Alexander E. Patrakov patrakov at ums.usu.ru
Tue Aug 23 20:29:36 PDT 2005

Greg Schafer wrote:
> Hi
> The explanation in Ch 5 Binutils Pass1 for --disable-nls says:
> "This disables internationalization as i18n is not needed for the temporary
> tools."
> This is misleading because only the Pass1's of Binutils and GCC are passed
> `--disable-nls'. For the statement to be accurate, `--disable-nls' would
> need to be passed to all pkgs during Ch 5 that can handle it.
> It's a hangover from the days when the Pass1's of Binutils and GCC were
> statically linked. (NLS was known to break compilation on some hosts when
> linking statically).

No, it was added deliberately in order to avoid binutils dependency on 
the host gettext.

> You need to either:
>  1) remove --disable-nls from Pass1's of Binutils and GCC (and explanation).

No, see above. We don't want to require that gettext is installed on the 

>  2) fix the explanation to reflect reality.

How about:

"This avoids dependency on gettext being installed on the host. Since we 
set LC_ALL=POSIX during the entire chapter 5, NLS would not be used 
anyway even if enabled."

BTW there are still (broken) packages that don't build in the et_EE 
locale at all (because they put "z" in the middle of Latin alphabet).

>  3) pass --disable-nls to every pkg in Ch 5 that can handle it.

This would achieve nothing in reducing the host requirements.

Alexander E. Patrakov

More information about the lfs-dev mailing list