Ask HN: Do you have startup ideas to share?

You can read the article of Paul Grahm in order to find some ideas by yourself: http://paulgraham.com/startupideas.html

Recently i thought about making to-do list that don't suck :) Interested to know more about it?

I have one but I think it pretty crazy and I am not really sure if it can work. I would love if you could find holes in my idea and help me to refine it or tell me if it can work at all.

So I have been thinking that the biggest roadblock to solar adoption is the high upfront cost of installation. If there could be an incentive to make money with solar a lot more people might be willing to "invest" in it.

So my idea is similar to carbon credits. It is a lot easier to measure the amount of energy generated by the solar panels. So the amount of clean energy generated by the households could be sold off the organisations with a bigger carbon footprint. Similar to how advertising networks sell page views to the advertisers.The unit of energy is well established. We could set the price of 1 watt of clean energy to a bidder.

So the question remains why the bidder would want to buy this clean energy? I don't have an answer to that besides that it could be a good marketing tactic. For very large cooperation it could be a chance to gain some good PR (if they really care about that).

The people participating in this program would have to install a chip that would monitor the generation of this clean energy. With networked chips like tessel and spark.io coming out soon it could become pretty easy for people installing solar panels to connect these panels to a local cloud where there could be an automatic transfer of funds to bank accounts of people depending upon the amount of clean energy that they have generated. Now for a give area this energy should be the same for a standard solar panel since the are will receive more or less the same sunlight. With some work standard pricing can be reached for a locality.

I'm working on a highly distributed cloud framework that implements a Bitcoin operated OpenStack deployment. Basically, you install OpenStack at your house or small data center, launch a small instance with some Open Source code that turns the instance into a virtual appliance controller, and then join a pool of other providers/users. The pool operator provides search functions (like geo) and Bitcoin addresses so others can launch a server by just paying coin to a given address. No usernames, no passwords. Secure, reliable payments for providers to get ROI on excess/hobby capacity. Fast, easy, and dirt cheap.

The initial pool run by myself will provide low trust, low cost, highly transient instances. Chaos Monkey model. Market prices will likely fluctuate based on instance launch demand.

The concept encourages multiple pools running the code, so there will be future opportunity in building up a brand that attracts a given set of infrastructure providers and/or service companies to band together and coop. This could be done to leverage features like geo location, excess capacity requirements, feature sets, or trust levels through accreditation. Think about a handful of Silicon Valley companies sharing infrastructure among themselves and being able to balance the costs through Bitcoin. They could also share cooperatively in the areas of operational competence and security, given they have a good relationship between the companies (which many do).

For me, it's been a long haul working on the concept, especially considering I had to get OpenStack accessible for the average user. Now I'm mostly done with my OpenStack scripts, I'm finally cranking out the code for the controller and have it launching instances by sending in Bitcoin to the instance addresses. It's pretty slick when you use it.

More work has to be done, so it'll be a few months before I do a general release of the first pool. If you or anyone else is interested in this concept, building a new pool brand, or building the services that need to be build around it, hit me up. My email is in my profile.