I was recently directed to this interesting paper on spam emails. In it Stefan Savage (who was at UW today but I missed him) et. al relate how they hacked a spam bot net. They give an interesting technical description of spam nets, which I found surprisingly sophisticated. But here's the real meat of their findings:

• ~25% of spams reach valid email addresses. Most of these end up in spam boxes, of course.
• 3.5e-3% of spams sent are clicked on. Most people are smart enough to avoid it, but that's still 17000 clicks per month.
• Only 8e-8% of spams (28 insecure men) sent resulted in purchases of fake pharmaceuticals, so there's not a huge amount of profit involved. Of course, sending email from a botnet is basically free.
• However, a much higher 3e-4% of people managed to get themselves infected, so the network is hard to shut down completely.
• Hotmail seems to have the most aggressive spam filters. Not a single message got through in the study.
Data was based on about 470 million spams sent over about a month, which is only a small fraction of what the Storm botnet sends.

 

Here's what worked for me:
Note that I was using Mac OS 10.5 and my iPod is formatted as HFS+.

This story is actually three weeks old, but I didn't have time to blog it before leaving on vacation. As you may know, Firefox 3 officially released June 17th. I tried it out a few days later, and posted a tweet of my impressions:

antidemagogue Firefox 3 is quite nice. I do wish that TabMix Plus was ported. I love the tags for bookmarks! 10:10 AM June 22, 2008 from TwitterFox

Tab Mix Plus is a nice little plugin which adds a lot of utility to tabs. I sorely missed having it on Firefox 3. Therefor I was delighted to receive this unsolicited reply to my tweet the following day:

firefox_answers @antidemagogue Tab Mix Plus dev build is compatible with Firefox 3, just not public yet. See here: http://tinyurl.com/38dzc9 08:20 PM June 22, 2008 from web in reply to antidemagogue

It seems that Firefox support has been tracking all posts relating to the browser and is trying to respond to any problems sited. I think that's a great way to use twitter—its like overhearing murmured complaints and then fixing the problem without even being asked!

The upper right cells have just finished mitosis. The cells are stained for DNA (red) and actin (green).
Cropped | Original

We used a tiny micropipette to inject dye into this single cell, just like a doctor injects patients using a needle.
Original

Two weeks of lab course really flew by. They were really busy, but I also had a lot of fun working in a lab. This particular course focused on mammalian biotechnology. Every day we started with a couple hours of lecture and theory, then dived into the day's experiments. We started with basic cell culture techniques and then flew through most of the major techniques in use today. I liked the in situ protein labeling techniques because they produced really nice micrographs like the one on the right. ETH is very well funded, so Fussenegger's lab has all the equipment you would ever like, including a really nice fluorescent microscope.

I was slightly surprised at how much antibodies are now used. We used them for purifying proteins, doing ChIP, all sorts of labeling experiments, etc. Furthermore, antibodies came up a lot in discussions of cell culture, since they still form the majority of the pharmaceuticals complex enough to need mammalian cells for commercial production. Antibodies are a huge industry. Another cool compound that came up a lot is nanoparticles. These are good for introducing DNA into cells. Some kinds will just push the DNA in by itself, but we also used 'particle bombardment' - literally shooting nanoparticles at cells with a special gun.

I don't know if I would want to actually work in a lab for long. On the one hand, it is exciting to touch the machines and really use your hands every day. But on the other hand, you are essentially just pipetting fluids from one tube to the next all day.

Perhaps the coolest part of the course was the field trips. We went to a small local startup, 'Red Biotec', to get a perspective on the business aspects of a small biotech firm. Then we spun 180° and went to enormous Boehringer Ingelheim (see Previous Post) where we encountered production techniques for cell culture at a huge scale. Finally we went to the Novartis plant in Basel. This facility focuses mainly on research - finding new drugs in nature and then figuring out how to produce them efficiently. At all three companies we got talks and tours from high-up lab heads, many of whom had studied at ETH. It was really eye-opening to see what the biotech field looks like commercially. I'm still not sure if I would want to have that sort of environment or if I'd be better in academics, but it was really nice to get an idea of what its like.

