Forums

>What is the mechanism to protect SBML (the license?) What forbid me to
>fork and develop another SBML, incompatible with the original? Or even
>to develop the Silly Biologist Markup Language, to describe the bad
>bench practises?
>
>Some could find the concern preposterous, but I'm submitting a grant
>application to the European Commission and they are picky about IPR.

It's a fair concern, one that the original authors of SBML
have wondered about before.

There are (at least) two levels of concern. One is the
protection of the name and acronym ("SBML"), and the other
is the licensing of the language, against attempts to subvert
them or legally take them.

For the first case (protecting the name), I think this is
usually done by registering a name as a trademark. The only
reason we might want to do this is to prevent someone else
from claiming SBML as their own in the field of biology
markup languages. However, it is unlikely that anyone at
this point could register SBML and claim it as their own.
According to the US Patent and Trademark Office FAQ, there
is something called common law rights:

Federal registration is not required to establish rights
in a trademark. Common law rights arise from actual use of
a mark. Generally, the first to either use a mark in
commerce or file an intent to use application with the
Patent and Trademark Office has the ultimate right to use
and registration.

Further, it is worth looking at other examples of markup
languages: it doesn't appear that HTML is trademarked, and
that certainly has more potential commercial impact than
SBML.

Finally, if one thinks about the possibility of some
organization attempting to hijack SBML in this field, one
has to conclude that it would be extremely unlikely to
succeed given the well-established nature of SBML at this
point. Who would be fooled? What success would this
organization have, if they antagonized the same user
community they would presumably be trying to gain from?

Thus, it appears reasonable to assume that we don't need
to trademark SBML.

The second issue is: do we need to stipulate license terms
for using the language SBML? This also appears to be
unlikely. First, we cannot make it more free than it
already is. Second, it has been published and disseminated,
so if someone *did* convince the US copyright office to
copyright the language, it would be easy to show prior art.
Third, the negative impact on the image of anyone trying to
take SBML this way would surely doom their commercial hopes
in the long run.

So in summary, our reasoning so far has been that we do not
need to take legal action in this area. In fact, it's quite
likely that registering SBML or copyrighting the language
would cause other people to suddenly worry about the
intentions behind the actions.

If any of you have experiences that would argue differently
from the arguments above, please bring them up. Also, have
not consulted the intellectual property attorneys at
Caltech, but we can if people feel we should to be on the
safe side.

MH