Lab computers
Posted Oct 2, 2002 16:47 UTC (Wed) by
gtb (subscriber, #3978)
Parent article:
The Case for Linux in Universities
<meta> Are you sure Dan wants feedback? His page doesn't say so, and it doesn't seem to feature his e-mail address either. </meta>
Anyway, one aspect I couldn't find in Dan's otherwise excellent document is that Linux is a big win on the computers that control scientific experiments. Linux's most compelling selling points in this domain include:
1) stability: Experiments have to run for weeks without interruption, often on buggy, exotic hardware.
2) patchability: The hardware in physical and chemical experiments is often so exotic that people need home-made drivers to run it. With Linux, you are likely to find a kernel patch on the Web, written by a fellow natural scientist who depends on its operation just as desperately as you do. Windows drivers, on the other hand, tend to be crappy for exotic hardware, if they're available at all.
3) Realtime: If you have hard real-time requirements in your experiment, just apply the RT-Linux patch. On the other hand, the closest thing Microsoft has to a real-time operating system is DOS, which isn't for sale anymore.
Hope that helps :-)
(
Log in to post comments)