Spam chongqing

I just got back and will be chongqing ASAP after recovering from my vacation. Got up this morning at 5:45 and got home about 12 hours later. Driving in rain sucks for us Texans who never experience much of it. It also sucks to be without an internet connection. But is also nice to know I am not so addicted to it I would go nuts without it. No internet means no email, which means no spam, except it was all still waiting for me when I got back. Luckily POPFile is amazing, only about 3 out of 829 were misclassified so after a quick skim through the spam folder I was done.

I just wish I was around when Manni was having all the chongqing fun by himself. Hakdata gone. A number of new submissions. More people linking. A new spammer unsuccessfully attempted to spam the POPFile Wiki. And Manni revamped the spammers list.

Chonqging, China is the city where the spammer that inspired us to start this campaign works/lives. We quickly discovered this spammer likes to spam lots of wikis, blogs, and guestbooks with his links and the keyword Chongqing. Once we realized this guy was a big time web spammer we decided the best solution (in addition to blocking him) was to wreak his Google PageRank for his favorite keywords. Now we are also actively going after a number of other wiki spammers (cases and nominees). If you or someone you know have had your wiki spammed let us know on our submission page.

Most people that don't see wiki spam as a problem don't understand what a wiki really is. Some say any open content on the net is just asking to be spammed. Often wiki owners do find the easiest solution is to stop allowing anonymous posting. But that is not in the spirit of a wiki, they are meant to be open to anonymous users. Being open encourages everyone to add their knowledge of the topic for everyone to benefit from. When posting to a wiki, the information or link should benefit other users, not your PageRank.

Many say add a robots.txt so Google won't index the sandbox page. They don't realize that spammers aren't limiting themselves to sandbox pages. No spammer ever hit the sandbox page at POPFile (maybe since its not named SandBox), they just spammed the main page and the FAQ index. On other wikis, most of the spam I clean is not on sandbox pages either. I do really strongly agree there are good reasons to use robots.txt or noindex meta tags though. Previous history versions, diffs, deleted pages, about user pages, and sandboxes shouldn't be indexed. But its not a total solution.

A few people even say you should block the entire wiki from Google. But many wikis have very useful information and should be in Google for people to find the information. Wikis are a very popular source for documentation for open source projects (either as supplemental or the primary documentation). Gnome, Appache, Gentoo, POPFile, and K-Meleon are just a few examples of OSS that make good use of wikis. By having the documentation in a wiki, it gives non-programmers a way to contribute to the software they use.

I just discovered what I have been calling webspamming is also apparently called spamdexing:

Wikipedia: Blog spam is a form of spamming known as spamdexing...

Slashdot has caught on to the problem of wiki spammers, they posted links to Philipp, the reformed Nigritude spammer, and to an article about it at Netcraft which happens to link to usemod.com's WikiSpam, which links to us. While its not a direct link, traffic here has picked up some. Hopefully more people will spread the word, and the word is chonqging.

I just discovered Mr. Chongqing now also spams for emmss dot net and emmss dot org, which are just redirects to emmss dot com. They appear to be new purchaces, whois says created 23-Apr-04.

Thats just days before he first spammed our wiki and began his inevitable downward movement for emmss and chongqing search rankings. So far we haven't beat him for either term. But we are right next to him at the top of the page for emmss and we are catching up on chongqing where he is the last result on page 1, we are on page 4 and 5 now.

They are setting up filters on outgoing mail from Hotmail and MSN. Supposedly it will be triggered when a user sends a large number of messages. They also ask other ISPs to do the same. Now this idea I like, for most legitimate email this will be no problem, but it will affect people that legitimatly email in large quantities unless providers are smart about who/how they filter. Newsletters like LangaList and Lockergnome are already frequently wrongly blocked by antispam efforts.

They are still pushing their "challenge in the form of a computational puzzle" idea, but that isn't going to help people unless everyone adopts their system. And unless their "puzzle" is really proccessor intensive, its not going to prevent spammers from sending spam, it just may mean they have to send a little less, or worse, get more virus infected computers to send it out for them. If its not their computer why would they care if it takes proccessor time.

I found out about this from winXPcentral.com, the original article is at infoworld.com.

