gituliar

One year is a quite long period; do you see some noticable changes since you started to fast: wieght loss, more/less active, etc?

How I love irony...

I think that no. GPL permits to keep one's changes private and doesn't force to make them public, that is bad in the case of scientific software. However, it forbids to distribute changes under the terms of other licenses, i.e. non-free, that is good.

I believe that an enforcement to share source code should be imposed on provate modifications of open-source scientific software.

I'm a particle physics theorist. In our field there are a lot of open-source (see http://www.hepforge.org/projects), as well as some private projects. I believe that source code should be published/shared in tandem with results it produces. Otherwise, those results have no scientific value and I personaly tend not to trust them, even though they are in agreement with well-known predictions. What matters is an algorithm results are obtained with, not results themselves. In other words, scientists should form a hacker-like community and don't get into the trap of a developer-user relationships where some develop while others use software.

I actually wanted to know your opinion concerning the migration of open-source to private software. The most common open-source and free software licenses, like MIT or GNU, allow this kind of migration. That is Bob could modify Alice's public project, produce results, and publish them without sharing sources with anybody. That looks unfair, since Alice did a great job and most likely would like to see the changes. How to protect Alice from such situations? Are there any kind of license which forces to share modified software when some results it produces are published or available to the general public?

What about WingIDE? Does it make sense to switch to PyCharm?

A key concepts beyond Mathematica are pattern matching and rule-based programming, but not a functional programming.

Yes, it has lambdas and first-class functions, but that's it. What about closures, for example? Therefore one would better not call Mathematica "a real functional programming language".

BTW, rule-based paradigm and pattern matching is a standard feature of languages from Lisp family, e.g. Scheme. From that point of view Mathematica is not something unique.