[RFC] internal patch implementation

[Rafael Villar Burke]

Johannes Stezenbach wrote:
> On Wed, Jan 10, 2007 at 10:17:07PM -0500, Chris Mason wrote:
>   
>> 2) I'm not allowing patches with ^Ms when it doesn't apply to the source
>> files.
>>
>> I can add this stuff, but how much magic do we really want in the patch
>> code?
>>     
> I'd say don't do it.  patchutils has some tools to repair
> broken patches, e.g. unwrapdiff and dehtmldiff.
> I'm also uncomfortable with mpatch fixing up things
> behind your back. The user should be notified that the
> patch is broken.
>   
Maybe that's worth noticing in the mpatch docs (or its docstrings), so
people could discover by themselves the avaliable tools to solve these
problems.

Regards,

Rafael Villar Burke