On Tue, Mar 17, 2020 at 8:20 AM Martin Michlmayr <tbm@cyrius.com> wrote:
>
> [ I wasn't sure whether to comment on Brian's platform or stay out of
> this, but I think it's important to scrutinise the plan. Please see
> my disclaimers at the end though. ]
Thank you Martin. You have a great amount of relevant experience. Your
feedback and questions are very much appreciated.
> Brian,
>
> * You write: "Debian Project Leaders should have more time to lead
> rather than be buried in the set of administrative tasks they
> currently face"
>
> It's not clear to that I follow this argument. Right now, the DPL and
> treasurer team only have to maintain a relationship with a third party
> (the TO). Isn't that *less* work than overseeing your own
> organization?
In the past, when we had a struggling TO, we'd have to engage volunteers
to help, and not just help for Debian, but help the overall TO, so we
needed to find volunteers that wanted to help a non-profit with a
wide mandate.
Speaaing only for myself, but one of the things that have held me back over the
years from volunteering for SPI, beyond the very narrowly-focused work I did to
establish Paypal accounts, were the requirements to support an ever-growing
number of non-Debian projects.
Also, it's not just maintaining a 3rd-party relationship. In SPI's case, half the
board is currently Debian members, which is why I initially couldn't understand
the reasoning behind the overall push to end special treatment for Debian.
I came to realize that as Free Software advocates, we have a commitment to
fairness and that is honorable. Thinking about it from the mandate of SPI,
I can't find fault in the choices that SPI's board took to seek equality between
projects.
However, I also came to strongly believe that Debian needs dedicated-purpose
entities that are aligned with Debian's special needs.
> * You argue that "history has shown that volunteers alone aren’t
> enough" and that "difficult to find enough people to volunteer to do
> these things".
>
> I would agree with this. And having done both volunteer and paid work
> in this area, I can attest that there's a limit on how much admin work
> someone will do as a volunteer.
>
> However, Debian has historically had a rather strained relationship
> with paid work. One could argue that the current Debian / SPI
> relationship works because Debian is paying a service provider. But
> if Debian were to have its own foundation, you could argue that the
> topic of Debian paying people will come up again.
>
> You make a good point that admin work is different. But will everyone
> agree? If Debian starts paying for admin work, why not pay for other
> activities where it's been hard to find work?
>
> Maybe some would agree that this is actually a good path to go, and
> that a Debian Foundation would lead to more paid opportunities in the
> future, but I think you could easily see this as a source of much
> disagreement.
>
> How would you address that?
You make a good point, and touch on the one main area that I think might
be controversial. However, I believe Debian has evolved over the years,
to where this might not be as controversial as it was in the past, especially
if we look at it in the narrow area of administrative work.
I believe out views have evolved for a few reasons:
1) We have now accepted 3rd-party paid help from SPI and it's definitely
improved service levels. It would be an artificial distinction to not
consider it for a Debian Foundation.
2) It's now largely uncontroversial that Debian funds are directed to "paid
work", in the form of Outreachy sponsorship.
3) It's been a long time. We've seen what works well with volunteer work, and what
doesn't.
I of course fully understand that having our own Foundation(s) could open up
future discussions about paid technical work, but that probably should be a
separate discussion, as that's not why I am advocating for this. (For the record,
I have mixed feelings about paid technical work, and only feel we should ever
consider it, where there is a strong consensus to do so.)
> Also, who is going to decide who to hire/contract?
The Foundation's board, which would include the DPL.
I'll say now that someone with your background would be near the top of my
list of candidates.
> * "the DPL is no longer a special member of SPI invited to all meetings":
>
> I have to give some context on this (BTW, I don't speak for SPI, but
> I'm a SPI member like anyone in Debian can become). SPI used to have
> 2 board advisors: a representative from PostgreSQL and the current
> DPL. At some point SPI said:
>
> * We haven't used these advisors in years
> * Why pick advisors from 2 big projects when SPI serves all associated projects?
> * SPI's meetings are open: let's encourage everyone to participate
>
> So let me ask this: why hasn't the DPL (or a representative) attended
> the public SPI IRC meetings? Registered guests are mentioned in the
> minutes and I don't see anyone officially representing Debian. Why
> didn't the DPL(s) actively pursue their advisor role?
>
> * "without informing us, after 10 years of de facto practice, SPI
> stopped waiving their standard 5% fees on DebConf sponsorship
> payments"
>
> So basically this is the crux of Brian's platform. He's upset that
> SPI is charging their standard 5% fee on DebConf sponsorship. (And
> there are important questions about this change, but I think that
> doesn't actually matter in this context.)
>
> What matters in my view: why do you think that 0 fees should be
> charged on DebConf sponsorship?
>
> And how does that square with your argument that this work can't be
> done by volunteers. On the one hand, you argue that this needs to be
> paid work. On the other hand, you are upset that SPI is charging for
> its work. Well, I don't think you can have it both ways.
>
> Of course, this issue would go away with a Debian Foundation, but how
> much is that going to cost to run? Definitely not 0 since you suggest
> paid staff. Less than 5%? More?
>
> Fundamentally, I think you underestimate how much work it is to run a
> non-profit properly, and therefore how much it would cost.
I think you misunderstand. I don't think it would be "cheaper" to run our
own foundation, and might even end up costing more. I hope I didn't
give that impression from my platform.
Largely I believe the Debian Project is large enough and mature enough
that we need dedicated purpose entities that we have more direct control
over.
The fact that SPI has made a number of decisions regarding the relationship
with Debian, without informing Debian, was really one of the things that made
me think deeply about the current arrangements, and what might be ideal.
> Have you done some numbers? How much income does Debian have each
> year? How much would an admin of this Debian Foundation cost? What
> other fees and expenses are there? Legal fees? Trademark fees? What
> about an external audit?
>
> This isn't explicitly in Brian's platform but he seems to believe that
> Debian could do it better and/or cheaper. Is there any evidence for
> this?
>
> Have you looked at other FOSS non-profits to see what they cost to
> run? Are you saying a Debian Foundation makes sense even if it's more
> expensive to run than paying a third party provider? (Maybe there are
> other benefits that would justify the cost of bringing this in house.)
I do think it would make sense, even if it costs more than paying a third-party
provider. Largely it's a matter of control. If we want better/different services
in a certain area, it's a decision that we as a project will be free to make.
Cheers,
Brian
> Martin
>
> P.S. I'm not saying there is no merit in a Debian Foundation. Maybe
> there is. Maybe there isn't. (And there are definitely things I
> don't like about the TO arrangement that would go away with a Debian
> Foundation.) My big concern is that in my view Brian completely
> underestimates the work and cost involved in doing this properly, and
> that it will just end up in yet another non-functioning non-profit.
> I've seen to many time people starting new non-profits to solve all
> problems and they just end up creating more (and similar) problems.
>
> Disclaimers:
>
> * I served on the board of the Open Source Imitative, Software Freedom
> Conservancy and Software in the Public Interest: I know that it's hard
> to run a non-profit.
>
> * I'm currently a paid contractor of SPI, so you can argue that I'm
> biased because of my income depends on SPI. You could also argue that
> I know better than many what kind of work is involved.
>
> * I am not on the SPI board and I do not speak for SPI.
>
> * I am part of the Debian treasurer team, but I don't speak for them.
>
> --
> Martin Michlmayr
> https://www.cyrius.com/
>
Attachment:
signature.asc
Description: PGP signature