darora

No, you're correct in that assumption.

In case of a complete server breach, that would indeed allow for de-anonymization of posts.

The friends' feed (or any other place, for that matter) doesn't show any identities.

For "rogue" friends, we have on our roadmap a mechanism to "flag" users for a review of their postings. That'd trigger an admin banning them if necessary.

The facebook component, is intended to be for a couple of things. Firstly, to allow you to restrict your discussion to just your friends (there are several scenarios in which you'd like their opinions, but you suspect they might not be completely honest with you if you knew who they were). Second, it allows us to be able to keep out bots en masse. The assumption being, of course, that Facebook polices its user-base to keep them out. In turn, the general quality of the discussions stays decent (well, for an anonymous platform, at least).

Well, quite simply, you can anonymously[1] discuss things with your Facebook friends (or the whole world, if you'd like).

[1]: That is, apart from the login and pulling in your list of friends so as to populate your "Friends' feed", we don't use Facebook IDs or names at all.

Hmm. I'm beginning to suspect that you're right. We'll be looking into implementing this. Thanks

To be honest, without the facebook integration of being able to talk to _your friends_, there are-as has been pointed out-far too many anonymous forums already.

True. Our reasons for using Facebook were twofold-

* It enables you to talk to your FB friends--a group whose opinion that you probably care about & respect--anonymously, so that they don't have to hold back or sugarcoat things.

* It enables us to ensure a reasonably decent quality on the service, by [hopefully] avoiding too many bots or trolls.