Use of UTF8 in X calls

[Fred Kiefer]

Hi Richard,


you did ask in the Changelog, if the switch from cString to UTF8String 
is correct for the X backend. I have been thinking about that change 
myself for some time and did not dare to do it. According to my Xlib 
Programming Manual (from 1995) the functions we are calling here only 
support the "Host Portabl Character Encoding", which is a bit more than 
ASCII and the client and the server have to agree on it. I would expect 
newer X implementations to be able to handle UTF8, but could not find a 
prove for this. Most of our current users wont have a host name, 
application name or parameter that is not inside of the supported range. 
And for the ones that use non-Latin1 (which is one of the more popular 
HPCE) cString would have failed them anyway and we now have at least the 
chance that their X supports this strings.
So from a theoretical point of view that change might be incorrect, but 
from a practical it is a clear improvement. In the long run we will have 
to question each single use of cString in GNUstep. It is almost gone 
from GUI already.

Cheers
Fred