For those already familiar with the free and open source software (FOSS) movement, Eric S. Raymond (ESR) is a man that needs no introduction. As a key figure behind both the practical and philosophical aspects of FOSS, his thoughts and opinions are as interesting to hear now as they were when first set out in his seminal book, "The Cathedral and the Bazaar" (which I reviewed early last year).
In a 2009 talk titled "After the Open-Source Revolution" (given at Villanova University) he covers some important and interesting points. The audience are a little short on questions, and there seems to be the annoying hum of a vacuum cleaner or similar machine running in a nearby room, but thankfully the recording is acceptable enough to summarise:
- Open source project repositories remain a problem. ESR was/is involved with a GPS-oriented project which used to be hosted on a site called Berlios. The site eventually went down, making them recognise the inadequacy of the underlying critical infrastructure supporting all of these projects.There are difficulties in making backups, exporting and transferring data. It's interesting to note that today many projects are heavily locked into repositories like GitHub, which wasn't very well known back in 2009, but has now exploded in popularity, providing yet another single point of failure for so many of these projects. GitHub faces a lot of other criticism, so perhaps a decentralized approach might be the best future option.
- Linux is happily used by ESR's octogonerian mother. Much of the ease of use is attributed to Ubuntu, which made a big difference and lit a fire under other distributions. Another big step-up in GNU/Linux based operating systems was the ease of use in installing software.
- Mobile/cell phones are an area where a clear and immediate difference can be recognised between ESR and more strict adherents to the FOSS philosophy, like Richard Stallman. The latter views them as surveillance devices (quite understandably) and further refuses to use one if any part of it is closed source. ESR, on the other hand, is a happy Android user, and while he mentions that it would be nice if some of the closed source parts of the system were removed, the extent of it being closed source doesn't bother him, as it appears to be leading a trend towards even more open source in that space. He hammers home this point, saying that "Victory is important, purity is not important".
- Microsoft and other similar software companies have to keep making changes to their products because they're in the business of selling software. It does them no good if you keep using the same software. MS thinks that it's most serious source of competition is its own old versions. They have a strong business incentive to create feature churn, but that also creates bugs and UI problems. This is an inevitable by-product of the closed source treadmill.
- On Software as a Service (SaaS): Not entirely comfortable with having important data on someone else's computer, but alleviated somewhat by easy exportability.
- Things we can learn from open source software: "Don't keep secrets. Secrets Suck. Secrecy is the enemy of quality." Human beings are really good at building complex systems that have lots of errors in them. Typically such errors can't be detected mechanically. The only hedge we have against systemic failures in these complex systems is peer review by human eyeballs. "If you keep secrets, you are cutting your throat, that's the lesson of open source."
- Final thoughts and Guideline: "You know you are doing the right thing when the effect of your actions is to increase the range of choices that people have. You are doing well if you are not narrowing their choices."