Re: Strange LT_INIT behavior

[Eric Blake]

Peter O'Gorman <peter <at> pogma.com> writes:

> As far as I can tell, the AC_REQUIRE is unnecessary, in the case where
> AC_PROG_F77 comes before LT_INIT, the compiler has already been tested
> before we get to _LT_LANG_F77_CONFIG (this is guaranteed by the
> AC_PROVIDE_IFLESE) in the other case, the compiler check is also
> guaranteed because we have redefined AC_PROG_F77 so that it has both the
> compiler check and _LT_LANG_F77_CONFIG.

Your analysis seems correct to me - thanks for tracking it down.  In other 
words, by making libtool never directly expand AC_PROG_F77, while still 
guaranteeing that the libtool hook of _LT_LANG_F77_CONFIG will always occur 
after the user's use of AC_PROG_F77, libtool no longer gets in the way of 
arguments supplied by the user.

> Also, I have no idea what to do when removing macros from libtool.m4, do
> I need to do something with lt~obsolete.m4?

I think that only matters for publicly exported macros; since you are deleting 
_LT_PROG_F77 from the libtool internal namespace, no one outside of libtool 
should have been calling it, so there is probably nothing else to worry about.

All that said, I am not the biggest libtool contributor, so if you want to wait 
72 hours for anyone else to also chime in with a review, that would be fine 
with me.  If the three days pass without further comment, then consider my 
review as sufficient.

-- 
Eric Blake