On 8/27/07, "Danai SAE-HAN (韓達耐)" <[EMAIL PROTECTED]> wrote:
> On 26-08-07 16:25, Gernot Hassenpflug wrote:
> > On 8/26/07, Gernot Hassenpflug <[EMAIL PROTECTED]> wrote:
> >> I am happy to say I could successfully convert the msgoth.ttf font
> >> from Windows XP (Japanese) to work under LaTeX. This is a test case /../
>
> Cool! If you don't mind, I'll put your HOWTO in the Debian documentation.
Please do! If you know what the correct steps are to activate a new
map file, please modify this... I don't think the texconfig route
(which comes up with a menu where I select REHASH and then SHOW
CONFIGURATION) is correct. But if I don't do that, then updmap-sys
--enable Map msgoth.map does not find the map.
> > I did the same with the HGSoeiKakugothicUB.ttf (extracted from a .ttc
> > bundle with fontforge) and notice that I seem to have messed up
> > something: only the japanese characters are correctly chosen from the
> > font---it seems the english text is all still set using the default CM
> > fonts.
To check, I carried out the same steps on my desktop at work. Same result.
> > Now I have no idea why that might be: for CJK there is apparently no
> > fallback if a font cannot be found, so if the roman letters are not
> > found in the converted font, should not simply blanks be displayed (or
> > nothing set at all).
>
> Hmm, weird. I would have expected this kind of behaviour if you set
> \usepackage[T1]{CJKutf8}, but you didn't use that line or use
> \usepackage[T1]{fontenc} in your example file.
Yes...I don't know what encoding is supposed to be used here. Reading
the documentation does not make it clear to me yet what characters the
CJK package acts on.
> I don't access to my CJK GIT repository ATM, but it might be that CJK still
> uses T1 by default.
Hmm, I thought \begin{CJK}{UTF8}{} decided what encoding to use. Now I
am even more confused. Back to reading documentation...
Regards, Gernot
_______________________________________________
Cjk maillist - [email protected]
https://lists.ffii.org/mailman/listinfo/cjk
