Web 2.0 desktop - continued

In my previous posting I have indicated the necessity to develop a Web 2.0 desktop. Whilst doing this, I discovered that there are various activities taking place in this direction. To name a few: eyeOS, gOS and Mozilla Firefox Ubiquity.
eyeOs offers a desktop for social computing. You can install various social computing applications like YouTube, LinkedIn and others on your desktop and access them easy. gOS is similar.
Ubiquity is yet different. It provides a command line interface in your browser with a number of predefined commands. These commands are related to the sites that you would like to access, e.g. your Google Mail, Wikipedia and eBay. Still, it is also possible to add other commands yourselves.
I use scribefire to maintain this blog. It runs as a plug in for Firefox and is probably easier to use than the command line interface. Let us see in which directions these developments move. To be continued.

Web 2.0 desktop

These days, a lot of people discuss Web 2.0. As it is stated, it will influence our daily working. Government for instance can be make use of Web 2.0 to activate co-creation of policy with citizens.
A lot of these initiatives are really including blogs like this, discussion groups, contacts and their activities, and so on. As an everyday person, Joe the plumber, I am also involved in those activities, hence this blog. However, it becomes fairly difficult to integrate those activities in your daily work. A new approach will be required, focusing on an individual. A personal Web 2.0 desktop is required.

How to achieve such a Web 2.0 desktop? One of its very first requirements is that is easy to program. It has to be programmable like 'Lego Mindstorm' as someone in Sweden informed me. This brings us to the basic features of such a desktop. These are for instance:

• managing your profile in various communities.
• managing contacts across different communities based on different interests, e.g. personal, hobbies, work, etc.
  
• publishing of information to existing sites like Flickr, Youtube, music, Twitter, Hyvess, and others, including your blog. Similar like this, that I have constructed using ScribeFire.
• accessing information from various sites based on feeds and other mechanisms like search. Search should access the sites that you are interested in, e.g. Flickr, Youtube, etc.
• communicating with various persons and groups, e.g. chatting, scribbling, email, discussions, and so on.

Possibly one could think of more functions of such a Web 2.0 desktop. This desktop should also run on mobile devices with their particular OS, like for instance the OS of the Apple Iphone/Itouch. If we are really able to integrate these functions on the future desktop, we achieve a new means of working. Development all depends on the existence of programming interfaces of suppliers of Web 2.0 functionality. If these exist, technical details can be hidden and an easy programmable platform can be made.

Next generation service platforms

The EU is initiating a call regarding next generation service platforms the coming year. It is not yet clear what they should do, however, there are some initial ideas. For example, they should support dynamic service configuration. What does this imply? What types of services should be configured? Are we looking at business or application services?

Most IT suppliers tend to look at application services. These will allow a user to dynamically create mash ups with mechanisms like supported by Google or Yahoo Pipes. Users can select services dynamically, but might have to enter data several times for each service. Service orchestration is done by the user.

What we really need is dynamic business process configuration of application services that support a business service requested by a user. The request may be interactively entered on a website or might be a message. A request needs to be matched with offered business services and a business process needs to be configured. It involves many aspects, e.g. business service mediation (matching a request to offered services), semantic mediation (semantics need to be mapped) and behavior mediation (a new choreography needs to be constructed for a user specific business service request). These issues are not yet addressed in current standards; those deal with other types of more IT oriented mediation questions linked to ESBs. We really need alignment of business services with IT services and dynamic processing of user specific service requests.

In case you agree (or really don't agree) with my point of view, let me know. Mash ups and pipes are really primitive examples of dynamic process composition.

ASP, cloud computing, SaaS

Some time earlier, I have discussed differences between open source and SaaS (Software as a Service). I argued that instead of investing heavily in tuning open source into an application for your organization, it would be better to explore the SaaS model. Why would you need to own software at all? You need to have it available.

Providers like Google and Amazon offer the concept of cloud computing. It has many advantages, namely scalability and performance. Furthermore, you pay per use, like in the SaaS model. The difference with SaaS however is that you own your application and that application needs to be developed on the platform offered by Google or Amazon. The main question here is: what is a possible exit strategy.

This same question is applicable to SaaS: what is my exit strategy. As you buy services offered by a software platform, the exit strategy would be to port those services to another provider with its particular platform. It might need adjustment of that new platform, because the particular provider offers services somehow different. Standardization of services should help portability.
Another part of the exit strategy in case of SaaS is the portability of data. You should be able to transport data from one provider to another. A good strategy would be to implement data services and port those services and the accompanying data to a new provider. Again, standardization of data services would help. However, do data services offer you sufficient performance?

When choosing a solution to your IT challenges, several models are possible. Each model has its challenges of its own. In the future, standardization of services would be the solution to choose the model you want. If you buy software or a service, in both cases the supplier must support the appropriate services. Who takes the next step in service standardization?

