Monday, November 29, 2004

Good morning sports fans! 

Like everyone else I've been following the weirdness at target.com with interest. I like shopping at Target. I didn't know you could get a blowjob there, for only $9.99. No picture available. 

I keep wanting to tell Marc Canter to stop jumping up and down, and (calmly and carefully) write a list of the data he wants to accompany podcasts, and we'll see what we can do. Now's a really good time to make specific proposals, not blanket condemnations. Marc sometimes misses things, like the Web for example. Podcasting will happen with or without the metadata. But it would be better with it. Drop the martyr act, it's borrrring, and get busy making a list. BTW, that's the first time, as far as I know, that a crucifix has appeared on Scripting News. Apparently it's also the first time the term blowjob has appeared here too.  

Tim O'Reilly will speak at the 10th anniversary celebration of the W3C, on Wednesday, in Boston. 

Kottke has audio of the end of Jeopardy uber-champion Ken Jennings' amazing run, on tomorrow night's show. "Too bad for Ken," says host Alex Trebek. Heh. $2.5 million. 

On this day five years ago, Manila shipped. 

Attention Frontier kernel developers in Seattle. We're having a meeting tomorrow (Tuesday) night at the Crossroads in Bellevue, to make face-to-face contact, and discuss various projects including the Python-in-Frontier work that David Brown is doing. So far it's Dave Luebbert, Dave Brown and myself. You're welcome even if your first name isn't Dave. Newbies are welcome too, but please C programmers only. We're going to start pretty early, 6PM, and go for a couple of hours.  

A warm-up job 

I need a quick programming warm-up job, so I decided to do something about referrer spam. Here's an example page. Almost all the pages it points to are spam. I'm going to create a blacklist, by checking each site, and see if it actually points at mine. If not, it goes on the blacklist, never to appear on this page again. Pretty simple. The purpose of the referers page was to show who was delivering the flow to my site. Then of course the spammers will include a link to my site and then I'll have to up the ante. Let's see what happens.

Before and after 

I just did the first pass, and it's nice!

I archived the old spam-ridden referers page here, so you can see the before. (Yes, I know the images are all broken.)

Now here's the after. What a difference. Geez. The cool thing is that all the sites in the list now actually are known to point back to us. So even if the spammers figure out how to circumvent this (they just have to point back to us) at least we share in the page-rank bonanza. They might as well work for us for a while.

BTW, this is Google's problem 

As I'm writing the code I keep thinking of neater ways to solve the problem, ones that would be harder to circumvent by the spammers, and still allow legitimate referers to show up. The problem is that they'd require cooperation from Google. Remember, the reason these guys spam us is to steal page rank from Google, by making it appear as if we're pointing to them.