Morning Surprise

Day 2 of Startup Weekend started out with a surprise. The night before a consensus had been built by a small group, and two members were going to combine those ideas into a 5-7 page powerpoint. Apparently they had their own ideas, because what was agreed upon by the small group, and what was presented were far apart.

The small group had agreed to focus on building social networks for community organization, and to focus exclusively on condo associations for the weekend. One of the main benefits of this approach was security - since user accounts would be created by a network admin (probably a condo board member), users could be assured that everyone in the group was an actual resident of the community. This approach also gave us a target group to sell the application to . We came up with a simple, but useful set of features based around the idea of 3 main type of information - People, Resources, and Notices.

What was presented did somewhat resemble what had been agreed upon - a location-based social network. Two key elements, however, had changed (along with several smaller details):

1. Instead of focusing tightly on community organizations, a social network could now be created by anyone that wants to. The network can be defined geographically using zip codes or by drawing a polygon on a map (this option was widely regarded by the dev team as not doable in the weekend.)
2. Instead of belonging to only one network based on belonging to a community organization , users could now belong to many networks, with a central page aggregating information from all networks

After this presentation I was starting to feel discouraged - I wasn't really interested in building a new Facebook, which is what this seemed to be moving towards. I was also bothered that two people had hijacked the whole project and basically made it what they wanted it to be, regardless of the consensus built by the small group the night before. This discouragement only grew as the User Experience team came up with their own idea of what was being built. Two hours into the second day, with no code being written, I had serious doubts that we'd be able to come together, and was considering packing it up and heading home.

Afternoon Delight

Once we actually started some coding, things started to improve. Reps from UX and Dev got together and hammered out a features list that both groups could agree on. For the first part of the day we worked on finalizing a data schema, writing specifications for models, and then starting to code those models up. The second half of the day we started to work on controllers - adding in small pieces of functionality at a time. The Design team was also working hard during this time, and was able to hand the design to a developer for conversion to XHTML/CSS.

We were expected to present a prototype of what we had at 9:00pm - and we would have met that if not for crippling network issues that left us unable to make subversion commits. After a drawn out process of doing commits one at a time, some of the design was integrated and we had some basic functionality to present to the group around 9:45.

Midnight Coding

Though I was not all that impressed with what we accomplished during the day, the demo was a major success. We were only able to show some basic things like creating an account, logging in, uploading photos, and some basic entry of resource, but the other teams seemed extremely pleased with what we had so far. The founder of Startup Weekend, Andrew Hyde, who has been at each of the weekends so far, said we showed more, by far, on Saturday night than any of the other weekends. Aw shucks, you must say that to all the development teams!

After the demo, after the network issues were resolved, and after a lot of people left, a core team of developers were left, and we really cranked through a lot in a short period of time. There is still a long way to go tomorrow, but I'm confident we can launch with a reasonable set of features by midnight Sunday.

Overall thoughts for day 2:

• There are a lot of very strong personalities in the groups, and early in the day there was a lot of head-butting, espcially between Development and User Experience.
• I don't think that breaking up into such tightly defined groups early on - Development, Design, User Experience, Marketing, Business Develpment - is that great of an idea. Many people have overlapping areas of interest and could have valuable input to give to several groups. I don't claim to have a perfect solution, but I think having at least one representative from every group embedded in the other groups could foster better communication.
• The dev team is so large that we have trouble communicating and getting stuff done when we're all together. It wasn't until most people left and only 6-10 developers were left that we really hit our stride.