Re: [Patents] computer.org article

["Nick Fleming" <Nick@bigfatpanda.fsnet.co.uk>]

----- Original Message -----
From: "Roberto Micarelli" <mi.ro@iol.it>
To: <patents@aful.org>
Sent: Sunday, June 02, 2002 4:34 AM
Subject: [Patents] computer.org article


> I found this reading really interesting and (sigh) totally on the side of
> great companies' interests. There is also a second page entitled
> "arguments against patents"...
> Why don't they say that, in order to write software, I need only a 1000$
> machine, ten fingers and my brain?
>
> http://www.computer.org/software/homepage/2002/03COTS/index.htm
>

Commercially, one of the biggest costs
in development is staffing costs. Companies in
England (according to the press) are currently
recruiting from Asia (India?) to keep costs
down.. but I digress..

From the article (page 2, "arguments against patents").

"Because software was not patentable early in its history, software patent
opponents argue that technology has evolved differently and that software,
therefore, does not fit the patent model. This argument will certainly lose
weight as time passes. One hundred years from now, it won't matter that few
software inventions were patented in the 1970s"

Given that the US has effectively prevented 80% of
copyrighted works from the last century from falling
into the public domain, what is to stop them extending
the length of a patent as well ?

I'd also object to the component model for
a number of  reasons:

You can't always guarantee components will work together.
(Competing companies might actively prevent interoperability).

If a patented component is withdrawn from the market,
you can't (legally) replace it.

It assumes components are the 'correct' way to
build software, but there is STILL "no silver bullet"
for software development.

Software usually has errors and varies in quality
between developers / companies. The article arrogantly
implies that all software is mostly bug free.

There is no guarantee a patent holder will fix errors.

Component based software tends to be extremely
bloated, slow and hard to locate where errors are,
and so do not always help reduce "time-to-market"

The learning curve for using components may
be substantial, incurring staff training costs in
order for them to learn (not mentioned
anywhere in the article).

But overall, all patents would do on software
components is just increase the overall cost
of development, with no actual benefits.

Without patents, you have the choice of
using components, (which can sometimes
save development time) or creating a
bespoke software solution for your clients.

I think overall, the article is a bit overly
optimistic on how business works, and
completely misses the impact lawsuits
have on development. (i.e. you can
sue someone out of business).

Finally, the article seems to imply that
there will be an international, patent-free
standard for component based interaction.
Otherwise you'd end up paying every time
you wrote a software component :-)

Is it me or are most of the arguments for
software patents just inconsistent ?


Regards,

Nick Fleming.

-----
Oh well.. time to watch England
get thrashed at World Cup Football :-)

(Thats proper football, not the American
game with lots of commercials :-P)