The Truly Scary Part About Facebook Open Graph | Hacker News Reader

The Truly Scary Part About Facebook Open Graph

jonathancavell.com

33 pointsBurgherJon220 comments

This is where the world is headed, though. We may not be used to it yet and we may not like it, but all the privacy we've been afforded in the past is due to the inability to amass this kind of data about people. Over time I think we'll get more used to the idea of someone knowing what songs we like, or where we like to eat (for the purposes of selling us those things). I seriously doubt our lives will become more private as we continue hand out our personal information left and right.

Yeah, good point. That is a kind of terrifying idea. They are, to a degree, accountable to their users, but other than that ... And it's quickly getting to the point where they have monopoly style lockin. I guess it feels to me, honestly, like they're already there. At that point they're barely even accountable to their users.

Like you said, they're not bad people, and they make a fantastic product, but it's still scary.

Those are my sentiments too. I'm scared for one company to have that much power, but they do make a great product and have great tools to leverage it. I'm most likely going to transition to facebook-only logins on my sites off the platform. The social discovery advantages are just too large.

yeah, i'm also moving toward facebook only auth.

Accountable to users is a HUGE point. What happens to a site when users feel alienated or violated or whatever? Those users begin to walk away towards an emergent alternative.

The facebook "lock in" isn't a lock in at all. It's more like a disincentive to walk away from the user base, media storage, and slickness of the service. But that can be reproduced if need be, I think.

I think there are better things to be worried about than the hypothetical situation where Zuck goes nuts.

Non-exportable data is pretty much the definition of a lock-in. There's essentially nothing that's impossible to walk away from, so the only thing worth considering is getting out what you put in.

In terms of a social application, it's legitimately harder to make a not-locked-in service. Who owns what, and how much of it is valid when you transfer data to another host? But it's still a lock-in, because there's no way to leave and retain things.

Imagine if your phone would only export contacts to a new phone if you stayed with your provider. Sure, you can still leave, but a data-lock-in is still a lock-in technique.

Soft lock in? I don't know... I guess I'm being a bit literal with the term "lock." It's not binary (literally handcuffed or not), but more of a continuum taking into account the cost of moving away. So sure, I guess I missed that.

And though I realize others may feel differently, I'm of the mindset that the costs are relatively low in walking away from facebook without my data. I have all my pictures anyway. Other peoples' pictures of me? Na... I don't own them. Statuses, messages, connections, likes, events? Sayonara without qualms.

exactly. It's not my data that binds me to facebook. It's the fact my friends are all there.

I think a lot of people define a lock-in as a monetary lock-in. You can't leave without spending X dollars to terminate early.

But given the time to re-do a large portion of your data on FB (assuming you've used it a fair amount, as many people on it do), at minimum wage, it's probably nearly as "expensive" to leave FB as to switch cell phone providers.

I know it's not a 1-to-1 comparison at all, and I've deleted my account and felt no loss. But my use of FB was incredibly minimal compared to most of my friends who are on it, and I hear "I hate Facebook, but everyone is on it" echoed from nearly everyone I encounter.

If there is a real need, wont scraping toools with easy interfaces be developed by competitors to allow a user to get personal data out while migrating? Its the user's data after all. Lock-in due to the usefulness of friend info might still be significant, though.

The constantly-changing UI could be considered a way of preventing such tools. Plus, I believe their TOS explicitly states that you can't use scraping tools, and a huge spike of activity would set off red flags. Running it for weeks / months is a massive disincentive to most potential users, so it effectively must complete quickly or it won't be used.

Besides, developing a robust screen scraping tool for a system as large as FB is inherently a very difficult task, and you still need to connect everything you get out, essentially re-building FB. For them to provide an export tool, however, would be ridiculously easy by comparison.

Either you trust an entity with your data or you don't. Whether that entity is an individual, a corporation with a bunch of grey-beards on the board, or a corporation run by someone who, presently, happens to be 25 isn't particularly relevant to the issue.

"Ownership" of this data is not clear to me. If FB does not allow large scale spidering/collecting of this meta data, then FB has the best use of it. On the other hand, if many other web sites use these meta data tags to be consistent with what is an ad-hoc standard, then anyone can spider 3rd party sites subject to their robots.txt file. Hopefully, by using the FB public APIs then developers can get sub-graphs of FB hosted data to work with.

I like standards, and I would like it if many people started supplying RDF for their sites. I am not sure yet if this advantage out weighs FB's lock on the best use of this data.

Somebody should start a company making fake, but credible Facebook identities to sell.

That's the part which confuses me about all this. It's the easiest thing in the world to set up a cut-out Facebook account.

ok great, so they know what movies and music we like and are gona target us with better ads, but isn't that a good thing? I mean I may actually pay attention to ads if they try to sell me things I want.

i just don't want to see and think about facebook's brand everywhere i go on the web. it's one thing to actively choose to post something to facebook, even if from another site...or retweet or digg etc. but that "like" thing ... there's no action there, or sharing of the thought of the user clicking. it's so banal, and so vague, that to see it spread to every site - keeping up with the joneses - just bums me out like seeing a gap show up on valencia street ...or somesuch.

i'll be interesting to see how regulators come down on FB b/c of their willingness to put user needs behind their own.

it's already starting to happen (not soon enough):

http://www.washingtonpost.com/wp-dyn/content/article/2010/04...

He keeps saying "no open", yet doesnt explain why he thinks its not open. Most of the data is actually available through API's.

So only browse Facebook in an Incognito tab. Got it.

Reply on news.ycombinator.com