how patents work (was: licensing of CIFS standards)

Alex Hudson home at
Wed Apr 10 14:06:50 UTC 2002

On Wed, 2002-04-10 at 14:29, Alessandro Rubini wrote:
> Hello Axel.

Hi Sallyandro :D

> On the other hand, most software patents and scientific papers do not
> include the implementation. They just describe the idea and some
> information on how it is implemented (papers have much information,
> usually, while patents has almost none of it).

In terms of the mechanics, yes. But in terms of the idea, no. The patent
brief is supposed to outline exactly how the process works - you don't
get a patent on the idea itself, but on the way you have achieved your
solution. For example, you can't patent a loudspeaker. But, if your
loudspeaker was designed in such a fashion that you made it 10% louder
than all the others due to some previously unknown acoustic property,
then you could get a patent on the device/contruction that allowed you
to do that. That wouldn't stop other people finding other ways of
harnessing the new acoustic property, they just wouldn't be able to copy
you exactly.

> Your quote is right, but not to the point that the skilled person must
> "implement" it, just "invent the same idea", as far as I know.

I think that depends on your definition of "idea". Certainly, if you
invent something that has been invented before, then that infringes the
patent. However, you can't pantent "ideas" - you can only patent
inventions, which are technological solutions to problems. If someone
else can solve the same problem a different way, that doesn't infringe. 

> But I agree that "skilled in the art" is used by patent offices as a 
> synonym of "a person with good knowledge but no inventive ability at
> all", which means "it must have been published".

Indeed. People often over-rate the amount of searching the patent office
will do as well - generally, it's up to the person applying for the
patent to do the search for prior art.

> But the point is that it's not about the implementation, only about
> the idea. Otherwise you can patent something (without disclosing the
> implementation) 

You can't do that. The point of the patent is that you disclose how you
solve the problem ;) From the UKPTO site:

"What must be included in the description

4. The description must be sufficient on its own to allow the invention
to be carried out by someone skilled in the field of technology to which
the invention relates. It should not normally refer to other documents
to explain important features of the invention and should never refer to
documents which have not been published at the time of filing the

A patent is concerned with how something works, it's not a land-claim on
a solution space. You have to outline your specific solution to the

BountyQuest talk briefly about prior art:


-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 232 bytes
Desc: This is a digitally signed message part
URL: <>

More information about the Discussion mailing list