Big Time Problem Statement, and Alternate Solution
Big Business tends to benefit more than smaller developers at a scale of massive transnational ‘cloud’ service provision due to the freerider problem in Free and Open licenses.
Smaller software developers are missing out on recouping research and development costs and project funding. This results in new features and maintenance for useful software going unplanned, and in some cases, useful projects being abandoned.
The freerider problem is well understood as a type of collective decision problem in economics.
Software is what is known as a ‘Toll Good’ or ‘Club Good’.
These are goods that are non-rival, in that use by one person does not prevent use by another. Toll goods are also excludable, which means persons and firms can be prevented from using them.
Excluding users or use cases is done through software licensing, normally for economic reasons although some exclusions claim to solve problems relating to technical contributions, (‘Copyleft’/‘Share-Alike’) or moral dilemmas (‘Ethical Source’).
By leveraging the excludable nature of software, a new licensing model could prevent cloud service providers and other large scale freeriders from obtaining benefits at the cost of smaller developers.
Not all Big companies will be freeriders. Some do contribute either labor time in the form of code or sponsorship, usually at their own discretion.
Also, many smaller developers and individuals will freeride too, but monitoring all uses of software is impractical and so proprietary software is the only solution if we want to prevent freeriding across the board.
The demand for immediate availability of computer system resources, especially data storage and computing power, without direct active management by the user has been growing rapidly.
These ‘clouds’ are distributed over multiple locations, each location being a physical data center, owned by, or at least controlled by a massive tech oligopoloy spearheaded by one of the Tech Giants like Google, Amazon Web Services (‘AWS’) or Microsoft, (‘Azure’).
Information about the scale of these companies operations at the technical level is hard to come by, and so proxy measures of size have to be used.
The requirement is for a license that allows free use to non-freeriders and excludes freeriders, but only at a scale that is non-trivial.
Insisting that an amateur or home user pay for software while a large scale cloud service provider gets the largest slice of the pie feels mean spirited, counter-intuitive and counterproductive.
So, we are left with two problems: How to identify Big freeriders?
The most efficient way to do this is to be able to identify, and exclude any Big businesses that don’t pay.
This requires a license that defines the properties of a business that make it ‘Big’ for the purposes of assessing the scale of freeriding.
A business that is ‘big’ in forestry, mining or agriculture, perhaps with a turnover of billions but just uses a program in their accounts department with has maybe one hundred personal machines is not the same as a company like Amazon or Microsoft who can upload the program onto their system and make it instantly avaialable for millions of consumers in return for a subscription.
The rationale for asking a Big business to pay, to limit the freeriding problem is to tie the license terms to the type of freeriding we want to discourage.
This suggests that the Big Time license, perhaps ought to consider the type of use rather than the type of entity.
The AGPL uses a definition that targets companies that provide ‘Software as a Service, (SaaS’) and seems to demonstrate some success in deterring use by SaaS providers.
By binding the license to SaaS providers only won’t eliminate all freeriding, but perhaps might be able to do a better job of eliminating freeriding among SaaS providers as the terms will be more specific and thus, less ambiguous and thus easier to monitor and enforce?
The current method of excluding all non-paying, Big businesses seems to be somewhat arbitrary and logistically complex since it uses multiple ‘tripwires’ to catch companies that slip through the net and yet it ignores freeriding among smaller entities - all the while perhaps being too ‘opportunistic’ in the case of large businesses that may well be able to pay for a license but would possibly be just as happy to use an alternative program should they be asked to pay under this license?