From: Johannes Schindelin Date: Thu, 29 May 2008 14:42:36 GMT Subject: Re: [PATCH] rollback index if git-commit is interrupted by a signal Message-ID: In-Reply-To: <483EBF1F.9000809@gnu.org> Hi, On Thu, 29 May 2008, Paolo Bonzini wrote: > > IMO it would make much more sense to _guarantee_ that the > > commity_style variable is set before the index is locked. It is > > feasible, and there is no good reason not to do that. > > No, it's not possible because the COMMIT_PARTIAL case first creates the > index_lock and then the false_lock. So why don't you fix _that_? > It would be curious at least to set the commit_style to COMMIT_NORMAL > after creating the index_lock, and upgrade it to COMMIT_PARTIAL later > on. I contemplated that, and my patch is the simplest code that's > needed and works. As I said, I think it is a regression, because you change code. Your argument as to why leaves me desiring another solution. Nuff said, Dscho