[RFC] generalized diff plugin - option name?

[Aaron Bentley]

John Arbash Meinel wrote:
> Aaron Bentley wrote:
> 
> I completely agree. I just didn't know of any source data to test it on.
> 
> I know in the past I found a bug or two by trying to convert all of
> bzr.dev to a cdv style weaves. So bzr.dev does exercise most of the code
> paths. (I do believe I added tests to prevent regressions).

Well, one obvious choice would be the entire bzr.dev ancestry.  'bzr cat
bzr -r1 > bzr; bzr diff -r1..2 bzr|patch; bzr cat -r2 bzr.new; diff bzr
bzr.new'

Aaron