Could you be more specific as to the "environment-modules" packages?
Do these produce a stand-alone executable, or do these require to be installed on both the development and target environments (machines)?
Part of the issue is having the environment for a build requiring versions or entire packages that are not available in the "stock" distribution, but part is having the environment available at run-time for the built program.
Does this also work for buildable environments that require libraries incompatible with the "stock" distribution, particularly after the build during run-time?
Yasha Karant On 10/16/2013 01:28 PM, Jeffrey Anderson wrote:
It seems to me that the "environment-modules" package available from EPEL can address many of your needs. We use it to provide access to multiple versions of python, gcc, boost, gsl and many other tools and libraries, all without interfering with necessary system-level operations. It is quite flexible, has a relatively low learning curve, and is widely used. It allows users to easily switch back and forth between different versions of these tools. Need python 2.7.2? 'module load python/2.7.2' Want to switch to 2.6.4? 'module switch python/2.6.4' Want to go back to the default system python? 'module unload python' and so on. Jeff On Wed, Oct 16, 2013 at 1:11 PM, Yasha Karant <ykar...@csusb.edu <mailto:ykar...@csusb.edu>> wrote: On 10/16/2013 09:12 AM, Jim Fait wrote: I run into this all the time, as we have a large number of somewhat incompatible software packages that we are required to have. What we have ended up doing is placing the real executable somewhere outside the normal path, and then putting a script with same name in /usr/local/bin or /opt/local/bin that encapsulates all of the foreign dependencies and environment. That way, the particular package can live with its requirements alongside the production system, with very few problems seen by the end user. Of course, this means writing a number of scripts, in our case a couple hundred, that stay fairly static with changes in the OS or the package in question, and that hide all of the nastiness that otherwise would happen, like a PATH environment variable 10 line long. Hope this idea helps. Jim Your example is one of the accepted methods for enabling the idea of polymorphism and encapsulation within an otherwise procedural imperative, possibly structured, environment. Assuming that what you are describing is for both the build environment and the execution environment of the environment/application being built, it should (in most cases) work. And -- it should be the norm when providing application building environments that do NOT require a virtual machine (e.g., maintaining a "more modern" Linux under VirtualBox under SL6x). I do not know which applications/environments you support in this way. A list of all that are not subject to for-fee or equivalent non-distributable licenses and for which you are willing to provide the scripts greatly would be appreciated. Maintaining such environments across new major OS environment releases often entails a large amount of effort. Yasha -- -------------------------------------------------------------- Jeffrey Anderson | jdander...@lbl.gov <mailto:jdander...@lbl.gov> Lawrence Berkeley National Laboratory | Office: 50A-5104E | Mailstop 50A-5101 Phone: 510 486-4208 | Fax: 510 486-4204