This effort is based on the following set of assumptions (which may not be a complete set, but there you go...):
Individuals support free software
In other words, while organizations such as the Free Software Foundation, Software in the Public Interest, The XFree86 Project, Federation for a Free Informational Infrastructure (FFII) and others may be of assistance in bringing together people, it is ultimately individuals that do the work and that get things done.
Individuals can be more responsible than groups
As soon as a group grows past having just one member, politics begins to "rear its ugly head."
It is unrealistic to expect FSEX to entirely eliminate the involvement of politics. It does, however, provide the opportunity for grateful individuals to directly deal with supportive individuals, thus reducing the number of layers of policies and politicians getting in the way.
This is not "democracy" in the sense of majority rule, but it certainly is "democratic" in the sense that it provides representation from the people.
This seems to me to be a Good Thing.
The Internet allows us to reduce dependence on "middlecritters"
A system where individuals send other individuals money by personal volition allows the contributors to allocate their contributions the way the contributor wants to as opposed to having to trust extensively in a "middlecritter" to decide things.
A direct contrast that I would draw here is with taxation. Richard Stallman originally proposed in The GNU Manifesto that there might should be a "software tax" to pay for the production of free software. He has some good ideas on how it might be administered, particularly in terms of providing "tax credits," but I consider there to be two fatal flaws:
He does not clearly define the tax base, which is always the fundamental question in taxation.
If there is no clear basis on which to assess a tax, the system breaks down. And I do not see any reasonable options that do not require extreme intrusion of government in order to evaluate the base.
Most of tax policy, and thereby the thick books sitting on tax lawyers' bookshelves, relate to do with describing exceptions and counter-exceptions to the loopholes that tax lawyers spend their time searching for. What tax lawyers do is to seek out ways of arguing that transactions are "not taxable" under some provision of the law. As time goes by, taxation law accretes exceptions atop exceptions to deal with loopholes that get abused. (As a result of all this, I have complete faith that "flat tax" schemes that would-be reformers blather about will do absolutely nothing to simplify tax codes. But I digress...)
In general, getting a tax set up means joining the "tax lawyer game," which is something I think we very much do not want to do.
He does not address the agency problem.
In order to ensure that funds that come from the universally assessed charge are used for legitimate purposes, this mandates an unfortunate array of bureaucracy and regulations that would by themselves consume much of the funding, thus increasing the tax rate that would need to be assessed.
This is why tax rates are so high, and governments are accused of inefficiency.
In effect, inefficiency is a direct price of trying to avoid corruption.
In contrast, if individuals send contributions when and where they prefer, this keeps responsibility in their own hands, and allows us to avoid the vast majority of these "mechanisms of government."
It often appears attractive to try to turn these sorts of projects into some form of "registered charity" in order to obtain favorable tax treatment, for contributors, of their contributions.
Unfortunately There Ain't No Such Thing As A Free Lunch is abundantly true, and charity policy is no exception to this principle.
The donors may be nicely treated by the taxation authorities, but this there are two other corresponding costs that occur at other levels:
The charitable foundation must go through some continuing paperwork in order to validate their charitable status, and
Payments to developers will generally be treated as taxable income, often with "employment taxes" and other similar charges. If payments cross regional or international boundaries, things get more complex still.
(I've worked on an HR/Payroll system project at American Airlines; the domestic complexity in the US is quite frightful; the international phase is not going to even attempt to handle payroll as trying to centrally manage payroll policies for over 50 countries would be far too much to cope with...)
The Free Software (Gift) Exchange Registry consists of a registry of the following things:
Developers vouched-for as being involved with Free Software projects,
A list of such projects, so as to provide some organization to the data,
Amounts pledged to developers, and whether or not pledges have been received.
The two critical things are the lists of developers, and the lists of pledges. Once joined together, people can examine the data to determine that, for example:
People working Linux kernel development might have been pledged $300,000, whilst those working on library development were only pledged $30,000.
Since libraries are important too, people may grasp this fact, and allocate later contributions more intelligently.
The database would consist of, roughly speaking, the following groups of data:
Anyone may register themself as a contributor.
This will contain address and naming information, as well as whether or not the individual wished to be publicly identified as a contributor. (Some people may prefer not to be known.)
This would also be used to represent identification information for would-be recipients, but more about that later...
It still proves useful to list projects. People developing free software tend to be associated with some particular system, whether it be a tool, application, or piece of documentation.
This would include information such as the name of the project, and contact information such as a primary email address and a URL.
It would be logical to set up a view of contributors and contributions based on this table, determining people and amounts associated with each system.
This is a very important table from an authorization point of view; it indicates what people are affiliated with projects, and thus appear worthy to receive gifts on behalf of their efforts.
My initial thought is that, in the ancient Hebrew legal traditions, people be considered "vouched for" if there are three witnesses willing to vouch for their address and for their involvement with the project/system in question.
One would hope that this would not be a matter of great contention; I however expect that most of the political problems that would arise would specifically arise in this area.
It would be reasonable for someone to request that contributions go elsewhere. For instance, I would suspect that Richard Stallman might prefer that contributions go to the Free Software Foundation rather than to him directly. Similarly, there might be a lot of people that would wish to express their gratitude to other "celebrities" in the little world that is the free software community; this might prove to provide overwhelming amounts to such luminaries as Linus Torvalds, Eric Raymond, Alan Cox, and others.
An unfortunate aspect of what might be called the "celebrity effect" is that these celebrities are likely to receive a disproportionate portion of funds. There are three salutory counter-effects:
These people generally seem fairly responsible, and I expect that it is likely that they would pass on excess monies to destinations that they would consider worthy.
These are gifts, and supposing a million people want Linus to be showered in money, it is not his obligation to deny people their desires.
Hopefully after making some such "millionaires," people would get some of the worshipfulness out of their systems, and consider more carefully where to send funding in the future so as to get a little utility from their encouragement.
Note that not all recipients will necessarily be programmers as such; people that support free software in other roles such as writing documentation or otherwise providing useful support may be well deserving of reward for their efforts.
This would join together people who are contributors with "vouched" recipients. Initially, this table merely indicates a pledge as opposed to a "validated" contribution.
Note that since no money goes to the Free Software (Gift) Exchange Registry, we have no way to directly validate that contributions are actually received. All we have is a voluntary reporting of amounts people are publicly declaring that they intend to give as gifts.
This data would be (voluntarily) filled in by recipients of gifts to indicate amounts actually received.
A logical approach would be to allow recipients to "log in" to a web page that would present a list of outstanding pledges that are directed to them. They could then mark off those that had been received, and perhaps indicate amounts received that had not been pledged.
Several "views" of the data would obviously prove useful:
Contributor: Project List
To a contributor, it would be valuable to have a list of projects, amounts pledged/received, and the ability to get, for each project, a list of possible recipients of gifts.
Having information on what projects appear to have received support would be valuable in allowing individuals to direct contributions (indirectly) towards whatever projects they deem "need their money most."
Contributor: Gift List
A contributor would find it desirable to determine if their contributions have been received, and to determine what other things they have contributed to in the past.
Recipient: Contribution List
This list would indicate what contributions are outstanding, and give the ability to indicate what gifts have been received.