EURIM Working Group Minutes
|
Working Party: |
Theme 01/ Working Group |
|
|
Ref: |
02-T02-Min08 |
|
|
Minuter: |
Emma Fryer |
|
|
Date: |
06/12/02 |
|
|
Circulation: |
Attendees and Apologies |
|
|
Queries to: |
Emma Fryer, Tel: 0191 384 0282 Mob: 07714 803 650 |
|
|
|
Minutes of the EURIM
Scoping Meeting on Open Standards and Open Source for e-Government 6th December
2002, kindly hosted by IBM |
Summary |
Meeting outline and
main points of agreement
|
|
The
objective of the meeting was to establish how EURIM could best help the
public sector evaluate Open Source and identify when it was appropriate and
cost effective to use it and when not. The
best way of achieving this was to help clarify the confusion and
misunderstanding surrounding this area, by providing definitions, information
and examples and spelling out the main issues. It
was agreed that it was not appropriate for the group to make any judgement. |
|
Actions Arising |
|
The group
agreed to prepare:
|
|
|
|
Papers Circulated
in advance of the meeting |
|
|
|
|
Tabled
Papers |
|
1. Total Cost
of Ownership for Linux in the Enterprise – Robert Frances Group, July
2002-12-18 2. Windows
versus Linux in enterprise Computing and assessment of business value in
selected workloads, IDC |
|
|
|
References |
|
1. ISO Website:
http://www.iso.org/iso/en/ISOOnline.frontpage 2. See tabled
paper 3. See tabled
paper 4. IETC –
Internet Engineering Task Force http://www.ietf.cnri.reston.va.us/ 5. ITSO –
Integrated Transport Smartcard Organisation: http://www.itso.org.uk/ |
|
|
|
Next Meeting
Date |
|
TBC |
Meeting Notes
|
|
|
|
Action |
|
1 |
|
Chairman’s introduction
|
|
|
|
1.1
|
The objectives of this exercise were defined:- to support the current
work of the OeE and OGC to help the public sector evaluate open source just
as they evaluate conventional software. Positions had become polarised and
entrenched and EURIM could best contribute by facilitating balanced debate
around a broad agenda and ensuring those that had to make policy or procurement
decisions were fully informed so that they could evaluate the technology
effectively within the marketplace, and make best use of it. There was no
intention to make any judgement about the benefits or costs of using open
source as opposed to proprietary software, but rather to help the public
sector make informed decisions.
|
|
|
|
|
|
|
|
2 |
|
Definitions |
|
|
|
2.1 |
It
was agreed that there was a difference between open standards and industry
standards. |
|
|
|
2.2 |
Although a definition of
interoperability was given in the e-GIF, there was no similar explicit
definition of Open Standards. The relevant open standards were implicit in
e-GIF but they were hard to articulate. It was easier, and more useful, to
define what they were trying to achieve through the standards rather than the
standards themselves. |
|
|
|
2.3 |
OGC
had reasonable working definitions of open source but the e-GIF definition of
Open Source was borrowed. |
|
|
|
2.4 |
It
was agreed that it would be helpful to differentiate standards from
specifications. The EU Procurement Directive in particular had failed to
clearly outline what was meant by standards as opposed to specifications. |
|
|
|
2.5 |
Standards
could be viewed as a top level, into which specifications, codes of conduct,
codes of practice, etc. fitted as subsets. |
|
|
|
2.6 |
It
was agreed that there was also an important distinction between patented
standards that required royalty payments and non-patented standards that did
not. |
|
|
|
2.7 |
It
was suggested that the ISO website would be a good place to look for definitions. |
Ref 1 |
|
|
2.8 |
It
was agreed that standards took several years to fix and so setting new
standards was not a useful objective. |
|
|
|
2.9 |
It
was agreed that the production of a short guide with simple definitions and
examples would be one way of adding value to the work already done on
defining open standards. This should differentiate specifications and
standards and be targeted at parliamentarians and civil servants for whom it
was agreed it would be particularly useful.
