Facebook Connect Will Be Game-Changing...and Dangerous | Hacker News Reader

Facebook Connect Will Be Game-Changing...and Dangerous

readwriteweb.com

6 pointsmarkbao8 comments

On the one hand, you have to admit this is revolutionary.

#1 No it isn't. Remember Microsoft Passport?

The web will be transformed from the still (somewhat) closed system it is today, to a massively social experience

No it wont. See #1.

Well, the problem they're addressing is going to be solved somehow by someone, it will probably be either this or openid and xfn. I think FB has head start so far.

Passport btw, offered no benefits really for app developers, not the same case here.

The problem with Grand Unified Login Theory is that it often goes against the primary goal of having your own web service which is to build your own user base. In the long run your user base is worth more if it is uniquely yours as opposed to just being a subset of someone else's user base. In this case if Facebook caught a cold then any web service heavily relying on Facebook logins would get a flu. No thanks.

"The problem with Grand Unified Login Theory is that it often goes against the primary goal of having your own web service which is to build your own user base. In the long run your user base is worth more if it is uniquely yours as opposed to just being a subset of someone else's user base"

Why? The value of your "userbase" is not in your DB table, it's the users who interact with your service and arguably find it useful or fun. For many types of apps, allowing the user to interact with their friends is extremely important in achieving the above.

I don't really get your "flu" analogy, maybe you can say what you mean, are you literally worried about FB's downtime affecting your web service? that could be a valid concern for some web apps, I guess, realistically not for most.

The difference between Facebook Connect and Microsoft Passport (now Windows Live ID) is reputation. With Facebook Connect you may be able to see that Bob friends with Jane, and since Jane is a good person to deal with then Bob is probably not too bad either. Social Networking + Payments is a killer combination if you can get people to participate.

If the average distance on the social graph between any two people is on the other of 6, 8 or even something larger like 20, is social graph distance a good predictor? Additionally, the prediction quality is lessened when people are not very discriminating in who they link up with, or all edges on the graph have the same weight or can not be tagged with useful, meaningful metadata. Who's going to spend the time to provide that metadata anyway?

Wasn't there some research a while ago that showed that six-degrees of separation is more true than is intuitively not?

So it's like turning your whole web site into a giant Facebook App, which as we have learned, has no business model.

Nothing Facebook has ever done has been game changing, ever; that'll likely continue to be the case. Email was game changing, the web was game changing, chat was game changing. "Game changing" is a big phrase, it shouldn't be thrown around so lightly.

Reply on news.ycombinator.com