One case where patience diff does much better

[John Yates]

John A Meinel wrote:
> 
> John Yates wrote:
> >
> > Before Bram Cohen identified the patience connection I posted a
> > description of the DSEE / ClearCase diff algorithm to the darcs
> > mailing list:
> > 
> >   http://www.abridgegame.org/pipermail/darcs-users/2005-April/006561.html
> > 
> > One feature I remember especially fondly was a merge tool could
> > explicate a diff in terms of moves as well as inserts and deletes.
>
> I know that Bram has made the comment that tracking moves was not worth
> the effort. I don't know if he is correct. But I can see how it could
> vastly complicate things.

Bram's opinion surfaced in  email I exchanged with him.  But in this
instance my point was not to advocate tracking moves.  I agree that
to do so would vastly complicate the weave/knit/whathaveyou format
and the corresponding algorithms.

The virtue of identifying moves is not so as to record them but to
present them in the ui of a difference viewer or reconciliation tool.
A move identified during differencing can always be recorded as a
delete/insert pair.

/john