Re: [Debconf-team] Information gathered for travel sponsorship
On 02/09/12 at 21:24 -0300, David Bremner wrote:
> Didier 'OdyX' Raboud <odyx@debian.org> writes:
>
> > Agreed. What's interesting is the ratio between the income and the
> > amount requested, not any of both absolute numbers. What I envision is
> > a list ordered by "days of work needed to produce the money I
> > request".
>
> I guess I'm not seeing how days of work is different than
> monthly_salary/20 (approximately). So you may as well just ask for
> salary. I can also imagine people with same salary having vastly
> different amounts of money left over to spend on travel. But maybe this
> is covered adequately by the open-ended questions.
>
> Anyway, I guess is all quibbling about details. The main question is if
> there is concensus that the privacy loss in revealing income information
> to the committee is outweighed by the benefits to the process. That was
> far from clear to me at the BOF.
I also think that this goes too far.
I think that the sponsorship committee should:
1) verify that the funding asked for is adequate (given the expected
travel costs from the originating country)
2) rank demands based on *benefit of attendance for Debian*.
If you quantify, for each applicant, the added value of the
attendance for Debian (not in $ or €, but in a "virtual currency"),
it becomes a simple problem (maximize benefit for Debian) with a
simple solution: order applicants using the "benefit/cost" ratio,
sponsor as many as possible => maximize the total added value of
Debconf for Debian.
I don't think that the "economical effort" should be part of the
ranking. First, it's very hard to quantify, because you need to consider
at least: income, general cost of living, family & other recurring
expenses, does attending debconf result in a loss of income (case if
working freelance), in a loss of vacations (if attending debconf during
vacations), or is the applicant attending debconf as part of your work,
etc.
Second, I think that it's irrelevant: I don't see why we should care
about how hard it would be for the requester to attend if we don't
sponsor him. Also, it turns the process into asking "OK, that requester
says s/he won't attend if we don't sponsor him 200€. But maybe s/he
will attend anyway? Let's see how much s/he earns."
I also don't think that the "current/past contributions to Debian"
should be ranked. Of course, they should be part of the form, to provide
context for the "project" of the requester and make sure it is realistic
(references are a good idea and help with that, too). But if you rank
past contributions and use that to award sponsorship, you turn Debconf
in a "thank you for your past work!" event. Someone who does fantastic
work in Debian, but plan to attend Debconf as a base camp for visiting
Switzerland, should not be sponsored.
Another idea: ask sponsored attendees to write a short report after
Debconf, on their actual work during debconf. It would probably be very
extreme to remove sponsorship after Debconf if the results are not good
enough, but it could be re-used during the next year's process.
Lucas
Reply to: