[IAEP] Sugar Labs Budget.

[Tomeu Vizoso]

On Tue, Apr 7, 2009 at 02:52, Benjamin M. Schwartz
<bmschwar at fas.harvard.edu> wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Christoph Derndorfer wrote:
>> Am 06.04.2009 um 23:28 schrieb David Farning <dfarning at sugarlabs.org>:
>>> Last week, I set the rather abstract goal of raising $100,000US for
>>> Sugar Labs.  $100,000 seems like a reasonable number for a project in
>>> its second year.  Last year, Sugar Labs proved that it could get a
>>> surprisingly large amount of development accomplished with very little
>>> money.
>>>
>>> As for direct spending, I believe that Sugar Labs should focus it's
>>> resources entirely on community and ecosystem development.
>>
>> Mmm, personally I'd say that at least a third of Sugar Labs' budget
>> should go towards actual development.
>>
>> All of that traveling and outreach becomes somewhat of a moot point
>> when the actual platform we're building a community around and
>> marketing doesn't work as advertised/expected.
>
> In my view, it's a question of amounts.  Some rough numbers from a US
> perspective: paying a software developer a salary of $50,000 a year, much
> lower than industry standard for the skill and experience we require,
> typically costs the payer about $100,000 a year due to payroll taxes,
> health insurance, and business paperwork expenses.  That's David's entire
> imaginary budget.

Well, actually OLPC spent less than $50k a year on me at some point. I
know that NN said that Sugar was a too expensive business for OLPC but
you cannot believe anything he says about Sugar...

Keep also in mind that about 90% (or more?) of OLPC's software
development budget must have been spent in non-Sugar stuff. In other
words, if OLPC hadn't invested in Sugar at all and had gone straight
to deploy plain GNOME, they would have had to pay anyway kernel
hackers, OFW wizards, linux "generalists", localization organizers,
etc. When OLPC said that they had "doubled" the Sugar workforce last
June it was just FUD because none of the people they contracted got to
actually work on Sugar.

And not only they had many times more people working on non-Sugar
stuff, but each of them must have cost them much more, by your counts
of how much costs a software developer in the US.

But I don't think that paying developers should come from SLs budget
anyway for the reasons I wrote in an earlier email.

> Also, I think it's important to remember that Sugar Labs has never
> received a development grant from any granting agency and Sugar has never
> been associated with any profitable entity.  Walter has diligently applied
> for many grants for educational software development but (as far as I am
> aware) none have been accepted.  OLPC and Red Hat spent way more money on
> Sugar development than they made back.  Sugar has not been a profitable
> enterprise.

True, but you may be missing the point that there are people in this
world that would actually be benefited if they spent money on Sugar:

- governments and other organizations deploying Sugar. They are
already paying big money in support costs because Sugar could be
better tailored to their realities. Also, they could get more bang
(education) for their money if they improved Sugar themselves.

- deployers of computers in education that are losing sales because
they don't have the selling edge they could have if Sugar was part of
their product portfolio.

- schools and other owners of computers that aren't being used or are
underutilized because the software in them isn't compelling enough for
being used in their educational context.

- more?

I have zero business knowledge and have no idea how money will finally
flow between those who have it and those who can use it to improve
Sugar, but I feel it like if enough potential accumulates, things will
start moving by themselves soon.

About grants, this may be of interest:

http://www.stormyscorner.com/2009/04/supporting-free-software-with-grant-money.html

> Currently we have no money, and no path by which we may particularly
> expect to get money, so all talk of spending it is a bit premature.

Agreed.

> Paying the core developers enough to work full time on Sugar would cost
> hundreds of thousands of dollars per year.
>
> The obvious question is: what can we buy for less?  I don't think we can
> reasonably buy the time of people who already have full-time jobs.  That
> leaves two types of people: contractors and interns.  Professional
> software development contractors typically charge very high rates,
> probably not affordable to us.  Student interns are cheap: perhaps $3000
> for three months of work.  It seems reasonable to me that Sugar Labs could
> offer summer internships, to go beyond the number of slots offered through
> Google Summer of Code.  Such a program would serve to enhance "community
> development" and potentially add some value to Sugar.  However, interns
> are unlikely to have much luck making Sugar "work as advertised", because
> core features and core debugging require familiarity with a large codebase
> that is beyond most undergraduates.

Well, yes and no. There's only one thing that is in they way of an
inexperienced hacker to get into the innards of Sugar and do really
good stuff: fear. We have this culture of praising the ultra-smart
hackers that master one codebase, and it misleads people in two
different ways:

- experienced developers don't want to leave aside their own tools and
start using the ones that are already in place for fear to being seen
as less experienced as they think they are, so they refuse to get
their hands dirty and shout instead for a rewrite,

- unexperienced developers see how people praise the ones that are in
the ring and believe they are much smarter than they really are,
increasing the fear to failure.

One of our challenges is to create a community were that fear is
reduced to a minimum. This means stop praising developers' smartness,
listening more to newcomers, making as clear as possible the stories
of those who managed to make substantial contributions, keep lowering
the floor, etc.

> My claim, then, is that Sugar is not sustainable on a professional basis
> until there are organizations with revenues from Sugar totaling at least
> ~$500,000 a year.  After immediate business expenses, this might be enough
> to hire one or two full-time developers.  So if you want to start paying
> developers, I recommend that you also start working on schemes to acquire
> Very Large Amounts of money.

Or start finding ways to get good developers for Not So Large Amounts
of money ;)

Regards,

Tomeu