On Wed, 2012-06-20 at 03:35 -0400, Richard Darst wrote:
> We could decide to allocate travel early, before we knew we'd have a
> surplus, and have confidence we'd eventually reduce costs. However,
> we are too conservative (=responsible) to want to risk having a
> deficit. Thus, the travel-last system remains.
> But there is something we could do: "Summon Bigger Bank". We would
> basically want something to even out the year to year uncertainty, and
> average out the travel sponsorship over years. And we do have a
> bigger bank: Debian.
Yes, I tend to think that travel sponsorship (or at least an initial
batch) should be somewhat separated from other DebConf costs and dealt
with early on.
I don't think it's even a question of a "bigger bank" and absorbing
costs, as DebConf is part of Debian. Equally, I don't think there's any
very good reason why travel sponsorship should be managed by the same
people who are focused on organising a conference venue etc.
> Of course, there would be a lot of implementation details to work out.
> I've watched four DebConf budgets now, and I think this would be
> workable. Historical travel costs are:
> dc9 - 27 kUSD
> dc10 - 23 kUSD
> dc11 - 40 kUSD
> dc12 - 18 kUSD
> which is much less than Debian's reserves, even subtracting a year's
> worth of new hardware according to the plan.
Yes. If this approach were taken, we might envisage allocating even 20
kUSD of travel funding early on. (At the implementation level, this
would then be treated as a fixed cost in the DebConf budget. In years
where there is spare money later, a second set of grants could be made.)
Making travel grants earlier would mean, among other things:
- The same amount of money would be able to fund significantly more
- We would avoid the current uncertainty in whether people allocated
travel grants can actually come any more or not.
- There would be more time to use a better (potentially fairer) travel
grant allocation process (e.g. the one fil has designed).
> And correspondingly, were we not to
> get some of our sponsors, we _were_ ready not to have travel
> sponsorship at all, if we needed our DC11 surplus to pay for the
> facilities, hotel and things.
This year I was perhaps the person most against allocating amounts to
travel grants while there was budget uncertainty, within the current
system. However, I would see doing early allocation, and losing some
budget flexibility, as having enough advantages that I would favour the
Equally I am perhaps the person in the team most in favour of trying to
make every DebConf end neutral or positive for overall Debian funds.
But if early travel grants were made with the cooperation of the wider
Debian project, then I would see the possibility of a small net expense
for Debian as a risk worth taking if an appropriate scheme were set up.
(From my point of view, the ideal scheme for deciding DebConf travel
grant allocation might remember not only the comparative utility for
Debian of taking different people to DebConf, but of spending money on
DebConf travel versus other Debian travel costs.)