[education-wg] Karen Cator

[Tom Hoffman]

On Thu, Apr 8, 2010 at 1:51 PM, Mike <mhuffman at comcast.net> wrote:
> “Karen Cator is the Director of the Office of Educational Technology
>
> for the US Department of Education.  She is very interested in the
> idea of open source, but she has questions.  "How do we invest in open
> source?"  "How do we know if open source is effective?"  "How is the
> model sustainable?"
>
> How do we figure out what Karen's goals are, and position open source
> as the best tool to help her achieve those goals?  If I should get
> another meeting with Karen, what should I say when I get into her
> office besides "open source is awesome"?”

"Pay for it once" is an argument you hear from people in other
countries that have a national open source strategy.  Instead of every
school or district paying licensing costs indefinitely, the federal
government, states, or consortia thereof pay to have an open source
solution created which can subsequently deployed nationwide
(worldwide!) at no additional cost.

Paid support will come from vendors.  All you have to do is put out an
RFP for paid support of an open source app.  Someone will decide
they'd like to take your money.  Or you can just pay your staff to be
trained to provide support.

You'll probably pay more for the initial implementation, because
you'll be more concerned about making it flexible, scalable,
standardized, maintainable, installable, etc. than if you were just
writing something for one school ("The Mythical Man-Month" tells you
it'll take 9x as long), but you save money in the longer run.

This is something we saw with CanDo -- the initial competency tracking
implementation at the Arlington Career Center probably was more
expensive than if they'd just hacked together something for their
local use, but by building on SchoolTool and pushing for high coding
standards they ended up with something that is now being used in five
VA counties with little additional expense to the state.

I wrote up some other observations on that process for a Gov 2.0
conference proposal (which didn't make the cut).  Obviously, there are
other ways to do it too, but this is a "bottom up" approach which
worked for us:

Title: Lessons from SchoolTool: Leveraging Local Innovation into
Global Collaboration

Challengingly short description: For almost a decade, from Capetown to
Arlington, Virginia, the SchoolTool project has explored the
intersection of K-12 school administration, open source software
development and philanthropy. We've had a few misses, but have hit
upon some successful prototypes of local/global collaboration,
including CanDo, an application developed by teachers and students at
the Arlington Career Center.

Abstract:

Schools can and have innovated to meet their administrative data
requirements. What they cannot do on their own is turn these
innovations into national or global scale open source projects. They
need help. Here's the formula that emerged from creating CanDo:

    * The core project must be initiated and funded by schools for
their own needs. They must be capable of executing the project on
their own.
    * The role of philanthropy or higher levels of government is to
provide additional resources to support elevating the project from a
local project to a full open source development effort.
    * The project must be built on a 100% open source software stack.
Ideally a platform with a well-established and professional
development community, accessible to the school's developers.
    * Developers must be trained in (if necessary) and held to
standard open source development practices from the beginning of the
project, including maintaining a public code repository and bug
tracker.
    * Be as Agile as possible.
    * Extra funding must be available for development sprints and
other opportunities to directly connect local developers with the
broader development community, in part to ensure that the project is
designed in line with the standards and practices of the community.
    * Packaging, release management, documentation, and code
maintenance are all expensive. Schools can't do it on their own.
    * Ongoing management of open source development is too expensive
for schools, and projects owned by businesses in the K-12 market have
not fostered community involvement. You need a non-profit neutral
maintainer with a separate funding stream.

In short, bootstrapping an open source project is not "free," but the
potential leverage to national or global makes it a valuable
investment for philanthropy or government.

--Tom