On Dec 10, 2007, at 10:31 AM, Matthew Knepley wrote: > I do not remember mail about this or I would have complained. You > talked > about moving other things.
It was an earlier mail > Python code nuder the 'python' directory makes > perfect sense to me. How the hell would I know what is in 'config'. > Also, > not all of the python code is configuration code. If there was python code that was not configuration/build code then I apologize about messing up that code. As far as I could see at that time python had the config code plus some no longer used code for generating python PETSc classes. I definitely believe all the config code should be in the same directory with a name that indicates it is config. Putting code in a directory named after the language it is written in makes no sense to me, the directory name should indicate what the code/ directory is FOR; so real people can get around the directory tree. Barry > > > Matt > > On Dec 10, 2007 10:16 AM, Barry Smith <bsmith at mcs.anl.gov> wrote: >> >> I sent email about a week ago and no one complained. >> >> As someone downloads petsc-dev they see directories config and >> python; >> this make NO SENSE at all, how are they going to know that most of >> the >> configure >> stuff is in the python directory? If I did not know PETSc and >> downloaded it >> I would be so pissed at the utterly stupid naming that I would simply >> delete the >> whole package and not even consider using it. Directory names are not >> for the developer (a developer can remember anything so long as the >> work on the package long enough), directory names are for the causal >> user and it is important that they make sense. >> >> Why does your finite element stuff care where the configuration >> code is stored???? At most you should have to change a couple little >> things. >> >> >> Barry >> >> >> >> On Dec 10, 2007, at 10:08 AM, Matthew Knepley wrote: >> >>> How does 'config' make any more sense than 'python'. And this >>> break EVERYTHING I do with finite elements. This should have at >>> least been discussed on petsc-dev. >>> >>> 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 >>> >> >> > > > > -- > 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 >