3 ms·
I'm one of the pioneers of peer-to-peer file delivery from the late 90s. It turns out that decentralization is a hammer looking for a nail. Almost everything
by orasis 12y ago
I'm one of the pioneers of peer-to-peer file delivery from the late 90s. It turns out that decentralization is a hammer looking for a nail. Almost everything is simpler and more nimble when you centralize. The future espoused by this article ain't gonna happen.
- jeffreyrogers 12y agoYeah, this is definitely true. You can see the trend even in networking, where the routing protocols have typically been very decentralized in nature. Now with the advent of OpenFlow and data center networking we're starting to see a shift towards centralized control of network routing.
- throwawayaway 12y agofile sharing? pretty big nail. a decentralised system supports easy access to files considered to be obscure or "long tail". unless you want to say that bittorrent is not a decentralised system.
- orasis 12y agoAmazon S3 would be a much better choice for files considered obscure. A well designed decentralized system will have to spend a decent amount of effort to ensure that the long tail files don't drop off of the network. If you want fast access speed to those long tail files, now you're going to need even more redundancy to overcome access bottlenecks.
- throwawayaway 12y agobittorrent allows a more ad hoc approach that gives results, whereas there doesn't appear to be financial incentives for someone to pay for amazon s3 to provide the same.
- dreamfactory2 12y agoWhat do you think of things like ostatus and technologies around http://www.w3.org/2005/Incubator/federatedsocialweb/wiki/SWAT1_use_cases http://www.w3.org/2005/Incubator/federatedsocialweb/wiki/SWA...? There are a couple of considerations for me: 3rd party (particularly commercial entities) vs personal ownership of user data and identity seems like a nail which is quickly coming into focus a) with the ways that people are wanting to use social media and b) in light of recent privacy and monitoring revelations. The kind of use cases above in SWAT, combined with p2p network effects seem to offer a glimmer of far more powerful functionality than any closed platform, however large, could ever keep up with (just as we see open source eat proprietary competitors through sheer scale and subsequent pace of innovation). My instinct is that the federated technologies suffered from being ahead of their time and will get picked up again at some point.