[asterisk-biz] Bounties.

[Matthew Rubenstein]

On Thu, 2007-11-01 at 14:09 -0400, Bill Michaelson wrote:
> I don't see why work-for-pay (bounties) are incompatible with
> open-source.  As Mr. Rubenstein points out, bounties are cumulative,
> and talent is busy.  At least the potential for large sums to be
> offered is there, and they aren't necessarily "half-assed", whatever
> that really means.  No offer to pay any amount need be seen as
> insulting.  One might just as well regard such derision as insulting,
> but the emotion serves no constructive purpose.
> 
> Perhaps if a more formalized or structured process for managing
> bounties was in place, cumulative bounties would grow larger and
> attract more developers.
> 
> Suppose you could contribute to a bounty by placing the cash in
> escrow, assign a deadline, and agree to an arbiter for judging
> completion.  Would such a system be workable?  Would it encourage more
> development of features desired by the broad community?

	Judgement of meeting the acceptance criteria is a transaction that is
rarely noncontroversial, even when the developer is working to their own
spec, at their own expense, in private :).

	But there is possibly a way to use copyright in the open process: any
claim on the bounty prohibits copying the submission for purposes other
than testing it. The prohibition is lifted, and the code released under
GPL, when the bounty is paid. That setup would bring both the contract
and copyright law to bear on enforcing claims on the bounty, once
accepted in fact by someone using the code.

	The strongest way is for a 3rd party to both hold the bounty paid by
its offerer, and to test the submission before paying its claim. But who
is that "trusted judge"? And should they be compensated, perhaps a
percentage of the bounty, for their work mediating the claims?

	Another way is to include the community as the judge. Code submitted to
claim the bounty can be tested by as many people as want, with test
results submitted with a vote of whether the bounty can be claimed.

	In any setup, the decision by the bounty offerer to pay the bounty
should be final. However, decisions not to pay are the part that will
never be reliably conclusive every time, as is the case in all business
transactions. That's the part of the system that no one has figured out
perfectly. And which each community using bounties has to figure out
what works best for itself.



> Alex Balashov wrote: 
> > On Thu, 1 Nov 2007, Jason Parker wrote:
> > 
> >   
> > > I have personally, on many occasions, seen somebody request a feature 
> > > which I think is neat/interesting/whatever.  I say "I think that's 
> > > great, and I can write that."  I do it *for free*, because *I* like the 
> > > idea.  If somebody is willing to pay me (hey, $50 will buy a pretty 
> > > decent dinner for 2) for something I enjoyed doing - even better.
> > >     
> > 
> >    Being an open-source developer myself, I can entirely appreciate that.
> > 
> >    The concept just seems confused, to me;  if you want to have things
> > developed in the "spirit of open-source" that you have elucidated, simply
> > request them as a feature in some kind of systematic way.  If you are
> > trying to create incentives of a commercial nature along with that,
> > then, in my view, don't make them half-assed.
> > 
> > --
> > Alex Balashov
> > Evariste Systems
> > Web    : http://www.evaristesys.com/
> > Tel    : +1-678-954-0670
> > Direct : +1-678-954-0671
> > 
> > _______________________________________________
> > --Bandwidth and Colocation Provided by http://www.api-digital.com--
> > 
> > asterisk-biz mailing list
> > To UNSUBSCRIBE or update options visit:
> >    http://lists.digium.com/mailman/listinfo/asterisk-biz
> > 
> >   
> _______________________________________________
> --Bandwidth and Colocation Provided by http://www.api-digital.com--
> 
> asterisk-biz mailing list
> To UNSUBSCRIBE or update options visit:
>    http://lists.digium.com/mailman/listinfo/asterisk-biz
-- 

(C) Matthew Rubenstein