Re: Font encoding vs. string encoding

[Fred Kiefer]

I am normally very much in favour of configuration files, bt for this 
one I can see no reason. Why should anybody want to remap iso8859-1 to 
anything else? It could be usable to add new aliases for already defined 
encodings. Yes, this might come in handy. But then again are there that 
many aliases used for font encodings? As Serg and Richard pointed out a 
different name is more likely referring to a different encoding, which 
than again needs to be supported by GNUstep. On the other hand the 
current code in GSFontInfo supports exactly all the encodings on my 
machine, so it is easy for me to think this is enough.


Sir Raorn wrote:

> Going back to my problem with mostCompatibleEcoding, I've
> looked into helvR12k.pcf
> (-rfx-helvetica-medium-r-normal--12-120-75-75-p-67-koi8-r,
> according to fonts.dir) and I saw, that CHARSET_ENCODING
> really set to "1". This is surely font's bug.
>
> I think best solution is not to use hardcoded mappings from
> reg-enc to NSEcoding, but congurable dictionary like the
> folowing:
>
> {
>   "iso8859-1" = "NSISOLatin1StringEncoding";
>   "iso8859-2" = "NSISOLatin2StringEncoding";
>   ...
>   "big5-1" = "NSBIG5StringEncoding";
>   "iso10646-1" = "NSUnicodeStringEncoding";
>   ...
>   "koi8-r" = "NSKOI8RStringEncoding";
> }
>
> It can be a file named for example Font.aliases and located in
> $GNUSTEP_ROOT/Libraries/Resources/Fonts. And of course it can
> be overriden by user's config just like DefaultKeyBindings.dict
>
> I can take care of it this weekend. Waiting for your comments
> and suggestions.