in reply to Re^3: (OT) nofollow considered harmful (was: CPAN::Forum opens its virtual doors)
in thread CPAN::Forum opens its virtual doors

All of these points are already addressed in the posts I linked in my previous node, but to be abundantly explicit, here goes.

Spammers operate by targetting mass use software. They take a blogging tool, forum engine, or whatever it is, figure out how to spam any site using it, and then look for telltale signs in URLs such as mt-comment.cgi. Often they don't even visit the associated blog or forum. Check the access logs of an affected site if you don't believe me: all they usually do is dump a lump of their links onto the posting script and move on.

To these spammers, checking for nofollow is just extra cost without any benefit. And there is no cheap defense against them either. If so many people required registration that profitable spamming necessitated dealing with that, then registration would quickly be dealt with by the spammers and would become useless.

Only as long as your URLs don't look like those of commonly targetted software are you reasonably safe from this sort of automated spam. Unless, that is, it became fashionable for everyone to use custom URLs. But then spammers would actually have to crawl sites to find post forms, in which case the presence of nofollow might make a difference.

The value of a spam comment *is* different for the spammer depending on whether this link is nofollow or not.

Sure. Now let's remember that in the time it takes to check whether there is nofollow he could also have delivered a batch of spam, and then explain to me what kind of economic sense it makes for a spammer to check for it.

The fact is, spammers don't care. If they can spam you, they will. nofollow is useless to you as a site owner. It only helps Google when they visit your site.

you'll see someone actually advertising for a product/service that alerts advertisers about "cheaters" who carry their ads but add nofollow to them.

Again this is something addressed by one of the linked posts. And it just goes to show that nofollow will make a much bigger difference for non-spammers than it will ever affect the spamming problem.

Like hashcash, and indeed most any technological mechanism against spam, the idea is to make it somewhat less profitable

No, you can't compare the two at all. Hashcash prevents spam from happening in the first place. nofollow makes Google less affected by spam after the fact. In other words, if you aren't Google, then nofollow is useless to you while Hashcash isn't.

Well, unless you're trying to deprive your competitors of PageRank despite linking to them.

Makeshifts last the longest.

Replies are listed 'Best First'.
Re^5: (OT) nofollow considered harmful
by gaal (Parson) on Feb 06, 2005 at 19:38 UTC
    Hashcash does not aim to prevent spam from happening; but to make it more expensive to generate. Increasing production costs may well be the most direct way of making spam less profitable, but it is not the only way to do so. Decreasing expected returns may prove more practical.

    Once again, it is not Google-the-corporation that solely benefits from nofollow. Advertisers who rely on spam ("SEO") are harmed by it. People who use Google (and like I said, other search engines that use inbound links for ranking) benefit from it. Last I checked, there were a few of those. Of the people you correspond with, how many use hashcash? (For me, the figure changes drastically if I count mail I send to myself. But hey, I'm optimistic.)

    Comment spam is worth more for the spammer if it influences rankings. Make that: the fact that comment spam can influence rankings is valuable for spammers. This is qualitatively different from email spam, where the recipient either responds with a purchase or doesn't. If a spammer can top the search engine rankings, he expects much bigger bucks than those he'd get from just N more recipients.

    Perhaps this problem was created by Google like one of your cited articles claims. Could be, and immaterial. This is everybody's problem now, and everybody stands to gain if it is mitigated.

[reply]

      Hashcash does indeed aim to prevent it. If you don't provide a valid hash, your message is rejected. nofollow tries to work around an accept-by-default policy.

      Yes, people who use Google may reap some benefit — at least, they might once the amount of abandoned but commentable blogs has sunk enough that nofollow actually makes a difference. Make no mistake, we're talking about a pretty large time scale here.

      Comment spam is not useful solely because it influences rankings. Does anyone rank anything based on the content of your inbox? Yet email is flooded with spam. Sure, the fact that comment spam can be used to game search engines is a nice bonus and currently an important reason to deploy it, but if that incentive disappears, comment spam will nofollow into demise. It's here to stay.

      The bottom line is: nofollow does not benefit a blog owner. Unlike other options which do. So by and large, nofollow is irrelevant on the spam front. It is only going to change the way the web works outside of spam: like the horde of bloggers who rejoiced that now they can link to the people they're deriding without improving the linkee's ranking. Sigh.

      Makeshifts last the longest.

[reply]
        One minute you're saying the individual blogger cannot benefit from nofollow, the next you're saying that bloggers are rejoicing because they have more control of ranking they indirectly award to sites they link. (Of course, maybe you think those bloggers are silly to think they are significant in any way.) Which one is it?

        I never claimed nofollow was some sort of panacea. nofollow makes spam less profitable, period.

        Hashcash users cannot reject an unstamped (or invalidly stamped) message outright. When everybody uses it and software becomes better trusted and debugged, they will; but until then the only reasonable way to use it is with a scoring filter, and never set the scores high enough that the binary decision about a message is made solely by its hashcash status. "Make no mistake, we're talking about a pretty large time scale here."

        I also never claimed comment spam was useful solely because it influences rankings. (I don't know why you insist to think that I did.) Spammers will spam where they expect return on their investment. Period. What I do claim is that comment spam's ability to influence rankings — and especially to "hijack" the top hits in large search engines — is an entirely new class of opportunity for spammers, which gave them hella lot of incentive to start doing it. Of course there's inertia, and of course they won't stop just because you take some of the profit away. If suddenly litigation rises a hundredfold, spam won't stop either. But all other things being equal, I believe that it will slow down.

[reply]

In Section Perl News