The foreword of the document tells us that Tom Watson, Minister for
Digital Engagement, has his head screwed on straight. In the very first
paragraph is a clear recognition and statement of the fact that Free
Software has brought enormous benefits and cost reductions, simply
through increased cooperation and transparency. The document is making
clear that the U.K. Government wants "in" on these kinds of benefits.
Also clear from the foreword and the rest of the document is an
implicit
understanding that the previous U.K policy on Open Source (last
formulated in 2004) may not have had the teeth required to take on any
large corporations pathologically-hell-bent on guaranteeing maximisation
of profits (at tax-payers' expense).
The new policy however spells out that the Government will not
only be
writing into new procurement contracts that it will expect to receive
the full rights to the software being developed, but also, stunningly,
it will actively seek the full rights to existing software under
existing contracts.
It's worth noting that the policy is not to immediately
release
the software so acquired under Free Software Licences, but instead to
evaluate "General Purpose" software acquired, and to release it as Free
Software "if appropriate".
In other words, the U.K. Government is endeavouring to
learn from
the success of Open Source Community development, and foster a
Department-wide culture akin to that of "Open Source". Civil servants
are not themselves allowed by law to compete with the Public
Sector. Given that they will be unable to actually perform any
development themselves of the software acquired, in combination with the
inevitable influx of Free Software Licensed source code as part of the
obvious forward-accumulation of cost savings, this policy will slowly
transform into one where the release of "acquired" software under Free
Software Licenses becomes the norm rather than the exception.
Even where the software is utilised for Military purposes, it's
still
possible to release software under a Free Software License, as it's more
often than not the case that the data is what is Classified. With a
little creative thinking and the removal of a few words, it's reasonable
to assume that much of military-developed software could be used for
business or educational purposes (Military-developed software isn't as
sexy as people think: much of it is simply organising or saving on
paperwork and bureaucracy, as Military organisations face many of the
same organisational issues as any standard business faces. They just
have a bit more money to spend. usually).
It's worth
noting, however, that due to the legal requirement that people working
under the Official Secrets Act are not entitled to "own" their work, as
they are acting wholely "as an extension" of their government, even
modified GPL-licensed Free Software does not "escape" from being
"Classified". In a straight fight between GPL and "Military
Classification", the GPL loses outright. The reason is simple: with
absolutely no rights allowed by law for individuals on a project, it's the
Government that is the sole recipient, the sole distributor and the sole
developer. Under such circumstances, the GPL is clear: there is no need
to release source code, because there is no other party involved
to whom the software has been distributed. About the only Free Software
license that is potentially problematic and would conflict with a
Military Classification is the Affero GPL which
requires that the software be released, regardless of its
distribution and deployment range.
The Way Forward
The following amazing committments have been made to us, UK
Citizens and
Free Software developers. The U.K Government will:
-
ensure that the Government adopts open standards and uses these to
communicate with the citizens and businesses that have adopted open source
solutions
-
ensure that open source solutions are considered properly and, where they
deliver best value for money (taking into account other advantages, such as
re-use and flexibility) are selected for Government business solutions.
strengthen the skills, experience and capabilities within Government and
in its
suppliers to use open source to greatest advantage.
-
embed an open source culture of sharing, re-use and collaborative
development across Government and its suppliers, building on the re-use
policies and processes already agreed within the CIO Council, and in
doing so
seek to stimulate innovation, reduce cost and risk, and improve speed to
market.
-
ensure that there are no procedural barriers to the adoption of open source
products within government, paying particular regard to the different
business
models and supply chain relationships involved.
-
ensure that systems integrators and proprietary software suppliers
demonstrate the same flexibility and ability to re-use their solutions and
products as is inherent in open source.
In other words, for example, there will be no more OOXML, no more
Microsoft Word or other formats.