I was amazed this afternoon when I checked my email and found an email from a spammer. Nope, its not the one of the normal 70% spam in my email box, it was a letter to me from Hakdata. It's a very well thought out explanation of Hakdata's plan to spam wikis. I guess its supposed to get us to leave him alone because they are "nice" about their spamming. I especially like how they thank me for blogging about them, which I hadn't even done (other than include a few of his keywords on my sidebar).

When Manni added them to the nominee's page I didn't really bother with them much since they looked like they were spamming mostly German wikis, which limits my effectiveness in reversing their damage. But lucky for Hakdata, Manni speaks German. Now that they spelled out their plan I think its time for our first new promotion to the worst spammer cases page. I am sure going to take them far more serious now.

Here is their email for anyone who wants to try and understand the twisted mind of a spammer:

From: "Hakdata" <mail@hakdata.de>
To: <joe@chongqed.org>
Subject: Friday, May 28, 2004 about us Hakdata

hallo joe ,

thanks for the blog about us. Really nice you found. Perhaps it
is better for wiki - owners to create a link page on witch we
can place only one link.

If it is your goal to remove all spammers from wiki pages. No problem,
we are working at the moment at scripts to find wiki sites and place our
links automaticly.

But this is not the thing we want to do. We need only 5 Links from Pages
with a Pagerank higher than 6. If we have them, we will stop to spam
wiki and never come back.

Kind regards
Hakdata

This guy just doesn't get it. Wiki owners aren't going to create a page for spammers to add their links to. The purpose of the wiki is to share information, not to help some pathetic wannabe SEO company boost their PageRank. Thier "promise" to not return after they have five links is stupid. If every wiki spammer in the world promised to only put five links on a wiki it would still leave wikis full of garbage and unusable.

I finally got a spam that complies with the CAN-SPAM Act. Here was the subject line:

SEXUALLY-EXPLICIT: The Naked Cast Of Van Helsing

Obviously since that is the first one I have seen (and I do get plenty of spam), I am still not convinced this act is really helping anything.

Well, only about a day since I posted about Nigritude Ultramarine wiki spam we got a very interesting email from Matt. Several wikis he is involved with have been hit by Nigritude spammers. This particular SEO contestant has realized he was spamming and decided to quit. And after some more discussion about links in SandBoxes really are spam he has finally begun to understand why it was wrong. Spamming only SandBox pages is still spamming!

Here are some of Philipp's quotes:

"I decided to stop posting backlinks in Wiki sandboxes, the SEO strategy previously explained."


"As Nigritude Ultramarine is only a game, I will not take it into areas some people take serious."

His original post telling people to spam Wikis.

His second post telling people NOT to spam Wikis anymore.

And there has been lots of discussion about the matter (including mine and Matt's).

Since I don't want to help even a reformed Nigritude spammer's PageRank I ran all his links through bloggers redirect.

Its surprising the number of links to my blog there are from Nigritude Ultramarine sites. If your one of those Nigritude nuts you should read this article.

Google is taking notice of this stupid contest, but instead of blocking or something drastic, they are caching pages and plan to study it to improve the engine. I guess there will be some good out of all this. I wonder if whatever changes they make hurt wiki spam effectiveness, though I am sure it will never go away.

Another effect of the Nigritude Ultramarine SEO contest is more wiki spammers. I haven't actually seen any yet, but the article says its happening and it doesn't surprise me at all. Its stupid though, whoever wins this dumb contest is unlikely to have been spamming, its going to be pure luck. What would be funny is if Google shuffles all the Nigritude results and those how tried hard and wasted lots of time and money go to the bottom.

Google Canada seems to have a special way of handling a certain Nigritude Ultramarine search, its got a shark logo instead of the normal Google logo.

I just hope this is the last big SEO contest like this. I don't think the good Google might get out of this is worth the mess.

Got this from a Slashdot story a few days ago:
The Chinese government finally might be doing something about the spam problem. Hopefully this "crack down" will involve Wiki spammers too.

Manni gave in to the blogging fun and recently blogged The "Buffalo Spammer" jailed:

If we caclculate the time he has to spend in jail for each spam mail sent, we end up with 0.13 seconds per spam mail. That's not so much, is it?

Wow, that's not enough time! ;-)

It takes more to delete each message. Spammers always say it only takes a second to delete their "targeted email" so I say a sentence of 1 second per spam would do it. That even sounds like your going easy on the losers, till they add it up. Sadly this guy only received the time he did because he was using stolen identities to send out his spam.