DK agreed to produce some definitions as a starting point. |
DK |
|
|
2.10 |
It
was agreed that a glossary of terms for non-experts would be very useful to
limit the scope for endless debates on semantics. AH agreed to propose to the NCC that they include a glossary on
the website of their Self Assessment Service which was just about to be
launched. |
AH |
|
|
|
|
|
|
3 |
|
Purpose of meeting and scope of activity |
|
|
|
3.1 |
It was agreed that this
meeting would be used to outline an initial work plan and allocate
responsibilities. |
|
|
|
3.2 |
The chairman tabled two documents,
which represented the polarised views and strong opinions held on the subject
of open source. One demonstrated that
Windows was cheaper than open source and the other showed that open source
was cheaper than Windows. The purpose
of the meeting was not to validate either of those positions but to try and
ensure that people were better informed so that they were in a position to
evaluate the assertions made from either perspective and make sense of
conflicting information. |
Ref 2 Ref 3 |
|
|
3.3 |
It
was agreed that EURIM’s remit was not to make commercial judgements but help
establish a framework for best value decision-making. |
|
|
|
3.4 |
It was agreed that it
was often of value to identify why people were disagreeing. Often those taking or recommending
decisions did not have sufficient understanding to discern that people were
talking at cross-purposes, for instance.
|
|
|
|
3.5 |
It was necessary to
look at the differences between open standards, interoperability standards
and those relating to open source.
The real issues for exchange of information over e-government websites
boiled down to what open source could and could not do and whether it could
be a means of delivering services.
Open standards and interoperability tended to be talked about and
strategised rather than implemented on the ground for meaningful benefit.
|
|
|
|
3.6 |
The
arguments over open source vs. proprietary systems had become political and
emotional and people no longer knew whom to believe. A credible matrix would
be far more useful than the arguments of the different camps. |
|
|
|
|
|
|
|
4 |
|
Graham Bird – the Open Group |
|
|
|
4.1 |
GB
gave a quick overview of the Open Group, which provided certification and
testing services and operated a membership forum for suppliers and consumers
of IT to solve their problems, particularly relating to
interoperability. |
|
|
|
4.2 |
It
was important to start by defining the problem, which was essentially the
ability to exchange information between the right people in the right format
at the right time, securely. Now that
organisations were becoming boundariless and multi-disciplinary teams were
talking the same language, systems were unable to support them because the
information was still held in stovepipes.
|
|
|
|
4.3 |
The
Open Group worked with suppliers, consortia and standards bodies to tackle
the issues of getting information to flow, structuring IT so that it did not
operate in stovepipes, and getting rid of boundaries. |
|
|
|
4.4 |
Boundariless information flow needed: ·
Formal standards ·
Best practice ·
Open source ·
Confidence It was not a question of whether the
technology was viable – excellent products existed - but whether businesses
had the confidence to put mission critical systems on open source software
without knowing how it would develop in future. |
JT |
|
|
4.5 |
The systems
that were going to deliver solutions needed measuring and certifying. Confidence in the IT industry’s ability to
deliver was very poor.
|
|
|
|
|
|
|
|
5 |
|
Feedback and further comments |
|
|
|
5.1 |
There were two
major factors influencing people’s attitude to proprietary software. On the one hand people like
parliamentarians did not want to lose the interoperability they already had
between systems in their different places of work so they were hesitant to
risk new solutions over proprietary solutions with which they were
familiar. On the other hand there was
concern over strategic dependency. It
was imprudent for government to be reliant on one supplier because of
security perceptions both in terms of vulnerability to failure and exposure
to controlled breaches.
|
|
|
|
5.2 |
Some large organisations were using open
source for major business functions but for commercial reasons were unwilling
to disclose what they were doing and why.
|
|
|
|
5.3 |
There was usually too much focus on
the pros and cons of purchasing open source:- the debate needed to be
broader. It was important to improve understanding of the development model.
Many big IT project failures were related to large proprietary systems and
open source had developed in reaction to that. Better understanding would
enable people, particularly those in the public sector, considering open
source to learn why it was there and how they might apply the open source
development model to develop their own software. The market could then
develop naturally. |
|
|
|
5.4 |
The biggest problem perceived by
government was the threat of lock-in.
