Re: mozilla thunderbird trademark restrictions / still dfsg free?
Michael K. Edwards wrote:
So the question is: is the right to call a bit of software by a certain
name an "important freedom"? That's definitely debatable. The name you
use to refer to a bit of software doesn't affect its function.
It can, especially in the case of a web browser; consider web servers
that verify that the client claims to be a sufficiently new Mozilla or
IE before sending DHTML.
That's a bit different - no one's arguing that the MoFo should have any
control over the UserAgent string of any browser, even one Debian ships,
just because it contains the word "Mozilla". Such an effort would be
both counter-productive and laughable.
Exactly what the app is called is a more difficult question. There's a
long tradition of ln -s /usr/bin/exim sendmail, but you could also argue
that if someone downloads and installs Debian or a derivative and types
"firefox", the trademark holder should be making sure they get a Firefox
they have checked for quality in a trademark sense.
It looks to me like there's a real storm brewing over trademark
enforcement in open source space. At least in most US jurisdictions,
trademark law applies an "enforce it or lose it" standard, and one of
the key criteria in judging whether a company takes its trademark
seriously is whether it exercises quality assurance over third parties
to which it has (explicitly or implicitly) licensed the right to
distribute goods or services marked with its trademark.
I think it's absolutely right to raise the wider issue.
In a hypothetical situation where Debian is the dominant distribution
channel for Software X, performs QA functions, and handles the bulk of
bug reports, the upstream for Software X could actually lose ownership
of the trademark to Debian. Even when the distributor relationship is
non-exclusive, a failure to exercise QA authority over the Debian
channel could weaken Mozilla's ability to enforce the trademark on
other channels. (Imagine "Mozilla Firefox, MS Authorized Edition"
with the crippling limitations of your choice.)
Or even just "Mozilla Firefox" distributed in an official-looking manner
rom www.firefoxbrowser.info with added spyware or bank login capture.
So the Mozilla folks are being responsible in setting out the limits
of the license to use their trademarks as part of the MPL, rather than
leaving the issue unaddressed and then springing it on people in
court.
We're not actually doing it as part of the MPL - we want to keep
trademark licensing separate from code licensing. The MPL doesn't speak
about trademarks except to say that it itself doesn't give you any
rights to them.
I think it would be a good idea to work out a modus vivendi
with them, such that the names of Debian-packaged Mozilla products are
unchanged, and designated persons from Mozilla have the right to file
RC bugs that the maintainer isn't allowed to downgrade. That at least
preserves the forms of trademark defense, at a rather minimal cost in
freedom.
One principle that we were originally working with in our trademark
policy is "QA in retrospect" - i.e. we let you do roughly what you want,
but if the packages are of a consistent low standard, we get to pull the
trademark and you have to change the name.
Now at the beginning of the thread, there were some objections raised to
this idea - but is it better than more intrusive forms of trademark control?
GErv
Reply to: