[Patents] "technical problem" == any problem

[John Tobey jtobey@john-edwin-tobey.org]

Steve Probert <steve.probert@patent.gov.uk> wrote:

>>Unisys claim to have patents in Europe (including the UK) for the LZW
>>algorithm.  I'm not sure how this has a technical effect, or qualifies
>>as 'software not as such', if IBM's homophone checker did not.
>
>>How is the criterion of technical effect applied in practice?
>
>Data compression is a technical problem.  Eg. How do you store large
>amounts of program/data on a disk that does not have sufficient
>capacity?
>
>On the other hand, IBM homophone checker looks at words two sea
>weather they are spelt correctly.  In the UK such an invention would
>probably be excluded from patentability as a method for performing a
>mental act, as well as a program for a computer.

This distinction strikes me as completely illogical, unworkable, and
open to abuse.  You (or the court or UKPO) have shifted the burden to
defining "technical problem".

Every patented thing solves some problem (for the sake of argument).
Whatever profit results from solving a problem can be used to buy more
disks, so every patent is practically equivalent to a solution to the
"technical problem" in your example.

In practice, benefit (or profit, or productivity) is amorphous;
different kinds of benefit are interchangeable with little overhead.
Typically, all kinds translate directly to monetary benefit, and vice
versa.  (The cases where the "vice versa" fails - with things that
money can't buy - are interesting but not very relevant to this
discussion.)  Therefore, it is impossible to make a meaningful
distinction between solving a "technical problem" and solving any
problem.

This is a major loophole in UK law, which I'm sure IBM and others
exploit to their great advantage at the public's expense.

-John