Interoperability with other systems was vital. Proprietary manufacturers did not
necessarily interoperate with each other, and if something went wrong the
file formats were neither available nor transparent – interfaces remained the
property of the vendor and were not always published. |
|
|
|
5.5 |
It was feasible for vendors to comply
with specifications but still produce products that did not always
interoperate. Mechanisms and
processes must ensure interoperability because customers of IT had now
developed an expectation that systems could work together - whether tube
passes or mobile phones- and supplier lock-in was no longer acceptable.
|
|
|
|
5.6 |
Government
standards were owned by the Public Record Office in Kew or available through
the e-Envoy’s channels or OGC. Different government departments had different
standards and collating them in one place was unfeasible. |
|
|
|
5.7 |
One
member observed that Microsoft’s new charging structures had created
hostility in the industry and caused people to consider alternative solutions
which was one reason why the arguments had become polarised, political and
emotional. |
|
|
|
|
|
|
|
6 |
|
OGC perspective – How can EURIM help? |
|
|
|
6.1 |
The Chairman
referred back to the agenda and asked OGC to outline the problems they had
and the kind of help they needed, so that they could assess how EURIM might
best contribute. EURIM could offer
networking, event hosting, fora and briefings.
|
|
|
|
6.2 |
OCG was running a government stakeholder group to
develop best practice guidance. They
confirmed that the report commissioned from Qinetiq to help formulate
government policy on open source had been very useful as an independent
assessment of the open source arena. |
|
|
|
6.3 |
OGC reported that keeping pace with open
standards as they emerged was particularly problematic. They therefore
planned to produce and adopt a mandate on open standards only when there was
a market to support them. Currently
people were apprehensive about making their organisations dependent on
products with no track record or underwritten guarantees.
|
|
|
|
6.4 |
It would be
helpful if open source and open standards were handled as separate issues.
This was agreed.
|
|
|
|
6.5 |
Things
would not change overnight despite the 2005 targets, which were effectively
aspirational. The target was to get from the present situation to transparent
transactions between citizen and government. In between a huge raft of legacy
systems had to be overcome. It was
not unfeasible to achieve success in government systems:- the barrier was
cost. The roadmap was in place but Treasury was not going to fund mass
migration overnight. |
|
|
|
6.6 |
The
chairman asked whether this meant that with e-GIF in place the best
contribution EURIM could make would be to raise awareness of this fact. |
|
|
|
6.7 |
The
chairman summarised – the problem was understood, the objectives were clear
and the means of achieving those objectives were understood, although not
resourced. |
|
|
|
|
|
|
|
7 |
|
Open Standards and e-GIF |
|
|
|
7.1 |
OGC
was leading work to define the e-GIF standards for e-Government and had
produced a paper on procurement.
Dealing with the public sector took time, there were historical
practices, legacy systems, legislation and cultural barriers to overcome. |
|
|
|
7.2 |
The
problem of ownership was a major one – who would be the guardian of an open
standard? Who could an organisations turn to for explanation or to effect
changes? Proper standards required
maintenance - from trivial elements like typographical errors to amendments
and new editions. Standards evolved and had to be maintained. No-one could risk a business on standards
that were not owned by anyone. It was vital to confirm absolutely whether
something was truly open source or open standard and provided pro bono
publico before making it available. |
|
|
|
7.3 |
By
definition an open standard had to be consensus-based and there was general
acceptance that the market would work towards that to achieve
interoperability. This was different
from one organisation or government setting a standard that suppliers had to
work to. This dichotomy led to
confusion:- some open standards conformed to e-GIF but others did not. Was it feasible to build on the work of W3DC or IETF in order to get government to
define what they required in terms of interoperability? |
|
|
|
7.4 |
IETF
had been working on standards for exchange of information in the public
sector. A meeting of the Committee on
International e-Government Standards was scheduled for the following week. The Committee’s task was to ensure the
right standards were achieved for international e-government by getting
standards bodies to collaborate and ensuring that existing standards were
modified to suit governments. This was replacing the IETF. |
Ref 4 |
|
|
|
|
|
|
8 |
|
Case studies
|
|
|
|
8.1 |
Real
examples were the best way of illustrating the need for programme
management - changes in behaviour,
practice and interest. They could
demonstrate in ordinary terms the organisational steps required and improve
information exchange between policy makers and implementers. The organisational barriers were usually
more problematic than the technical ones. |
|
|
|
8.2 |
The contract between TFL (Transport for
London) and TranSys would make a good case study, to illustrate the problem
of not having interoperable open standards.
