[prev in list] [next in list] [prev in thread] [next in thread] 

List:       koffice-devel
Subject:    FW: [office-formula] OpenDocument Fellowship will fund formula
From:       "Bastian, Waldo" <waldo.bastian () intel ! com>
Date:       2006-08-24 22:40:31
Message-ID: 1B47D24854C7BC4FA8DA28BEBB59B0BA5EA647 () orsmsx419 ! amr ! corp ! intel ! com
[Download RAW message or body]

People working on Kspread might find this an interesting opportunity to
earn some money. Please contact David Wheeler if you are interested.

Cheers,
Waldo

-----Original Message-----
From: David A. Wheeler [mailto:dwheeler@dwheeler.com] 
Sent: Thursday, August 24, 2006 3:21 PM
To: office-formula@lists.oasis-open.org
Subject: [office-formula] OpenDocument Fellowship will fund formula
standards work!

All:

I've been notified that the OpenDocument Fellowship
is providing funding to aid development of the ODF formula standard,
thanks to a directed grant from an anonymous donor.

This is great news; it validates that this is important work
in many people's eyes, because people don't fund work they view as
unimportant.  It also validates that we're doing good work, because you
don't contribute money to active groups doing bad work!

My understanding is that the donation is A$10,000.
There will an extremely tiny auditing overhead,
and currency exchange rates over time are very hard to predict
(so being a little conservative is wise), but I understand that this
will translate into about US$7,100 / EUR5.500 or so.

I propose that to properly use this money, we create a list of "tasks
left to do"
as of yesterday (August 23, 2006) - which should be easy, just look at
the
marked TODO/TBDs and the undefined functions.
We can then assign a money figure (a "bounty") for how much completing
each function/task is worth. Since there are several functions left to
define,
we could define all other tasks in terms of how many "functions" of
effort it is.
To "complete" a task (for money purposes), some other trusted person
must AGREE
that it was completed (yes, even if I do it!), and that would act a
quality check.
I think we should fund the reviewers also,  to encourage careful review;
perhaps we give the original author 90%, and the reviewer 10% of the
money
(I expect reviewers to fix any problems they find themselves, if there
aren't too many problems, since this would greatly speed the process).
If we pay a "bounty" of US$25 or so for each function definition,
I think this would greatly accelerate the process without sacrificing
quality.
The bit operations and the byte-position functions are probably too
easy; I'd treat the bitops functions as 1 function,
and the byte-position functions as one function.
And at least for functions, what is required is already defined in
detail - see:
http://wiki.oasis-open.org/office/Formula_Work

I think it'd be great if we consider paying "honorariums" to a few
additional
distinguished reviewers as well.  That's in addition to the
extraordinary
expertise _already_ in this group.  In general, I want to use this
money to speed development and increase the quality even higher
if that's possible (though our standards are already very high).
This is in addition to the widespread review, and use of generated
test suites, that are already in place to produce high-quality results.
If the honorariums won't be taken, we can use the money to increase
the per-function bounty.

I am NOT interested in the "lowest bidder" - quality CONTINUES to
matter!
I think the people who have proven themselves already should
have the first crack at this.  If others are interested in working with
us,
please let us know!  With this funding, we will be able to pay a few
others with the necessary expertise, and that will really help
us complete the process even more rapidly, yet with high-quality
results.
If anyone on the formula SC can recommend anyone, please let me know.
Of course, any such person needs to prove themselves with their work.
We want to continue to have very high quality results.

Some people may not want to receive money, or may not be able to because
they're already paid to do this work.  For example, many companies are
already
involved in this effort, and I suspect most have rules against
double-payment.
But we also have a number of people who have great expertise, and have
contributed their considerable knowledge, but are limited in what they
can provide without any compensation. While this is certainly not a
life-changing amount, this small compensation should make it easier
for these people to contribute more of their time.

This seems (to me) to be something rather new in the world. We are now
starting to develop a new economic model in standards development, where
outside experts (who often go unpaid for their time) and traditional
companies
(who ARE paid) can work together in a much more focused way.
This gives us a FAR broader and MUCH more inclusive
approach to standards development, certainly in comparison to situations
where one commercial company pays the bills and unsurprisingly controls
everything
as a result. We now have commercial companies and not-for-profits
working
side-by-side, working together to produce a standard for spreadsheets we
can ALL use.
Users come from a variety of views, and don't want to be controlled by
any
one supplier -- they deserve standards whose very process ensures that
many views
are represented.  We already have people from a variety of countries
(such as the US, UK, Canada, Germany, and China), and from a variety of
business models (e.g., both open source and proprietary projects).  The
only
way to produce the best standard is to have this extraordinary
inclusiveness
of expertise, and I'm delighted that it's happening in this case.

Any ideas/suggestions will be greatly appreciated.
It's a pain and expensive to pay in little pieces, so I think we need to
tally up,
and then pay in a lump sum.  I'm sure other mechanism issues will show
up too.
We want this to be a SIMPLE process, not a complicated one, so that we
can concentrate on producing a high-quality standard.

Note that while this funding is appreciated, NOTHING HAS CHANGED in
terms
of standards rules.  This is still an OASIS project, with OASIS rules.
In particular, all work must be (non-exclusively) contributed to OASIS
so that the
results can be distributed freely on the web and converted into an ISO
standard,
per the OASIS IPR rules.  We are still a subcommittee, charged with
creating
a draft to submit to the OpenDocument TC.

Companies often fund development of standards,
so in that sense this is nothing new.  What's new is that we can
now fund those with expertise, even when they are not working directly
for
a company who is interested in supporting this work.  Companies try to
do
this sometimes via consulting arrangements, but those often fall short.
This additional funding will let us be far more inclusive than anyone
else,
and that's really exciting.

To handle these donations, the OpenDocument Fellowship has formed
a "Project Committee" to oversee money distribution - its job will be
to make sure that money is distributed only when services are actually
rendered.  I've agreed to serve on that, to help get it going. More info
here:
 http://opendocumentfellowship.org/development/projects
This is not the only project being funded; we have around US$7,100,
and two other projects total US $30,000.

The legal name of the organization actually handling the money is
"Friends
of OpenDocument Inc.", an Australian not-for-profit organisation that
handles
money for unincorporated groups (including those with members around the
world) that are working in areas related to its mission.
Details about them are at "http://friendsofopendocument.org/".

I mention all these legal mechanisms for those who are curious about how
the heck the non-profits and for-profits are managing to work together.
It's quite amazing how commercial companies, technical experts,
philanthropists,
and those with legal and non-profit-establishment knowledge are working
together to bring open standards to everyone. I wouldn't be surprised
if this becomes a model for future open standards work.

--- David A. Wheeler
_______________________________________________
koffice-devel mailing list
koffice-devel@kde.org
https://mail.kde.org/mailman/listinfo/koffice-devel
[prev in list] [next in list] [prev in thread] [next in thread]