Introduction is a bit too narrow, emphasizes

[cboettig]

Intro has 5 paragraphs. Para 1 defines reproducible research w/ appropriate references (though maybe credit Clarebout with the phrase).

Para 2 discusses open science, feels a bit out of place. Git can be used without any openess at all. Need to be more clear why openness + git is particularly powerful for making science more reproducible, efficient, and scale-able.

1. reproducible science needs open science. This isn't a given for many people. (indeed, the ideal that I describe in general / pseudocode what I did, and you repeat it from scratch, has much to be said for it. If only it could scale beyond trivial exercises to practical use cases).

Para 3 discusses version control. Should make use cases clearer. Are we talking about publications, code, or something else (yes, we're talking about all at once, but that's a lot to handle in an intro. Perhaps start with the publication, the way it is traditionally version controlled with draft.doc, draft_v2, revised_draft_v2, revised_draft_v2_cb_edits, etc...

Para 4 emphasizes distributed version control. a bit technical for an intro?

Para 5 maps the rest of the paper, allbeit too vaguely. The paper itself has a really solid outline that divides the use of git into several arenas, (Which are mostly independent of each other and should progress from easiest to more difficult.) These should be named more clearly.