Re: [Fsfe-uk] Patent article for review

[Richard Smedley]

On Wed, 2003-05-21 at 18:38, Ciaran O'Riordan wrote:
> Hello all,
>   I was contacted (through AFFS) on Monday to write an article for
> Electronics Weekly on Software Patents.  The audience is a mostly
> patent loving crowd that views patents as something that protects
> them the way we believe the GNU GPL protects us.

Just the audience we want to reach :-)

>   Please let me know if there is something terribly wrong with
> this.

Nothing terribly wrong at all - some small points below:

> > America has allowed Software Patents since the mid eighties, the European
> > Patent Convetion specifically dissallows the patenting of "programs for
> > computers" but some software patents have been slipping into the system in
> > Europe, brining their patentability into question.
> >
> > Are Software Patents good for Europe?

Any reason for the cap S and P?

> >   --by Ciaran O'Riordan
> >
> > Some proponents of the patent system think it's natural to extend it to
> > other fields but software development is a lot different to hardware

different /from/

> > development.  Software development can be done on any general computer, no
> > significant investment is required for equipment or machinery, and the
> > software can then be replicated infinitely without using up resources.  In

Compare development costs for eg. pharmaceutricals

> > this way, it is not just an industry regulation, it can affect individuals

not quite clear - perhaps for the intended audience,
a better point would be SMEs and start ups being affected
as much as individuals...

> > too.  Software is just a sequence of instructions, similar to sheet music or
> > a recipe.

Good point. Also a mention of software as simply applied 
mathematics (algorithms).

> > Traditionally, the software industry has been regulated by the copyright
> > system.  This works quite well for software since copyright is an automatic
> > law, you don't have to apply for it or wait for it.  In contrast, using the
> > patents system is a long, drawn out process.  Avoiding copyright violations
> > in software is easy, it's a choice you make, but avoiding patent
> > infringement can be impossible.  A patent may not be publicly viewable until
> > 12 months after it has been applied for, 18 months in America.  In the
> > software industry that is sufficient time for a complete development and
> > release cycle.

Important point :-)
Reinforce with ``the entire cost of your product development
will be lost when a patent on your process is revealed just
before you come to market''?

> > The number of software components in a finished product makes it quite hard
> > to check for infringements.  MySQL AB is a company that develops a database
> > product, their main business is selling components of the MySQL database to
> > third parties who integrate it into their own products.  This style of
> > development is common in the software industry and generally leads to better
> > software since components can get tested by many companies rather than many
> > companies developing their own components.  MySQL AB cannot check their
> > millions of lines of code for patent infringements so when a company asks
> > them if there are any patent issues with the database, the only truthful
> > answer is "we haven't been threatened with any yet".  So working together
> > becomes a legal minefield that should be avoided if possible.

Good point - uses a Free Software example to give
a general point, and advertise MySQL :-)

> > The 21 year term of patents is also at odds with computer software.  Not
> > many software ideas are still usable after 21 years, the public will not
> > benefit from the eventual disclosure so it may as well be 200 years.  21
> > years is also long enough to out-live many software companies.  A recent
> > phenomenon in America is "IP firms" approaching companies hit hard by the
> > economy and offering to buy some of their patents off them.  The companies
> > either say yes or go out of business, the IP firm then makes it's money out

makes /its/ money

> > of licensing these ideas that they have not contributed to.  A system
> > designed to promote innovation then funds the complete opposite.  Very large
> > companies are mostly immune to these firms since they can afford to
> > challenge them in court.  

Is there room to mention cross-licensing?
IBM et al. hold enough patents to tie any small
innovator into a cross licensing deal - thus 
perpetuating a system where the big guys will
always make money from whatever the ``little guy
cranking out inventions in his garden shed''
comes up with - it's a very powerful argument
against introducing them into a supposedly 
competitive market.

> > Small companies are mostly immune because they
> > don't have any money to sue for.  Medium enterprise becomes the target.  At
> > some stage, all companies must be "Medium" businesses.  A common path is for
> > the company to start small and get a venture capitalist to fund them so that
> > they can grow into the medium business sector.  This works well for all
> > involved but how does a venture capitalist know that their investment will
> > not simply make the small company a target for patent infringement cases?
> >
> > Finally, for software to be usable, it has to behave as a users wants and
> > expects.  This means that one great new idea must be combined with many
> > known ideas in order for it to benefit it's users.  A word processor has to
> > be able to read and write Microsoft.s Word documents because they are a
> > standard.  In America, Microsoft have patented their latest video format.
> > Microsoft.s dominance is hard to compete with already.  Software patents can
> > make competition illegal.

s/Microsoft.s/Microsoft's/

Thanks for writing this, Ciaran, if smaller software
houses wake up to the dangers to them, and lobby their
MEPs, we stand much more chance.

Can you but a box at the end with a couple of web links, and
something about contacting their MEP?

 - Richard

-- 
Richard Smedley

http://www.linuxuserexpo.com/
24-26 June, NEC, Birmingham,UK

``Any attempt to brew coffee with a teapot should result in the error
code "418 I'm a teapot". The resulting entity body MAY be short 
and stout.''                            -- RFC 2324

This email message may contain privileged or confidential information. The 
information is intended for the use of the individual or entity named above. If 
you are not the intended recipient, you are hereby notified that disclosure, 
copying, distribution or the taking of action in reliance on the contents of 
this electronic information is strictly prohibited. If you have received this 
email in error, please notify address@hidden immediately via reply email. Any 
personal content in this email is that of the individual author and should not 
be interpreted as being endorsed by the company in whole or in part. Whilst 
care is taken, it is the responsibility of the recipient to ensure that this 
email is free from virus infection, and no responsibility is accepted by Live 
Publishing Group for any disruption, loss or damage arising from its receipt or 
use.