Month: April 2006

I really like LaTeX Beamer

As I previously wrote, I’ve discovered the Beamer class for LaTeX some months ago and I really like it! It’s very easy and straightforward to use (provided you know a little bit of LaTeX, of course). I’ve also noticed it forces me to actually prepare my slides and illustrations before beginning to create the slide show. This is a good point since 1) it forces me to stress the structure (rather than doing it as one goes along the slides) and 2) it allows me an easier and better re-use of illustrations and slides previously shown (in Powerpoint or Impress, you had to think of what objects you have to copy ; here, you have only text to copy).

I have to prepare a small seminar for tomorrow (well, technically, it’s today but I need some sleep before going to work). Because this seminar is the continuation of a previous one, it’s very easy to re-use the structure and some slides. Now I see the power of LaTeX and Beamer 🙂

Since I use LaTeX on my laptop, I’ve always noticed that the pdflatex command is fast but uses a lot of CPU ressources and turn on the fan. Tonight, I was curious on how this was possible. I wrote this small-and-dirty bash script to try to see what whas happening:

#!/bin/bash
c=1
while [ $c -lt 120 ]
do
        ps aux | grep [p]dflatex >> proc.dat
        ps aux | grep [p]dflatex >> tmp.dat
        cat /proc/acpi/thermal_zone/THRM/temperature >> temp.dat
        cat /proc/acpi/thermal_zone/THRM/temperature >> tmp.dat
        sleep 0.5;
        c=$((c+1))
done

It’s not optimised and it certainly uses a small amount of ressources by itself. I am not sure each line is written exactly 0.5 second after the previous one. And it needs some post-processing in order to be ready for Gnuplot. But, anyway, here is the result:

CPU usage and temperature

After 7 seconds, the pdflatex command is launched but we can already see that the script in itself increases the temperature. For a 20 slides presentation, we see that pdflatex is using up to a bit more than 50% of the CPU and its duration is about 7 seconds. And, as it can be expected, the temperature has a small delay compared to CPU usage (temperature increases only some time after the rise of CPU usage). Temperature is increased above 60°C ; this turns on the laptop fan.

Nice addresses for holidays

After one year of hard work, we finally took some holidays. We spent three days in Middelkerke, at Hotel Were-di. It’s a nice hotel, clean and small. It’s not too far from the beach and the city is much less crowded that other big cities on the Belgian coast. The restaurant on the ground floor is nice, food is good but I still don’t understand the need to pay so much for the “nouvelle cuisine” (I know: I am a greedy fellow).

If you want to have good food and nice wine, I suggest “De Kloeffe”, a few steps from the Were-Di hotel. We were really delighted by the Easter Menu (schrimps as starter, lamb roasted in front of you for the main course, nice ice cream and profiterolles, … mmmm!). The address is P. de Smet de Naeyerstraat 5, 8430 Middelkerke.

Finally, if you want to escape from the crowd in Brugge, I suggest “De Proeverie” (Katelijnstraat 6, 8000 Brugge ; unfortunately their website is only in Flash). It’s a nice and cosy tea-room where you can sit and relax for a few moments. If you are not too hungry, simply ask for a cup of tea or coffee and you’ll receive what you want with chocolate and other delicacies (of course, you can still order cakes, pastries, etc.). 🙂

A new Jabber ID

I’ve recently learned that the Jabber.org server was an experimental one but I was too lazy to change my Jabber ID. Now that the Jabber.org server is down (apparently since 23.00, Belgian time), I am forced to change. My new Jabber ID is “jepoirrier at jabber.fr”. I will still use the old one (“jepoirrier at jabber.org”) but I will gradually switch all my contacts to this new ID.

BTW, the Jabber.org status can be found here. The Jabber.fr server is hosted by APINC and its status are here.

Finally, I am beginning to develop a software that will help scientists to take note of animal behaviours while they occur. The main scientist will be on a computer, using the software. And the observed behaviours should be displayed on the main computer but also on another one. I think I’ll use XML to send these observed behaviours as messages. So I am wondering if it will be possible/feasible/easy to use the Jabber API to send thoses messages (instead of writing the whole message-handling part from scratch). Stay tuned for more info!

Changing the hosting company for poirrier.be

I am changing the hosting company for my main website (poirrier.be). I am leaving Amen because they became very costly (more than 80 € per year for only 50Mb). After considering B-one (problem: they forbid videos on your website, even if you are the producer), I finally chose Online (a little bit more than 30 € per year for 500Mb, they are not the cheapest but I know they are reliable). In less than an afternoon, DNS from Online to the university were already up-to-date. I hope that the transition will be as transparent as possible for everyone.

P.S.: the only problem I encountered with Online is that they don’t allow any dash (“-“)in e-mail login. I will then be obliged to switch my e-mail address from “jean-etienne” to “jeanetienne” at poirrier.be. Other family members don’t have this dash-problem and you can use their e-mail address as usual.

Nices curves…

Every morning since the beginning of my Ph.D., I check the temperature and relative humidity when I am arriving in the lab. Now, after nearly 4 years, I have really nice curves (ok: nice charts, if you prefer) …

T and HR charts

As you can see, temperature for the animals is always between 20°C and 25°C (this is good) and relative humidity in their room usually fits in the 20-80% range (ideal range is 30-70%). For a lab where we are only able to monitor the temperature and relative humidity, it’s good. I am sure that if we had the possibility to control it, it will be better (we only have a pot of water if the relative humidity is too low or an additional heater if the temperature is too low).

What’s really interesting is that relative humidity follows a 1-year cycle: we have high levels of humidity in summer and low ones in winter. Temperature seems to follow another cycle: hotter in summer, cooler in winter. I checked (but did not the statistical tests) and it seems that our experiments were not influenced by these changes.

Finally, these graphs show that I did not take any holidays, except for Christmas 2004 and my honeymoon (end March 2005 – mid May 2005). And some people are suggesting that I didn’t work hard for this Ph.D. … Notice we rarely had more than 20°C in my office … And, yes, these charts are made with Gnuplot, a free software (but not GPL)!