Next steps for e-Government

Several countries have done a lot of work in the past years that result of useful applications for both citizens and companies. Multi-channeling has been adopted, the service oriented architecture explored and possibly web services are already operational.
Most of these efforts are on the technical level. It is still difficult to convince business people of local and national government agencies of the possibilities and the impact on their processes. Technical solutions are not yet translated into business perspectives, which makes it difficult to apply those solutions.

Such a middle out approach, trying to map business issues into existing technical solutions is always difficult. On the other hand, a top down approach seems very time consuming: specifying business issues and needs and supporting them by technical solutions. I am of the opinion that current technology and tooling is getting there. Lacking are integrated tools and techniques that support a top down approach. One is able to specify semantics of business needs and there supporting business processes, but transforming those specifications into technical solutions is still hard labor.


Although a government CIO and other responsible managers can possibly be convinced to standardize business semantics and processes, I think it is still difficult to close the loop. If anyone is aware of integrated tools, approaches, etc., please react.

What about open standards?

There is a great many standards under development for supporting the Semantic Web, specifically to support web services in an open environment. Of course, we are all familiar with XML Schema, WSDL, OWL and others. We have a web service stack and a semantic web stack for showing how these standards all relate. These days, Semantic Web Services are fairly new, but do we have a Semantic Web Service Stack? New standards need to be developed with their accompanying tool support to support Semantic Web Services. SAWSDL, WSML, and WSMO are but a few examples of these standards. Do we all have to use these standards, what is their overlap if any, are some of the questions that we could ask ourselves.
To be brief, we need a framework to identify the functionality of open standards. Do frameworks already exist? What do they cater for and are standards already grouped according to those frameworks. I know of some frameworks, most of them being specified for organizations and not tailored for behavior between organization (or any systems). From science, some parts are already available, e.g. those focusing on process aspects. My suggestion would be to develop a framework to specify external behavior. If any one is familiar with such a framework, let me know.

Government architectures

Lots of effort is put in establishing government architectures. Interoperability frameworks are listing various technical standards and ICT governance is also put in place. People transform existing documents into different parts, like strategy, interoperability, and so on. Any good guidelines for IT architecture, including government architecture, should consider generic aspects like abstraction and decomposition, paradigms like SOA, and so on. Existing principles like 'multi-channeling' and 'no wrong door' are widely accepted by most governments. New technology to support these principles should be explored. Is it possible to create one virtual government (per country, for the EU, and global?)?

Besides changes in technology, other aspects need to be considered, since the environment is constantly changing. How to deal with Web x.* for instance. What does it mean for government behaviour? How to cope and integrate standards from the gaming environment for disclosing graphical information and interacting with citizens and companies? In this context, governments still consider graphical information as geographical information.


To be brief, we should set general principles for a government architecture and try to implement them in existing and new projects. Furthermore, we should consider new developments and integrate them in our approach. We have to keep an open mind. As IT architects, we should provide the means to government and business to do an possibly change their work with new means.

Structuring government systems - public services

There are many initiatives for creating an infrastructure for government organizations. Basically, these initiatives should cover two issues, namely customer effectiveness and process efficiency. Customers, citizens and companies, should get the services they require and processes should support those services in a proper way.
A number of national (Dutch) and international (EU) projects considers modeling of public services according to the laws they have to support. This is basically a correct assumption, because one of the tasks (or the only task) of government organizations is to execute one or more laws and possibly local regulation. However, there are some aspects lacking, namely:

• A law is a (legal) model of restriction with respect to behavior of citizens and companies. An IT system should not model a law, but the underlying system on which the restrictions of the law are modeled.
• Derivation of public services from laws gives an ideal input-output model of that public service. In practice, the behavior of citizens, companies and government organizations is more dynamic. These behavioral aspects should be considered, described and published when implementing public services.
• Technically, public services and their input-output can be specified in many ways. It is advised to specify them as ontology with open standard, that can (automatically) be transformed into technical representations. Separation of functionality and technical representation should be the guiding principle.

Of course, there is more to it to implement this approach to public services, e.g. case management. We will discuss those aspects another time.

DGO 2008 - the final day

In my previous post, I did not yet incorporate the final day. This one was in fact one of the better parts of the conference, although it was only half a day. It included discussions on citizen engagement as the next phase in e-government, e.g. aspects like e-rulemaking (although I did not grasp the technicalities presented the previous day), e-townhalls and, of course, Gov 2.0 (every itself respecting sector has a 2.0 strategy). This particular presentation combined with the others raised ideas with me for a next generation government: Government as an ecosystem. In such a particular situation, the government should consider its particular function. Currently, most local municipalities focus on so-called support operations and administration. They have no particular knowledge and means for policy making and incorporating new policies easy. They should focus on the latter one. Support operations can be shared, examples were presented, and administration can be handled by local offices with central IT facilities (like financial institutions).

