Free as in Freedom,
- not free as in beer*
Richard M. Stallman’s quote, well known to free software advocates, brings clarity to an ambiguous term – “free” can refer to freedom, or can mean “gratis”; both can be on-topic as far as software is concerned. It has also become, in a way, the motto of the free software movement.
Many initiatives draw inspiration from free software philosophy – libre culture movement, Wikipedia, open educational resources, and many other, base on ideas floated by and tested within free and open source software projects. The “free as in freedom, not free as in beer” thought is also present outside of the freedom-loving software developers’ world.
Usually it’s the first part of the quote that gets the most attention and focus. It is about freedom, after all, and not about whether or not something is available gratis. This focus was (and is) required to clearly demarcate software, culture or educational resources that give and preserve freedoms of their users from those that are just available cost-free (allowing for access, yet denying the rest of the Four Freedoms); the priceless from the zero-priced.
We might need to change that accent, however. Software developers, artists and educational resources creators, libre or not, have to eat, too.
Richard Stallman had introduced a simple yet effective criterion of whether or not a given software (or any other resource, for that matter) is freedom-preserving – its license has to guarantee:
- freedom to run/use the program without any restrictions;
- freedom to examine how it works and to modify it;
- freedom to distribute it further;
- freedom to distribute one’s own modifications of it.
To make extending the set of libre software easier, in the first free software license, the GNU GPL, one more trick has been also used – copyleft, the requirement that all software based on GPL-licensed software will also have to be distributed under the same terms.
Copyleft clause has since become a point of contention within the free/libre/open-source software community. The debate between its detractors and proponents is as vivid today, as it has been 30 years ago.
The MIT/BSD crowd argues that copyleft denies developers of derivative works (in this case, software based on a GNU GPL-licensed project) the freedom to close their project or change the license.
The GNU GPL side points out that even if that particular freedom is denied in such a case, it’s for the greater good – others, including the users of the derivative work, have their four freedoms preserved.
The debate, then, concerns the freedom of the derivative work’s author to close that work, versus the four freedoms of all users, ever. And of course, this is relevant not only to software.
Within the software development world and outside of it the copyleft clause tends to be considered “bad for business”. Derivative work authors would like to be able to close their works regardless of the licensing of the originals, so as to earn a living on them – after all, how can one make money on something that is free to copy at will?
The answer lies with new business models, compatible with the culture of sharing (and sharing of culture). Crowdfunding, voluntary payment-based models, making money on merchandise (like band t-shirts) or concerts, and (in the case of software) selling services like feature implementation, support, or deployment, allow the creators to thrive and earn a living even though – or, as often is the case, precisely because of – fans sharing of their works.
These are not obvious and seem uncertain – and yet more and more often they finance productions, large and small. On the other hand, the “tried and tested” ways of making money on creative work are not a guaranteed way to make a profit. Even more so with the market being saturated by huge companies.
Preference for non-copyleft licenses might stem from lack of trust to new models: “I might want to sell a closed product based on this, what then?” However, if I can close something, others can, too. We’re all worse-off.
The Heartbleed debacle illustrates this well. A trivial software bug in a popular free software library used on the Net by big and small alike to provide secure transmission had huge consequences for the whole FLOSS ecosystem, and broader: for the whole Internet. It also remained undiscovered for years.
Software involved – the OpenSSL library – is available on a non-copyleft license. It’s being used by companies, including most of the heavyweights (including Google, Facebook, and Amazon), in their products and services.
They use it, but do not really help develop this crucial piece of software. OpenSSL developers did not have the funds for regular code audits that would have discovered the bug long before it caused any harm.
Large companies also do not share their modifications. OpenSSL’s license does not require it, so why would they? Turns out Facebook modified their OpenSSL version in a way that (inadvertently, probably) made it insusceptible to the bug.
Had OpenSSL used a copyleft license, requiring sharing modified code with the community, Heartbleed might have been discovered much earlier, causing much less harm.
Not free as in beer¶
Free software, libre culture, open educational resources development has its cost. Thousands donate their time and expertise, and share effects of their work. It often is overlooked, usually when while arguing for use of FLOSS the “it’s gratis” argument is being used.
It is not. Time to start properly valuing the work put into those initiatives. And to support them, also financially.
Copyleft, turns out, can help here too: if nobody can close my work, I myself can also use their enhancements. We’re all better-off.