Jan 7, 2010

Predictions: Future of the Web in 2020

My summary of a long article at:
13-slide slideshow with data on web usage. That page has links to related content and whitepapers

My comment
Most of the research projects mentioned in the article will fundamentally change the architecture and premise of the TCP/IP protocol. I hope Vint Cerf (now with Google) will improve our understanding of the relative impact of these research projects and the risks and opportunities for commercialization. Or give the Google point of view on which vendors might take which critical roles to bring which of these innovations to market.

Article Summary
These innovations might be implemented on the Internet in 10-15 years. They will be tested now on a virtual networking lab being built by BBN Technologies called GENI for the Global Environment for Network Innovations.

Key Topics with brief opinions by moi
  • Virtual networks like GENI
  • Content-centric networking where content types are assigned names so end users can choose content types to 'subscribe' to, thereby getting more relevant data and avoiding some spam.
  • Software defined networking intends to shift intelligence to an external, user-programmable controller, giving users more control. Bad news for router folks.
  • "Floating Cloud' concept where only the tier value is sent, to manage immense routing tables. The problem with all abstractions is information loss.
  • "Sometimes On" wireless networks, aka 'opportunistic networks' where one device that does not have access can leverage a peer, connected device to relay messages.
  • Davis Social Links intends to substitute the universal addressability of IP addresses with social links to provide a layer of security, like in Facebook where the propagation of a message is authenticated by 'verified friends status.'

1 comment:

  1. Sounds like a number of revolutionary networking innovations.

    Given the money invested in the existing infrastructure, if these networking technologies are offered by anyone other than the incumbent network equipment brands, will they be adopted? On the other hand I could imagine a Cisco talking their customer base into these sort of technologies.

    Still, it has taken a long time to get to the level of IPv6 deployment we have now. As of late 2008, nearly 10 years after IPv6's definition, Google found only about 1% of WW Internet traffic was using IPv6.

    [See: http://www.ripe.net/ripe/meetings/ripe-57/presentations/Colitti-Global_IPv6_statistics_-_Measuring_the_current_state_of_IPv6_for_ordinary_users_.7gzD.pdf]

    It's not enough to be a better mousetrap.

    Where the Internet was innovative and still provides opportunities for innovation, the incumbent networking infrastructure will be resistant to disruptive change.

    The potential, however, seems staggering.