some practical ideas

[Peter Vandenabeele <peter.vandenabeele@mind.be>]

Hartmut,

Below are some ideas I personally have about the problematic 
situation regarding SW patents in Europe. I also send this 
text to some national and international journalists and some 
members of the JURI committee of the European Parliament that 
are in the political party that I am a member of.

Any critique or additions to this text are welcome.

I have the impression mainly the major USA companies are 
lobbying through their importers, local subsidiaries and Vallue 
Added Resellers (VAR's) for the adoption of the European SW 
patent in order to be able to extend and enforce the USA dominance 
of Software manufacturers in Europe. E.g. the industry 
association for large companies in Belgium (Agoria) promotes the 
European SW patent on behalve of its large members (Telindus, 
Microsoft, HP/Compaq ...), but 90 to 100 % of their business is 
made in reselling/integrating HW and SW that is manufactured in 
the USA. So, I really have the impression, the net effect is to 
promote further dominance of USA based SW companies in Europe, 
rather than to promote European businesses, while that should be 
the honest argument that the proponents of SW patents in the EU 
parliament should follow (and hopefully not only the USA 
software industry funded lobby).

The cynical effect is that the European Commision now promotes a
measure that will effectively further reduce the capabilities for
small start-ups in Europe (SME's) (be it in Free Software or 
proprietary SW) to grow a succesful business. As soon as such 
a European SME becomes somewhat succesful and would become a threat 
to the established SW companies, they can grab one of their 
somewhat related trivial SW patents out of the cabinet and 
kill-off the small competitor. Independent of the merit or relevance 
of the patents that are used for the attack, the mere cost of the 
legal defence, is enough to kill-off a small company like ours. As 
soon as our visibility becomes somewhat larger, e.g. by trying to 
bring a product on the world-wide market. This is really a game of 
the large companies: they use it to maintain some balance between 
their R&D investments and to suppress any new entrants in the 
market.

Full disclosure: When I worked at Alcatel R&D a few years ago, I 
was granted 5 USA/IPO patents on electronic devices. So, I have a 
clue on how the system works. The theory was very clear: "we need 
to obtain as much as possible patents at our European company to 
have enough counter weight against the large USA companies in our 
market (Lucent, ...). Also, we need to protect our investment so 
that no one else (read "a new entrant in the market") can profit 
from our heavy R&D investments and reduce our capability to earn 
back this investment, with a profit.". At the time, I could 
understand the reasoning for such specialized developments in e.g. 
datatransmission over fiber, that is only accesible to a small 
group of large (or smaller, but heavily VC funded) organisations. 

A fundamental problem with software patents in comparison with the
kind of patents of which I am an author myself, is that Software 
Development is advancing at such speeds, is developed on such a 
massive world-wide scale, uses such intrinsic complexity in bringing 
together thousands of simple "Methods" in a single program, that 
there is no way anyone could correctly judge if a certain "Invented 
Method" is new or already used by some-one, somewhere on this planet. 
Nobody can claim he can check correctly, realisticially that no 
one else used the basic methods of e.g. "Instant Messaging" 
somewhere before in the world (independent of the actual claim
itself, it is simply unrealistic to prove that no one in the world
ever used similar technology with a buddy list and instant 
messaging to them).

And you could go on with a long list of very basic, trivial patents
that really stand in the way (_not_ promote) further development.
In the current world, you just cannot know if a certain software
"Method" has not already been used by someone else. One other
reason is that SW largely relies on _secrecy_ as one of the two
other methods of Intellectual Property protection. So, in many
cases, the source code is hidden to the outside world, and you 
don't know what exact method is used in other people's software.

What to do next ?
=================

I see 4 steps:

- try to avoid the granting of the EU software patent law

- if it is accepted: practical defense against it

- the bloodshed

- flight-or-fight ?

Step 1: avoid the acceptance of the EU patent law
=================================================

The current law proposal for EUROPE is mainly inspired by 
the BSA (an industry association to check into the correct 
application of the conditions under which SW from it's 
members is used; mainly USA based SW companies).

We really need to open up a serious economical debate on what 
will be the dominant overall effect of the acceptance of this 
law on the EUROPEAN SOFTWWARE INDUSTRY. 

