Actually, the Private Use Area is from $E000 to $F8FF. The conventional way to express that in discussions about Unicode is as U+E000 to U+F8FF.
Those values are not indices to the glyphs. They are codepoints in the Unicode codepoint map.
Here is an analogy which might help. Suppose that a building company is building a new housing estate. On the plans the building company numbers the houses. The building workers refer to the houses by those numbers. They are just internal numbers used by the builders.
When the houses are to be lived in by people, they need addresses for mail deliveries, deliveries by grocers, and so on. The builders do not allocate those addresses, they are allocated by the Local Authority of the Government. The numbers used by the builders are of no interest to the mailperson delivering mail, the grocer making deliveries, and so on.
The analogy may not be perfect, but I hope that it will give an idea of the fact that the glyph index and the Unicode codepoint are two different things.
For when you press just one key, that might well be correct.
However, some software packages allow entry of Private Use Area characters, as well as regular Unicode characters, by special methods. For example, Microsoft Word has an Insert Symbol facility. Serif PagePlus has a similar facility.
There is a method, sometimes called the Alt method, which is used in Microsoft WordPad, whereby if one holds down the Alt key and one then keys the decimal value of the Unicode codepoint using the digit keys which are at the right of the keyboard and then releases the Alt key, the character will be included in the document. For example, please consider U+E001. Using Microsoft Calculator in View Scientific mode one can calculate that the decimal value is 57345 so using Alt 57345 causes a character to be included in the document. It may well show as a black rectangle in this post as the font may not contain a glyph mapped to the U+E001 codepoint.
However, sometimes it can be useful to produce a non-Unicode font for a special purpose. However, if a codepoint position which Unicode uses for, say, D, is used for something else, then it cannot be used for a glyph for a letter D.
Yet it can be a useful technique if one wishes to use a font for some special purpose and the program will not accept Private Use Area characters.
For example, keying a Private Use Area character directly into Microsoft Paint using the Alt method is not possible, though I have just found out that keying it into WordPad and then copying and pasting the character works in the version of Paint which I am using. So, when I made some experimental fonts some time ago I added the glyphs into both the Private Use Area and some of the places normally used for the English alphabet.
The fonts are linked from the following thread. If opened in FontCreator with Postscript name chosen in the Tools Options… Overview Caption section, this may be helpful in observing how one person used the Private Use Area for his own project.
http://forum.high-logic.com:9080/t/view-augmentation-logos/1811/1
I recently saw an interesting font which might help to illustrate the use of a non-Unicode glyph within an otherwise Unicode font.
http://fontstruct.fontshop.com/fontstructions/show/71705
In that font, the left-side part of a heart shape is encoded where Unicode places a ~ character.
I hope that this helps.
William Overington
4 August 2008