IPFS is a cool tech, for sure. But it looks like IPFS is focused on "high throughput content-addressed block storage model, with content-addressed hyperlinks".
The file-transfer needs of DIST is simply for a decentralized method of firmware updates--as outlined in the IBM ADEPT paper.
As Melanie's co-founder for ToVieFor, it's true that they painted her in a bad light. She's super intelligent and very very good at grabbing onto a vision and propelling it forward. She may have hung on too tight to that vision at times, but I'd take that any day over a co-founder who was aimless and did not know what they wanted. We were up against a very difficult, entrenched industry, and she made as good of progress as I could expect any new entrepreneur to make.
I too am a bit disappointed in Bloomberg. The fact of the matter is that you have a lot of strong, driven entrepreneurs in the same program, and while some of what you saw is accurate, Bloomberg really made an effort to polarize and make extreme the various personality types. Reece does not always talk about sports or use athlete analogies. Jason Baptiste doesn't walk around constantly saying he'll own the internet (although if you prod him for it, he'll probably say it), and Melanie is not clueless. I'm super excited for her current startup, she definitely shows more wisdom in leaner methodologies.
As far as the idea goes, I think it's great, and very polished. Nice work!
I had a bit of a bad initial reaction when I heard the name, however, as it feels very imperialist and negative to me. Might just be me though, so just take it as a single datapoint. But my suggestion would be to look into names that are a little more positive.
We attended the Foursquare hackathon yesterday as participants and ended up winning with our thedealio.at private messaging app.
To be honest, the only thing we had completed prior to arriving at 1:00p was that we knew the idea. We had no name, no code written, had not heard about the new API features released, nothing. So at least it's one datapoint that the winners aren't all pre-developed, polished products.
For me, the biggest benefit of hackathons are that you are absolutely forced to use all sorts of skills, not all developer-focused. We had to discover a problem that was worth solving for a lot of people (P/M fit, I suppose). We had to battle with stupid issues like buying an .at domain name and having my credit card get locked up due to an overseas suspicious purchase, then having to wait for DNS to propagate so we could get an SSL certificate issued as required by a new feature of the Foursquare API. Also how the look and feel should be, the UX--simplicity, fun, and engaging. We spent the first three hours writing no code, just getting our game plan in place.
Then of course in the actual development we broke all kinds of rules like code duplication, no testing, editing code directly on the production server, etc. We focused squarely on speed, knowing that time would be our biggest threat.
Which brings me to my main point. Building any software product is all about compromises, and knowing when to make them. It does go against the spirit of a hackathon to bring a polished product to demo, while just networking all day. We didn't do that and did okay. But only by using every means available (including pretty much skipping dinner).
I hope that the possibility of polished products doesn't hurt the hackathon spirit, and this post is mostly to offer encouragment that single-day hacks can still hold their own. There were some really great hacks there, and the ones I know that were built just yesterday were among my favorites (4squareand7years being my particular fave).
I would probably go the highest-level route I could, using tools that I am familiar with (and won't get caught in nasty workarounds due to my unfamiliarity with the platform).
In my case right now, I'd probably put together a web mashup using a container of useful platforms--Like Apache ServiceMix using Camel, ActiveMQ, CXF, etc. to grab services, munge them, and output something useful.
Then again, going super low-level might be fun too. Go to the depths of the OS or a compiler and fix an annoying persistent issue.