[petsc-dev] metis config error on XT4

[Matthew Knepley]

On Wed, Nov 30, 2011 at 4:49 PM, Barry Smith <bsmith at mcs.anl.gov> wrote:

>
> On Nov 30, 2011, at 4:44 PM, Jed Brown wrote:
>
> > On Wed, Nov 30, 2011 at 16:40, Jack Poulson <jack.poulson at gmail.com>
> wrote:
> > No, but perhaps I should. My understanding is that those lines only need
> to be changed when ParMetis is included as a subproject, as they seem to
> have configured the build system to use relative paths.
> >
> > That's a misuse of CMake. Normally you build from a separate build
> directory (so that you can clean reliably and independently between
> configurations). That ParMetis misuses CMake by default isn't an excuse for
> not allowing it to be used properly.
>
>   A well designed buildsystem would not allow itself to be misused.
>

I am adding code to BuildSystem to electrocute users who try to screw with
it.

   Matt

-- 
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
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20111130/516fe8e9/attachment.html>