I pursued higher education with the goal of being able to work on the questions I was genuinely interested in. As a doctoral student, I would bring proposals to my advisor with the questions I wanted to study and he would always rebuff them with various objections. He often told me, in asides, about various questions he was interested in, but it took me more than a year to realize that this was his way of telling me what kinds of questions I was supposed to study for my dissertation.
When I was visiting one time, my brother Phil said, "I've got to show you this cool thing". He took me to his office where he had a Sun workstation with a huge CRT and opened up a window with a grey background. He said, "This is a file on a server in Switzerland, but if you click these blue words [click] NOW we're looking at a file that's here in Champaign. And if you click these words [click] NOW we're looking at a file on a server in Minnesota." It was an early version of Mosaic, because he knew Eric Bina and some of the other people that were building the first graphical web-browser at NCSA.
When I got back to Western, I set up perhaps the first webserver on campus and created some webpages for the laboratory activities we were building in our computer lab. I wrote up a proposal and took it to my advisor and said, "Now I know what I want to do. I want do do my dissertation on the educational uses of hypertext." He made a grimace and said, "Well… I don't know much about that and this whole 'world-wide-web' thing? Nothing might ever come of that." Eventually, I got the hint and worked on what he was interested in: An Account of Expert Phylogenetiic Tree Construction from the Problem-Solving Research Tradition in Science Education. It was a good dissertation and I learned a lot. And it got me my position. But I think if I'd graduated in 1996 with a dissertation on the topic I was originally interested in, it might have been more timely and relevant. And taken me farther.
So when I write a proposal and someone with an administrative role tries to tell me how, rather than working on what I'm interested in, perhaps I should work on what they're interested in, it doesn't go over so well. Nope, nope, nope. No.