[Linux-aus] Dealing with the patent process

Shane Stephens shans at research.canon.com.au
Wed Aug 4 07:25:02 UTC 2004


On Wed, 2004-08-04 at 08:28, Arjen Lentz wrote:
> Hi Pia,
> 
> On Tue, 2004-08-03 at 15:02, greebo at pipka.org wrote:
> > [...] I'm thinking we get the Open Source 
> > community (including companies of course) involved in the patent process. 
> > Apparently patents have a public period where people can speak up about any 
> > prior art. I suggest we tie into this, and start getting involved with 
> > speaking up about patents. This means they won't necessarily get passed (if 
> > we have good cases) and then after say 6-12 months, we will potentially have 
> > a good paper trail of the issues the FTA has locked us into concerning 
> > patents. Other ideas are much welcome :) 
> 
> I haven't checked for AU specifically, but traditionally the entire
> process is public. A patent is not a secret thing. Everybody can look at
> the whole story, they're just not allowed to implement it without
> licensing. This protection in essense comes into force the moment the
> patent application starts, because the review process of an application
> depends on feedback from experts in the field.
> In practise, this feedback loop is broken in many countries (most
> notably the US) but the process itself has not changed. So the "hooks"
> are there for anyone who is paying attention.
> 
> So I think you are very right. We basically need to track patent
> applications, identify software patents and other relevant issues,
> inventorise them (website) and find the prior art that inevitably
> exists.
> 
> I know that you can set up triggers on the AU trademark site to be
> notified when item with certain keywords are added to the database,
> likely something similar is possible for patent applications. Or just
> having people take a daily peek to identify stuff that needs adding to
> the list. Once identified, handling by a broader group of people becomes
> possible. Particularly if we can categorise.

There is definitely a process in the US by which members of the public
can protest a patent application.  The applications themselves are
published 18 months after first filing, and are granted at some later
date after publication.  There's a relatively easy way to find recently
published applications in specific fields on the uspto website
(www.uspto.gov)

As long as the protest is filed between the publication of the
application and the granting of the patent, then it will be considered. 
At least, that's how I read the legal mumbo jumbo that they provide
detailing the process.

Filing a protest consists of identifying the application in question
clearly, getting together your prior art, consisely describing why each
piece of prior art is relevant, and filing a copy with BOTH the patent
office and the applicant.  This can be done by first-class mail.

I'm sure a similar situation exists in Australia, and I might
investigate later today.  However, as far as I can tell, our real
problem at the moment are patents filed in the American patent system,
as there seems to be a much greater proportion of spurious patents
emanating from here.

Cheers,
	-Shane

> 
> Regards,
> Arjen.





More information about the linux-aus mailing list