Open Web Definition (Version 0.4)

[Update: did a slight update of the definition]

Here's another stab at a (short) Open Web definition. I put out a call for folks to help craft a one or two sentence concise definition of the Open Web awhile back. I'm not quite sure if we are there yet, but here is one based on boiling down alot of the feedback from many people. In the comments below give this a +1 or a -1 with a rewrite to make it better (and hopefully shorter). Expect to see more rewrites until we get this tight and have general agreement.

Open Web Definition:
"The Open Web is an interoperable, ubiquitous, and searchable network where everyone can share information, integrate, and innovate without having to ask for permission, accessible through powerful and universal clients."

Here is a litmus test definition we can apply when wondering whether something is part of the Open Web and actions to take.

Open Web Litmus Test

Does it have:
  • Composability - The ability to innovate, link, contribute, search, and integrate without red tape, fear of a lawsuit, or having to ask "please?"
  • Interoperability - The ability for developers to interoperate without having to know of each others existence
  • Ubiquity - The ubiquity of a set of open technologies and services agreed upon by the widest possible community
  • Universal Client - Empowering and evolving the browser and web technologies as a universal client
+1 or -1 on this litmus test? If you give it a -1, try to not have the bullet points expand; we want this to be a concise list out of which other things can fall out, not a laundry list.

The goal is to get something as tight as the Open Source definition and work of the Open Source Initiative. We also want to have a short litmus test of characteristics that can help to determine both whether something is part of the Open Web as well as to guide actions and results to make it a reality.

Thanks to everyone who helped either by directly answering or by blogging about what the Open Web is in general:

Jeff Bailey, Jeff Fisher, Kevin Marks, Dion Almaer, DeWitt Clinton, Anthony LaForge, Jason Chen, Jason Robbins, Christopher Keene, Jeffrey S Miller, W3C, Jeff Scudder, Justin Mattson, Talin, Logan Hanks, Tim O'Reilly, Rod Chavez, Anthony LaForge, Glen Lipka, Ray Cromwell, Chris Dent, Joseph, booch, Tom Trenka, Keif, Alex Russell, polterguy, mojave, ialexi, Joe Walker, David Eaves, Mark Pilgrim, Kevin Mullins, Simon Willison, Sumit Chandel, and many more.

As an aside, I liked Jeff Fisher's tongue-in-cheek definition of the Open Web:

Jeff Fisher: "The open web is the great mystery at the heart of all of us"


Brad Neuberg said…
From Dylan Schiemann, who doesn't have an account to comment:

"I'd love to leave a comment, but I don't seem to have an account handy
to login.

The only part I don't like is the "powerful and universal client" which
implies only one client. That said, I can't think of something better
at this point, so +1."
Anonymous said…
Brad-where does "view source" fit into this picture, if at all. I think it's a very powerful proof-point for an "open web" that it's essentially open-source-by-default, at least to some extent, and we can all learn from each other's work this way. Also, part of the Open Web to me is that it's based on standards and specs that aren't owned by any single company (cf. flash/silverlight). Do either of these issues belong in the short summary or litmus test?

Great effort, thanks for pushing this forward and helping make it concise! Thanks, js
cdent said…
I wrote a response on my blog. Here's an excerpt:

This is nice, tight, hits the right parts and works well enough, so +1. But for sake of discussion: like some of the other comments I find the use of the term “universal client” potentially misleading. It implies, I think unintentionally, one client uber alles. What’s great and critical about an open web is that the client is anything that can talk HTTP: a web browser, an extensive bit of code, a little shell script, telnet to a socket.
Brad Neuberg said…
There were several comments about the universal client part:

I wanted to make sure to capture what is happening in Ajax, which is to strengthen the browser as a universal client. In general, this has actually been a trend on the web from the beginning. The Open Web vision includes strengthening this component of the web, including having better vector graphics; more programmatic tools for building things like GMail, Google Maps, etc; better audio-visual support, etc. I would say the Open Web vision definently has the browser as an important component, with the ability to access the web in other ways as you suggest, such as through telnet. However, there are components of the web that you could never access through Lynx, for example, that only a more powerful client such as a modern browser can work with.

I agree with Dylan though that it should say "powerful and universal client_s_", with the s added to connote that there should be multiple clients. I'll update.

In terms of View Source, I think thats important but don't want to have the definition expand more than it is. In terms of standards, that falls out of earlier parts of the definition without needing to be explicity stated.