Python Trademark Status

[Tim Peters]

[David, turns up a CNRI application for a service mark on Python,
 filed 9-May-2000]
> ...
> My own notes:  The trademark was filed on May 9, 2000.  First use is
> documented as July, 1995.  In truth, Python was used as a name for the
> programming language *years* before that date.

I am unhappy right now and should not be posting.  I'll calmly note that
trademarks are much messier than copyrights, and particularly across
national boundaries.  The "first use" claim appears to coincide with Guido's
move to America to work at CNRI, and it appears CNRI filed the application a
month after they knew Guido et alia were leaving.

> My impression of trademark law is that it is eminently sensible.
> It exists to prevent consumer confusion, and judgements are made based
> on the likelyhood of confusion arising from trademark misuse.
>
> I daresay that if CNRI splits their own version of the language from
> Guido's development, it will cause extensive confusion.

I've seen no evidence that CNRI wants to continue Python development, and
they've certainly not announced any such plans that I'm aware of -- indeed,
BeOpen PythonLabs is producing CNRI's Python 1.6 release on its own dime.
They're very keen on intellectual property law, though, and there are
several possible reasons for wanting to own property other than to make
constructive use of it.  Some are nicer than others, but I won't speculate
about that here.

> Certainly *I* expect "Python" to be the language that was invented
> and is directed by Guido; not a language that is coming out of CNRI.
>
> I don't want to cause headaches for CNRI and Guido,

You're not!  The existence of this particular headache apparently predates
your reporting of it <wink>.  It's good to know it; someone else had
volunteered a search earlier and didn't turn this up, but that may have been
in a different country; we didn't know about it.

> but I suggest that some consideration be given regarding who should
> hold the trademark for the name "Python" as applied to programming
> languages.  If the general feeling is that CNRI should *not* be the
> trademark owner, then perhaps the TM office needs to be contacted
> and the true history of Python revealed to them.

Guido certainly wants the Python name for himself (or an independent legal
entity created for the purpose), and for God's sake he's surely earned it on
all ethical and moral grounds!  As he mentioned on Python-Dev a while back
in response to a question, he has in fact already contacted an attorney to
start the wheels turning.

> The fact that they have lied regarding "first use" may be enough
> to stop the application process.

They're certainly not lying about their *own* "first use" of it.  CNRI is
lawyer-happy, and they wouldn't try such a boneheaded stunt.  I expect this
will be just another case where three lawyers will give three opinions, and
it will end up as fodder for yet even more "negotiation" rather than risk a
protracted battle.

[Aahz]
> I think there may be a legal definition for 'trade' that is used here.
> My understanding of USA law is that a trademark can be applied for any
> time a 'product' is involved.

Note that CNRI didn't apply for a trademark, though, they applied for a
service mark.  If we were lawyers, I bet we'd find that fascinating <wink>.

i'm-sure-someone-does-ly y'rs  - tim