Through running Gizbuzz, I probably sign up to at least one new Web 2.0 service every day. Some I never use again, but a few are genuinely useful, and I think would genuinely be useful to someone else. However, unlike me, that other person isn’t at all interested in the technology (the service itself), but whether it is of any benefit to them (ie. whether it is fun, useful, interesting etc). What’s going on in their mind is a ‘cost-benefit analysis’ if you like, and the irritation of signing up to a new service which they might well not like is a significant barrier to entry.
It is why, when collaborating with friends, I suggest Google Docs and Spreadsheets rather than Zoho - however easy signing up is, it is still a barrier and an irritation when friends already have Gmail (and therefore Google) accounts.
That’s where OpenID comes in; simply put, OpenID removes the need to sign up to a service. I sign up to an OpenID server once, and am given a web addresso. When I want to use an OpenID enabled service, such as Zooomr, I simply give that service my OpenID address, and am redirected to the OpenID server, where I then confirm that I authorise the service to use my ID. I am then redirected back to the service, and am able to use it instantly.
There are obviously many advantages to OpenID, such as privacy, security and interoperability, but the headline grabber for me is ease of us, and it is why I think OpenID could be a major force in preventing the Web 2.0 boom from becoming a bubble. If people are able to signup to a new service in probably less than 20 seconds, without any privacy concerns, confirmation emails or other annoyances, they are far more likely to adopt a less conservative attitude to trying out new services, and as such, are probably likely to find themselves using more. The more people that use a web service, the more money it makes and the less likely it is to go under. Thus the bubble is less likely to pop.
I’m wondering why you include privacy and security in the list of OpenID’s advantages. OpenID, as I understand it, is simply about establishing identity. Whether or not the OpenID provider is secure is outside the scope of the protocol.
I’m implementing OpenID in a personal project, and I love it (once I got my head around it, it made my life as a developer extremely easy), but I don’t think security and privacy are really in the list.
Also, how/when do you forsesee the developmen of the requisite critical mass for OpenID to be broadly adopted? The fact that AOL gave every AIM user an OpenID doesn’t seem to have sparked much adoption from web sites. Maybe I just haven’t seen them?
“Development” ends with a “t.”
In terms of privacy and security, I had in mind APIs. For example, IMified currently takes your username and password for all the services which it plugs into. With OpenID, this wouldn’t be necessary, because as I understand it you would only have to give IMified your OpenID address (assuming all services being used also utilised OpenID), and then, since the other services are already approved, IMified would never need to be given the user credentials. I might have got this wrong though.
Critical mass - an interesting one. For OpenID to be successful, I think it is initially crucial that it has adoption from services rather than users. Which makes AIM’s distribution of OpenID’s useful, but not crucial. I initially got an OpenID because Zooomr required me to have one. Now I use it on other services.
ironically but hopefully this site be OpenID enabled.