Re: [Quilt-dev] [RFC] compatibility layer

[John Vandenberg]

On 9/18/05, Andreas Gruenbacher <address@hidden> wrote:
> On Sunday 18 September 2005 02:42, John Vandenberg wrote:
> > On 9/18/05, Andreas Gruenbacher <address@hidden> wrote:
> > > On Sunday 18 September 2005 02:10, John Vandenberg wrote:
> > > > The attached patch takes it one step further.  Each program
> > > > can have many functionality checks, and configure will fail, or fall
> > > > back, if any check fails.
> > >
> > > Sounds useful, but no attachment as far as I can see.
> >
> > Sorry; here it is.
> 
> Entirely removing the failure message doesn't sound like a good idea.

Do you want compat.m4 to use AC_MSG_ERROR when any of the tests fail?

> ... I think
> the reply to AC_MSG_CHECKING should be AC_MSG_RESULT(yes) or
> AC_MSG_RESULT(no), not AC_MSG_WARN. If the error message comes from outside
> the test, the calling wrapper can print the test result itself.

Using AC_MSG_RESULT(no) instead of AC_MSG_WARN(no) is fine for one
test, but $4 could include multiple tests.

When there are many tests that need to be run, 'no' may not represent
a failed test, so tests need to explicitly alert the user to which
ones are causing the problem.

Or, a cleaner approach would be for QUILT_COMPAT_PROG_PATH to iterate
over, and expand, each test in $4.  Then it will be able to
communicate with the tests via variables. The m4 to do that wont be
pretty.

--
John