Input on anticompetitive characteristic of public code

Carsten Agger agger at
Thu Jun 21 19:04:27 UTC 2018

On 06/21/2018 02:33 PM, Erik Albers wrote:

> * How can we oppose the argument that publicly financed software released as
> Free Software is anticompetitive?
> '
 From a Danish perspective, the notion of free software as 
anticompetitive is completely ridiculous. The Danish public IT 
infrastructure is completely dominated by two HUGE American-owned 
companies, KMD (once owned jointly by the municipalities, now sold off 
to Advent International) and CSC (now DXC Technology).

They oversee non-interoperable monolithic systems, many based on a 
legacy infrastructure dating back to 80s mainframe programs, which are 
of course proprietary - all interaction with these companies cost the 
municipalities and the government huge amount of money, but on the 
bright side they are tied hand and foot to these vendors - until very 
recently, these companies were practically speaking the only owners of 
public data in Denmark (not even joking - through the CPR system, KMD 
for years charged the municipalities about € 0.33 per pop for accessing 
*their own* citizen records - which created a use case for my company's 
free software product "CPR Broker" which caches these records. The rest 
of the market is owned by a few other big players, Microsoft Windows on 
all desktops and nearly all servers, Microsoft AD for identity 
management and ADFS for federation, and a few other large proprietary 
vendors such as SD Løn for salaries processing.

The reality is that within this ecosystem there's *no* real competition, 
there's only monopolies and complete lock-in ("oh you want us to 
integrate to vendor X, even though we can provide product Y at double 
the price. Yes, we can do that - just wait one moment while I calculate 
the price ... oh, pity! Looks like you're not going to be saving any 

In the case of free software, "open source" as they consistently call 
it, the public authorities order what they get from an independent 
vendor and are at any point free to take the code and go to another vendor.

And there is, in fact, a small but burgeoning sector of free software 
vendors, some more commited than others, but delivering e.g. through the 
OS2 collaboration which I mentioned in a previous mail:

Between these vendors there's real competition, as we can in fact can 
and do sell each other's products. And by introducing open standards and 
co-funding, we're in fact very slowly getting into the creation of free 
and municipality-funded competitors to the monoliths. But that this 
should be *impeding* or preventing competition is risible - as I said, 
this very new development is *introducing* competition where previously 
there was none.

> * What can we bring up on the other hand in favor of publishing as Free
> Software from a competitive point of view?
> (except the usual non-dependencies)
I think I covered that in the above.
> * What other arguments can be made in that context to balance an even
> anticompetitive decision pro Free Software (like public duty to supply,
> binding public money with public goods etc)?
Open standards; loosely couple integrations, i.e. in an infrastructure 
with N systems you don't have to make N*(N-1) integrations between 
proprietary systems, you can make N integrations to open standards and 
find a common way for them to communicate.

In fact, the Danish association of municipalities (KL) formulated such a 
protocol, based on AMQP and the Danish OIO standards for public data, 
and called it MOX and later commissioned my company to make a POC for it 
- speaking of deployments we're just getting started, but the system can 
be seen here:


Disclosure: I'm working for a company which is dedicated to creating 
free software and currently serving the Danish public sector and thus 
not without interest in this question.

> * Are there more examples in Europe in that - like in Switzerland - national
> courts decided in favor of publishing publicly financed software as Free Software?
> Looking forward to your input
> Best regards,
>     Erik

More information about the Discussion mailing list