[XeTeX] Accented gliphs

Mojca Miklavec mojca.miklavec.lists at gmail.com
Thu May 15 15:11:14 CEST 2008


On Thu, May 15, 2008 at 2:43 PM, Ulrike Fischer wrote:
> Am Thu, 15 May 2008 12:59:38 +0200 schrieb Mojca Miklavec:
>
>> But the next question raised usually turns out to be, why doesn't
>>    \usepackage[T1]{fontenc}
>> work as expected. And those old hacks are never going to work in Xe(La)TeX.
>>
>> Convincing maintainers of fontenc and inputenc to fix the packages to
>> work with XeTeX (in the sense that they would simply be ignored in
>> XeTeX)
>
> While it is ok if inputenc is ignored in XeTeX, disabling fontenc is
> absolutely not a good idea. E.g. it will break all the chess fonts.

I don't know how chess fonts work, but I meant not necessary disabling
it, but at least ignoring the "T1" the part which says that \ss should
map to this and that position in the font (there is probably some
other character at that position anyway).

But to be honest - it's better if users complain that it doesn't work
now, and fix things, than trying to make them believe that T1 is still
needed and let them happily widespread their XeTeX code with
\usepackage[T1]{fontenc}.

inputenc could use \XeTeXinputencoding primitive, but again: I'm not
sure that it's a good idea to support that on high level. One should
get rid of old encodings as soon as possible anyway.

> Also there is no real problem in using T1-encoding (or other encodings)
> together with XeTeX as long as you don't use non-ASCII-input which will
> break in a lot of cases -- not because of some deficiency of fontenc but
> because of the missing inputenc-step.

Not only that. Even if ß, ÿ and every other accented character would
be mapped to \ss or other corresponding slot in T1 encoding, at least
the hyphenation patterns would still be broken.

> \documentclass{article}
> \usepackage[T1]{fontenc}
> \begin{document}
> ß \ss
>
> ÿ \"y
> \end{document}

Mojca


More information about the XeTeX mailing list