September 2004

You are currently browsing the monthly archive for September 2004.

Half a year

These days it is six months since I started working at Laerdal Sophus. It is amazing how quickly that time has passed. So far it has been just as educational and fun as I had hoped for.

Ida has left the country

Today my littlesister left Danmark to travel around in New Zealand for three months. I already miss her.

The birth of Naja

A bit delayed, but here is the account of Naja’s birth.

Maria started to have contractions Monday (the 13′th) morning. While the contractions where painful, they where also short of of duration and irregular in frequency. In the evening Kamille showed the first symptoms of chickenpox.

Tuesday morning Kamille was showing quite a few chickenpox. Maria continued to have irregular and short of duration contractions during the day.

The night to Wednesday was horrible: Maria’s contractions got worse and so did Kamille’s chickenpox. Thankfully, my parents was able to come and pick up Kamille in the morning.

At 11 o’clock Wednesday (the 15′th) the water broke and the contractions got worse, longer in duration, and with higher frequency. At 12:30 we went to the hopital and at 15:10 Naja was born. The birth went well with no complications, and the midwife and midwife student were really nice.

Naja latched on to Maria

Naja latched on to Maria without any problems and started to suck away immediately, within 10 mins after the birth. Naja’s weight was 3806g and she was 54cm long.

Naja and Maria

At six o’clock Kamille, our parents and siblings came to greet Naja welcome to the world. Kamille demanded right away to hold Naja, and that wish was granted.

Kamille and Naja

Maria, Naja, and I spend the night at the hospital in a so-called “family-room” (Kamille wasn’t allowed at that ward because of the chickenpox. At 9:30 Thursday morning we left the hospital and I picked Kamille up at my parents. Dispite the chickenpox Kamille had an enourmous reserve of love for her littlesister. The first few days was a bit stressful because of the chickenpox (Naja was an angel). On saturday the chikenpox finally got better, and so did family life.

Kamille kissing Naja

I have uploaded an album with 43 pictures from the birth.

Tucking in early…

We are tucking in early tonight. Because, judging the signs, we will have to get up during the night….

Update 2004-09-14: No we did not need to get up during the night.

Changing website

It is just a small change, yet it feels like a big step. This evening I made a new (sketchy) webpage in the friislarsen.net domain, and redirected my old page at the IT University to the new page.

Now I just need to fill the new page with some content.

Shared snapshots

Yesterday I bought an external hard-disk for backups. To make the backups I’ve rolled my own script (snapshot.sh) that uses rsync to make a snapshot of my home-partition and transfer it to the external hard-disk. The interesting part is, that the script takes advantage of the rsync option --link-dest to share a large part of the snapshots using hardlinks.

So after making three snapshots du -sh tells us that the snapshots takes up 25.3 Gb of disk space:

8.5G    /mnt/external/vips-backup/snapshot.0
8.4G    /mnt/external/vips-backup/snapshot.1
8.4G    /mnt/external/vips-backup/snapshot.2

However, df -h tells us that only 8.6 Gb of the external hard-disk is used:

/dev/sda1             187G  8.6G  179G   5% /mnt/external

Nifty, and with just 40 lines of bash scripting.

I have uploaded a new photo album with pictures from Kamille’s birthday and two short movies with Kamille saying “thank you” to her friends from Utah who sent her a lovely dress.

Now with Java

Following up on the benchmark from yesterday, I’ve produced a Java version of the queens benchmark (queens.java). Below is the updated graph. I have a suspision that the reason Mono and Java is performing so bad is not only due to exceptions, but rather it has something to do with garbage collection.

Graph showing running times for the different executables

I also tried to improve the C++ version based on some hints I got from Asger and Søren (colleagues at Laerdal Sophus). The C++ improvements worked really well when we tested them at work with Visual Studio 6.0. But the improvements had absolutely no effect when compiling with gcc 3.4.1. Either gcc is very smart or very stupid. Something suggest the later as a quick test reveals that gcc version 3.4.1 is producing worse code than version 3.3.4, which in turn is producing worse code than version 3.2.3 . I guess that I’ll have to install Windows and the free Microsoft compilers, so that I can compare them against gcc and Mono.