.yahoo.com and other search engines using the two words Petra Sancta together for the search. I have tried some offline experiments with a Java applet and the results are good. I have also produced a font with 51 glyphs which includes those 19 glyphs and others for four sizes of type, various object replacement characters, various wait for push button push codes and various markers for producing a programmed learning package encoded within a text file using WordPad. Precision is essential for such an activity and the Private Use Area is used to provide that precision. Actually, I was rather hoping that the start of a Private Use Area encoding might be produced by a few interested people fairly quickly, perhaps in this thread or in some email correspondence. Once that is done, then font support could gradually be produced. William Overington 29 March 2003