Re: The merge from hell...

From: Linus Torvalds <torvalds@osdl.org>
Date: Feb 2, 2006, 08:08 UTC
Message-ID: <Pine.LNX.4.64.0602020002110.21884@g5.osdl.org>
In-Reply-To: <Pine.LNX.4.64.0602012353130.21884@g5.osdl.org>

On Wed, 1 Feb 2006, Linus Torvalds wrote:

> 
> I _think_ your version of the rule ("pluses and minuses in all the same 
> columns") ends up being exactly the same as my rule ("only two different 
> versions").

Actually, I take that back.

My version of the rule was not just that there were only two versions: the merge _result_ had to match one of the versions in one of the branches too.

That's important. It's important because you _can_ actually have a merge where both branches have the exact same content, but some evil merger ended up editing just the merge result.

So there are literally just two versions, but one of the versions _only_ exists in the merge result. Such a thing must always be considered a conflict, since basically the merge result is obviously "interesting".

That requirement can still be rewritten as a slight variation of your version, I think: "the pluses and minuses are all in the same columns, and the final column has to match one other column"

Maybe I'm not thinking clearly. Somebody else should double-check my thinking. My brain is tired, and I'm going to bed.

			Linus