On Mon, May 21, 2012 at 12:34:52PM +0200, Janne Grunau wrote:
> On 2012-05-21 12:14:44 +0200, Diego Biurrun wrote:
> > On Mon, May 21, 2012 at 10:18:39AM +0200, Janne Grunau wrote:
> > > On 2012-05-18 19:05:05 +0200, Diego Biurrun wrote:
> > > > --- a/doc/git-howto.texi
> > > > +++ b/doc/git-howto.texi
> > > > @@ -346,6 +346,58 @@ git checkout -b svn_23456 $SHA1
> > > >
> > > > +@chapter pre-push checklist
> > > > +
> > > > +Once you have a set of commits that you feel are ready for pushing,
> > > > +work through the following checklist to doublecheck everything is in
> > > > +proper order.
> > >
> > > I honestly think the wording and the checklist is way too strict for
> > > most commits. Strictly following it for every commit is nonsense and
> > > I'd guess nobody will do it.
> >
> > Really? I do it for all batches of commits that I push. Now I certainly
> > do not mind common sense being applied, but such a checklist should cover
> > all bases and not teach you how to cut corners from the outset.
>
> the writer of the checklist excluded ;)
>
> I usually don't make a 'make clean' before pushing unless I have Makefile
> or configure changes. I do a push -n and look at the git log -p of the
> reported changes.
You're advocating a fast and loose cowboy attitude towards pushing ;)
> > > > +@itemize
> > > > +@item @command{make distclean}
> > >
> > > When is a make clean not enough?
> > >
> > > > +@item @command{./configure}
> > >
> > > without distclean this could be a simple make config which would handle
> > > out of source tree builds and other required configure options just
> > > fine.
> >
> > Supposedly 'make distclean' returns you to a vanilla source tree. I like
> > it because it forces you to do a fresh configure run and stray options
> > you might have in there will not throw off your test results.
>
> Some might have configure options which are required to test the changes
> cross builds or a non-standard CC come to mind first.
Not sure which side of the debate you are advocating now..
I'll send an updated patch in a moment.
Diego
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel
