Track your scientific scripts with Mercurial

If you want to publish your scientific scripts, as Nick Barnes advises in Nature [1], you can very easily do so with Mercurial.

All my stuff (not just code), excempting only huge datasets, is in a Mercurial source repository.¹

Whenever I change something and it does anything new, I commit the files with a simple commit (even if it’s only “it compiles!”).

With that I can always check “which were the last things I did” (look into the log) or “when did I change this line, and why?” (annotate the file). Also I can easily share my scripts folder with others and Mercurial can merge my work and theirs, so if they fix a line and I fix another line, both fixes get integrated without having to manually copy-paste them around.

For all that it doesn’t need much additional expertise: The basics can be learned in just 15 minutes — and you’ll likely never need more than these for your work.²

Update 2013: Nowadays I include the revision of scripts I use in the name of their output files or folders, so I always know which version of my scripts I used to create some result.