Jyte should, if legal liability is a concern, choose to disclaim responsibility for what authors write, rather than censor.

I imagine they're more concerned with image than legality, since obscene language usually isn't considered libellous. Jyte is a high-profile OpenID consumer and a great PR tool for JanRain, so it would be unfortunate if the featured claims on the front page were always full of naughty words.

Note that "libellous" refers to libel, a particular offense; whereas "liability" means something similar to "vulnerability".

If that wasn't your misunderstanding, then I don't know what you mean.

I didn't misunderstand you. I just couldn't think of anything other than libel that JanRain might be held responsible for...

I think just about any Internet site is vulnerable to suits of "distributing offensive material [to minors]", or "distributing hate speech", or whatever. While these suits are unlikely to succeed, the legal fees would be annoying.

A disclaimer that Jyte exercises any responsibility for what authors write would be far more defensible, and therefore end such suits far more cheaply, than attempting to censor and yet letting clever people say "fuck" regardless.

This claim is offering a way to reduce the risk without censorship.

Please see my comments on your other claim about censorship. However, I/we are open to suggestions on this.

Other possibilities: copyright infringement, child pornography (most likely to happen through some idiotic 15-17 year old girl wanting to know if her breasts are hot and posting a picture), privacy violations (posting someone's contact information), harassment (posting someone's contact information and telling people to prank call the person being an example), and the things above already mentioned.

I may have forgotten something, but that sounds like the likely legal dangers. The examples are not meant to be limiting, I can come up with very different ones, just likely common ones to provide a feel.

Sure, but those examples aren't being prevented by the offensive language filter. Any of those things could happen right now, even with the filter in place.

In fact, the argument could be made that having a filter in place makes you *more* liable because you have demonstrated a willingness to take responsibility for eliminating some content.

D'A

Yes, I wasn't saying the filter helped with any of those. I was responding to Ryan's comment, " I just couldn't think of anything other than libel that JanRain might be held responsible for..."

I thought it might be of use or interest to some to know the likely possibilities as best as I know them and can think of them.

The stop word list was introduced after Jyte was abused with hateful and racist claims. The intent was not to censor, but to prevent mass additions of offensive claims. The claims were not added by people interested in being part of the jyte community, just people wanted to be malicious. The stop word list helps prevent this kind of behavior. You are correct that users can choose to ignore these claims or offensive users, but mostly they got offended and chose to leave the site without coming back.

And yes, the implementation could be better. We're working on that, so prepare for a flood of raccoon claims.

Suggestions?

Suggestion: Provide a way for the *community* to choose what should and should not appear. Perhaps a "flag this claim for deletion" button.

It's far better to let the Jyte Spy crack monkeys react quickly to content that *is* offensive to the community, than to entrust it to an inhuman algorithm which places responsibility squarely with Jyte.

That's a great idea.

What happens when it's flagged and who ultimately decides if it's deleted?

That's what I was going to say... you don't want a majority clique to have the power to remove things they don't like. You do want to get rid of truly worthless or inoffensive content (or hide it or downgrade its noticeability).

The way I have seen this work decently on other sites is through a team of trusted individuals who look over flagged material (not the entire site, but just things brought to their attention) and make a final call. To balance community choice with human checks. And only things flagged to a certain threshhold.

The problem with this system is you need to pick trusted people and those people need to spend a fair amount of time on it. I think both of those drawbacks are pretty serious drawbacks, and may well not work for Jyte. Plus, since those people have power, they are often disliked. In any dispute, at least one person won't want something censored and at least one person will (or it wouldn't have been brought up). So, someone is guaranteed to dislike the verdict. There is no good solution to that except to make it easy for public oversight of what is going on to occur.

This might work if flagged claims when to a special category. They couldn't appear in the normal ones, and could only appear in "potentially abusive" claims. Then people could read over them if they were curious or wanted to check what was being shunted off Jyte, and they could catch and report any abuse of power from the people making the decisions of what to shunt.

Not a perfect system, but potentially a workable one.