two recent NEWS entries

[Bruno Haible]

Hi,

I'm trying to understand the NEWS entries for Autoconf 2.72.

1) I don't understand this paragraph:

  *** New macros AC_SYS_LARGEFILE_REQUIRED and AC_SYS_YEAR2038_REQUIRED.
    These act like AC_SYS_LARGEFILE and AC_SYS_YEAR2038 respectively,
    except that they require large-file and year-2038 support respectively.
    As with AC_SYS_YEAR2038, application and library builders should
    configure consistently.

What does the "except that they require ..." mean?

Does it mean that the configure run aborts when the requirement cannot
me met?

Or does it mean that AC_SYS_LARGEFILE_REQUIRED requires large-file
support without year2038 support, and AC_SYS_YEAR2038_REQUIRED requires
year2038 support without large-file support?

Or what else?

Can the text be clarified, please?

2) I think this paragraph is a bit misleading:

  *** AC_PROG_MKDIR_P now falls back on plain 'mkdir -p'.

    When AC_PROG_MKDIR_P cannot find a mkdir implementation that is
    known to lack race condition bugs, it now falls back on 'mkdir -p'
    instead of falling back on a relative path to install-sh, as the
    relative paths now seem to be a more important problem than the
    problems of ancient mkdir implementations with race condition bugs.
    See <https://savannah.gnu.org/support/?110740>.  The only ancient
    mkdir still in common use is Solaris 10 /usr/bin/mkdir, and for that
    platform AC_PROG_MKDIR_P falls back on /opt/sfw/bin/mkdir which
    should work.

because it sounds like on Solaris 10 there is no problem. But actually
on the 3 Solaris 10 machines I have access to (including gcc210.fsffrance.org),
there is no /opt/sfw/bin/mkdir. There is only /usr/bin/mkdir. So, on Solaris 10,
parallel builds will be unreliable (i.e. prone to failures), and the advice
should be to avoid the option '-j' of GNU make on this platform.

Bruno