Comment by freehunter | Hacker News Reader

Comment by freehunter | Hacker News Reader

It must be tough being a search engine these days. On one hand, you've got people like this complaining that your software is working exactly as intended and just linking to a website that is publishing reports that, at first glance, link the person to a crime (which he is technically "linked" with the crime, being a victim). On the other hand, once you start filtering results to avoid these types of complaints, people start complaining that you are no longer an impartial aggregator, and accusing you of potential bias.

I'm not saying the man was wrong for demanding Google take down the results. I'm also not saying he is right. Google was working as intended. The site they had crawled made an unfortunate implication. This isn't Google's fault, this is the downside to being involved in a crime. You have limited control over what others say about you or how they use your information. Google is in the sticky situation of legally needing not to offend, but also legally needing to be as impartial as possible. The courts are making this extremely difficult for them. If anything is directly to blame, it's a lack of understanding the nature of search engines.

This further goes to show startup founders the troubling legal situation that exists in the world. Bring a lawyer onboard, but even then you have to recognize that you will likely find yourself in a sticky situation at some point.

I'll go ahead and say that I feel bad for him, but he's still wrong. He shouldn't have the right to have it taken down, but it does get into the interesting area of how "free speech" is defined around the world. In theory, Google (web search) just helps you find what other people are talking about.

"What are the sites that mention 'Milorad Trkulja'?"

"well, one of the sites is this melbourne crime site..."

It's kind of interesting to consider that Google -- by finding that result and giving it to you (especially with a snippet) -- is in fact repeating the statements contained in the result. However, it seems like they should be able to argue that it's not libel, because it is a fact that some site wrote that content. If that's not upheld, search engines become useless or sued out of existence (as does, say, the news).

This is the dangerous water we're wading into with any "right to forget" focusing on search engines (because they're the big fish and can bear it financially, essentially) instead of the actual source of whatever a person wants forgotten. You can pretty easily imagine large sections of the internet that aren't allowed to be indexed based on arbitrary submissions by people, presumably unevenly applied if there's not a central clearinghouse of "statements of which we shall not speak", but instead, notices are sent to whatever search engines/major sites that the person happens to notice (DMCA takedown notices fit that bill too well already, but this would be even worse, as it's not specific links, but "don't associate my (somehow disambiguated) name with this subject").

It's hard to know exactly what to do in this case, though, because the original offending site really does appear to be "defunct" and the only links I can find to it lead to a domain registration page. The search results are also now flooded with this story, of course, so, is he only winning on past defamation? The procedure should be to get the offending site to take down the content and then have Google remove it from its cache, but that appears to be already done.

> The procedure should be to get the offending site to take down the content and then have Google remove it from its cache

I agree. What happens when the original site takes down the content, but it stays in Google's cache even after they've been contacted?

What if the content was very much worse? (eg, 'NAME is a convicted pedophile', with images of the person's face; where that allegation is totally false and malicious.)

EDIT: Looks like I'm mostly wrong about this particular case.

(http://ozsoapbox.com/melbourne/yahoo-held-liable-for-search-...)

> I agree. What happens when the original site takes down the content, but it stays in Google's cache even after they've been contacted?

I believe that is the way you do it. In theory, it will fall out of results on its own, but you can request it be removed from results (but only after it's been removed from the site):

http://support.google.com/websearch/bin/answer.py?hl=en&...

(appears to require a google account, which is shitty, but maybe they get a lot of fraud here or something)

> What if the content was very much worse? (eg, 'NAME is a convicted pedophile', with images of the person's face; where that allegation is totally false and malicious.)

That's pretty close to the Bettina Wulff case[1]. Yes, you want that down immediately, but I don't think that should change the procedure. Again, if you think of web search as "what's out on the web", it becomes just a statement of fact. I will grant that it gets pretty abstracted for people, especially in that case, which is more about search suggestions, and is instead "what are other people searching for when including these words?", but there's no indication of the actual source of the suggestions.

> EDIT: Looks like I'm mostly wrong about this particular case.

Can you be more specific? Nothing lept out at me. Was it that he hadn't gone after the sources of the story, just yahoo and google?

[1] http://www.bbc.co.uk/news/technology-19542938

The BBC article makes it seem like he's unrelated to any criminals, and was a victim of a shooting. It seems that he was the victim of a shooting, but that he was associated with a criminal, and a newspaper reported that when they reported the shooting, and a website covered that report.

So it's not malicious people creating false stories that have no basis; and the "link" between him and crime isn't something accidentally falling out of Google algorithms. Someone put it in print; a website covered it; Google indexed it; I think the guy then got the newspaper to take their page down, and got the site to take their page down, and then made a mistake when he asked Google to remove their pages?

Hey nodrama, you seem to have gotten hellbanned for insulting QR codes, despite usually making good posts.

In reply to your comment of:

it didn't work exactly as intended: "As a result of the attack Mr Trkulja said that entering his name into Google Images brought up images of other people beneath which his name appeared. He said some of these figures were allegedly murderers and one a drug trafficker."

If the internet thinks that his name is associated with murderers and drug traffickers, then I think it's part of the job of a search engine to reflect that. Search engines don't just help you find information, they also help you find conversation.

Regarding hellbanning, I just checked nodrama's history as well and there are a lot of posts by nodrama since he/she was hellbanned. I tend to think before I post so even a small reply by me may have taken me five or more minutes to think about before composing.

It's sad to think of all the wasted time this person is spending posting here, all because he/she called someone's post "stupid". Hellbanning should be reserved for true trolls.

Moderation around here is hardly transparent, I doubt we'll see anything change any time soon.

This seems like an especially egregious case. Where does one complain/appeal round here?

You send a very short email to pg with links to your user name and comments, and asking for an unban.

Or you just create a new username.

Except the point of hellbanning is that the accused never knows they've been banned.

Which is why it seems particular unjust in borderline cases. It should be reserved for intentional trolls and spammer only.

It was a comment worth downvoting, but every other commenter on the post agreed with him, and the post itself was flagged into oblivion. Hellbanning is a pretty harsh sentence for something that probably only deserved negative karma.

Even worse, qrafti.com is a joke/troll site. It's not particular surprising that nodrama fell for it. The joke is not revealed until you try and use the site. nodrama cannot be faulted for only reading the cover copy.

If you have showdead on, you can check it out at https://news.ycombinator.com/item?id=4217358

Moderators, please correct this injustice. Thanks.

There seems to be something of a problem with hellbanning here, in the last few months. I've actually turned showdead on because comment death -- at least in the middle of the page -- no longer seems a reasonable predictor of comment quality.

Image search is a bit different, due to the decontextualization of separating the image from the page when presenting results. Image search doesn't show snippets like "here is a picture of someone who is not XYZ", it just shows a problem future of not-XYZ when you search for XYZ

it didn't work exactly as intended: "As a result of the attack Mr Trkulja said that entering his name into Google Images brought up images of other people beneath which his name appeared. He said some of these figures were allegedly murderers and one a drug trafficker."

Reply on news.ycombinator.com