How many MIT students does it take to change computing?


A new SIPB chair and vice-chair are taking office tomorrow, and the other night several of their predecessors took an evening to give them an orientation.

SIPB has two priorities: people and projects. Each active project has its own organizers, maintainers, and/or developers who move it forward and make its decisions, so the role of the chair and vice-chair is about keeping track of how things go, helping connect the project to outside resources and connect new contributors to the project, mediating shared resources like the machine room, and making sure that key projects get passed on from year to year.

It makes sense, then, that we spent most of our time talking about people—bringing people in the door at SIPB, making the office a welcoming place for them, drawing them into our community, and electing them as members. We hear in almost every membership election about how the organization could do better at this. Here’s a quick version of why it’s so important:

Every year, about 1/3 of student SIPB members graduate.

Put another way, in steady state:

Size of SIPB = 3 * (# new members / year)

For example, right now SIPB has 26 student members, and by my count 9 are planning to leave MIT in June. So the only way SIPB can stay as strong as it is is to get 9 new members this year, and about as many again the next year, and the next year, and so on. Fewer new members ⇒ fewer members ⇒ fewer awesome projects, fewer people to learn from, fewer people to hire away to Ksplice (ahem, maybe not everyone shares that motivation).

Fortunately, we built a good track record over the last few years:

    academic year          freshmen &
      starting     total   sophomores

        2010        ???       ???
        2009         7+        4+
        2008         9         6
        2007         8         2
        2006        10         3
        2005        10         2
        2004         5         3
        2003         4         2
        2002        10         3
        2001         8         2
        2000         3         2

From those numbers in the last five years, it’s not hard to see how we got the organization to the point where three strong candidates stood at the last election for chair, and where the office is full to crowding at nearly every Monday’s meeting. It’s also clear how it wasn’t always this way—the numbers from the 2004 and 2003 academic years led directly to the election of 2005 in which the nine-member EC comprised every student member of the SIPB.

But my favorite aspect of these numbers is in the column on the right. When I was the chair in 2008-9, I put an emphasis on getting people involved in SIPB in their first and second years. I’ve heard a lot of people’s stories over the years of showing up at SIPB as a freshman or sophomore, going away for a variety of reasons, and finally coming back two or three or more years later and becoming members. Some of them went on to become highly active and valued contributors, and it’s too bad for everyone that we didn’t succeed in bringing them in the first time around. With the record 6 freshman and sophomore new members in the 2008 academic year, I think we succeeded in turning a lot of those stories around into members who will be active students for a long time. Edward and Evan have gotten this 2009 year to outpace 2008 so far, so the new team of Jess and Greg have the chance to finish it at another record. 2010 will be theirs to create, and I wish them the best of luck in outdoing 2008 and 2009 both.

Written by Greg Price

February 15th, 2010 at 1:58 am

Posted in Uncategorized

Tagged with ,

3 Responses to 'How many MIT students does it take to change computing?'

Subscribe to comments with RSS or TrackBack to 'How many MIT students does it take to change computing?'.

  1. Hmm…I somehow just didn’t quite parse that table when you sent it to us by e-mail.

    But 4 more successful elections and this becomes SIPB’s best recruitment year in the last 10? I think I might have to take that as a challenge :)


    15 Feb 10 at 6:54 am

  2. Very interesting post. It’s good to compare your priorities to mine when I was vice chair in 2000-2001 and chair in 2001-2002.

    First of all, I think the patterns of behavior around projects, where projects have a team that recruits new members, are a healthy change. When I was vicechair we had a handful of projects, some entirely without maintainers, and were often in the role of begging for someone to take on a project, rather than asking people to join the current maintainers to keep a project going. That said, the new pattern hasn’t yet demonstrated an ability to end of life a project healthily. It will be something to watch for.

    In terms of membership and recruitment, those were always things we talked about, but in 2000-2002 they weren’t things we worked on. SIPB at that time was a very social organization, and we recruited our friends.

    The bigger focus for me was on repairing some of the organizational challenges I inherited. These included angst-baggage from previous inter-membership conflicts, and a resulting aversion for conflict. The passage of time helped with some of this, as did members graduating and leaving SIPB who had been on one side or the other of conflicts. Other aspects of conflict I tried to deal with more directly. And in general I worked to bring personal disagreements either into the open or out of SIPB. Sometimes this made the environment more contentious, but I think it prevented unacknowledged issues from building up.

    The other organizational problem, which is more endemic to technical organizations and service organization of all sorts, was stop energy ( Well meaning members, out of concern for resources or just to avoid redundant work, would often discourage other (generally younger) members from doing work they wanted to do. The stopping sometimes looked like helpful suggestions that expanded the work required, or outright discouragement for people who were reinventing wheels, or just additional requests for details before resources would be allocated.

    Some of what I did to remove stop energy was oriented around behavior change, some of it was about process change. For behavior change, I mostly worked to call out members who engaged in excessive stop energy, to make clear what they were doing and in some cases to shame them. When it came to process, I tried to create pools of available resources, like having a vmware server where virtual machines could be allocated to new projects without require purchases of hardware or authorization of access to existing machines. At the time there were only ever a few projects taking advantage of it.

    Fixing these two organizational problems was my priority. The size of the SIPB was small but stable at the time. And my assumption was that no matter how many members we had, we would not be accomplishing much until we resolved these organizational issues.

    Richard Tibbetts

    17 Feb 10 at 7:45 pm

  3. Thanks for both your comments. Evan, I hope you win that challenge. :) Don’t forget the extra points for setting new records in both columns.

    Richard, very interesting to read how you saw things in your era as chair and vice-chair. (To be perhaps clearer for others reading, your year as vice-chair began in early 2001 and your year as chair in early 2002.) I think today’s SIPB has learned pretty well the lesson of stop energy that you taught—people catch themselves and others applying it and then hold back. And I’ve heard about the vmware service’s value from people who started projects there in the years just after you were active; today, as you know, SIPB operates a similar Xen virtualization service, XVM, for the whole MIT community, and it continues to be highly fruitful for basically the same reason.

    I don’t doubt your choice of priorities for the situation you found SIPB in. While the two problems you focused on will probably never go away—and I certainly spent my share of time on them as chair six years later—I gather from many people that they were much worse in the era you worked in than they are today, so I’m glad for the work you did to reduce them. And one of the impacts of both problems, in fact, is that they make recruitment hard; too much angst and conflict will make most people find someplace else to be, and stop energy drives people away along with their ideas.

    On the other hand, once both problems are diminished enough that effective recruitment is possible, I find that it provides a way to reduce them further. Historically, a new person who shows up at SIPB has been much more likely to end up a member if they are loud, thick-skinned, and comfortable with producing conflict. (This includes me, to be clear, though I like to think I apply my conflict-comfort wisely and resolve more than I create.) One nice aspect of active recruitment is that it disproportionately brings up the membership chances of people who are not that way, to more nearly equal the chances of those who are. I think we have a lot more non-conflict-happy active members today than we did just a couple of years ago, and from what I see, I think it has created a more friendly environment with some norms to back it up.

    We’ll see how stable that new environment turns out to be. I suspect that it will continue as long as SIPB’s leadership continues to take the quality of the SIPB environment seriously. For the near term I think that’s the case. Six more years in the future, maybe the chair from 2014 will share their thoughts and we’ll have another discussion about our comparative priorities. I bet theirs will be different from mine.


    18 Feb 10 at 2:37 am

Leave a Reply