I recently came across a blog post by Prof. Amy J. Ko on how tenured faculty can accelerate the demise of tenure. Applying the same sardonic tone, here are five ways you can ensure OpenSees will also be gone sooner rather than later.
1. Don’t learn – A firm grasp of linear analysis is sufficient to get you through nonlinear analysis. Furthermore, data structures, algorithms, and programming are for computer scientists, not earthquake engineers. OpenSees is wrong for not doing what you think it should do and difficult for not making what you want to do more simple.
2. Don’t evolve – What you’re doing now is fine. OpenSees version 2.x.0 with a bespoke MATLAB post-processor gets the job done. Your scripts are inefficient, but they work. What’s the incentive to change?
3. Don’t help others – You definitely made OpenSees your own all on your own.
4. Don’t say thank you – Employees at the OpenSees call center should not be compensated for simply walking you through a diagnostic decision tree. A blanket “Thanks in advance” will suffice.
5. Don’t respect intellectual property – OpenSees is open source, so you can do whatever you want with it and any other associated products. Copyright laws and acknowledgments only apply to real creations like books, music, and commercial software.
If we make a concerted effort to do all of these things, we will continue to function in silos and never help anyone do anything with OpenSees while discouraging others from sharing their work. After all, when someone else figures out how to do something before you, that’s one less bean you can count.
As Prof. Ko says in her post, in reality, it is hard not to do these things. Learning can be frustrating. Evolving exposes you to risk. Helping others takes time. To some, saying thank you too often shows weakness. Understanding intellectual property can be difficult.
But all these things matter.