[Bug-gnubg] Translation issues again

[Petr Kadlec]

Hi all, it's me again...

I still have those problems with the charset in typewriter-styled text. I 
have tried to use the gtkrc file, set the encoding in the language code, 
use UTF-8, windows-1250, and iso8859-2, all with completely no success. (I 
have found that gettext probably converts all messages to some proper 
charset, as it is all the same whatever charset I use for the messages. 
Nevertheless, I tried to use manually converted messages, just for sure.) 
Everything runs fine in either case, EXCEPT those two dialogs: Eval and 
Gammon values, which share the typewriter-styled font, and where some 
accented characters are omitted.

After that attempts, I tried to search the source of gnubg for font names. 
I found these (not guaranteeing completeness/correctness):
-b&h-lucidatypewriter-medium-r-normal-sans-14-*-*-*-m-*-iso8859-1
-B&H-Lucida-Medium-R-Normal-Sans-12-*-*-*-*-*-*
-*-times-medium-r-normal-*-64-*-*-*-p-*-iso8859-1
-*-helvetica-medium-r-normal-*-8-*-*-*-p-*-iso8859-1
-adobe-utopia-bold-r-normal--[~%d 0 0 ~%d]-*-*-*-p-*-iso8859-1[49_52 54 56]
-adobe-utopia-bold-r-normal--[0 %d ~%d 0]-*-*-*-p-*-iso8859-1[49_52 54 56]
-adobe-utopia-bold-r-normal--%d-*-*-*-p-*-iso8859-1
-B&H-Lucida-Bold-R-Normal-Sans-12-*-*-*-*-*-*
-B&H-Lucida-Medium-R-Normal-Sans-12-*-*-*-*-*-*
-Misc-Fixed-Medium-R-Normal--12-*-*-*-*-*-iso8859-1

As it seems to me, the problem may be caused by the fact, that some of the 
font names have iso8859-1 hardcoded, so that gtkrc does not affect them, or 
am I wrong? Wouldn't it solve the problem if those iso8859-1 were changed 
to * ?

If that is the case, I would also ask to search the source for all 
"iso8859-1" (and "iso-8859-1") strings and change them to either "*" in the 
case of GTK fonts, or let them be configurable, as in the case of HTML 
export, where <meta http-equiv="Content-Type" content="text/html; 
charset=ISO-8859-1" /> is always output, no matter what language+charset 
gnubg uses. The simplest solution is IMHO to use gettext on that 
"ISO-8859-1" string, which allows the translator to change it to whatever 
charset he/she uses.

P.S. In the unlikely case that it could help, I have attached a screenshot 
of the problem. It shows the Gammon values window; the left half of the 
window is completely correct, see e.g. the "Bobr~i dovoleni" (r~ means r 
with hacek) checkbox. The right half of the window is screwed up. The first 
line should read "Hodnoty pro gammon ve hr~e o peni'ze". That same r~ 
character which is correct in the left half is missing in the right half.

Regards,
Petr Kadlec

bad_font.png
Description: PNG image

--------------------------------------------------------------
Petr Kadlec, student
Faculty of Electrical Engineering
The Czech Technical University
address@hidden, ICQ #68196926