Techniques:
Cell culture, transfection methods (CaPO4, electroporation, nanoparticles, various viral vectors), IP, ChIP, western blot, RT-PCR, Reporters (SEAP, SAMY, Luciferase, ELISA, GFP), FACS, microinjection, particle bombardment, microencapsulation

By chance I rediscovered a book I read at age ten or twelve. It made a lasting impression on me at the time because it has some really dark themes which were way too mature for a fifth grader. The surprising thing was how much the book had changed in my memory. Some scenes I recalled in precise detail, but had them totally wrong chronologically. In others I had transposed characters, or changed the setting entirely. For instance, in one section they are trying to cross a magic barrier. I remember much of the struggle, but in my memory it took place in the more mundane setting of a mountain pass. It is funny how we modify memories over time.

Wizard's First Rule (Sword of Truth, Book 1)
by Terry Goodkind

This is a well-crafted story with a complex and interesting plot. Goodkind adds some novel elements to the standard fantasy fair, creating a compelling world. Sometimes it shows that this was his first book - a few dialogs are awkward and poorly written, especially during the exposition. But it still left me wanting to read the next book in the series.

Boehringer Ingelheim

8 photos | Map

I'm currently taking a lab course in biotechnology. This year the class is focusing on biotech in mammalian cells. Today we had a trip to a world leader in mammalian culture, Boehringer Ingelheim in Biberach, Germany. This is one of the biggest manufacturers of biological pharmaceuticals (read antibodies) in the world. Basically, they start with cells which produce a particular drug, usually in hamster cell cultures in a test tube. They then grow the cells up from a single clone to huge 12000 L tanks three stories tall, then extract the antibodies secreted by the cells. Everyone complains about drug costs, but after seeing this facility I can understand the price a bit more. The design of processes to produce drugs is truly a feat of human ingenuity, and the measures needed to implement the technology are extreme.

The day started with lovely drive through rural Switzerland and southern Germany, including a ferry across Lake Constance. After driving through beautiful vineyards and orchards we reached their sprawling campus. I was never that interested in pharmaceutical companies, but now I understand the appeal. The campus is impressive - nice architecture, plenty of trees, and huge. The cafeteria food was good and we were served coffee afterwards as we sat and chatted. The labs were spacious, light, and ultramodern. Money is not a consideration - whatever is needed to make the deadlines and quality standards is implemented. The flip side of this is that jobs there are secure and a bit static. Most of the people grew up in southern Germany and there doesn't seem to be that much international mixing within Boehringer. Overall, it struck me as a good place to go after you have kids and want to settle down, not a hip, lively biotech firm.

The biggest problem facing their antibody production is contamination. Since they are amplifying the cells something like a billion fold, the smallest bit of contamination ruins the whole project. FDA rules force them to keep a very close eye on possibilities for bacteria or especially viruses to enter the cells, since they would eventually show up in patients. To solve this problem they have automated almost all their procedures, so that humans never have contact with the cells. Computers control all the things needed for cell growth - nutrients, air, pH, etc. It is amazing how different their facilities feel from the university's, where human involvement is constant and time is often sacrificed to save money.

The scale of the reactions was constantly bewildering. In our lab we consider a process to be pretty big if you can measure it in milliliters rather than using the usual micropipettes. At Boehringer they have to use exponential notation for some of their measurements! They have chromatography columns two meters wide and "glassware" I could barely cover with my whole hand. Protein isn't produced all that fast - think about how slow your hair grows - but they can produce a years supply of some of their drugs in just two weeks. And as they churn this stuff out they still find time to sit at the cutting edge of research into ways to improve the process.

KNIE

8 photos | Map

Uetliberg by Night

7 photos | Map

For Niklas's birthday last week we went to see the circus, KNIE. This is a traditional circus with a big tent and a dirt ring. The only circus I've been to recently is Cirque du Soleil, which had quite a different feel. It feels very old fashioned to see trained animals or midget clowns, but I guess it is part of the tradition.

They had some really cool acts. A man juggled seven mirrored pins which flashed hypnotically as they flashed up almost to the ceiling. Acrobats slid smoothly between perfectly poised poses, high in the air (unfortunately the tent wasn't big enough for swinging acts - only vertical motion). A man turning rapid cartwheels on the feet of his partner gave the crowd a scare with a nasty tumble, but really won them over with a graceful recovery and an astounding finish. A team of tumblers did flips and aerials through spinning lassos. Two body builders stayed frozen in unbelievable angles for long periods of time.

Between acts clowns played tiny violins and cracked bad jokes about swiss culture. There was also a parade of trained animals worthy of a small zoo - horses, elephants, Shetland ponies, camels, and even a llama. The ringmaster's young son also seems to fit into this category of entertainment - I have to admit he was cute. The training was pretty impressive. I've never seen so many species walk on two legs before.