> > [ GPL tug of war, and other half-informed statements ]
> If they never issued a written offer to anyone, and have distributed their
> binaries, then they have not complied with GPL.
Slow down, everybody.
My understanding is that this company planned to introduce their
product at Comdex. That probably means they have postponed all
responsibilities they could, as long as they could get a passable
demo unit onto the floor. I think RT-Control, in particular, can
relate to this circumstance.
I also imagine this means that they have not shipped any product yet,
and thus have not had opportunities to violate the GPL yet. If they
have passed out beta test copies of the hardware without the written
offer, they are on shaky ground, but one that does have precedent.
My correspondence with them leads me to believe that, when things
settle down a little, they will post their mods to their web site.
While this approach is not the minimum required under the GPL,
it will satisfy everyone (assuming they follow through).
Their processor is definitely not an ARM derivative, at least according
to their documentation. They have info on their processor on the web
at http://developer.axis.com/hardware/etrax/ .
I doubt that they are using uCLinux, although many concepts and
possibly even some code is shared. A quick trip through their
processor data sheet showed no mention of an MMU.
> The problem now, is that the only ones that, legally speaking, have a
> position where they can demand them to comply are the people whose
> copyright has been infringed, i.e. Linus and others.
If you find out for sure they are violating the GPL, let me know.
I own the copyright for big chunks of Boa.
> This is IMHO a significant weakness with the GPL concept. And one reason
> why assigning copyright to FSF may be a good idea, since they have at
> least *some* legal resources available to put at least a *certain* weight
> behind their demands.
So far, even the threat of legal action has been quite effective.
In the age of instant mass communication within the community, I see
that phenomenon strengthening. Remember the Pentium Divide bug?
Ever heard of the Slashdot effect?
My biggest concern is not with the letter of the GPL, but how it
applies philosophically to an embedded product. If you buy a camera
with GPL'd code in ROM, what good does it do you to have a copy of
the code under the GPL? One of Richard Stallman's prime motivations
in writing the GPL is to empower the user of a piece of software to
fix bugs (or hire someone to fix bugs). No individual empowerment
takes place when, to fix a bug, you need to fork over $25,000 NRE
for a mask ROM run, followed by surface mount rework. Even if the
code is in Flash (and means to update the Flash are available to the
user), suppose compilers for the processor were unavailable (and
information to write one was unavailable)?
- Larry Doolittle <LRDoolittle@lbl.gov> and <email@example.com>
This message resent by the firstname.lastname@example.org list server http://www.uClinux.com/
This archive was generated by hypermail 2b30 : Sun Apr 07 2002 - 00:01:33 EST