Re: [PATCH] Make git-branch a builtin

From: Petr Baudis <pasky@suse.cz>
Date: Oct 24, 2006, 11:39 UTC
Message-ID: <20061024113924.GC20017@pasky.or.cz>
In-Reply-To: <7vy7r6qkmg.fsf@assigned-by-dhcp.cox.net>

Dear diary, on Tue, Oct 24, 2006 at 08:43:35AM CEST, I got a letter where Junio C Hamano <junkio@cox.net> said that...

Show 17 quoted lines

> Shawn Pearce <spearce@spearce.org> writes:
> 
> >> > Wouldn't it make more sense to just include the full SHA1 of the
> >> > file we are deleting rather than the entire 131 line negative diff?
> >> 
> >> How would you do "git apply -R" on something like that?
> >
> > Uh, you have the full SHA1 in the index line.  So you just have to
> > reattach that blob to the named path...  pretty simple actually.
> 
> Bzzzt; wrong answer.
> 
> Think of a future when you can shallowly clone near the tip of
> git repository that does not have shell-script git-branch.sh
> anymore.  You cannot expect to already have the preimage of the
> patch in such a case.  You would still want to be able to revert
> the change with "git apply -R".

Hmm, how is this argument not applying to binary diffs you can't revert either?

-- 
				Petr "Pasky" Baudis
Stuff: http://pasky.or.cz/
#!/bin/perl -sp0777i<X+d*lMLa^*lN%0]dsXx++lMlN/dsM0<j]dsj
$/=unpack('H*',$_);$_=`echo 16dio\U$k"SK$/SM$n\EsN0p[lN*1
lK[d2%Sa2/d0$^Ixp"|dc`;s/\W//g;$_=pack('H*',/((..)*)$/)