[A2k] Re: [Patents] Re: [Upd-discuss] SCOTUS Overrules Rigid US Appeals
Court Approach to Patentability of Combinations of Existing Elements
Jonas Bosson
jonas@illuminet.se
Thu May 3 05:42:14 2007
Richard Stallman wrote:
> This decision may reduce the number of bad software patents
> (if it results in a real change in the conduct of the patent system),
> but it will not make software development safe.
> That requires eliminating patents from the software field.
>
> (There are two kinds of bad software patents:
> those that are obvious, and those that are unobvious.)
>
I think the AT&T vs MS ruling is much more interesting as it could
nullify program claims (=record on a carrier/publication). This is
quite a step, as it would define a border against information or
"blue-print" infringements. I find that a development in this direction
is quite workable as more hardware related software patents could remain
on the basis of process or machine claims. Still, software giants could
sue computer manufacturers that deliver pre-installed machines,
cellphones or companies running the software - but software publication
in itself could not be an infringement. Its a good start for a knowledge
sharing economy...
Perhaps I am missing something, but in the case of obviousness in
combining elements I find improvements from this a bit far fetched. It
could perhaps make it easier to combine prior art, but just using prior
art as text with thin knowledge is quite entropic and misses a few
lessons in AI.
One problem with most software or business patents is that they claim
quite abstract descriptions involving communication of information
between different entities over certain medias. We could have prior art
for web-user but not for a general customer in a certain service. Its
simply very hard to kill a moving semantic/abstract target, so the
patent stays even if you find prior art to save yourself from
infringement in your specific implementation. This way the infringement
risk to the market remains. On the positive side... this could perhaps
give the examiners some integrity to use the knowledge in their fields
instead of just text searches. Lets see if USPTO makes any moves.
/jonas bosson, software developer.