The other day, David Mertens asked about using Pinto & Stratopan to help manage the Perl modules he uses for physics research at Northwestern University. I had never really thought about it before, but this turns out to be one of the best use cases I could ever imagine.
Reproducibility is a big concern for scientists like David. Most projects have a revolving crew of graduate students, and everyone needs to run the same experiments and get the same outcomes. Also, other scientists will verify the work and must be able to independently reproduce the results. “It works on my machine” is definitely not an acceptable defense in an academic peer review.
So when David publishes a paper on synchronization and stimulated emission in an array of mechanical phase oscillators on a resonant support, he needs to provide his colleagues with the same tools he used. And that’s exactly what Pinto and Stratopan are designed for.
David can create a Pinto repository with all the Perl modules he used for his research. When new grad students join the project or another scientist attempts to verify his work, he can tell them to install modules from his Pinto repository (using
cpanm, or whatever they like). They will get precisely the same versions that David used, rather than whatever is on the public CPAN or in his VCS at that moment.
Many research projects go on for many months or even years, and David can use a Pinto repository to manage change over time. If he publishes a revision or a does related study using newer modules, he can create a separate stack in the Pinto repository with a different modules while still preserving the old stack.
Stratopan will take all of this one step further by hosting the Pinto repository in the cloud. That way, David can share his repository with anyone, anywhere, and without having to install any local software or maintain a server.