I'm creating a piece of software which will be used by in-house users. My code will all be written in pure Python; however, it depends heavily on a number of third-party Python modules, many of which have C/C++ dependencies (numpy, scipy, etc.) Installing these packages on my machine involved a morning of several serial "./ configure;make;sudo make install" steps. I'd prefer to automate all of this for my users, but I'm not clear how to do this - I'm in an environment of mixed Mac (desktops) and Linux (servers) and I'll need to install on both platforms, and would prefer not to run around to each computer making sure the dependencies are met.
I've looked through the documentation for distutils and setuptools, and it's not obvious to me that there are hooks in either one for passing in configure and make calls. The fallback, I suppose, presuming that the correct version of Python is already installed, is to write a custom Python script that "knows" about each dependency, and takes the appropriate action. I should be able to depend on everyone having gcc and make installed, I think. Does anyone have any suggestions on the best way to approach this? Thanks, Mike -- http://mail.python.org/mailman/listinfo/python-list
