Font problems in BAN logic encoding

Some of the characters used in the BAN.jt encoding aren't in widely-available Unicode fonts. All the machines Bernard and I have tested are missing or not correctly displaying at least the symbols used in the first four menus (believes, sees, says, and whatever the bar-with-double-arrow is called). That's true even when using the best Unicode fonts we can find (see below), and on all the platforms we know about: Linux, Windows, Solaris and MacOS X.

On MacOS X (which I know best because that's where I develop Jape) at least two of the glyphs are garbled, and font-fall-through somehow makes lines appear below their designated position.

So when am I going to fix it?

These problems won't be properly fixed until there is a widely available freeware Unicode font with full glyph coverage. My hopes are pinned on the STIX font (see their home page for information on their progress -- but beware! they are two years behind their first promise of completion). Watch the skies!

How about a workaround?

BAN_ASCII.jt uses multi-character sequences to represent the BAN glyphs. Proofs you produce with it won't be readable under the normal BAN encoding (but if you ask, I guess I could produce a translator at short notice). 'Believes' is |=-; 'sees' is <|; 'says' is |~; and the other thing is |=>.

What's the best Unicode font, then?

The best I know for Jape purposes is distributed by Microsoft and it's called Lucida Sans Unicode. It is sometimes available on the web (Google lsansuni.ttf). If you have Windows you probably already have it. Otherwise, even if you download it, you need a Windows license, or perhaps an old Word license or two would do. Let conscience be your guide; roll on STIX to wash away our sins.

Email me!

If the BAN_ASCII encoding is still a bit too rich for your machine, send a screenshot to bugs@www.jape.org.uk, and I'll go the extra mile for you. But no complaints about split infinitives, if you want a helpful reply.

Richard Bornat 9.vi.2005