Re: Re: Re: [patch pasky] update gitcancel.sh to handle modes as well

[Petr Baudis <pasky@ucw.cz>]

Dear diary, on Fri, Apr 15, 2005 at 01:39:06AM CEST, I got a letter
where Martin Schlemmer <azarah@nosferatu.za.org> told me that...
> On Fri, 2005-04-15 at 01:15 +0200, Petr Baudis wrote:
> > Dear diary, on Fri, Apr 15, 2005 at 01:04:50AM CEST, I got a letter
> > where Martin Schlemmer <azarah@nosferatu.za.org> told me that...
> > > Rather use checkout-cache to sync our tree, as should do the right thing
> > > instead of diffing (cancel imply just blow away everything).
> > > 
> > > Signed-off-by: Martin Schlemmer <azarah@nosferatu.za.org>
> > > 
> > > gitcancel.sh:  839b3c58f20f6eb8412f499a891e007e2e67d114
> > > --- 839b3c58f20f6eb8412f499a891e007e2e67d114/gitcancel.sh
> > > +++ uncommitted/gitcancel.sh
> > > @@ -10,9 +10,8 @@
> > >  #
> > >  # Takes no arguments. Takes the evil changes from the tree.
> > > 
> > > -# FIXME: Does not revert mode changes!
> > > 
> > > -show-diff | patch -p0 -R
> > >  rm -f .git/add-queue .git/rm-queue
> > > +checkout-cache -q -f -a
> > > 
> > >  update-cache --refresh
> > 
> 
> PS, shouldn't we add a read-tree $(tree-id) before the checkout-cache?

A correct tree should be in the index. And if for any completely weird
and smelly reason it isn't, I'm pondering if we should back that out and
always reread-tree (takes some time, but do you do git cancel all the
time?), or rather comply to it and adjust to the loaded tree, whichever
it is.

-- 
				Petr "Pasky" Baudis
Stuff: http://pasky.or.cz/
C++: an octopus made by nailing extra legs onto a dog. -- Steve Taylor