[petsc-dev] horrible lies about PETSc that need to be fixed

[Smith, Barry F.]

> On Feb 4, 2018, at 10:37 PM, Fande Kong <fdkong.jd at gmail.com> wrote:
> 
> 
> 
> On Sun, Feb 4, 2018 at 7:13 PM, Smith, Barry F. <bsmith at mcs.anl.gov> wrote:
> 
>    How dare you defame us in such a way. PETSc never ever "needed" cmake and certainly never will
> 
> This is because we build PETSc with Metis and Parmetis (by default) that requires CMake. We usually need Metis/Parmetis for partitioning an unstructured mesh. 

   Sure but that does not mean that PETSc needs CMake, see the better wording 

	• CMake. A modern version of CMake (>2.8) is required to build some of the meta packages we need to include in PETSc.

       http://mooseframework.com/wiki/ClusterInstructions/ClusterSetupSingleUser/

       Probably a lot of text for the various cases could be combined to a single case, resulting in less text to maintain.

> 
> 
>  
> 
> 
> http://mooseframework.org/wiki/BasicManualInstallation/OSX/#0-pre-reqs
> http://mooseframework.org/wiki/BasicManualInstallation/Linux/
> 
> 
> 0. Pre-Reqs
> 
>         • CMake. CMake is required to build PETSc.
> 
> We could say CMake is required to build Metis/Parmetis?
> 
> 
> Fande,
> 
>  
> 
> 
> ;-)