This is particularly important for a Government to do, as demonstrated
by one U.K. Government-developed web site asking users to download a
.doc and fill
it in and email it back. I sent a stinking letter to them explaining
that it would be necessary to purchase Microsoft Office, at a retail
cost of £500, and then to throw away my existing machines and purchase
replacements capable of running Microsoft Office - a further
£500
- and was it the U.K. Government's policy to proactively support
business cartels in this way, oh, and what about blind people? Having
had a bit of a laugh, and thinking nothing further of what I wrote, I
was stunned to hear back eight months later saying "we have updated our
web site after receiving various feedback, please could you review it".
The U.K. government's new policy ("Action 8") therefore
proactively makes such issues a thing of the past, on a
government-wide scale, not just a small Department's Web Site, but right
up to and including multi-billion-dollar NHS contracts spanning decades
of development.
This section is also worth noting:
The Government will, wherever possible, avoid becoming locked in to
proprietary software. In particular it will take exit, rebid and
rebuild costs into
account in procurement decisions and will require those proposing
proprietary
software to specify how exit would be achieved.
There's a phrase you often see in novels, where the characters are
described as doing a "double-take" or they "can't believe their eyes",
or "blink in surprise". Now I know what it actually feels like
to "do a double-take".
This kind of wording makes it clear that there will be no
"wool-eyes-pulling". "Sweetener-deals" such as the illegal provision of
software at reduced cost with the intention of profitting at a later
date will be out of the question (as was done to the London
Metropolitan Police: the Free Software deployment of Desktop machines
across thousands of workstations using OpenOffice, with which the users
were quite happy, were replaced with Microsoft Systems, based on
"guarantees of cheapness of price". Half-way through the deployment,
after the Open Source Server Infrastructure had irrevocably been
replaced by Microsoft servers, a blackmail call was placed with demands
for further money to complete the installation, claiming that the
much-reduced quote was only for the work done up until that point).
Later on, section 6.8 says that the U.K. Government will look to
acquire
the full rights to all existing proprietary software, so that it can
have the same benefits afforded to it that it has with Free Software,
but not only that, it says that where appropriate, "General Purpose"
software that it acquires the rights to will then be released under Free
Software licenses.
Without such pro-active decision-making on the part of the U.K.
Government, it is clear that the U.K. Government's arm could be twisted
so that I.T. suppliers could state that they have complied "to the
letter" of the I.T. policy, by "evaluating" Open Source alternatives
(Action 5: "Supplier Challenge, page 8) but concluding - always - that
Open Source is inadequate.
By having a clear and unequivocable policy that the software
must
be procured and/or developed under an OGC (Open Government Council) OJEU
clause (Action 9: "appropriate release of code", page 8) then the ball
begins to roll, because the U.K. government will slowly acquire the
rights to the software infrastructure used to govern the country, and
will inexorably be able to build and improve on that infrastructure
without being tied into a single monopoly supplier, directly
through a contract or indirectly through a proprietary standard for whom
there is only one supplier.
The most stunning committment of all is this, the last section
of the
document:
Action 10: Communication, Consultation and Review: Government will
communicate this policy and its associated actions widely and will
expand it as
necessary. It will engage with the Open Source community and actively
encourage
projects that might, in due course, develop into "Government Class"
products. It will
keep the policy and progress on the actions under review, and report on
progress
publicly.
Remember that double-take thing? The U.K. government will actively
encourage "Government Class products"? What can I say - er, Welcome to
the Club? We take American Express? nonono, sorry, I mean, Fantastic!
Great! Perhaps you should consider talking to Google, to learn from
how it sets up the "Google Summer of Code" Programmes, and set up a U.K.
Government Inititative "OCG SoC", with a view to sponsoring students and
bright school children to work on developing "Government Class" capable
products, with mentors from the Free Software community guiding their
progress. The success of GSoc should leave you in no doubt as to the
extent to which an OCG SoC programme will likewise be successful.
Also you should look at setting up a specific web site or web
sites - an
entire Sourceforge-like or SchoolForge
site, with clear and immediately-accessible goals distinctly laid out.
