Re: Netgear missing build tools preclude GPL firmware build

On Fri, 04 Mar 2011 10:01:59 +0000, Joseph wrote in message 
<4D70B897.2020204 <at> heenan.me.uk>:

> On 04/03/2011 09:24, Arnt Karlsen wrote:
> > On Thu, 03 Mar 2011 19:43:55 +0100, Armijn wrote in message
> > <4D6FE16B.9010009 <at> uulug.nl>:
> >
> >> On 03/01/2011 09:20 PM, Paul Parsons wrote:
> >>> Thus Netgear seem to admit that their GPL firmware cannot be built
> >>> without Conexant proprietary software.
> >> Which is quite common. I suppose that you were expecting a tarball
> >> with Makefiles and everything as part of the "scripts to control
> >> compilation and installation". Right now there is no consensus
> >> whether top level Makefiles (for a firmware for example) should be
> >> included. Some people say those should be, others say it is just
> >> for individual components, because the firmware itself is not a
> >> derivative work, just "mere aggregation".
> >>
> >> Personally I don't know if those scripts should be released. I
> >> would like them to release it though. That being said, I did get a
> >> mail from NETGEAR that they are looking into this but that this
> >> takes time. I fear that you were just being stonewalled by their
> >> support desk :(
> >>
> >> armijn
> >>
> > ..it would be useful to imagine a copyright case, where Netgear etc
> > would want to prove its case by making the court able to verify it
> > gets the same files and the same file check-sums, by following the
> > same recipe.
> 
> I wonder if it would? At least in the UK, I would imagine such a case 
> would be covered by expert witnesses and evidence that didn't get 
> released to the public.

..in closed source litigation, that is exactly what the parties will
want, so no problem there.

..this list is about GPL violations and the potential associated
litigation over code etc evidence, that _should_ be released to the
public.

..obviously, stolen closed source falsely released under the GPL, 
is not a GPL violation, it is a copyright _etc_ violation.

> Anyone that brought such a case would have to be sure the code wasn't 
> correct, or they'd risk becoming liable for Netgear's legal costs.

..exactly why we want an easily compliable firm legal standard. ;o)

> (Besides which, checksums often differ between binaries compiled from 
> the same source with the same tools for a variety of reasons - it is 
> notoriously difficult to reproduce an identical binary from identical 
> source for many toolsets. For starters, random things like file 
> timestamps, compile time/date, absolute source tree locations, exact 
> compiler patch version, compiler install location, processor type,
> local username, local hostname and just random uninitialised data are
> often thrown into .o files.)

..which is exactly why we and the judges will want an easily 
enforceable standard recipe, to enforce the laws _fairly_. ;o)

--

-- 
..med vennlig hilsen = with Kind Regards from Arnt Karlsen
...with a number of polar bear hunters in his ancestry...
  Scenarios always come in sets of three: 
  best case, worst case, and just in case.