Organization

FreeTable is dedicated to the public good. That said FreeTable probably will need to raise money to pay substantial hardware bills and other operating costs. For this reason, and the fact that creating public datasets does not seem to be one of the permitted activities of a 501(c)(3) non-profit, FreeTable will when it incorporates likely incorporate as regular corporation.

The need to raise money is a pressing problem. There are two routes, issuing debt, and issuing equity. The problem with issuing debt is if the debt can't be repaid, the issuers can come and take control of the business, destroying FreeTable's pubic good mandate. Issuing equity is in this respect safer.

If we are successful, we plan on dual-class stock structure, similar to that used by Google, News Corp., or Berkshire Hathaway. Class A shares would initially be freely issued to non-profits and subsequently to for profit actors in order to raise capital. They could be used by the non-profits as a source of dividend income or sold to raise additional income for the non-profits. Class A shares would be entitled to one vote per share. Class B shares would be issued to actors hoped to represent the public good, and entitled to ten votes per share. Class B shares would be convertible into class A shares , or transferable as is.

Stakeholders

Consideration should be given to what would happen if one day we are able to run at a surplus, and therefore choosing as shareholders organizations that can do the greatest good with any such surplus.

Proposed initial Class A shareholders:

  • 20% Gordon R. Irlam Charitable Foundation - established by project founder/funder
  • 20% International development NGOs
  • 20% Internet organizations eg. EFF, Wikimedia
  • 5% for volunteers and potential future employees, but not contractors
  • 15% in exchange for additional paid in capital

Proposed initial Class B shareholders:

  • 10% Gordon Irlam - in exchange for $100k initial capital
  • 10% additional class B shareholders TBD

There is a tension between the ability of users of data to make maximum use of data, and the desire to operate at a surplus so that others can use the funds to do great things. The proposed division amongst shareholders has been chosen with this is mind.

Revenue model

On a per user basis costs are very small, but in the aggregate they could be substantial. It will thus be necessary to charge for database access. This will be done in a manner similarly to Google's App Engine, in which the first so many operations per month are free, and then for high volume users we charge $0.00001 per database operation. Certain popular database tables may be charged at premium rates.

The goal is for this revenue to be sufficient to fund our operations as well as provide some funding to non-profits. This is a reasonable goal. Ebay has a single database that is used to match buyers to sellers, and they take in $9b/year. So even allowing for a far greater proportion of the transaction costs to benefit users, if things go well, and we are truly successful, a revenue of $100m or more isn't out of the question.

Founder/funder

Gordon Irlam Gordon Irlam runs the Gordon R. Irlam Charitable Foundation which is active in global public health. He was technical product manager at Cygnus Solutions, the world's first free software company, and technical founder of Postini, an anti-spam company sold to Google for $620m.

He has committed $100,000 in funding to this project. This funding will be released in stages subject to milestones being met. It will be used to pay for the hardware resources needed for an extended period of time, sample data, organizational establishment legal/accounting fees, marketing, and some development expenses.

He wrote the initial FreeTable implementation.

Developers

Sergey Krillov Sergey Krillov is an extraordinarily talented Python developer affiliated with Rainboo Software.

He rewrote the initial FreeTable implementation, and implemented the FreeTable browser in one weekend.