- Will there be better possibilities for European companies to 
  get better defense of their IPR on a world-wide scale (as 
  expected by the good-faith proponents of the law that want 
  to advance the European SW industry) ?

- or will there be a suppression of any smaller upcoming European 
  SW initiatives by the USA dominance in the current SW market ?
  And on a similar line, will Software patents be used to suppress
  the further growth of Free Software development.
  
  (I learned that Belgian Banks, that want to set-up Instant 
  Messaging within their organisations, with entirely different 
  SW on different platform, are still afraid they will have to 
  pay license to AOL; would we have to pay a license now to use 
  such a basic feature as Instant Messaging ?)

What could be arguments in the debate ?

I believe in the theory of "lazy consensus". Anybody that votes
against some idea, must have a better alternative. We all agree 
that there must be possibility for IPR protection for those
authors that want to protect their work through technical, legal,
and practical means, even if it is to defend the Freedom of the 
work, so as Copyright is used by the authors licensing their work
under GPL. But, what is the best implementation that does have 
the least side effects ? It is commonly known that there are 3 
methods for IPR protection:

- Copyright
- Secrecy
- Patents

Copyright:
==========

Copyright is a system that is used to _publish_ a work, and still 
forbid people to make further copies, without having the agreement
of the Copyright holder to do so. 

Copyright has been used very succesfully, e.g. by Microsoft, to 
begin as a very small start-up with minimalistic funding and a 
clever CEO, to build a SW emperium. Microsoft did _not_ require
SW patents to become succesful. Even more, if SW patents had existed, 
the large players of that time (e.g. IBM that made the fatal business 
mistake of allowing Bill Gates to retain Copyright on DOS and sell 
his own copies) would most probably have sued Microsoft using 
Software patents to avoid Microsoft competing succesfully against 
PC DOS and OS/2.

Copyright is also used by the Free Software movement to protect
the Intellectual Property Rights and only allow it to be used under
the conditions stated by the authors, namely that the Software be
Free for all users and with varying degrees of "Inheritance of 
Freedom" aka Copyleft.

A MAJOR advantage of Copyright is that is clear-cut. If I know that
I write my code entirely myself from scratch and I know that I only
use pieces of code for which I was granted the right by the original
Copyright holder (be it for a fee or gratis), I have full assurance
that I am operating in an honest, legal way and not violating any 
one else's rights.

This is also scalable. Even if there is 10 time more software in the
world in 5 years from now, I still have full control of Copyright:
if I write it myself, or I use code that I have a guaranteed right to
use it from the person that gives it to me, I know I am covered and 
not stealing anything from anyone.

Secrecy:
========

Secrecy is a system wherein typically the source code, the recipe, is
_not published_. All people and organisations that come in contact 
with e.g the source code have to sign Non Disclosure Agreements to 
avoid that the code becomes generally public.

Secrecy has been used very succesfully by many SW companies that make
proprietary Software (Microsoft again is a prime example). In many cases 
it is of major importance for those companies that the SW remains hidden, 
so they create an incentive for their customers to buy a source license
(if they want to adapt it themselves) and a barrier to their competition,
not to be able to easily copy all the implementation details that are in
the source code.

As a side note, in certain cases, if you want to base you work partially 
on GPL licensed software, you actually abstain from the possibility to use 
secrecy as an IPR protection model. But you knew that at the time that 
you decided to base your work on GPL code. If you want to use secrecy as an 
IPR protection strategy, use BSD or proprietary software. You have a free 
choice there and no one can blame GPL for asking this condition in return 
for free use of the SW.

The fact that Microsoft is heavily arguing against GPL, proves that secrecy
is an important and succesful method of IPR protection for Microsoft.

Again Secrecy is clear-cut and scalable. If I write an independent piece 
of work (not based on GPL) and I decide not to disclose it, than that is 
my entire right and this is not dependent on 1,000 or 1,000,000 other 
implementations of my idea. Even as a small company, I can contain the
consequences of using secrecy. I just sign NDA agreements with all parties
involved and I am not touched by the size and complexity of all the rest
of the world developing sometimes similar software.

