Page:Aaron Swartz s A Programmable Web An Unfinished Work.pdf/59

From Wikisource
Jump to navigation Jump to search
This page has been proofread, but needs to be validated.

7. BUILDING FOR FREEDOM: OPEN DATA, OPEN SOURCE 47


This is undoubtedly useful, but again, a rather pale notion of freedom compared to the four freedoms that free software provides. If Amazon was truly free, you wouldn’t just be able to write programs to automate your usage of the application, you’d be able to change how the application actually works.

The obvious solution to this challenge is simply to release the software on the Web server under the GPL or some other free software license. Then anyone could download a copy and modify it to their heart’s content. And a new version of the GPL has been released, AGPLv3, which requires that people who use its software in web applications make their software available to the application’s users under its free terms.

But only a completely asocial web application consists purely of software. The vast majority of them are interesting because they give you access to data contributed by other users as well. For example, the software that lets people edit web pages is just about the least interesting thing about Wikipedia. The reason the site is so popular is because so many people have put their accumulated knowledge into that software.

Wikipedia has addressed this by going one step further—not only is the source code free, the data is too. Anyone can download a copy of the Wikipedia database (excluding users personal information) and start up their own copy of Wikipedia based on it. And then they can modify their copy of Wikipedia’s software to work however they please.

It’s beautiful in theory, but in practice, of course, nobody does this. Even if your version of Wikipedia was full of fantastic new features, it would still be nearly impossible to get anyone to use it. People use Wikipedia because that’s where all the other people are; it’s practically impossible to get everyone to switch.

For Wikipedia, the problem is somewhat ameliorated by having some pseudodemocratic control over the site. So Wikipedia is run by a board elected by (a tiny subset) of its users and the board has nominal control over the software and modifications that get made to it. But this is still a far cry from the freedom GNU/Linux users have in the non-networked world. Running for office, getting elected, then pushing your patches through a change-resistant bureaucracy is a lot more difficult than modifying some source code files on your computer and restarting.

And so, the hard-core partisans of software freedom propose that we will see the pendulum once again swing away from centralized server computing and back to a world where we all run applications on our local machines. Only this time, instead of being applications that don’t use the network or only talk to a distant