Extracting the text from google's cache because the site is down at the
time of writing, SchoolForge's site states:
SchoolForge's mission is to unify independent organizations that
advocate, use, and develop open resources for education. We advocate the
use of open texts and lessons, open curricula, free software and open
source in education.
In order for "Action 10" to be met, it is necessary to have a
site which can be the focus of the efforts of Free Software development,
accumulating the results and actively encouraging participation.
Code.Google.com on its own does
not quite fulfil these criteria, because:
- although it provides individual projects with the tools to create
and manage their source code, the site does not include discussion
forums, expecting users to go to groups.google.com or other resources.
- the per-project Wiki is more focussed on allowing developers to
"make a few notes" than it is about presenting an attractive and
informative [i.e. wikipedia-like] face to the rest of the world
- google gives its own projects that use code.google.com or where
google itself would benefit "high priority precedence", including better
URLs and deploying documentation. the remaining projects are treated as
second-class citizens.
Setting up a decent Free Software "Forge" that helps attract
potential
developers - and keeps them - is a tough call, and, as can be seen, even
Google shies away from providing the full suite of functionality that is
ultimately required. Sites like Sourceforge rely on advertising revenue
to sustain them, whilst Google very wisely decided to keep
code.google.com entirely ad-free and resource-focussed. Other than
sponsorship logos, placing advertisements on a government-initiative-led
web site would probably not be a good call: the wikipedia funding model
would probably be best.
Royal Chartered Society for Software Architecture
This incredible document clearly shows that the U.K. government has
the interests of its tax payers at heart, by clearly showing that
it wants to learn from the lessons of Free Software development
and Free Software Community cooperation and effectiveness.
There is a second lesson that can be learned from historical review
of a suitable formal body: the Royal Charted Society of Architects.
Architecture and the construction of buildings is an
historically-recognised vital service to citizens, recognised
as such by ensuring that there is a legal framework for the
safe construction and adaptation of buildings. Where bringing
in an Architect is not appropriate, "Building Regs" apply which
are enforced at a Local Council level. So there is a specific
framework for both small and largescale efforts.
What is the purpose of these efforts?
Well let's ignore the somewhat
insane aspects, such as requiring that even in a change of the
number of dwellings in a property (flats back to a house, or vice-versa)
the ground floor must be "wheelchair" capable, requiring internal
brick walls to be moved - no matter the cost - because the corridors of
older houses are often only 80cm or 90cm, and the requirements for
"wheelchair access" are 120cm - let's ignore these insane aspects for
now, and concentrate on the relevant ones and the overarching goals
instead of the bureaucratic political-correctness-gone-mad overlay that
so often gets placed on top of well-meaning and initially
well-thought-out ideas.
The purpose of the efforts of are to ensure:
- Safety.
- Aesthetics.
- Value for money and cost-effectiveness, short and long-term.
- Fitness for purpose, short and long term
For example, a new Public Building is constructed, or a new Olympic
Stadium, and its expected lifespan could be thirty to fifty years, with
renovations planned for and documented.
Not only is it blindingly obvious as to why there are
procedures, rules,
laws and regulations governing the construction of buildings, but it's
been blindingly obvious for such a long time. RIBA
gained its Royal Chartered status in 1837.
So the point is this: whilst Architecture in some form has been
a vital
part of humanity's development and progress for millenia, Software has
only become a
vital part of society in the past fifteen to twenty years. Yet the very
same principles of safety, fitness for purpose and value for money both
short and long term apply equally well to software as they do to
architecture.
In fact, with the fiasco of the NHS
twelve billion dollar budget predicted to reach a whopping £31
billion as just one example, let alone that of the unwelcome
"National ID Card Database Bandwagon" on which so many Civil Departments
have jumped, causing exponential spiralling costs as the software is
rewritten each time to incorporate the new department's requirements, it
should be clear as daylight that there's something drastically wrong.
Imagine, therefore, the role that a British Royal Chartered
Institute
for Software Architects could play, in mitigating the risks and the
spiralling costs. A BRISA certified member would be legally held
accountable for the completion of the project, or lose his Chartered Status.
