Re:Practices in source code sharing in astrophysics
by ylem on Friday April 26, 2013 @12:54PM
(#5613942)
I'm sorry, but this attitude is just wrong. In order to trust results, scientific source needs to be verified. I'm a physicist and I've seen people not share code for a few reasons:
1) The code smells bad--I've had code that's useful, but that I feel is really ugly--but even that I've released when people have asked because it's better than having it just sitting around
2) To gain papers--for example, once I needed maxent for something and a guy wouldn't release his source and wanted to run that portion of the problem to be a co-author--so, I just wrote my own
3) Competitive advantage--like the OP says, he wants to keep an edge by having code that other people don't. Personally, I prefer to keep an edge by doing better science. I think if you're coming up with software that others use, then you still get some indirect credit from it, but I think that things progress better if the code is available and everyone isn't wasting time reinventing the wheel--it also helps for error-checking.
[ Parent
| on HN ]
There are no comments on this comment yet.