Patents:
========

Patent protection is a system where a Method or Apparatus is _published_
and in return the inventor is granted the right to forbid others to use
that Method or Apparatus for a period of 17 to 20 years.

Patents have been used succesfully, essentially in large businesses to 
balance the R&D investments of the different companies. The story of the
lone inventor that finds this magic invention and earns a billion dollars
with it is very, very rare. In practice, it is much more a weapon for large
companies to fight against now opponents that want to enter a market (e.g.
the RIM case in USA/Canada, where the entire mobile e-mail market seems to
be dominated by a single company, based on the general concept of clever
communication and synchronization of mail between a hand-held and a central 
server; this will cost the consumers millions of dollars of higher than 
normal market prices, that will be partially consumed by RIM to pay patent 
lawyers ...). Same for Amazon. It just means Amazon has a weapon to hold-off
other competitors, defend a monopoly and charge higher fees, based on that
de facto monopoly based on a very basic idea.

In practice, it is obvious to all software people knowledgeable in the
matter that most of these patented software implementations have been used
many times before (and it could even be that we do not know it, but still
they where used somewhere on the globe), but for some reason, these patents 
on basic ideas are still granted and used to kill-off the competition. This
is where we really see a defunct system. It is intended to promote technical
innovatiom, but instead it moves the entire debate an survival strategy from
making the best product to getting the best patent lawyers and best patent
strategy.

So, for small companies, patents remove the cleanliness and scalability of 
the other IPR protection systems. Suddenly, we (as a small company), could be 
attacked at any time, by anyone because we use some patented method (e.g. the 
mouse cursor implementation on my screen is patented, the X Windows redraw 
mode 
is patented, the Flash file system that we build into embedded devices is 
patented, the hyperlinks on my web site are patented, ... and with a little 
search, I could find hundreds of small elements of our daily operations that 
are patented by someone, somewhere, and the impact is growing day after day). 

In theory, we would have to follow-up all patented stuff in Europe to know 
how many additional patents we are violating with each new day the sun comes 
up. This is not realistic for an SME. And if we get attacked, paying the 
legal 
fees is not realistic either, so that is the end of the company, over-and-out.

Thank you Brussels !


Step 2: if it is accepted, practical defense:
=============================================

Due to the heavy lobbying from the rich USA based companies to enforce the
SW patents also in Europe (we have discussed above, why that is in their
interest), I am prettu pessimistic and I fear there is a major chance this 
law will pass.

What to do next. In Belgium, asking a patent without anteriority research 
costs only 75 EURO and including the anteriority research it costs about 
500 EURO. 

It is essential to understand for my reasoning that these amounts are 
"sponsored" by the government. This research is in fact be a _lot_ more 
expensive, but government sponsors this, because they believe it is good for 
the local economy in Europe to foster as much as possible patents. 

So, what I suggest is that with the movement of people that are opposed to
Software patents for the reasons argued above, we start also asking massive
amounts of software patents on anything that we believe could be somewhat
inventive.  After some of these patetns are granted, we could assign the 
ownership of these software patents (trivial and less trivial) to some sort 
of "Free Patent Foundation".

What would be the effects:

- we would clogg the entire system so that the number of actual patents
  that could be used against us would go down.
  
  Please understand that this clogging is an artifact of the system; the 
  current wave of Free Software that is written by many people does not 
  preclude any one else from writing proprietary software independently if
  they wish to do that; but it is only in the _patent_ system that we 
  are forced to enter a common "legal cage" and we could be embarassing 
  each other (which _we_ prefer not to enter, but the legal system forces 
  us into that cage with other people that have different ideas on how 
  to manufacture and protect software).

- politics would see that there is a problem that an increasing amount
  of money is used to sponsor anteriority research into trivial and
  non-trivial software patents, with the only intention to inhibit other 
  parties from effectively using their rights from their SW patents.

- some of the patents we request would pass the examination and could 
  be used as a counterweight if any large, foreigh company attacks a 
  small SME in Europe or attacks a Free Software implementation. If the
  companies that are opposed to SW patents put some money in this Free 
  Patent Foundation, it could also sue those other companies that try 
  to use SW patents to sue small companies of even Free Software 
  initiatives that are not companies, but just people movements.
  
