[darcs-users] Why darcs? What does the 'physicists proof' actually prove?

[Alex Shinn]

The following paper is old but an interesting read.  They define a 
selective undo operation (e.g. to undo just your changes from a group 
project) by defining inverse operations on individual changes and 
commuting where possible to "rotate" the change you want to undo to the 
front and then applying it's inverse.  It's essentially the same as 
what darcs does.

   http://www.eecs.umich.edu/~aprakash/papers/undo-tochi94.pdf

-- 
Alex