On Wed, Oct 23, 2019 at 11:24 AM Faibussowitsch, Jacob via petsc-dev < petsc-dev@mcs.anl.gov> wrote:
> Hello All, > > How feasible would it be to accept user input during the configure process > when installing a new package using —download-package when the dependency > is missing? For example say you are configuring with the following options: > > ./configure —download-pragmatic > > Which depends on Metis, so halfway through configure you will receive the > following error: > > > ******************************************************************************* > UNABLE to CONFIGURE with GIVEN OPTIONS (see configure.log for > details): > > ------------------------------------------------------------------------------- > Did not find package METIS needed by pragmatic. > Enable the package using --with-metis or --download-metis > > ******************************************************************************* > > Why can’t there instead be a prompt that states “[Package] depends on > [Dependency], would you like to also download [Dependency] (y , n)?” > whenever PETSc detects that you are missing a dependency. Configure would > then restart from the appropriate section (wherever that may be) in the > configuration process instead of all the way from the beginning. Use case > for this would be for configuring on systems such as Theta where the > configure process takes roughly 13.772 billion years to complete. Having to > restart from the beginning every time is tedious. > > As I am largely unfamiliar with the internals of the configure process, > this is potentially more of an involved change than I am imagining, given > that many libraries likely have many small dependencies and hooks which > have to be set throughout the configuration process, and so its possible > not everything could be skipped. > We had this many years ago. It was removed because the benefits did not outweigh the costs. Thanks, Matt > Best, > > Jacob > > -- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener https://www.cse.buffalo.edu/~knepley/ <http://www.cse.buffalo.edu/~knepley/>