[Numpy-discussion] arrayobject.h (and friends) install location

[Sasha]

On 1/16/06, pearu at cens.ioc.ee <pearu at cens.ioc.ee> wrote:
> ...
> Basically, the main argument behind the current decision is that there are
> systems where people don't have permissions to install header files to std
> python include directory but they can install to python site-packages
> (e.g. when installing prebuilt distributions).
>

That's the argument that I mentioned.  It also seems to be not just
the "main", but
the *only* argument.  I am still not convinced.  What kind of systems
do you have
in mind?  If you are talking about Linux distributions that prepackage
python, there
is no reason for a sysadimin to let users write to
/usr/lib/pythonX.Y/site-packages
and not to /usr/include/pythonX.Y/. It is much easier to mess up a say Red Hat
system by placing wrong files into /usr/lib/pythonX.Y/site-packages
than by messing
with /usr/include/pythonX.Y/.  A site that would like to allow users
to install python packages should either recommend them to use
--prefix or provide an installation in
say /usr/local that has more liberal permissions to pythonX.Y directories.

> Also note that not all potential numpy users are able/willing/.. to use
> --prefix option. Why we should scare them away when we can provide
> defaults that work for all situations, though the defaults are not
> standard and require some tolerance from developers to follow numpy docs.

Personally, I can tolerate any location less that 10 levels deep under
install prefix
and current location is only 8 :-). However, if there is a valid
reason to change the location of include files, this functionality
should be implemented as a (possibly default) option to distutils, not
by treating includes as "data_files" rather than "headers".

If the alternative location is specified as an option in site.cfg and
is accompanied by a comment describing how to disable it, I am sure
there will be fewer posts asking where did the headers go in the
future.

-- sasha