OK I sent each of you a new version of the lib.
There are three ways to refer to a character.
1) using a char --- for ease of use (this is the common case) - should support
all ascii characters
2) By code value. User may change the default to some other code and use that
if it is convenient.(or neccessary).
3) By glyph index. Ultimately this is the important one. The others are
provided for convenience only.
I added a couple more conversion routines. These are to allow the user to
convert what they have to glyph indices and then use the (faster) *Glyph
routines.
Some issues you all brought up:
1) Using long for character codes. Using unsigned long is a possibility. I
kind of like the specificity of uint32 though. The user knows exactly whats
going in (platform independent). Chris you are free to change my mind on
this. I won't use signed long, though.
2) The if(code > (2^ 16)) trick. This makes me uncomfortable for the
reason Chris mentioned. A code has 32 bits. I want to assume the user is using
all 32 of them.
3) The whole font server API business, independence from Freetype and
generalization of fonts. There are pros and cons to making such a library. My
concern is whether or not this particular library is useful. I think it is.
It could even be used to make the library Chris was speaking of. I wanted to
make an api that allowed Freetype to be easily used in conjunction with GGI.
Anything more is beyond the scope of my original intentions.
4) The conversion library. I downloaded it and I have to say that it is
pretty neat. Maybe someday it'll make it into the standard linux distros. I
think it should. I don't see what stops an advanced user from dowloading this
library and using it in conjunction with ours. Again the functionality of char
conversion (although related and convenient) seems a bit beyond the scope of
what is intended here.
Try the libs I sent you. I think the changes are a good step forward. They
only happened because you guys cared enough to email about it.
As you can see differnent people have different ideas of how a library should
be made. Ultimately, the usefulness of a library determines it's value.
--
Lee Brown Jr.
[EMAIL PROTECTED]
