Hacker News new | comments | show | ask | jobs | submit login

I guess I'm in the minority, but I'm not really buying that this is abusive.

Yes, it would be nice if SO removed the nofollow from known-good posts (which, indeed, it seems they are now doing), but adding nofollow is a pretty simple and reasonable way to make the site a whole lot less attractive to some of the most obnoxious sort of spammers. I would hope nobody is adding links to their SO answers with the expectation of an SEO benefit.

I thought that the entire point of nofollow was to apply it to links that come from user content so that spamming your site becomes less useful. Before this article, I never even heard of the idea that you're supposed to apply nofollow selectively based on your own judgment of how spammy a post was. I always thought it was a simple dichotomy: links created by users get nofollow, and links you create yourself don't.

That's typically the strategy for blogs. Links in a blog post are known good and don't have nofollow. Links in comments have nofollow to deter comment spam. That works for blogs where the site owner is creating the main content and the comments are secondary.

Sites like StackOverflow only have comments from users; there is no original content from the site owners (except inasmuch as they might participate as users). Using the blog model, it does seem reasonable to not use nofollow on content created by users who have already been vetted by your reputation system.

I agree that it does seem reasonable to be more granular with nofollow. But refraining from doing so and just applying nofollow to all user content hardly seems like abuse, even so.

StackOverflow has the philosophy that high-ranked users are just about indistinguishable from staff. So in that sense, high-ranked user profiles are similar to the staff page and the strict dichotomy of user-generatedness is blurred.

If there's a link in a stackoverflow answer of mine, it's there because there's really good information at the end of the link. That's the sort of information Google & Bing use to make their searches useful, it's the sort of information I want Google et al to know.

Google and Bing are free to decide to just disregard 'nofollow' tags on stackoverflow.com if they want.

Remember the tag was only put in place to help search engines in the first place. (Well, to stop people from spamming to get good search engine results. I think you get my point.)

I remember a court case where a website's terms of use precluded search engine bots for indexing the site. The fact that the site owner did not use the industry standard of the nofollow tag ultimately worked against them.

So yes google can disregard nofollow on a case by case basis and face the associated potential legal consequences. I am sure that this would not be a problem in the case of stackoverflow, but I am not sure this is true in general.

I'm going to assume you actually mean robots.txt instead of nofollow, since that's the only accepted way to automatically state "do not index this site" (or portions of it). The big difference here is that the site in question controls the content of robots.txt, while it's the owners of the sites with inbound links that control the presence of rel=nofollow.

As such, I don't think nofollow could have any possible legal consequences. It's intent is to indicate that you haven't vetted the links in question and as such are specifying a lack of trust in their content (i.e., user-submitted links); though like the article says, many people attempt to micromanage SEO through them which dilutes their usefulness (I've been told we have some links that do the same thing; no doubt someone thinking they were smarter than Google... so now we have to track them down and undo these pointless additions)

I almost turned my head all the way upside-down to figure out how Google is under some kind of requirement to obey nofollow when deciding what pages to recommend.

"You can browse my pages but you can't get useful information from them" is bizarre. I'm not saying that no site anywhere will ever sue over it, because this is America and you can sue over anything.

"google can disregard nofollow on a case by case basis"

This is somewhat correct. While Google won't disregard noindex/nofollow directives (or robots.txt Disallow) it can - and will - show "blocked" pages in SERP if the quarry is specific enough or a the page is "strong" (lots of inbound links/strong social signals and etc...)

However, it will not show in page info (hide description and title) and use alternatives sources of information to fill those out instead.

For example: it can use DMoZ info instead of the real one.

Source: http://www.incapsula.com/the-incapsula-blog/item/395-new-sni...

They can disregard the nofollow tags, but they don't have as much information about the links as SO does.

Yes, I agree. StackOverflow is harming the quality of results in Google because they want people to filter through them for the information. Anyone that's used SO knows that answers can be misleading, incomplete or straight up misinformed. At the very least they're generally less comprehensive and less keyword-full than source material that would appear as links.

I suppose it's selfish of me to want SO to share them and selfish of SO to mark them as nofollow and not allow the source articles to rise in ranking.

Original topic is not explicitly about content, but speaking of SO's content:

> StackOverflow is harming the quality of results in Google because they want people to filter through them for the information.

Well, no, actually. StackOverflow may be shooting itself in the foot. SO cannot dictate the rules of the search game to Google, and Google adjusts its SERPs pretty quickly, especially nowadays, when user behavior gets to have bigger influence on SERP with every ranking algo update. Poor quality content leads to higher bounce rate and lower avg. time on site and visit depth, which will inevitably lead to lower ranking and less traffic for SO. This (and perhaps a thousand of other factors) works for every site, and even more so for heavy traffic content projects like StackOverflow. So don't be afraid, the system will adjust itself.

Overall, I agree with your sentiment: I love SO, but nofollowing the Web looks kind of lame and selfish. After all, doesn't SO have its staff of moderators and admins to fight spammy comments?

yeh right that's going to scale NOT - UGC spam can cause your site to tank I have seen this happen to major authority sites (I work for FSTE 100 publisher)

Our suspicion is that when we clamped down on the spammers by banning accounts they reported our site for spam that they had created we lost 20% of our traffic - major brands like us can tough it out this could kill smaller sites without the runway to survive this or have "friends" inside the wire at Google.

I have had to help a small company completely rename and start again on a new domain after some one hacked the site and inserted 1000's of pages with links to porn sites.

Yes, I would be happy with the traffic from SO though the SEO benefit from a "follow" link may not hurt. I have seen the tricks that spammers do to get follow links and am sure SO will be spammed crazy if it allowed them.

I agree. The article incidentally misses the most obvious reason which is spam prevention.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | DMCA | Apply to YC | Contact