(Update: Opera chief development officer Christen Krogh’s response: the centralised and decentralised Web will coexist. See below.)

There is something very uplifting about Opera’s vision of a Web that turns every user back into a node on the network, with all the rights and responsibilities that implies (this is the blog post today that explains the idea, and this is an inspirational video.)

The idea behind Opera Unite, in brief: every PC would act as a server on the Web. There would no longer be any need to upload your data to some internet company’s giant datacentre, it would stay under your control and be shared with other PC users through a peer-to-peer arrangement.

Isn’t this what the internet was meant to be like?

It suggests an online world where social networks emerge naturally out of interaction between groups of people, rather than being the artificial constructs of social networking sites. Where users control their personal data, rather than being caught up in the clash of giant clouds run by the likes of Google and Microsoft. Even the trolls of the blogosphere seem to be having a hard time picking it to pieces.

Nice as the idea sounds, though, there are good reasons why a more centralised Web has emerged over the last decade.

Part of it has to do with economics. It simply isn’t as efficient to store and process data in such a distributed fashion. And can you imagine the damage to the climate from keeping all those connected PCs powered up?

There are also security and reliability issues to deal with. Like it or not, your data is probably safer in a datacentre than on your own machine: you’re less likely to lose it to some technological failure. And once your PC is a Web server, do you think you would be better at preventing the bad guys breaking into it than Google is at defending its own datacentres?

Copyright enforcement would also become a headache. Opera says its own servers would play a role in connecting peers in the network. This sounds like the sort of role Napster played in its own file-swapping network. Would that make Opera liable for copyright enforcement, and how would it fulfill such a task?

I’ve put these questions to Opera and will report any response.

Update: I spoke to Opera’s Christen Krogh, who made a good case for why a more decentralised approach might work in certain cases. Masses of personal data already sit on local devices and will never get uploaded to a centralised server. These take the form of photos or other information you might want to share with members of your family or close friends, or access from a different computer. This informal sharing doesn’t require the same service level as that provided by a “cloud” company like Google, so issues of reliability and constant availability aren’t as important.

This more modest vision is far less all-encompassing than the one set out by Lawrence Eng in the blog post mentioned above, and might solve an unmet need on the internet - provided Opera can make it easy enough to install and use, and can attract third-party developers to write applications that take adavantage of this new platform.

Krogh’s response on other issues (and my comments in italics).

Cost: Local storage is now cheap enough for this not to be a real issue. And it is so convenient, it’s hard to see people giving it up in favour of some all-encompassing “G-Drive” in the clouds.

Security: The technique of creating separate “sandboxes” for applications to play in, insulated from other data, is well enough established to resolve security concerns.

Copyright: Any internet user can already download the open-source Apache Web server and set themselves up on the Web, so this doesn’t break any new ground. (If Opera’s technology made it so simple that millions of people installed its new browser/server software, however, the scale of file sharing could be hard to police.)

Tags: browers, Opera

June 16, 2009 10:03pm in Internet | Comment