RT @ITprofligate http://ping.fm/PtxAy election UK politics conservative labour here is part two
In case you can't see it on conjoint here it is.
David Cameron Follow On
Published on Saturday, April 10th 2010. Edited by Lab Zipzipace. tag
I have made my points about what would need to precede a serious change in the manner of public sector IT procurement.
The necessary first steps are repeated below. Form a powerful, focused and independent team who are able to model in economic terms desired goals and chart a path in legal terms. This team, and perhaps a few other teams that devolve from it, should have a remit such as this:-
1. dismantling of reliance on large suppliers.
2. scrutinising contractual obligations and finding points of non-fulfilment.
3. in conjunction with 2. wholesale renegotiation of contracts that cannot be terminated.
4. concentrate on de-scoping large projects into far smaller cooperating units.
5. create a component database and monitor supply costs against key elements.
6. emphasise the technology for cooperation and reporting.
Notice that this presupposes the benefit of 1.
This point, and others such as the major point about the role of Open Source need to be independently established.
In what follows I take them as a given as the ideas I advocate would only be able to be built on from such a consensus.
I am going to avoid an analysis in minutia, but there are some useful observations that can be made here.
Varney has failed
Supply into the public sector that has aimed at replacing back office activities with on-line capability has been developed as a series of monolithic islands.
This is the basic thrust of the Varney report which recommendations the present government has tried, in its last round of IT procurement, to act on.
Not only has this failed, the situation has actually been made worse by further embedding managerial practices and software encoded procedures (or you might say lack of) into the broad system.
It is true, as previously discussed, that some large percentage of promised savings will be made, but the whole system remains at risk due to its extraordinary fragility, major possible points of failure include:-
i. not being able to communicate data cross department. ii. inflexibility with respect to new or evolved requirements. iii. its reliance on existing suppliers.
It is certain that Varney's vision has not been realised and cannot be realised given the current state of arrangements with major suppliers.
Varney's vision
On a high level Varney's vision has been an opportunity to deploy new data handling techniques and precise techniques in classification of assets and content to enable flexible information surfacing, depending on context.
What this means is that there are techniques, in which this country could have taken a lead, that would solve the problems presented by Varney.
In particular, these techniques could be used to mitigate against the risks of future changes in requirements.
The solution to unanticipated changes of context can and should be designed into a system.
Flexible data sharing
I have previously drawn on the metaphor of ships in the English Channel.
We may imagine that each ship is a government department.
Whereas in the case of ships they may only rarely need to communicate with each other (although it would have to be with 100% reliability) in the case of government departments it may be the norm.
What if a government department were to need to change the way or to which other department it needs to communicate? Perhaps like a ship having a different destination.
It can be seen that this would be highly desirable to be possible and was, in fact, part of the original intention of Varney that is not being fulfilled by suppliers.
It is crucial that this be instituted at the earliest possible opportunity, to avoid future disruption and costs.
This begs the question of how this might be achieved as IT spending is going to be radically cut.
Without the research of the independent team as I suggest, I can only offer an opinion which is that it is both important and feasible to do.
Contracts must be withdrawn and as they are some new contracts can be let. The focus of these new contracts would be precisely on small, manageable projects and the absolute necessity for them to share data with other projects.
By which I mean by the most appropriate means, given the project, with standard based protocols and practices the priority.
Why the GCloud is Totally Wrong
The design of process as embodied in software is the most important aspect of application development.
If existing ineffective and poorly communicative applications are redeployed into a central cloud the only benefit is the material one of housing and centralised maintenance. But there is a huge disbenefit in terms of increased inflexibility in the ability to identify failing to purpose applications and re-engineer them. This could introduce horrendous barriers to change and would also introduce huge barriers to the policy I am advocating which is split up and relet contracts to much smaller parties.
If there ever were to be a design of castle, moat and drawbridge to protect existing suppliers, this would be it.
The Oligopoly Fights On
Sensible application development takes place in small co-located teams that have the ability to work through large amounts of fine detail on a daily basis.
Large suppliers do not care about the basic facts of human interaction necessary for successful project development.
They spend their money on convincing the customer of the complexity of the tasks, and therefore its high risk and cost.
Then they spend their money on creating other layers of complexity that enable them to change the requirements to suit what they want to deliver, finally striping out their costs at the delivery stage.
They do every thing in reverse to how delivery should be made because they are not incentivised to get delivery right, while the customer buys services from those whose interest it is to shroud delivery in secrecy and complexity.
Small suppliers would not be able to behave like this so long as a sensible procurement process were to exist.
This would be along the normal lines of competitive tender, possibly with greater openness than in the private sector.
How this might be done in detail and other issues such as whether making Crown Copyright
Seco 0.8.0 Release
-
There is new release of Seco - 0.8.0. This is yet another incremental
improvement of the first of its kind JVM-based notebook environment.
Link: https://gi...
2 months ago
