Subject: Re: GiftNet, distributed credit card processing
Date: Mon, 1 Oct 2001 15:10:09 -0400

Seth Gordon wrote:
> I don't see how GiftNet would provide a service that anyone would
> actually want to use.
I missed the original proposal, but considering that I had thought of
something a long time ago that might reasonably be called "Gift-Net"...

> Anyone who wants to reward the authors of a particular open-source
> program can usually find out their email addresses, and use PayPal or
> e-gold to send them money.  If the authors want to make this job
> easier, they can put a PayPal or Amazon Honor System link on the front
> page of the program's Web site, or they can use something like Cafe
> Press to make associated T-shirts, coffee mugs, etc.  If a third party
> wants to encourage its clients to support a certain project, they can
> put the appropriate PayPal/e-gold links on their own site (crediting
> directly to the accounts of the program's authors).

I am a satisfied user of Linux, Perl and Samba.  If I wanted to give a
hundred dollars to any of these three projects, and have the feeling
that it was somehow fairly divided among the contributers, that would
be rather hard.  Oh, I can give money to, say, Larry Wall.  But he
said that until last summer he would have no idea where he should send
money for the overall benefit of Perl.  (The appropriate place to
donate is now an organization called YAS.  Strangely enough this does
not appear in the list of contributers to Perl.)

> All of this is already easy to do, with no significant transaction
> costs, and the mediating institutions (PayPal et al.) have done as much
> as any such institutions can do to encourage strangers to trust them.
> So I don't see where the value proposition (I believe that's the
> proper biz-speak term) for GiftNet is.

If I followed your suggestion with Perl, the transaction cost of
trying to track people down, and then deciding (for instance) what
portion to give to Simon Cozens, would be nontrivial.  In the US it
would also be possible to set up a registered charity.  As a charity,
people would get tax benefits on donations, and it would get tax
benefits on its awards.  Plus if you have a charity, employees of
many large companies would be able to take avantage of matching
donations to donate much more efficiently.

So an organization to facilitate giving money to open source
developers can indeed have substantial advantages over PayPal.

> The hidden assumption behind the GiftNet proposal is that many people
> don't give money to support open-source programs because *it's too
> hard for them to do it*.  I don't think this is a significant reason.
> The most significant reason, I think, that many people don't give
> money to popular open-source products is that *they can get away with
> it*.

I believe that there are people who would donate to larger projects
but cannot figure out who they should send a donation to.

For the record the proposal that I had thought of looks like this.
Set up a charity with a list of key projects, organizations, and
individuals who are influential in free software.  Each person who
registers themselves as being involved in free software should
indicate roughly what portion of free software related activities that
they are involved in depend on which other people who are "lower
profile" than they are.  Then any person who wants can donate to this
society and mark the key person they want their donation to start at.
1/4 (picking numbers out of the hat) of the donation goes to that
person, and the remaining 3/4 works down to the people that person
named.  Those people each get 1/4 of that, and the rest continues to
trickle out.  And so on until the donations become too small to bother
tracking.  So that way a single donation to, say, Linus Torvalds would
result in a web of donations that roughly corresponds to "The Linux

I think it could be set up.  It might or might not attract donations.
But I think that the exercise of quantifying how much others matter to
you would cause more fights than the money is worth.

All in all I suspect that funding charities like the EFF, FSF, and YAS
is a better option than trying to find a comprehensive and fair way to
compensate open source developers...