There is a valid point in this argument, namely that bandwidth is limited, despite the advertised claims of numerous ISPs, and that network load brought about by constant peer-to-peer filesharing impinges on those users who do not participate in this type of activity. What kinds of solutions are there to this problem?
- prevent people from being able to share files by disconnection them?
- introduce a bandwidth 'capping' scheme?
- improve the network capacity by laying more cables?
- replace the exisiting cable network with fiber optic?
The second option has already been introduced by a number of ISPs (once again, Virgin Media, I'm looking at you). Traffic shaping seeks to cap the amount of data a user can upload and download within a given time of day. Once a threshold has been reached, the ISP essentially throttles the connection making larger downloads very slow. This is a distinctly unpleasant practice for families who own multiple computers in their homes and have large connections in order to cope with the demands of, say, a family in which one person may be downloading a game via the legitimate Steam service or others like Xbox Live and WiiWare. Combine this with the legitimate purchase of music, TV and film content from the iTunes store, as well as visiting video streaming sites like BBC iPlayer or the HD channels of YouTube and Vimeo, and you can easily push at the limits of the usage cap. I haven't even mentioned upgrades to operating systems like Vista and OS X (the latter is not known for small download sizes) yet. The idea of unlimited broadband is looking like a far off dream now...
It seems like networks are reluctant to spend more money on digging up roads and upgrading/replacing the existing infrastructure, especially when the business case for providing these expenses seem weak when compared to replacing dial-up with broadband. Will customers really want to pay high premiums for faster data transfer? It seems like the UK broadband infrastructure has been a victim of its own success in recent years. Penetration is high which is good for ISPs in a competitive market but broadband is less exclusive and taken from granted now. Customers expect a certain level of service, especially those customers who have paid for premium packages for many years.
Clearly, there are problems with all of these scenarios. In relation to the improving the network infrastructure Schofield make the analogy that "it's like building more roads to 'solve' the car problem". I'm not sure this is quite so simple especially given that the eco-impact of building more roads for CO2-producing cars doesn't play out so dramatically when alternate solutions for fiber laying exist (such as H2O's decision to use the sewer system to house 100Mbit/s cables). The problem seems to be that the growth of broadband penetration has not been matched by structural reform. 100Mbit/s connection are already de rigour in some nations (eg Sweden, France, South Korea, etc), and the UK may be lagging behind.
There is an argument which suggests that faster technology breeds innovation, which Schofield attributes to Google's Vint Cerf. When usage is restricted, businesses who offer excellent services like Vimeo's HD content or the BBC's iPlayer may find themselves unpopular, especially when customers may have to watch their data use.
It seems like the BitTorrent protocol is being targetted in particular. BitTorrent is a fantastic protocol which allows for the rapid dissemination of files across ring-fenced networks and the wider web more generally. It works so well because of its decentralized nature which males it more efficient than if every person wanting access to a file was to try and get it via FTP. Just look at the problems Mozilla had when attempting to break the official download record for Firefox 3 on June 17th. Demand was 'overwhelming', servers ground to a halt and crashed temporarily. BitTorrent distribution would have circumvented this problem.
There is an assumption that all peer-to-peer traffic is driven by illegal activity when this isn't always the case. Schofield goes so far as to label the protocol as 'poisonous'. There are grounds for criticism when "10% selfishly grab around 75% of the internet's bandwidth", but is it fair to blame the protocol for its ease of use and speed of distribution or those users who want to get the advertised connection speeds they are paying for?
I suspect that there is no easy answer to this question. I'm not even convinced that it (or Schofield) asks the right question. It may be that ISPs need to included a caveat or two in future proclamations, similar to stock market share prices - something along the lines of 'your broadband speed may go up or down in line with the user base of the network'. Can you imagine ISPs offering a sliding scale feature on their front pages showing the impact of having more customers on their network? Then they really would be victims of their own success...
Blogged with the Flock Browser