RMS is fighting the "noble fight" again:
But what does it mean to "use free software"? Does that mean escaping from proprietary software, or merely installing free programs alongside it? Are we aiming to lead people to freedom, or just introduce them to our work? In other words, are we working for freedom, or have we replaced that goal with the shallow goal of popularity?One of the most basic freedoms we can fight for is the freedom to choose, and by extension, the freedom to make our own decisions. Swapping a tyranny of proprietary software for a tyranny of of ethically-correct non-proprietary software is still living under a tyrant.[...]
For instance, what should we say when the non-free Invidious video driver, the non-free Prophecy database, or the non-free Indonesia language interpreter and libraries, is released in a version that runs on GNU/Linux? Should we thank the developers for this "support" for our system, or should we regard this non-free program like any other--as an attractive nuisance, a temptation to accept bondage, a problem to be solved?
We should welcome software like nVidia video drivers, Oracle databases and Java environments running on top of free software. Sometimes, it's important to solve a problem today, not labor today to produce solutions seven generations hence.
Playing these absolutist word games completely ignores the fact that the software industry is a dynamic, evolutionary environment. It's a conversation. We are certainly better off today than we were 20 years ago, thanks in some part to free software, but also thanks to proprietary software, and hybrid models that mix both. Today, some needs are adequately met with 100% free software stack. Some needs demand a 100% proprietary mixture. Many needs can run with some mixture of both.
Twenty years from now, we'll still see free software, proprietary software, and mixtures of the two. The relative proportions will certainly change. But proprietary software will never go away, and that is not a failure of the free software movement. If anything it is a success, because it insures that we have freedom -- the freedom to choose our taskmasters.
There is a huge amount of free software for a large number of purposes. I'd estimate that about 25% of all specific applications have a free software program capable of directly filling the application, and another 25% that are largely filled but would require some customization. The other 50% are mostly specific to a particular domain (industry-specific, tax law jurisdiction specific, etc.) and less general purpose.
A mixture of free and proprietary favours the free, the most widely-used appliations are met first but others follow. Eventually, proprietary is left fighting over ever-smaller ever-more-specific areas. It soon becomes easier for a small group to extend free software that already does 90% of the job for those areas than for a proprietary company to do so - the proprietary company has to compete with the free 90% as well as the new 10%.
At the moment, 75% of Microsoft's software is necessary for them to break even against free software - they must keep apace in its development but they are getting an ever-decreasing amount of revenue from it. This will accelerate - much of their revenue now is from momentum dating back to when free software could match only a tiny portion of their capabilities.
Proprietary software companies other than Microsoft have even less ability to compete against free software in general - they only thing the can do is compete in specific areas. So, for example, Oracle competes against MySQL and PostGres for now, but only in the higher end of the market - the low end is becoming ever more completely in the free softwre camp, and the boundary between "high-end" and "low-end" keeps moving higher.
The only place I see where proprietary software will thrive is when it finds a new application area and gets patent rights to that area. Just think, Microsoft might someday only be able to make money from places where it truly does innovate in a significant way - and if it doesn't keep a large enough set of such innovations coming, it won't be able to afford to support the rest of the support code to build those innovations upon.
(Warning, of the numbers given in this message, 55% are guesses on my part, the other 45% I just made up.)