Such responsibility immediately changes the whole ball-game.
Suddenly there is a person who is required by law to put their foot down
when it comes to irresponsible and ill-educated decision making. They
are responsible for ensuring that the procurement phases are
well-thought-out, well-structured, costed and planned. Without the
BRISA member's approval, then just like for a large Building Project,
the project cannot legally proceed.
With the BRISA member's assistance required by law, the Civil
Servants'
job is made infinitely easier.
It's often the case that Civil Servants are asked to plan I.T.
systems
which they are simply not equipped or well-enough-informed to ask for.
It's often the case that a "dummy contract" phase is first set up -
often by accident - where it is the Bidders who actually help the Civil
Servants to crystallise their ideas, knowing full well that the initial
bid can be dramatically underpriced, because they cannot "contradict"
the Civil Servants by pointing out the flaws in their initial
requirements (as to do so would lose them the opportunity to make money
on the "real" contract, and making money is the sole exclusive purpose to which
corporations are pathologically dedicated, through their Articles of
Incorporation). So, the Bidders "help" the Civil Servants to
"fulfil" the initial contract, where the actual bidding process itself
shows up the flaws in the initial plan, and helps educate the Civil
Servants.
In other words, Civil Servants are not knowledgeable of the
"plan to do
three versions" rule, because they are not Software Engineers. Their
first version, therefore, is drastically under-budgeted, and the Bidders
know damn well that that's the case. And they also know that when they
have won the first bid, the use of proprietary software and proprietary
standards gives them a 100% chance of winning the subsequent development
rounds, which will be dramatically higher than the original contract.
This alone is why the U.K's new policy on Open Source and Open
Standards is so absolutely vital.
Not only that, but as the "ID Card" database example shows, the very
fact that the Civil Servants ask, and a demo is produced, often results
in "me-too" behaviour from other Departments. This is natural: Software
development shows us that it is only when a product or a demo is placed
into people's hands that they actually understand its
implications and its uses.
The problem comes when the whole development process -
higgledepiggledy
or well-planned or otherwise - is kept "under wraps", with no
transparency or explanations as to why the costs have spiralled out of
control. A BRISA member would be responsible for ensuring that the
documentation be made publicly available via Town or County Hall Web
Sites, along with the actual source code where appropriate [again,
the "Forge" principle comes to mind].
Documentation outlining exactly why certain design
decisions were
made, cross-referencing them to the recordings of the conversations or
documents where the design requirements and functional specifications
were first laid out and subsequently adapted. Digitally signed by the
BRISA member as being a true and accurate record of the development of
the project, as per legal requirement, under Royal Charter.
Transparency.
Learning from the Architecture Profession's shining example.
Summary
The U.K.'s new policy, to both retrospectively and proactively
fit an
Open Source internal culture on top of its I.T. procurement and
development is just absolutely astounding. Concrete examples are cited
of public money already saved through small-to-medium-scale deployment
of Free Software, as an encouragement of the way forward, and a way to
ensure that proprietary software "advocates" are well-informed.
The U.K. government is stating that it can clearly see that Free
Software development principles of cooperation and transparency are
beneficial and save money, and the U.K. Government is stating that it
wants more of that, in no uncertain terms, when it comes to the
infrastructure that is now running the Country and on which its citizens
now so critically depend.
The U.K. Government is unequivocably stating that it wants to
own the
rights to the Software which runs the country (and for the software
standards used to be "open"); that it wants to engage with the Free
Software community to help develop that Software; and that it is
prepared to back that up by making sure, where appropriate, that the
software it acquires the rights to is then Free Software Licensed.
Effectively, the U.K. Government is putting its foot down and
stating
that it is unacceptable that its citizens pay to be taken advantage of
by corporations, over something as complex, misunderstood and so
absolutely vital as the Software that is used to run the Governance of
the Nation.
The future will perhaps bring much bolder steps in the right
direction,
with not just policy documents but also a legal framework in place for
the design, development and procurement of Software - perhaps even
underpinned and safe-guarded by Royal Charter.