Re: I want "fast forward my workdir to upstream if it's safe"

From: Eyvind Bernhardsen <eyvind-git@orakel.ntnu.no>
Date: May 11, 2009, 20:11 UTC
Message-ID: <EDAB1A7F-628A-47F6-BAEB-89935E86546A@orakel.ntnu.no>
In-Reply-To: <7vk54rvb8x.fsf@alter.siamese.dyndns.org>

On 8. mai. 2009, at 17.57, Junio C Hamano wrote:

Show 11 quoted lines

> Eyvind Bernhardsen <eyvind-git@orakel.ntnu.no> writes:
>
>> I read it as a request for pull/merge --ff-only (based on "abort if
>> the workdir is dirty or is not a fast-forward update").  This feature
>> has been implemented twice, but never included:
>
> Interesting.  Do you mean twice they were both found lacking,  
> substandard,
> useless, uninteresting, buggy, incorrect, or all of the above?  Or  
> was it
> just somebody simply forgot to apply a perfect patchset twice?

Sorry, I should have summarised.

I guess Yuval explained what happened to his patch himself. Sverre's was dropped after a fair amount of discussion and several rounds, but his series added multiple fast-forward strategies, not just "ff-only".

The last reference I can find to it in the archives is Dscho asking for the feature to be held off pending the builtinification of git- merge (http://article.gmane.org/gmane.comp.version-control.git/ 82623). Sverre probably lost interest in following up after that.

It looks like there's some interest behind the "--ff=only" feature now, at least. I remembered the thread because of it; I wanted to replace a "git reset --hard" in my workflow with something slightly safer, and --ff=only would have fit.

-- 
Eyvind Bernhardsen