8 Aug 2011

So how did they do? The #DarHackathon results show

The judging panel waits while participants
prepare to present. Photo from @simplyluca
Two days ago, an enthusiastic bunch of young Tanzanians set off in teams on a 48-hour software development marathon. Or rather a sprint, since it would usually takes weeks or months to develop the kind of applications they were putting together. After a couple of largely sleepless nights (judging by the kind of chatter on twitter overnight and the conversations going on at COSTECH this morning), the 48 hours were up, and it was time for the teams to show us what they had managed to do.

So how did they do?

Well, we had several shopping- or trade-related apps - a map-based system for finding second-hand items for sale near you, an accreditation system where satisfied customers earn a good reputation for fundis and other service providers, a tool linking online bidders to offline auctions, and an online shopping centre for local shops to show off their wares. The current problems with the availability of fuel suggested very similar ideas to two groups - a resource showing where fuel is available and at what price, one using the web, the other using SMS. And there were a few social-impact apps as well - a security system for university students to register and protect their property, a system for the police to keep records of problems with vehicles and drivers, an SMS-based health information service and an online voting system for general elections.

Three groups were awarded certificates and prizes. Team Reig, the group responsible for the online voting system and the only group with any female participants (see here), came in third. Team Kidogo, with their map-based system for locating second-hand items for sale, were second. And the overall winners, fully deserved, were Team Four, for their reputation system for service providers, the Alternative Certification Environment (ACE).

This was well presented, paid attention to aesthetics as well as functionality, and most importantly, had a good and useful idea at its core. It builds on the idea that word-of-mouth from satisfied customers is often a service provider's best form of advertising. If a programmer (or electrician, or mechanic, or tailor) does a good job for you, you will tell you friends about it. So why not tell people online, and over time that programmer (etc) earns a reputation based on how many recommendations they get.

Overall, for a 48-hour exercise, they all did pretty well. Nothing was in a fit state for launching immediately, but you couldn't expect that in such a short time. I was very impressed by the skills of the developers - there were clearly some very bright and capable people taking part.

On a more critical note, I felt the ideas behind the apps (with a couple of exceptions) lacked imagination and didn't really push the potential of the web and mobile phones very far. Both the fuel station apps, for example, rely on fuel station owners updating the system whenever they changed their prices or stopped selling. Why not crowd-source that information instead? There was no form of crowdsourcing involved in any of the apps, no use of social networks, nor anything media-related, which was a shame.

Another concern was the apparent willingness of participants to reinvent the wheel at every opportunity. There are open source online shopping platforms that are easily available online and ready to go, so why develop a new one? There are user-management systems you can take wholesale, so why bother developing your own login system? Enthusiastic amateurs can do a lot online these days without needing software expertise, so software developers need to raise the bar in order to stay in business.

But please don't get me wrong, the skills on show were very impressive. With a few good ideas of ways to put those skills to use, there clearly potential in Dar's developer community. And after all, this was a software development exercise, not an idea-development exercise.

And with more events like this - I'm told there are many more to come - with developers are showing off their skills and learning from the expertise of others, fresh and powerful ideas are bound to emerge before long. Pick up half an idea here, combine it with a thought from over there, and work out how to package it together with some inspiration from somewhere else, and you're away. That's how new ideas happen, and the Dar Hackathon was a great way to get things started.