mattbelcher.com

That last post was pretty depressing, so I'll try and end this saga on a good note. After a depressing Spring semester, Kim and I decided to take a trip to Europe that summer. Neither of us had ever been and it was something we had always wanted to do. Since we were both running out of free time as graduate students, we decided to just go ahead and do it. Five weeks traveling Europe definitely did the trick. I had a lot of time to read and reflect while riding trains and I decided that taking the theory qualifying exam wasn't going to work for me. A good 30-40% of the exam was devoted to complexity theory which is something I had never had a class in and there was no way I was going to get those questions. Another chunk would have been written by the hardest professor I know, so I wasn't going to get a lot of points there either. I estimated I had less than a 50% chance of passing the second time.

However, I persisted. I decided since my primary interest was the application of algorithms theory to network protocols, I could do that equally well in the systems and networking research area. The networking qual consists of evaluating and presenting two research papers in networking to three faculty members, with an optional oral quiz component. Of course, this meant I needed to find an advisor in networking ASAP. I covered the details of this previously so I won't go into it again now. After much drama I got a pass to start doing some real research in wireless networking.

The part of grad school following the qual was great. I loved working on cool networking technology. hacking the Linux kernel and networking drivers, and just generally being a researcher. I submitted papers to major conferences and got some good reviews, although no publications. I was working a lot, often 14-16 hours in a day between my research and teaching duties. I would be in the lab all night since that was the best time to try out my wireless mesh. I have some fond memories of this time, even though it was a lot of work.

Unfortunately, the clock was quickly running out for me. Because I wasted so much time before the qual doing non-research related activities, I had a lot of catching up to do. My advisor told me I'd need 4-5 publications to graduate, preferably in good conferences. Since he only wanted me to submit to good conferences, I guess they would all be in good ones if they were accepted. I'd need to crank out some papers fast and get some lucky reviews to do that and still graduate in 5-6 years. Unfortunately for me, it didn't happen. The reviews of my work were mixed. Some reviewers obviously didn't understand the paper - or basic statistics! - and so rejected it out of hand. When I finally began implementing my system on real hardware, I discovered that the problem I was trying to solve wasn't as widespread as I had been lead to believe. My simulations weren't scaling up very well, but I had so much invested in the project that I felt like I should just keep hammering at it.

Eventually my failure to get a publication in the networking area, combined with my short clock, lead me to start thinking about leaving. The real kicker came when, on a Friday after just submitting a paper to one of the top conferences in my area, my advisor told me I was doing a good job and would be back on track to graduate soon. Then when the paper was rejected, he changed his opinion of me, told me I needed to work harder, and said he doubted I would ever graduate. Kim and I were also going to have our first child soon, so I didn't think I'd have a lot more time to work.

So I decided to leave. I started looking for a job and had some amazing success. Even though I failed in academia, I was apparently very desirable in the commercial world. I received lucrative offers of employment from Google, Motorola, Volition, and my current employer, Allston Trading. I'm making more money that I probably could have as a professor, I have more time to spend with my family, and I don't need to worry about trying to get tenure. Quitting was hard, but in retrospect it was probably one of the best decisions I ever made. I probably should have done it sooner.