Maybe I drop a word about why.
CS was an option for me, but I would have done it only to learn more about programming (and maybe how the hardware works). Since I already knew a lot about programming when I was 18, I decided to go for physics, because Astronomy and Quantum physics were both also strong interest of mine.
Basically this decision opened a bunch of very interesting job opportunities in science programming that would have been hard to get into without a scientific degree. I can wholeheartedly recommend to not study CS, but a different science, especially if you know that you want to write software. You can easily learn programming by yourself and get detailed knowledge in another topic. This combination is hard to beat (most scientists are very bad at programming).
It was a lot of fun to program in my own lab environment. Maybe because I controlled the whole pipeline starting from requirements to finally using the software for my own scientific work.
Its a pity that this is rare. Currently I develop software for an ESA space observatory. Sounds incredibly cool, but sucks a lot because of the size of these scientific projects (and because of the bureaucracy of ESA and NASA).
How rare is this? The caliber of mind required for physics should be more than enough to make a decent programmer. What's preventing your colleagues from just picking up what they need to know?
