Skip to content

Front About Search
  You are not logged in Link icon Log in Link icon Join
You are here: Home » Members » Edd Dumbill » Old WtW Stories » P2P may succeed where tired web apps fail

Log in

P2P may succeed where tired web apps fail

by Edd Dumbill -- 2000/10/24

Despite the nagging feeling I get that there's something not quite right yet with so-called peer-to-peer applications, exemplified by the recently-released Groove, there are many ways in which they could achieve the aim of a writeable, collaborative medium before the Web gets there.

All of a sudden in Groove we have a user-level application that makes secure communication, independent file-sharing and structured information a snap. For years now the few who have blazed the trail for such technologies have been laboring under poor user interfaces, and poor interoperability from applications. That landscape is changing, but--somewhat unexpectedly--not on the Web.

The Impasse of the Broken Web

It's an old story. To move forward, a break has to be made with the past. It happened with the Web--those Who Knew looked down upon the Web's transport protocol, HTTP, as flawed. Yet, the empowering effect of the simple HTTP and HTML created the face of the Internet as we know it today, despite its technical weaknesses.

In turn, the Web itself has fallen prey to a laziness from those who create the tools. The lack of imagination in the common email client or web browser is staggering. Seven years on from the first browsers, and the most detailed information a computer can reliably extract from a web page is still the <title>. For searching, we're restricted to brute-force full-text indexing from tools like Google.

Many who author professionally for the web prefer to code their HTML by hand, or roll their own content management solutions, as the quality of editing tools has been so dismal. That basic technique separation of content from presentation is still being discovered by some as a "revelation." A lack of adherence to or encouragement of standards means there are still very few documents about which one can reliably discover important information like a summary, keywords, author details, etc.

Actually publishing content to the web, and sharing with friends or colleagues, is also still frustratingly difficult. The web's unsophisticated security infrastructure means that it's both agonizing and ineffective to protect your data--and even more so to do that for diverse groups of contacts and content. Add to that the mechanics of obtaining an ISP account and maintaining web space and dependable URLs, and you can't blame the users for the tangle that is the web.

The only folks who have achieved these things (accessible documents, sensible URLs, embedded metadata) are the die-hard minority. Everyone else has either used the broken tools provided by vendors who don't care for standards or interoperability, or just not bothered at all.

Email, though somewhat more basic in nature, hasn't fared that well either. Basic hygiene factors of communication such as digital signatures and encryption have been painfully slow to emerge, and difficult to use and understand. Valuable groupware features have been overlooked in favor of colored text and HTML-mail.

The humble email has also become the general beast of burden to overcome the deficiencies in the Web's infrastructure. Documents get emailed around an organization, maybe to hundreds of recipients, rather than published once on a web server. Programs, pictures and other files all get shared the same way. Web tools lack the directedness and convenience of email, making file-sharing by mail the only practical method (thereby losing out on any help the web may have brought, such as organization, classification, etc.)

P2P Makes the Leap

Without any history on its shoulders, the array of custom applications that make up the nebulous "P2P" cateogry are free to break out from the fustiness of email and browsing applications. They can change the rules and do things "properly," given the chance.

Want proper metadata about files? Read it from the filesystem or make the user add the data in before they can go on. Want secure communications? Don't offer any insecure methods. Want decent file permission management? Bind file-sharing into a user's contact list.

By changing the rules of engagement, P2P applications are able to embrace the more sophisticated requirements of today's multi-layered online user base. When critical mass is reached, as we've seen with Napster and instant messaging, then the extra granularity and control pays off.

However, it still seems that many P2P-type applications struggle to prove as compelling as the Web. Broken as the web may be, it has the advantage of a low technical level of entry, an implicit openness, and limitless flexibility. New P2P technologies hide both the mechanics and often the data from their users. The Web won't go away, but it must change.

The Web Must Learn From P2P

The short-term advantages of P2P thrills are a substitute for a more advanced Web. Ultimately, these myriad file-sharing programs can be replaced by effective metadata distribution and access control on the web. Proprietary communications will be superseded by XML-based transports and discovery mechanisms. Custom clients will stumble on interoperability issues.

Yet the P2P "revolution" is a sharp jolt for the Web, and may be the way that users are shifted out of the status quo into secure, distributed communications and computation. As P2P apps take off, the Web must learn and incorporate these applications into its own infrastructure.

The end-goal is both the usability and immediacy of applications like Groove, but with widespread accessibility, openness and democratic access afforded by the Web.