also sprach Lucas Nussbaum <leader@debian.org> [2014-10-14 17:46 +0200]:
> It seems that if we change the process to:
> - budget team sends the budget for review to debconf chairs + DPL (who
> will give the final approval)
>
> It addresses your major concerns:
> […]
> Would that work for you?
Yes, with the exception that I don't think the DPL should be the one
to give final approval of the budget.
The DPL will always be in charge of the amount of money Debian
commits to DebConf, but this is a different decision from approving
a budget, and it's a decision that predates each budget approval.
By the time the budget is approved, the chairs are acting within the
powers delegated to them, so there is no need to involve the DPL.
My motivation is not to cut the DPL out of the loop. I think we
should strive to make decisions as close as possible to where the
knowledge is. There might not be much delay introduced when both DPL
and chairs need to get involved and understand the budget. But it'll
be more work for the budget team, and it'll mean the DPL uses up
time s/he could spend elsewhere for things s/he's already delegated.
> > If it looks like we'll fail to meet the approved budget (which may
> > well already include Debian funds as income), then we might need to
> > investigate the possibility for Debian to commit more funds ahead of
> > time. Until more funds are committed, the budget won't be met, other
> > sources of income have to be sought, and expenses cut. This is the
> > whole purpose of budgeting.
>
> I disagree with that. Expenses should be decided based on their
> importance, not based on whether money is available at a given time.
I never said anything to the contrary. If we consider travel
sponsorship important (and we do), then we won't just slash it e.g.
before slashing the truly optional stuff.
Expenses *must* be decided based on whether you can pay for them.
If you don't meet your budget, you need to either increase income
(which could mean talk to the DPL and try to get more money approved
from Debian), or lower your expenses, and write a new, balanced
budget. There is no other way, unless you have your own central bank
or a lot of influence therein. The budget is a forecast and if your
forecast says you want to spend more than you have, you will not
just magically end up with enough money to avoid running a negative
balance.
> If you do the latter, you end up with situations where travel
> sponsorship cannot be allocated because there's not enough
> sponsorship money received *yet*.
The approved budget contains a certain amount allocated for travel
sponsorship. For the sake of simplicity, let's say that is 10k out
of a total of 100k expenses. A budget now *binds* you to not spend
more than 90k on all the other expenses. So provided that you
generate 100k income, you are guaranteed to have 10k for travel
sponsorship.
If you only manage to generate 90k income and all efforts to fill
that hole didn't pay off, then you need to do something. Since
travel sponsorship for DebConf is an important aspect, we won't just
slash those 10k. We might be forced to reduce travel sponsorship to
9k and reduce other expenses by 9k as well. This is the sort of
prioritisation you speak of. I will get back to this later.
The only time we'd really get into trouble is if (a) all other 90k
expenses were already committed on everything other than travel
sponsorship, and (b) only now do we find out that a 10k sponsor is
bankrupt, so we don't have the money to refund people's travels. To
guard against this, you can buy insurance… or count on Debian. But
this is *highly* unlikely to happen, especially since we are
*trying* to raise funds early and be done with it by the time we
start spending stuff.
> > If DebConf e.V. fails to pay bills in the end, then it is
> > primarily our responsibility to deal with that, not Debian's.
>
> It might primarily be DebConf e.V.'s responsibility, but it's
> primarily Debian's problem if we fail to organize a DebConf that
> meets Debian's needs. So it's also up to Debian to ensure that
> DebConf will be a success.
Absolutely, and we'll gladly increase the amount of money Debian is
willing to commit in our budget, especially in the light of any
surpluses going back to Debian anyway. So what is DebConf worth to
Debian? 20k? 50k? 100k? We can budget accordingly. And even if you
are going to commit 200k, you can rest assured that this will not
impact our fundraising efforts, nor drive up the expenses.
In other words: I think we should work towards a budget including an
amount X that comes from Debian, and this amount X has already been
granted by the DPL.
I don't think we should even consider working on a budget that
contains some amount Y that needs to come from Debian, and in asking
the DPL to approve the budget, coerce Debian into committing this
amount. The difference might seem cosmetic. I think it's fundamental
instead, actually.
> > understanding of the whole picture, and I think
> > worst/base/best-case scenarios are best for that.
>
> I would prefer to have a rough common agreement on priorities in
> terms of cost reduction.
Sure, and I think that there is no better way to express that than
using worst/base/best-case scenarios.
If we don't meet our budget, then we need to cut costs, and which
costs to cut is — to me — best expressed by the "worst-case"
scenario of a budget: the minimal case in which the conference will
happen. Between base and worst case, there's a multi-dimensional
descent ("there are many many ways to slash costs") and the ideal
path to take might be very different from when you need to save 5k
than when you need to save 10k.
Concretely, it might make sense to cut starter and desert out of the
conf dinner and save on C&W by cleaning up ourselves, if this means
we save 5k. But if we need to cut expenses by 10k, then maybe the
better answer would be to cut the conf dinner entirely (saving e.g.
12k) and pay 6k for C&W instead.
Point being: worst vs. base case already defines the playing field
within which we can move around to make ends meet. If travel
expenses for base case are 20k and 15k for worst case, while conf
dinner might be 12k and 0, then this is a statement on priority.
Granted, it's much harder to get agreement (by whom, btw?) on these
scenarios, as they are rather complex. But they are much more
flexible later when you need to make decisions that aren't so nice.
A list of priorities is probably easier to agree upon, but it would
limit your choices later on.
It might also be that I simply don't understand what you are
envisioning. Would you be able to provide an example?
> It seems that you have some unspoken fears about this review from
> DebConf chairs and the DPL. It might be useful to do this review
> work rather sooner than later, so that we all know if your fears
> were justified :-)
I am all in favour of sooner rather than later, I think that's clear
by now.
I am surprised you call my fears "unspoken" because I thought I just
composed two rather long e-mails about my these. In one sentence,
that would be: we should work towards an organisation where
decisions can be made as close as possible to where the knowledge
is, and avoid involving more people than necessary.
--
.''`. martin f. krafft <madduck@debconf.org> @martinkrafft
: :' : DebConf orga team
`. `'`
`- DebConf15: Heidelberg, Germany: http://debconf15.debconf.org
DebConf16 in your country? https://wiki.debconf.org/wiki/DebConf16
Attachment:
digital_signature_gpg.asc
Description: Digital signature (see http://martin-krafft.net/gpg/sig-policy/999bbcc4/current)