The card was not ITSO (Integrated Transport Smartcard Organisation)
compliant so either a new standard would have to be written or bus operators
would have to adopt TranSys’s standards. It raised questions of system
ownership and setting boundaries and was a good working example to illustrate
practical problems.
|
Ref 5 |
|
|
8.3 |
Two
further case studies of successful interoperability projects delivered to
tight deadlines were suggested:- the Atlanta Olympics and the Manchester
Commonwealth Games. IBM and Microsoft agreed to provide case study
material. |
ML, PB |
|
|
8.4 |
Although
these were to an extent greenfield projects and so some of the problems
associated with large organic systems that had evolved over time could be
avoided, they had incorporated old systems, which they had ring-fenced,
standardised and then interfaced. |
|
|
|
8.5 |
The
e-procurement targets were another case, where a catalogue system built to
e-GIF standards was causing concern about the exclusion of SMEs and local
tradesmen. This was being addressed
by the production of “SME packages”.
Once everyone had to conform to the same standards the playing field
would be level. |
|
|
|
|
|
|
|
9 |
|
What form will EURIM’s
contribution take?
|
|
|
|
9.1 |
The
chairman observed that the contents of a useful briefing paper were emerging
from the discussion. It could
outline what government was intending to do, identify the structures and
organisations in place to achieve that, provide a series of short case
studies that illustrated the things that could prevent it getting there and
raising issues like programme management or finance. Such a briefing would serve to reassure
people that the policy was good and the programme was essentially in place,
but also identify issues that needed to be resolved. |
|
|
|
9.2 |
It
was agreed that the programme would benefit from more publicity and more
political priority. There was also an opportunity to publicise the next
version of e-GIF (VI) which was scheduled for April 2003. |
|
|
|
9.3 |
Specifically,
OGC needed assistance in publicising policy to the supplier community to
involve them in the debate to find the right solutions. It was therefore agreed that a succinct
document that publicised the information already available would be very
useful. |
|
|
|
9.4 |
It
was agreed that EURIM could also be very useful in helping OGC raise the
profile of what they were doing and inform understanding in the public
sector, particularly explaining it to other government departments. |
|
|
|
9.5 |
The
chairman proposed that the group consider what it could constructively do to
create publicity for the OeE and OGC’ activity, formatted in a way most
effective for the target audience – in paper form for MPs and in electronic
format for civil servants and industry. |
|
|
|
9.6 |
The
content would focus on what OeE and OGC were doing to achieve the required
interoperability objectives for the citizen in his transactions with
government. |
|
|
|
9.7 |
It
was agreed that to attract political attention papers needed to include a
call for action. |
|
|
|
|
|
|
|
10 |
|
Target audience
The
target audience would include parliamentarians, civil servants and suppliers.
