Interesting Idea: Design Patterns for Scientific Software

OpenScience Blog on Design Patterns for Scientific Software.

Interesting Observation about Microsoft and Science

Kitware Blog post on What Scientific Journals Can Learn From Microsoft. Seems like scientists can learn from this too.

Supercomputing by country, app, processor, company, ...

Interesting distributions for Supercomputing power

Agreeable Quotes!

Quotes from Nature's

Computational Science ... Error ... why scientific programming does not compute. By Zeeya Merali.

"There are terrifying statistics showing that almost all of what scientists know about coding is self-taught," says Wilson. "They just don't know how bad they are."

"In the long term, though, Barnes says that there needs to be a change in the way that science students are trained."

"Science administrators also need to value programming skills more highly, says David Gavaghan, a computational biologist at the University of Oxford, UK."

"Gavaghan now uses the software industry's 'master–apprentice' approach to train graduate students in his lab."

And this one is really fitting if you work at a national laboratory:

"To all scientists out there, ask yourselves what you would do if, tomorrow, some Republican senator trains the spotlight on you and decides to turn you into a political football. Could your code stand up to attack?"

---

I couldn't agree with some of these quotes more .... see Where Is Software Development Really Learned?

Nature Article: Publish Your Scientific Code! Really?

Recently, Nick Barnes wrote a one page article in Nature encouraging scientist to publish their code pretty much as-is. He makes many good points especially ... if the results from your code are good enough to publish, then your code should be good enough to publish too.

My main concern is the possible unintended consequences that scientists will think they never need to improve their software developing skills since what they do is "good enough." When in reality, often, their skills are not good enough and that many scientists are publishing results that are not good enough due to the poor software they have developed. We do not want scientists to lose the sense of urgency to do a better job in writing good software that is a real scientific tool and that produces good science. Yes, I could see where once they publish their code and if peer review deems it flawed, or the like, that they might be embarrassed and want to do better.... but highly unlikely. Besides scientific software should be peered reviewed with code reviews and/or pair-programming just like any other type of software before results are published from it or the code is released to begin with.

Again, "Be a better scientific software developer today than yesterday!"