RAND Licenses

Dear LWN readers,
 
        There has been much discussion on the net - and particularly within the
free software / open source community on the issue of Reasonable and
Non-discriminatory Licenses. It has been pointed out that these
Licenses are discriminatory when it comes to free software. I also
believe that they are more discriminatory to smaller software publishers
than they are to the large ones. Especially the ones that are large
enough to have an arsenal of patents that they can use to manipulate
better arrangements with the holder in question.
 
        I have a suggestion of what I would call a truly nondiscriminatory
license: Let's base it not on dollars or dollars per unit but on a
percentage of profit. For example if a new protocol were to be
developed to serve WWW pages and it became so prolific that all of the
major WWW servers (IIS, iPlanet, Apache, AOL, etc.) needed to
incorporate it to stay competitive then they should all be able to
license it in a way that made since to their business plan: whether
motivated by profit or not. If all of the WWW servers were forced to
license it for say 2.5% of their net sales then it would work out.
Apache is free. Two and a half percent of nothing is nothing. Problem
solved. Even AOL would score under this proposal as I believe that they
made the source code for their WWW server open source long ago. However
companies like Micro$oft - who charge lots of $$$ for their software -
would have to pay 2.5% of that some to the patent holder of said
protocol.
 
        On the same note suppose someone developed NaI-HTML (New and Improved
HTML), patented it, and licensed it using this new RAND License.
Mozilla would obviously be able to use it for free (they charge nothing
for their software). Opera would be able to include it in the free
version of their software but may have to pay a small percent to include
it in their commercial version (or should I say
commercial/advertisement-free version) because they charge for it. The
tricky part would be Micro$oft's IE. If on the one hand it is solely
contained in IE and IE is given away for free I suppose that they would
be able to not pay. But, on the other hand if they embed the protocol
in Windows, so their other applications can understand it, then it
becomes part of a non-free system and they should have to pay. If they
chose to add the functionality to the Office suite then they would for
sure have to pay for it.
 
        The tricky part is something like Mandrake distributing Mozilla: they
do charge for Mandrake but not for Mozilla; should they have to pay for
including a free program that uses the RAND protocol in their non-free
distribution? What about the people that download the ISO off off the
Internet and don't pay them anything for that - surely Mandrake cannot
afford to pay the RAND fee in that case.
 
        Micro$oft is documenting many protocols at this very moment in its
attempt to comply with its anti-trust settlement. Further they are
trying to claim that by using a (current) RAND proposal that their
protocols are open to everyone. We, in the open source community,
understand that we are being discriminated against but we need a way to
articulate that point to Judge Colleen Kollar-Kotelly and others that
have the political power to affect change. Especially to organizations
such as the W3C who are getting stuck in the middle of a bad situation
by overpaid patent litigators.
 
Regards,
Tres Melton
 
P.S. This, by no means, means that I agree with the prospect of
patenting software. This is just an idea that would make it taste a
little bit less bitter.