Bug#176661: locale.1 does not mention LOCPATH
On Sat, Mar 08, 2003 at 01:20:29AM +0900, GOTO Masanori wrote:
> At Fri, 07 Mar 2003 23:51:21 +0900,
> Junichi Uekawa wrote:
> >
> > > > --- locale.1.orig 2003-01-14 21:30:42.000000000 +0900
> > > > +++ locale.1 2003-01-14 21:31:13.000000000 +0900
> > > > @@ -233,6 +233,11 @@
> > > > .Vb 1
> > > > \& Metadata about the locale information.
> > > > .Ve
> > > > +\&\s-1LOCPATH\s0
> > > > +.PP
> > > > +.Vb 1
> > > > +\& The directory where locale data is stored
> > > > +.Ve
> > > > .SH "AUTHOR"
> > > > .IX Header "AUTHOR"
> > > > \&\fIlocale\fR was written by Ulrich Drepper for the \s-1GNU\s0 C Library.
> > >
> > > It's easy to add this description, but why is this needed? Who do use
> > > it? Please explain to me.
> >
> > I've been looking for a way to specify the directory.
> > And this information is available elsewhere also.
>
> And, still my question is: why?
> No reason, no accept.
>
> Locale data should be the same exactly. Or are you hacking locale
> system or locale data? BTW do you know glibc-2.3.1 uses only
> /usr/lib/locale/locale-archive ?
>
> > If there is an alternative POSIX defined locale environment to do a
> > similar thing instead, then please document that.
>
> OK.
>
> > It is not obvious how to use a different directory than
> > /usr/lib/locale for the locale data from reading the manual pages.
>
> Well, it's not worth while describing.
I disagree. I've needed to use LOCPATH a couple of times. For
instance GCC's testsuite wants to use it to make sure some locales are
generated before running the libstdc++ tests.
--
Daniel Jacobowitz
MontaVista Software Debian GNU/Linux Developer
Reply to: