Re: 2.6.14.5 to 2.6.15 patch

[Greg KH <greg@kroah.com>]

On Wed, Jan 04, 2006 at 03:31:01PM -0800, Randy.Dunlap wrote:
> On Wed, 4 Jan 2006, Alistair John Strachan wrote:
> 
> > On Wednesday 04 January 2006 23:13, Greg KH wrote:
> > > On Wed, Jan 04, 2006 at 10:58:24PM +0000, Alistair John Strachan wrote:
> > > > On Wednesday 04 January 2006 22:31, Greg KH wrote:
> > > > [snip]
> > > >
> > > > > > > The issue I hit was we have a 'latest stable kernel release
> > > > > > > 2.6.14.5' and under it a 'the latest stable kernel' (or words to
> > > > > > > that effect) on kernel.org.
> > > > > > >
> > > > > > > Then when 2.6.15 came out, that was it!  No patch for the 'latest
> > > > > > > stable kernel release 2.6.14.5'.  It was GONE!
> > > > > >
> > > > > > Yes, I brought this up a couple of weeks ago, but I was told
> > > > > > that I was wrong (in some such words).
> > > > > > I agree that it needs to be fixed.
> > > > >
> > > > > How would you suggest that it be fixed?
> > > >
> > > > It's difficult, but perhaps providing a link to the latest "stable team"
> > > > release in addition to Linus's release would solve the problem.
> > >
> > > But what happens when we release a 2.6.14.y release and a 2.6.15.y
> > > release at the same time (as people have requested this in previous
> > > threads...)?  What would show up where?
> >
> > You're right, it's complicated. In that case I'd still opt for showing
> > 2.6.15.y, as the vast majority of people manually installing vanilla kernels
> > will either be on the latest-ish kernel, or have a clue about what they're
> > doing (who doesn't know the ftp URL off by heart now).
> 
> I agree.  I think that one previous -stable patch version should always
> be listed there, even if we think that 2.6.N is stable.  :)

I don't at all.  If we do that, people will assume that they need to
wait till 2.6.N.1 before trying that kernel (as it wouldn't be "stable"
otherwise.)  So no one will test it, to really generate the bug reports
that we need to get to that .1 release.

Or should we just throw out a .1 release with the first simple patch
that comes along just to make the kernel.org page update properly?  I
don't think so...

thanks,

greg k-h