You're blocking ads, which pay for BlenderNation. Read about other ways to support us.

Google Summer of Code Summit - report from Kent Mein


Here is a report on Google Summer of Code Summit from Kent Mein.  Go ahead and read what he has to say

Hello Fellow blender heads.

It's been a couple of weeks now, but while the annual blender conference was going on, fellow Blender Developer Chris Want (Hos) and I were hanging out at Google with a bunch of other open source developers.  We had a great time, so I thought I'd share a little about the trip with you.

The summit was for the Google's Summer of Code.  Which I'm sure you all know about by now. I've been a mentor for the program for a few years now.  This year I had the privilege of being a mentor for Davide Vercelli (UncleZeiv's) Lightcuts project.

For those of you that do not know Chris, he tries to keep a low profile working behind the scenes.  He has been the administrator for the project, and he's one of the main reasons things work so well.  He keeps tabs on all of the projects, making sure things are running smoothly. He deals with the problems when they show up, and in general just makes everything work. So if you ever get a chance to meet him buy him a drink. ;)

This is the second time I have attended the summit.  And we all had a great time.  This year I went with Scott Johnson, a CS grad student here at the University of Minnesota where I work. Scott was a mentor for Crystal Space.

The first thing I thought you might find interesting was how the conference had been organized. Google calls it an "unmeeting".

At the beginning of the conference they explain what an "unmeeting" is.

In the front of the big conference hall, they have a white board that contains an empty chart with headings similar to this:

Room A                  Room B

They explain that its time for people attending the conference to propose things to talk about. For about 15 minutes it's utter chaos. It's a free for all where the people attending the conference pick a room, a time and a topic to talk about for an hour (longer if needed).  In general, the people that pick the topic are in charge of the room for that hour.

When I asked the organizer if there was a session on "general improvements to the GSOC", she responded with, "Yes there is, and you are going to give the presentation. Please pick a time and a room and write it up on the chart." Once everyone had a chance to put something on the board, the Google staff did some quick shuffling where needed to make sure the popular topics had enough seats.

I don't know the exact figures, but there were roughly 500 attendees, a little over a dozen rooms of various sizes and 2 days to fill.

The chart was complete in approximately 20 minutes. This was my second time attending the summit and I'm still amazed at how well this process worked.  I think the first year we had the conference, they had a lot of Google employees ready to give a talk if people were afraid to volunteer.  This time around, that was not the case.  There were a couple of talks on Android but other than that I think most of the talks were created by non Google people.

It was by no means perfect, there were a couple of instances where rooms needed to be shuffled at the last minute to fit the number of people that showed up for a given session.  But it worked surprisingly well.  Especially considering how unorganized it seems.

I won't bore you with the details of the session I wound up hosting. There were a lot of great conversations going on though, both during the conference and during the off hours.  For example on the bus ride over to goggle on the second day, a bunch of us started talking about how to get more women involved with open source development.  In a session on Game Development we talked about how to get a commercial company into jumping on the open source bandwagon.

I got to talk with a bunch of great people from interesting projects. Some of them were: Crystal Space, SDL, Audacity, K-3D, tux paint, Music brainz, Battle for Wesnoth.  And of course I showed Big Buck Bunny at least a dozen times to different people.  Basically anyone that was willing to sit down and watch it.

Probably the most memorable session I attended was titled "The Human Super Computer."

This session could probably be classified as a workshop or an experiment in audience participation.  Pretty much everyone attending had no clue what the session was about, but it sounded interesting, so we showed up.

We were given the following information as an introduction: Supercomputers are cool.
Humans are in many ways smarter/better than current computers.
How would a human supercomputer work?
There are a couple of key things to making a human super computer.
1. It needs to be self organizing/adapting.
2. It needs to increase bandwidth.
3. Human super computer has 3 roles:
a. workers       - work on a project.
b. observers     - look at the work being done.
c. facilitators  - help the worker stay on track.

This made little sense to any of us, and some people decided to leave at this point.

We broke into groups of 3.  (based on the number of people we had in the session at this point) We were told our first task for our human super computer was, to talk about what a human super computer was and to organize our group so we could concentrate on a task.

The only direction we were given was the above and it was mentioned that the roles did not have to be assigned to a specific person, the roles were to help clarify key elements that were needed for the experiment to work.

We had people trickling in and some leaving, so the first task was odd to say the least. I was in 4 different groups before the this task was over.  The last group I was in, I had to explain the whole thing to two new people who showed up late.

In this group, we talked about how the 3 roles were similar to how you might manage a group of people but slightly different.  You maybe have a manager, project leads and then people that work on a project. It was mostly speculation about what we were suppose to do and what do the roles mean.

After about 15 minutes we were told to migrate to brand new groups. One person stayed in place, another person moved left, and the last moved right.

The second task for our human super computer, was to talk about ways to improve the GSOC program. We wound up talking quickly about the first task and what each of our groups did. Then we moved on to this new task.  We talked about a couple of things and wound up focusing on how to improve communication between the various open source projects that were participating in GSOC. After 15 minutes again we were asked to migrate to brand new groups.

The third task for our human super computer was the same as the second but with a new group.  We again wound up talking quickly about our last group and what topics we were working on.  Each group had different ideas and we wound up focusing our discussion on how to encourage more women to get involved in the program.

After our 15 minutes were up, we had a wrap up session. We came back to a large group of people and formed a circle. We then went around the circle talking about how the experiment went and made a number of observations.

Observations from everyone involved in the session: Everyone that stayed for the whole thing thought it was a great experience and that it worked a lot better than we would have thought it was going to. It was also something we felt you had to participate in to really understand it.

We all thought that at the beginning were not getting enough information.  Once the process was over though everyone felt that the lack of direction was actually a good thing and part of the process.

Everyone was amazed that at each group change, we were able to bring the new group members up to speed on what our previous group was working on very quickly.

We mentioned how 3 people seemed to be just enough to keep the discussion going, and noted that pretty much everyone was participating.  Which is very different from one person up in front giving a presentation and asking for the audience to participate.

We talked about the roles and how none of the groups said, Person A you are going to be the worker, B your the observer etc...  But, people seemed to switch roles naturally.

It was a very unique and interesting way to tackle a problem. We talked a little bit about how it would be interesting to see this in a business setting, or even in a larger group of programmers.

After the Conference was over, Scott and I had another evening to kill.  So we decided to hunt down a "In N Out Burger" from another bus ride conversation.  We wound up walking a couple of miles more than we probably wanted to but it was worth it in the end. The burger was great, I'd stay away from the fries though. On the way back we Decided to play mini golf at the craziest mini golf course I've ever seen.

That covers the majority of what I wanted to say.  Hopefully you found this report interesting.

Feel free to drop me an email if you would like any additional info.


Image courtesy of (see image here.)

About the Author

Benjamin Bailey

I am a young creative artist with a dream to conquer the world with my ideas. I love singing, writing, drawing, and modeling in 3D - and I'm ready to put these skills to use.


  1. Im still not quite clear on the human super computer thing. From what i gather it sounds like a group of 3 people organized into a worker observer and facilitator fashion. Is that right? Is it for group dynamics and organization purposes? Is that what the talk was about?

Leave A Reply

To add a profile picture to your message, register your email address with To protect your email address, create an account on BlenderNation and log in when posting a message.