  In this way, this movement could also obatain the "patent pool"
  that is so typical for large SW companies. Thsi could be used to
  actively fight against abuses of SW patents by others organisations.

I know this does not sound like fun, but if SW patents are reality, then 
we may have to adapt our strategy and live by the rules; but also use
these rules in a fully legal manner to defend our own interests.

Step 3: the bloodshed:
======================

If the SW patents are accepted, there will be a lot of bloodshed amongst
smaller SW companies. People will work hard to build up a company, build
SW on an honest basis, not stealing any code from any one, but still
as soon as the product of such a start-up becomes important and threatens 
an established player, this start-up will be killed-off or forced to sell 
too early, before reaching maturity and full shareholder value under threat
of litigation based on SW patents. Similar attacks will occur against Free 
Software projects, that are not in a company, but just an association of 
individuals, working together to promote and develop some common piece of
Free Software (e.g. Microsoft started using its patents on some trivial 
CIFS techniques to attack the Free Software implementation of that 
protocol (one popular project under attack is called "Samba")).

I predict it will take approximately 5 to 10 years before a large enough
percentage of the population will eventually be personally hurt enough by 
the situation (assume e.g. a massive switch to OpenOffice.org that would 
be countered by Microsoft using patents on all kinds of details in an 
Office package to forbid anyone from publishing the openoffice code base
on his web server). We are not talking of illegal copies of music here, we 
are talking of authors that have deliberately granted the entire right to 
anyone to use and publish the code, but are still not allowed to publish 
the code due to SW patents on certain aspects of how an Office package 
typically operates.

Within 5 ot 10 years, there will be enough people that are personally 
hurt (lost their jobs, lost all shareholder value of their company,
where directly attacked because of publishing Free software under attack,
disallowed to use Free software programs that they were using previously, ...)
that this will come on the agenda of the EU again and after a lot of 
bloodshed, the European Parliament, will under local proessure of its 
inhabitants, revoke patentability of SW in Europe.

Step 4: fight-or-flight:
========================

The pessimism on this Software patent issue is so deep with some people 
of our SME, that some of the brighest people are loudly contemplating to 
move to more Free countries (outside of the EU and USA), that do not have 
the intention to include SW patents in their legal structure, until the
decision is revoked. For personal reasons, I do not have this option to 
"vote by feet" so I will have to live with it. One clear consequence is 
that I much less prepared to take risk and reinvest my profits from 
ongoing business into new developments, because I fear a realistic chance 
that as soon as we would launch a product that is succesful and related 
to Free Software, we would be the victim of an attack with software 
patents that would wipe us out immediately (we have a capital of 
170,000 EURO, so by no means sufficient to see through a serious 
litigation case on SW patents). So in our case, this is strongly 
stiffling the innovation risk that I am prepared to take as CEO of my
company.

Conclusions:
============

For my personal case, software patents mark a serious step back in
the way we do business. Up to now, I knew that if I was running an
honest business, not stealing code (i.e. only reuse source code for 
which I received the right to reuse them) and not violating the NDA's
that I have signed, I could safely grow my business and focus on
productivity, development and customer satisfaction, not heavily
worrying about the legal aspects of the business. With the advent 
of software patents, I know that as soon as I make a succesful product
as a small company, I am subject to random litigation by any of my 
larger competitors that can use some more or less related software
patent to kill my company with a single signature on their side (and
yes, my major competitors are USA based, just as in many other parts
of the SW industry). So this clearly reduces my motivation to take
risks in innovation and my potential to attract the best developers.  
The best developers will either move outside of EU, or they will join 
larger companies which face less direct risks from software patents.

SW patents will seriously harm the European SME software industry.

-- 
.--------------------------------------------------------.
|       Mind: Embedded Linux and eCos Development        |
|     Mind is looking for Embedded Linux Developers      |
|--------------------------------------------------------|
| Peter Vandenabeele, CEO          email:  peter@mind.be |
| Mind ( http://mind.be ;)          tel:  +32-16-30.96.66 |
| Vaartkom 11                      fax:  +32-16-30.96.44 |
| B-3000 Leuven, Belgium           gsm: +32-478-27.40.69 |
'--------------------------------------------------------'