Re: Sun Java available from non-free
On Wed, Jun 07, 2006 at 05:45:27AM -0700, Mike Bird wrote:
> On Wednesday 07 June 2006 04:30, Wouter Verhelst wrote:
> > On Wed, Jun 07, 2006 at 12:51:25PM +0300, George Danchev wrote:
> > > On Wednesday 07 June 2006 12:34, Wouter Verhelst wrote:
> > > > What I cannot imagine is a case where an upstream change would result
> > > > in only Sun's Java to break rather than a whole bunch of applications
> > > > (so they would most likely be noticed before the release), and/or to do
> > > > so on Debian only, rather than on every Linux distribution out there;
> > > > and it would seem that for any case where the effects are much wider
> > > > than just Debian, it can reasonably be argued that the problems are,
> > > > not under our control, which would free us from the burden of having to
> > > > idemnify Sun.
> While some people cannot imagine that a contract will be
> enforced as written, judges can.
I didn't say I could not imagine that a contract would enforced as
written. I said that I consider the chance for something to actually
occur to be small enough that it can be ignored.
Remember that this is not about Freedom. If it were about Freedom, I
would agree that there is a problem here. But that is not the case; it
is about avoiding to get sued.
I don't think the demands they are making are wrong per se, nor do I
think that the examples of where they /would/ be wrong are realistic, or
have any chance of actually occurring.
> > > And you are not to be liable for that only if the modifications made
> > > to the underlying systemm are not under your control. If a new
> > > upstream version of glibc or the kernel breaks Sun java to function
> > > properly or as documented then I believe (according to the license)
> > > someone should be be held liable for that break. Who's that? Upsteam?
> > That's Not Our Problem(TM). We're only to indemnify Sun for the things
> > we are directly responsible for. It doesn't mention /anything/ about the
> > stuff for which we are not directly responsible.
> Debian can argue that it is not responsible for software
> not in Debian archives. However, all software in Debian
> archives is signed in by a DD, a member of Debian's web
> of trust.
> A new upstream bug does not affect Debian until Debian is
> changed by the DD's incorporation of the upstream version
> containing the upstream bug. When that change is signed in
> to Debian, that is a change to Debian made by and authorised
> by a DD. At that point, Debian becomes responsible for
> incorporating the upstream bug into Debian, and Debian
> becomes responsible for indemnifying Sun.
That's one way to look at it.
Another way would be to say that if there would be a bug where glibc
does not work as documented, which appears on _every_ glibc-based
platform, then Sun did not do a proper job in testing their software
(since it occurs everywhere, remember), so it's really their fault, not
Perhaps that depends on the time when the bug is actually introduced.
Fun will now commence
-- Seven Of Nine, "Ashes to Ashes", stardate 53679.4