[Discuss] Self-introduction and more on software patent

[bogstad at pobox.com (Bill Bogstad)]

On Fri, Sep 30, 2011 at 4:52 PM, Hsuan-Yeh Chang <hsuanyeh at gmail.com> wrote:
> On Fri, Sep 30, 2011 at 3:34 PM, Bill Bogstad <bogstad at pobox.com> wrote:
>
> (100 patentable ideas) != (100 patent applications).? Often times, one
> patent application includes many patentable ideas, some of them are included
> in the allowed patent claims, while some of them are disclosed but
> unclaimed.? So, a patent application serves two functions: 1. describe the
> invention; 2. tell the public the metes and bounds of the patent right
> sought.? Anything described in a published patent application (19 months
> after filing and before grant) will become prior art against later filed
> patent applications.? So, if one describes his/her inventive concept in
> detail in a patent application, that description will be time stamped by the
> patent office and stands for prior art.? As long as you don't care about the
> scope of your patented rights (as you are not really seeking patent
> protection anyway), I see no reason why you cannot write up a patent
> application yourself (by borrowing a copy of Patent-it-yourself) and get it
> on file with minimum costs.? Ultimately, with a few hundred dollars of
> official fee, you can do a "patent application" including hundreds of pages
> (thus including hundreds of patentable ideas) and get it on file and later
> published by the patent office.? This will make patent examiners very happy,
> because patent publication database is normally the first place they search
> and find prior art to reject a patent claim.

So you are suggesting filing a single application per program and
throwing every idea that wasn't in your CS textbook into the
application?  I'm not sure that would make the examiners happy as it
would be a mess of unrelated ideas, but okay.  I did a quick check of
US patent fees and it looks like there are extra fees if you have more
then 3 claims, more then 20 claims, multiple dependent claims, every
50 pages over 100 pages, etc.  It looks like it would be easy to go
over $500 even if one is a 'small entity'.   My actual experience with
filing patents is (almost) non-existent, so maybe I'm misinterpreting
the charges that would be involved.

You also seem to be suggesting that not using correct language/style
in the application will still be sufficient to get protection.  Does
this mean I could just print out copies of my manuals, design docs,
and source code and ship them to the patent office with a check?   My
impression has been that to get anywhere with the patent office you
have to very carefully words things in a language that most
programmers don't know.   Maybe that's not needed if you just want to
make sure that the patent office has a record of your ideas rather
then actually having a patent issue.   It would be interesting to see
how the patent office would react if it started receiving what were
essentially document dumps where the filer had no intention to ever
follow through on their application.

Still, I would like to hear more about this idea.

> I probably shouldn't compare the importance of liberty and property
> interests with patents and copyrights.? My point is, something is written in
> the constitution for more than 200 years, and it would take an
> overwhelmingly prevailing argument to get it changed.? So far, I don't see
> the open source community has made a case on why Congress should change this
> part of the constitution.

You persist in calling current patent legislation part of the
Constitution.  It's no more a part of the Constitution then the
legislation creating NASA, the FDA, Social Security, or the Interstate
road system.   The Constitution authorizes but does not require there
to be patent or copyright law in the USA.  I am not familiar with
patent law changes in the US, but I know that in the last 100 years
Congress has changed the law or ratified treaties to change copyright
law something like seven or eight times.  So it would appear we make
changes like this all the time.   Look at recent changes to increase
the length of copyright.  Opponents pointed out that Congress keeps
moving the bar farther and farther out and suggested that this was a
violation of the requirement for "limited times".  As I understand it
the courts decided that as long as Congress gave a specific number
they could change it to anything they wanted.   If Congress wanted to
change the length of a patent to 30 seconds, I don't see the
Constitution standing in their way.  Now clearly your are right in
that the open source community hasn't managed to get Congress to
change the law, but please don't make it sound harder then it actually
is.  Also don't forget that the patent office denied applications for
years which involved software and it was a series of court case (not a
change in legislation) which changed this.  It would have been trivial
at the time for Congress to pass a new law saying, no that's not what
we want.  (Now, of course, there are entrenched interests with lots of
lobbyists to oppose this.)

>
> International treatise is another factor.? But don't forget who is the
> biggest player behind those treatise?? It's the United States.? Over the
> years, the US government has constantly challenging China and many other
> countries for "stealing" intellectual property from the US.? If the US
> abolishes patent law before anyone else, how would the administration
> challenge other countries for not protecting intellectual property?? I don't
> believe the US government would slap on its own face...

You are using the word "intellectual property" here which RMS warns
about.   I'm not suggesting eliminating copyright, nor am I even
suggesting eliminating patents.  I am suggesting that changing US
patent law so that software went back the way it was before those
court cases might be a good idea.   All the US would be saying to
other countries is that we don't believe that patents are the way to
protect software, use copyright instead.   i.e. Feel free to use any
ideas you want when you write your own software, just don't copy any
of our code.  I suspect that for most countries this would be a net
win for their local programmers as I think most software protected by
patent probably originated in the US.   Eliminating patent protection
for whole classes of programs would make it easier for their local
industry to compete with the US.  (And if you believe RMS, it would
help the US as well as it would increase innovation in software.)

Bill Bogstad