Marijke van Gans was the first person I met on the internet who I came to think of as a friend. We both used to spend time in the Compuserve Science and Maths Forum when I was a teenager. She must have been in her forties. I learnt a whole lot about maths and science there, particularly physics, and a good deal of that was because she explained things so well and patiently - and a good deal more was because she inspired me to try to do the same thing, and explaining things is probably the single best way to make sure you understand them yourself.
My experiences there made it clear to me that shared written spaces on the internet had enormous potential for teaching people of all ages about science. While I was busy learning science - I was there all through my GCSEs and A Levels - I also picked up plenty about how people learn science, and the kind of easing in and perspective shifts that can help them to understand it.
When I posted about the Curlicue fractal it was Marijke who put her finger on why it's a fractal, and followed that up by producing a tiny curlicue fractal generator in assembly language, that animated in realtime. Her explanation of special relativity in terms of rotations in spacetime clicked for me in a way that nothing else did.
We were both in London when I first came to the forum, but by the time it occurred to me that it might be worth trying to meet up with someone I only knew from the internet, she had already moved to the remote Scottish island of Bute, where she didn't seem to be putting her mathematical genius to use much outside of the forum and a few web pages: She co-ran a site called 'Maxwellian Demon', about information theory and thermodynamics, and made a start on 'A Mathematician's Paradise' after Wolfram's Treasure Trove of Mathematics was eaten by corporate interests. Her personal site largely focused on her love of cats, coffee, cigarettes, and her beautiful island home.
She joined the London Mathematical Society in 2003 and in 2004, buoyed by the respect she received from her peers online, she moved to Birmingham to pursue a PhD on combinatorics under Robert Curtis. Her thesis was completed in 2007, titled 'Topics in trivalent graphs'.
On the 21st of April 2009, Marijke died of cancer, having been sick for some months but with nobody realising just how bad it was until it was too late. We had been out of touch since just after she completed her thesis, and it wasn't until seven months later that I looked her up again and discovered the bad news. She is missed.