|
|
|
|
10.1 |
It
was agreed that EURIM could provide links to suppliers through the membership
and through organisations such as Intellect who were also members. |
|
|
|
10.2 |
NCC
was due to launch its e-GIF Advisory Service in January. This would supply an assessment service so
that organisations could see how well they complied with the standards. This was an important part of implementing
e-GIF and needed support. EF agreed
to approach Intellect’s e-Gov group to see if this was a suitable topic for a
meeting. |
EF |
|
|
10.3 |
It
was agreed that the priority was to encourage people to buy into the
processes around e-GIF to remove fragmentation and share knowledge. |
|
|
|
|
|
|
|
11 |
|
Content and scope of EURIM’s
Message
|
|
|
|
11.1 |
It
was agreed that open source and open standards should be dealt with
separately, since they involved different issues. Open source should also be differentiated from open systems
and interoperability. |
|
|
|
11.2 |
It
was agreed that a simple glossary that differentiated open source and open
standards should be included in the material they were planning. AH agreed to start this process |
AH |
|
|
11.3 |
The work had to help people
understand what they were choosing between and how to set up evaluations. It was therefore agreed that, rather than
concentrate on existing products it was better to focus on how the public
sector could make use of the open source development process. There was very little understanding of how
open source software was developed, and the pros and const of adopting that
development model, particularly for large bespoke public sector systems which
had such bad track records both in terms of failure and cost. These failures
had been the catalyst for the development of open source that enabled
interoperable multi model systems. |
|
|
|
11.4 |
It
was agreed that the group would prepare a paper on open source, explaining
how it was put together, but not making any value judgements. |
|
|
|
11.5 |
It
was agreed that open standards were poorly understood and clarification was
essential. There was general confusion over the roles of standards bodies,
their respective value and how they contributed to interoperability. They should seek to clarify these areas. |
|
|
|
|
|
|
|
12 |
|
Next Steps - Allocating actions
and areas of responsibility
|
|
|
|
12.1 |
The
chairman observed that EURIM had a unique position at the interface between
industry and government and proposed a three phase forward plan: |
|
|
|
|
·
Firstly a simple status report giving an overview of the
current state of play and helping to publicise the good work already being
done. ·
Secondly a briefing on open standards issues to help
improve understanding. It should
incorporate recommendations to government – (some of which would already have
been implemented) and outline the things that people responsible for policy
decisions should consider when setting frameworks ·
Thirdly a briefing or paper on open source which included
an explanation of the different business models involved (when contracting
for proprietary solutions support might well be included, but when using open
source, support might well have to be contracted for separately) |
|
|
|
12.2 |
At each stage an initial draft would
be prepared from comments and inputs and circulated. The material would be revised in the light
of further comment and feedback.
Refinement and editing would continue by email and a further meeting
would be called to address the issues that could not be resolved, before any
material was circulated externally. |
|
|
|
12.3 |
The
chairman proposed that the group be broken down into sub-groups to tackle the
areas of activity that had been identified. |
|
|
|
12.4 |
Volunteers
agreed to help as follows:
|
|
Attendance – 6th
December 2002
|
Fname |
Surname |
Organisation |
|
Susan |
Bide |
DTI |
|
Graham |
Bird |
The Open Group |
|
John |
Borras |
Office of the e-Envoy |
|
Paul |
Butler |
IBM UK Ltd |
|
Earl of |
Erroll |
House of Lords |
|
Emma |
Fryer |
EURIM |
|
Chris |
Godwin |
IBM |
|
Nigel |
Greenaway |
Fujitsu Services |
|
Caroline |
Halcrow |
Office of the e-Envoy |
|
Andrew |
Hopkirk |
National Computing Centre |
|
David |
Keech |
BSI / DISC |
|
Matt |
Lambert |
Microsoft |
|
Roland |
Merrick |
IBM |
|
Samantha |
Mitchell |
BBA |
|
Simon |
Moores |
Zentelligence |
|
Mick |
Morgan |
OGC |
|
Lord |
Renwick |
Former Parliamentarian |
|
David |
Rose |
The Open Group |
|
Richard |
Taylor |
Qinetiq |
|
Roger |
Till |
e-centre UK |
|
Philip |
Virgo |
EURIM |
|
Steve |
Winslade |
Chartered Institute of Arbitrators |
Apologies
|
Helen |
Dickinson |
POST |
|
Andrew |
Hardie |
|
|
Malcolm |
Herbert |
Red Hat |
|
Garry |
Harney |
GM Fire Service |
|
Martin |
Leverington |
OGC |
|
Malcolm |
Lowe |
Accenture |
|
Arlene |
McCarthy |
MEP |
|
Tom |
McGuffog |
SITPRO |
|
Richard |
Nethercott |
CMG |
|
David |
Rippon |
BCS |
|
Walter |
Smith |
DEFRA |
|
Ian |
Taylor |
MP |
|
Dorota |
Warren |
Individual Observer |
|
Gregg |
Wilkinson |
Fujisu Services |
|
David |
Wright |
EURIM Staff |