Martin Geisler Online

Yesterday, at the exercise class for the course on Security and Fault-tolerance in Distributed Systems, Christian Cachin booted his laptop to show us some graphs explaining how the sophistication of onlin attacks have gone up, while the level of knowledge necessary to execute these attacks have dropped.

His laptop was running Debian and I noticed that the image of the Debian Swirl in his version of GDM has a little white border around it. And this is exciting for me because I made that border! :-) Or at least I filled wishlist bug #201303 against the gdm package, which the maintainer then accepted. It’s very cool to see something like this — even though it’s a purely cosmetic thing.

Last week I attended a number of diffeernt courses, some of them were really interesting, some of them not so. Since there were no exercise classes last week I had some more time than usual, so I litterally jumped into different lectures as I found out about them at the course catalog.

During so I discovered some new interesting lectures that I will be following: Security and Fault-tolerance in Distributed Systems in which we’ll take a more detailed look at the difficulities in keeping large distributed systems online. In the course on Language-Based Security we will look at stuff like proof-carying code and static security analysis. It’s going to be interesting to see how this stuff actually works — so far I’ve only heard about it as some sort of theoretical result, but it seems that it can actually be implemented. The last new course is also about security, specifically about E-Privacy: Privacy in the Electronic Society. The first lecture I attended was a bit dull, dealing only with the laws and regulations in this field (standards and recommentations for privacy policies on websites) but the lecturer promissed me that it would become more “hard-core” later on.

I’ll update my calendar later when I get time — it’s quite boring to make (X)HTML tables by hand when you’ve tried making them using the PhpWiki syntax or something similar.

I’m have finally brought my computer back online again — so now the title of this website is true once again :-) The first thing to do was to update my Debian installation with about 400 MiB of updates. I’m tracking the soon-to-be-stable “testing” distribution of Debian, which means that I receive bug fixes pretty quickly after they are found and committed to the “unstable” distribution. But it also means that a lot of updates pile up when my computer isn’t online.

I’m online through a Netopia ADSL router, which we could buy from Bluewin for 200 CHF. That included the PCMCIA card for Stéphanie’s computer as well, so I think the price is reasonable. The only minus is the fact that there doesn’t seem to be any way of making the card run under Debian — I’ve searched a lot and tried all the tricks I know but nothing worked. So I guess Stéphanie will be limited to Windows for now.

Today was my first day at ETH, and the two lectures I attended were quite good. The first was on Convex Optimization. I’ve previously taken at course dealing with optimization in general, so I was not totally lost when we quickly went through the Simplex method and the weak and strong duality theorems. But I must admit that it’s been a long time since I’ve looked at this kind of stuff…

The second lectures was on a completely different subject: Parallel and Distributed Databases. In the course at DAIMI about programming in the large we were introduced to the relational model for databases. In this course we will look at how queries performed against this model can be parallelized so that they can still be answered quickly. Here we’re talking about queries made against databases with gigabytes or terabytes of data, which is spread over many disks attached to different computers. I think this course will be very interesting!

In this first week there wont be any exercises, but in the coming weeks there will be… It will be interesting to see how they are compared to what I’m used to from DAIMI.

While looking at the statistics for my little site, I noticed something strange in the section about referers. This month I’ve got a large number of hits from German ensurance comparison websites?! We’re talking around 15.000 hits from more than 25 different websites.

Visiting a random set of them reveals that they all look the same and only contain some Google adds. So I guess they are part of a scheme where a lot of phony websites try to increase their Google PageRank?