[darcs-users] darcs patch theory

Ian Lynagh igloo at earth.li
Mon Sep 8 09:48:27 UTC 2008


Dear patch theory enthusiasts,

As most of you know, I've been working on a patch theory for darcs. The
new theory is similar to darcs 2's, but simpler, and should solve the
correctness issues.

There is now an implementation, minidarcs, of my proposed new theory. It
only implements a subset of the darcs commands, and in a restricted way
(e.g. "minidarcs record" non-interactively records all changes and
invents a patch name for you), but it should be enough to do tests and
experiments with.

Note that minidarcs is not suitable for use with real projects, comes
with no disclaimer or warranty, etc.

You can get it from this darcs repo:
    http://code.haskell.org/darcs/minidarcs/

(please don't push to that repo without talking to me first).


The theory description has also been slowly improving since my last
mail. The build-up, sections 1-7, should now be much more readable;
following discussions in #darcs-theory, I plan to add something about
minimal contexts, and perhaps also posets, which should help with the
proofs. Section 8, and in particular the definition of conflictor
commute, still needs some more restructuring. Formal proofs are still
conspicuous by their absence.

The PDF here has been updated:
    http://urchin.earth.li/~ian/theory.pdf

and LaTeX source is in the darcs repo here:
    http://urchin.earth.li/~ian/theory/


[please let me know if you don't want to be BCCed on future patch theory
mail]


Thanks
Ian



More information about the darcs-users mailing list