Content-Centric Networking & the Next Internet 153
waderoush writes "PARC research fellow Van Jacobson argues that the Internet was never designed to carry exabytes of video, voice, and image data to consumers' homes and mobile devices, and that it will never be possible to increase bandwidth fast enough to keep up with demand. In fact, he thinks that the Internet has outgrown its original underpinnings as a network built on physical addresses, and that it's time to put aside TCP/IP and start over with a completely novel approach to naming, storing, and moving data. The fundamental idea behind Jacobson's alternative proposal — Content Centric Networking — is that to retrieve a piece of data, you should only have to care about what you want, not where it's stored. If implemented, the idea might undermine many current business models in the software and digital content industries — while at the same time creating new ones. In other words, it's exactly the kind of revolutionary idea that has remade Silicon Valley at least four times since the 1960s."
Re:Magnet links? (Score:5, Informative)
Did he just reinvent magnet links?
Closer to a reinvention of freenet.
Or maybe reinventing mdns
Or maybe reinventing AFS
Its been a pretty popular idea for a couple decades now.
Sounds like the principle behind URNs (Score:5, Informative)
See http://en.wikipedia.org/wiki/Uniform_resource_name [wikipedia.org] . This is a very old [and good] idea.
For example: urn:isbn:0451450523 is the URN for The Last Unicorn (1968 book), identified by its [ISBN] book number.
Of course [as the dept. notes] you still need to figure out how to get the bits from place to place, which requires a network of some kind, and protocols built on that network which are not so slavishly tied to one model of data organization that we can't evolve it forward.
Re:Magnet links? (Score:4, Informative)
" (In Jacobson’s scheme, file names can include encrypted sections that bar users without the proper keys from retrieving them, meaning that security and rights management are built into the address system from the start.)"
It sounds like he made them worse; but otherwise pretty similar to magnet links or the mechanisms something like Freenet uses.
Perhaps more broadly, isn't a substantial subset of the virtues of this scheme already implemented(albeit by an assortment of nasty hacks, not by anything terribly elegant) through caches on the client side and various CDN schemes on the server side? URLs haven't corresponded to locations, rather than to either user expressions of a given wish, or auto-generated requests for specific content, in the majority of cases for a while now(and, on the client side, caching doesn't extend to the entire system, for security reasons if nothing else; but it already covers a lot of common web-resource request scenarios).
Now, in a perfect world, "we have a pile of nasty hacks for that" is an argument for a more elegant solution; but, in practice, it seems to be closer to equivalent to "we already have stuff that mostly works and will be cheaper next year", which can be hard on the adoption of new techniques...
Re:Magnet links? (Score:2, Informative)
more like CDN servers, except smarter.
There are already mechanisms for this.
What needs to exist is a hybrid approach where the end users are the origin servers, and the CDN notes operate as capacity supernodes on their local ISP, in turn these ISP supernodes talk to each other. If a piece of content needs to "disappear" the end user removes it from their system, and it will tell the supernodes that the content is no longer available, leaving only users who already have it to talk to each other if they still want it. If a piece of content is meant to be long-lived (eg movies, tv shows) then the originator simply has those data files on a dedicated host node.
What happens today is you get torrent/magnet links which get all the bits from everyone, but when people get bored, or disconnected, there goes your seeds. The other side of this is CDN, where not everyone does this (think most blogs, webcomics, and self-hosted podcasts.) Youtube for example has edge nodes at most ISP's, where as Ustream, certainly does not. This gives a preference to Youtube.
So the hybrid approach is to borrow the EDGE server part of the CDN, and make these a type of torrent seed that expires when the originator says so. This keeps mistakes to a minimum. It also acts as a weak DRM, in that you won't know what the origin server is to try and pull it directly, only from the supernode edge. The supernode edges keep from having to saturate expensive connections like transatlantic/transpacific/wireless links.
But this isn't solving the fundamental problem. Capacity and bad caching practices.
Ads... never cache, because they want tracking
PHP... never caches because the content is dynamic
But these are only small parts of bandwidth, but sometimes they make up large pieces of web pages, for example, having FaceBook and G+ widgets can add 2MB per page, of which only a small portion is cached, due to using cache-busting techniques like affixing ?v=123 to the end of the script, or setting cookies
Re:CCN is not $other_technology (Score:4, Informative)
You need to be able to find things somehow. This requires either some set of central servers, which somewhat defeats the purpose, or a method of broadcast communication that isn't blocked by your ISP. There's a good reason your ISP blocks UDP broadcast and multicast packets - on a large network broadcast leads to exponential packet growth.
For most of us the most limited part of the internet infrastructure is the link from the last router to our house. Picking up my youtube cat videos from my neighbour rather than from a cache server on my ISP's backbone may seem like a good idea, but in reality you're switching traffic from a high-capacity link between my street's router and my ISP, to a low capacity link between my neighbour and our router.
If you're going to cache things on my computer you're going to be using my hardware. That hardware isn't free, and neither are the bits you want to use my internet connection to send. How am I going to be compensated?
Re:Magnet links? (Score:4, Informative)
I think the whole thing falls under the "I have a great idea, but I actually don't have the foggiest idea how infrastructure works now, but hey, I need to a BIG SEXY CONTROVERSIAL headline."
Imagine if even a tenth of the fucking morons out there who pontificate on subjects for which they had no real knowledge at all actually did have that knowledge. My God, we'd probably be terraforming Pluto by now!
The irony is strong in this one.
Anyone pontificating about internet infrastructure who doesn't know Van Jacobson [wikipedia.org] is a fucking moron.