Re: RFA: next gettextization step

[Andrew Cagney <ac131313@cygnus.com>]

> Index: ChangeLog
> from  Tom Tromey  <tromey@redhat.com>
> 
> 	* main.c: Marked all strings with _().
> 	* configure: Rebuilt.
> 	* configure.in (PACKAGE): New subst.
> 	(AC_OUTPUT): Create po/Makefile.in and po/Makefile.
> 	(SUBDIRS): Added po.


> 	* po/gdb.pot: New file.
> 	* po/POTFILES.in: New file.
> 	* po/.cvsignore: New file.
> 	* po/Make-in: New file.

BTW, I've hit problems doing GDB releases in the past due to these files 
and their makefile rules (unfortunatly I've lost the details) - from 
memory a make clean was zapping the files and it tried to do strange 
things when the source was on a read-only file system.  I note H.J. is 
currently raising what look like the problems I encountered on the 
BINUTILS list.

Anyway, looking in the opcodes directory, the CVS repository appears to 
contain the following generated files:

	opcodes.pot (aka gdb.pot)
	*.gmo

I'm wondering if it would be better to handle these generated files the 
same wah as for LEX and YACC output - the release/snapshot process 
generates those files.

thoughts?
Andrew