But the reasoning that is advanced is rarely as forthright. It is always couched in devious language but the arguments are generally as mendacious as they ever were.
The latest "why we don't need the GPL" argument comes from one Donnie Berkholz of an analyst firm called RedMonk. It begins with a falsehood: "In the early days of the GPL and copyleft software, it played an important role in forcibly training companies how free/open-source development worked." No, it did not. The GPL, being a free software licence, had nothing to do with open source at all; it was about ensuring freedom for users, freedom of the political kind.
The man who wrote the licence, Richard Matthew Stallman, was keenly aware of the one human factor that nobody could control - greed. Hence he put in the requirement of share and share alike to ensure that nobody could profit from another's efforts without contributing back something themselves.
The concept of open source was born in the late '90s, an initiative by Bruce Perens and Eric Raymond to make what was called free software acceptable to businesses. It is a development method. It has nothing to do with ideology or freedom. There are numerous open source licences that do not even qualify as free software licences.
Berkholz contends that tech companies understand the benefits of GPL-ed software at this point. If they do, one wonders why is he arguing that the GPL isn't needed any more - if something is beneficial, why do we need to get rid of it? It's something like saying that since people now understand what a seatbelt does, we can do away with it.
Berkholz takes it for granted that use of the GPL is declining, referring to another RedMonk article as evidence. There we find the author is relying on figures from Black Duck Software, an organisation which keeps secret its methodology for coming to such a conclusion. Nevertheless, Black Duck's figures are often cited as gospel. It reminds me of the Goebbelesian principle - repeat something often enough and if people have no access to an alternative point of view, they will come to believe it.
A couple of months ago, Free Software Foundation executive director John Sullivan presented a paper titled "Is copyleft being framed?" wherein he pointed out that, "Some of the most common licensing numbers used are published by a company who does not share their methodology. That's not science. FLOSSmole is an interesting project to generate the data in a veriï¬able way."
As Red Hat developer Matthew Garrett wrote some months ago, figures from FLOSSmole show that the type of licence is dependent on the type of developer; web projects tend more towards permissive licensing while projects on Launchpad, for example, tend more towards GPL licensing.
Nobody knows for sure whether GPL licensing is declining or not. But if you talk to a developer and ask whether he or she is agreeable to having their code locked away by X or Y proprietary outfit, the answer is predictable. Developers want to benefit from their code and that, in proportion to its usage; nobody wants a neat little utility that he or she devised to end up making a fortune for some proprietary company without them getting a decent share.
If Berkholz's article cited facts and figures to justify his conclusions, one would be more inclined to accept it. But he makes vague statements such as, "But today, open-source software is much more familiar, and tech companies generally understand how the model works, if not always when to open their own software. So the initial mission served by the GPL of force-feeding open-source development isn't as necessary as it once was, at least among educated tech companies. I suspect the lifecycle has now shifted."
To me, this is one more case of repeating something and finally hoping that it will become the accepted wisdom. This happened in the run-up to the 2003 invasion of Iraq too; the method is the same.
Open source is a great marketing point these days. It carries with it a connotation of honesty and accountability - and after the scams that led to the global financial crisis of 2008, anything that seems to even vaguely connote openness is welcomed.
It is instructive to note that the GPL does not make it impossible to build proprietary products; it only stipulates that derivative works of a GPL product need to adopt the same licence when the joint codebase is distributed. The Linux kernel itself has several proprietary firmware blobs which are not derivative works and hence the code for these blobs does not need to be distributed.
But people who are in a hurry to make money, those who have only green dollar signs in their eyes, see the GPL as a hurdle. The fact that everyone benefits from such a licence is not of great concern; no, an individual company's bottomline is all that matters and the rest be damned.
It is somewhat ironical that Berkholz has chosen to spout this argument barely a fortnight after Red Hat Linux announced its first year of billion-dollar-plus revenue; this is a company that sells GPL-ed software and hasn't done too badly going by the balance sheet. But then such facts would get in the way of specious arguments, I guess.
We need the GPL more than ever now because human beings never learn from history; greed was the undoing of the world economy three years ago but all the signs are that the same trend continues. Anything that gets in the way of our doing so is something we need to preserve. (Indeed, if history was the guiding principle, Western troops would have left Afghanistan long ago. But then, as I said, we never learn from the past and are doomed to repeat our mistakes.)