When you connect to a website or service on today's Internet, connections to servers hosting site contents are established. If a server is not available at the time, some part of a website or all of it may not be displayed in the program you are using.
While there are some options to grab the data anyway, caching for instance, you would still rely on a limited amount of servers and companies to deliver those contents to you.
Torrents on the other hand work in a different way. While they are seeded initially by one user or a group of users, the data is made available by downloaders as well, so that the pool of potential content distributors increases with every user downloading files from a torrent.
The basic idea behind SyncNet is to introduce a web browser that uses a similar system when it comes to making available web contents.
Whenever you access a website using the browser, it is stored in its entirety on the local system. The next user who accesses the website will download it from the original server and you.
The more users access a resource, the more it is spread across machines and devices, and the less likely it is that it will become unavailable.
A side-effect of this is that it will decrease the load on the original server.
There are caveats however as the author mentions. One of the biggest issues is the loading time of websites. When you load a website the "normal way" in a web browser of choice, it takes no more than seconds usually to download, render and display it on the local system.
SyncNet in its current state downloads all files for a requested site according to the author. He has plans to improve that behavior though to speed this up in future releases.
Another issue is that only static contents are supported. While that ensures that many websites can be accessed using the browser, dynamic ones cannot be or only partially.
The method to make a new site available is also less than ideal. You need to add a directory of HTML files to Bittorrent Sync. The author has plans however to change that to make current websites available via the browser.
Last but not least, SyncNet is only available as source code and not a binary that you can run on your system.
SyncNet is an experiment more than it is something that many users will start to use soon. It may be the way to go forward on the other hand, with more and more countries starting to censor Internet resources and service providers trying to undercut net neutrality to make a quick buck.
What's your take on Internet decentralization? Know of any other projects that aim for it?