I was recently in [[Google>Brighton]] for [[Google>SOSP]]. For those of you who don’t know about SOSP, its the premiere conference in systems and networking (more systems, less networking). Its very competitive (20 papers selected out of 150 odd submission or so) and the who-s-who of the community are there to attend. Naturally its a great opportunity to be there and see all these people in action.
This was my first SOSP and I think the biggest lesson I took away from this conference was that you should never attend a good conference unprepared. Some things to keep in mind:
* true, you might be on a paid vacation, but you can always go back to Hawaii or Cancun or whatever. But good conferences (SOSP happens every 2 years) are hard to come by, and you’ll attend only so many of those before its time for you to market yourself for a job when you’re about to graduate. So you do want to make the best of it.
* Never, //ever// sleep in a talk. You might be jetlagged, tired, hung over from last night’s booze — whatever the cause, don’t sleep there in front of everyone. If you are sleepy, go to your room and sleep there.
* Try to attend all talks. Its hard, but not impossible. While you’re at it, try to //listen// in all talks. SOSP turned off wireless during the sessions, and I think it worked out pretty well. When you don’t have a laptop in front of you, there’s not much choice but to listen.
* Prepare, prepare, prepare: I can’t emphasize this more. Its critical to be prepared. You don’t want to sound like jack-ass in front of all the big shots there, do you? Try to read the papers. Or the abstracts at least. Definitely read the papers that are relevant to your research. Do some background work on the authors — what have they worked on before, what are they working on now, who are they working with, where are they from etc. Knowing a little bit about the people you might/want to meet will definitely pay off.
* //Do not// go and blabber garbage with someone famous, just to get them to talk to you. As Stefan says, its not important how many people you know. Its important how many people know //you//. While some may disagree with this, but personally I don’t like going up and talking to someone unless I really have something meaningful to say. Maybe about their talk. Or about their paper. Or about their student. Or about their birthplace. //Something//.
* Network, network, network: This is easier than it sounds. Some people are born naturals. Others have to work at it (I’m latter). But hopefully you won’t have to do too much. Two rules of thumb help: avoid sticking together with your own flock; and when in company of strangers, at least introduce yourself and try to get to know them. Hopefully over a few dinners and social events, you would have built enough critical mass so that conversations will flow naturally.
Just some things I realized I should keep in mind for my next conference, and approach it in a more systematic way than I did this time. The 3 days at SOSP were quite useful — I met a lot of people, I got to talk about my work a little, and I got the first “taste” of the systems community. I came back from SOSP a little more inspired and a little more motivated :-)