🚀 go-pugleaf

h9210876@hkuxb.hku.hk (Lee Sau Dan ~{@nJX6X~}) wrote:
>
>> 
> Now, what if I write a program which, when fed with the same input
> data, will produce the same output as the LZW algorithm?  
>
     You see, what you really would need is a program that could
encrypt a picture in such a way as to be readable with a GIF reader.
That might not even help, since as I read the actual patent, its the
encryption that is protected.


 
> A method patent only prevents others from using that method, but
> never prevents others from obtaining the same product using a
> sufficiently different method! 
 
   Not really true in this case.  If you built a tape drive that 
wrote LZW tapes, you'd be infringing no doubt, regardless of the
actual circuit design or microcode. So, no program that produced a
LZW encrypted picture would be "sufficiently different"

> So, if my program, say, do it
> using some magic mathematical formulae rather than the code-table
> lookup in the LZW algorithm, it should not be infringing, should it?

   Yes, it would. LZW is not a method of producing an encryption, it
is a method of encryption. Subtle, but major difference.


   My major beef with Unisys here is that when LZW was first patented,
a sort of "verbal license" was given by them stating publicly that 
they would never attempt to enforce the patent against "software only"
uses. I'd have to say that the current reversal probably will never
stand up, given that the so-called enfringining programs have existed
for nearly as long as the patent, with Unisys' tacit and public consent.
To cause such a disruption on the world wide web at the moment is 
really ..... (four letter word deleted).  I can only assume that 
their legal department is worried about renewing the patent if some
uses are shown where Unisys made no attempt to protect its rights.

                        Best Regards,
                           Bob