Re: AC_PATH_PROG ignores the variable

[Alexandre Duret-Lutz]

>>> "Ralf" == Ralf Corsepius <address@hidden> writes:

[...]

 Ralf> An ambiguity in the doc's wording.

Seconded.  I believe all we can do in this area is it to change
the documentation to reflect what the code actually does.  Any
change to the code will break one setup or another.  (For
instance the suggested change will break Automake's
AM_PATH_PYTHON.)

The current behavior does not seem really far from the
documentation anyway.  AC_PATH_PROG is about computing the
absolute pathname of a maintainer-supplied program; presently if
the user wants to override the result of the check s/he must
give a plausible result, i.e., an absolute pathname, otherwise
the check needs to be performed.  The doc only omits this last
constraint.

Besides the code breakage, there is another reason I'm not fond
of the suggested change.  Basically it says that when the output
variable is an absolute path it overrides the output of the
test, but when the output variable is not absolute it overrides
the *input* of the test.  This I/O mix sounds awkward to me.

I'd rather keep the current semantic where setting an output
variable can only affect the output: either the test uses the
input of the maintainer, or it skips the test and uses the
output supplied by the end user.
-- 
Alexandre Duret-Lutz