Re: Patches for OS/2 (1)

[Andreas Buening]

Bruno Haible wrote:
> 
> On Wednesday 13 November 2002, Andreas Buening wrote:

[snip]

> > > It is wrong to make libintl depend on additional libraries.
> >
> > LIBS might be used to add extra libraries that extend the libc of
> > the system. You may have noticed that Andrew Zabolotny (the former
> > OS/2 porter) had written a libiconv wrapper for the OS/2 NLS API.
> > If I want that gettext uses that lib I use LIBS="-luse_native_iconv"
> > so that configure doesn't even try to find any other silly libiconv
> > that might be available.
> 
> That's convincing enough that I've added $(LIBS) to intl/Makefile.in.
> But it would be cleaner if either
>   - Andrew Zabolotny's library were called usable with "-liconv",
>     this would mean no changes to iconv.m4 are needed, or
>   - if it is called differently, you provide an iconv.m4 patch that will
>     make use of this library if present and if the "real" libiconv
>     is not found.

I'd like to keep it up to the user because
- there may be reasons when it's prefereble to use the native OS/2 API
  or the GNU iconv.
- the code in iconv.m4 can't know whether the user wants to use the
  native OS/2 API or GNU iconv for this purpose.
- If a package uses an external (shared) libintl then -liconv is
  redundant because libiconv is already linked into the shared intl
  library.

So I think the current behaviour is okay: -liconv is choosed if iconv()
is not present in the libc (or in $LIBS) and if libiconv is available.

[snip]


Bye,
Andreas