Another presentation was on an E-Government Framework, which was composed of aspects like purpose and role of government, societal trends, technology changes, information management, human elements and interaction/complexity. The framework is based on global research. The basic message is: interdisciplinary research is required.
The final presentation on urban simulation was especially of interest for one of my assignments. It gave insight in separating value indicators and simulating urban changes to get new values for those indicators. The open source simulation environment UrbanSIM was adjusted to support a number of value indicators; the value indicators could not be configured. Furthermore, the next phase in research for urban simulation is citizen participation.

International conference on Digital Government

Currently, I am staying a couple of days in Canada for a conference on Digital Government. The basic idea is that eGovernment is out and we have to move to an information government. As within private business, the Internet will have a big impact for the government: they have to change processes and become user centric. This is the great challenge for government organizations for the next decay.

The government and society faces a number of issues to be solved. A first important issue is that the government has a lot of public information of citizens and companies. Although there are strict government policy rules with respect to publication in the context of Regulatory Privacy Policies, this might not be sufficient. Currently mashups like those that are possible with Google maps are feasible. Although the public information can be anonymous, mashup of that information may reveal relevant aspects of citizens and companies. Thus, Private Policy Rules mechanisms must be put in place and have to be maintained by each citizen/company: which public information allow they to reveal to whom.
Similar mechanisms can of course also be installed for applications like Hyvves, Facebook, and so on. The younger generation publishes lots of information on the web, which they later can regret. A Private Policy Rules mechanism might help. These developments are still new and it is not yet clear how they will work. Research is ongoing.

A second issue is also in this domain: identification and authentication. Lots of governments are seeking for the holy grail in this respect. US has the RealID program which requires to have a unique identification for all US citizens by 2017. In the meantime one of the basic problems is verification of such an ID. During Katerina, lots of government records on which an identification is based, were lost. The state of Maine decide to use the records of a trusted authority as a new basis, namely those of the Baptist church. This example indicates that federation of identity needs to be implemented. Liberty Alliance is looking into these aspects. Open standards are a necessity to get things started.
I have also seen a study which indicated that citizens were only interested in establishing their identity for simple services like applying for a drivers license. In case of complex services like tax declaration citizens and companies were not to worried about identity; they were moreover worried about correct data entry and data processing by the government.

A third aspect is innovation by governments. A number of talks were on this subject, e.g. collaboration of governments for development of software like in the open source community. Some municipalities drive these developments, although there are many forces against it (people in the government think they loose their jobs; the normal process is by RFPs whilst open source processes are evolutionary, and so on). Like always, in the US everyone is looking at a proper business case for these types of developments (there are 50 states and 20.000 municipalities, so currently high IT investments that are under pressure). A representative of the OECD however stated that government should do its job properly and leave these kinds of developments to others (the OECD?).

Like open source, also open standards were discussed. Furthermore, examples of business process re-engineering were given, e.g. in the Beer Lively Lab of Heineken were for a pilot situation, customs is looking at a pull mechanism instead of the current push (declaration) mechanism. Finally, I would say. This is a revolutionary change for customs: they need to have access to trader systems based on web service standards and have to audit those systems and the internal administration! There is still the aspect of risk management: which goods/containers have to be selected for inspection in for instance a port. The fraud with current procedures is still estimated in 2005 to be identical to the figures of 1985 for the EU: 1.5 BEuro per year.

Besides these more general presentations, there were also quite specific ones, e.g. text analysis and annotation by learning algorithms, how to organize regional for handling incidents, and so on. These types of presentations are not easily transferred into a Dutch situation.

Basically lacking in this conference, I would think, is a general overview of the state within several countries regarding Information Government. A number of stage models were presented to show how a government organization could move forward. Quite general aspects with respect to architecture were discussed; there is a lack of a common semantic model for government information and thus the development of application (web) services. I have presented a poster in this respect, which was especially with the Dutch delegation (a rather large one) good received.

Open Source and SaaS

Quite a number of government organizations focus on software development with open source. Main advantages, they say, is no license fees and no vendor lock-in. What they forget, is they will still have a vendor lock-in, namely with the software developer that has adjusted the open source software to fit the needs.

Internationally, there is a strategy of Software as a Service. Main advantages is, that a customer pays per use and no investments in developments have to be made. The software is ready for use. Quite a number of examples are already available, e.g. there is a growing number of service suppliers of Content Management Systems with this business model. Examples are Webnode, Moonfruits, Plone, Virtuoso, but also Microsoft supports this model.

Question will be: do I invest in open source or do I apply SaaS? Other questions will come next, e.g. in the context of CMS as SaaS: how can I switch my content from one SaaS supplier to another? One could state that this is a developing business model, worth investing in in case an organization has a relative small IT department and IT is not the primary business of that organization.

IT architecture and projects

In my projects as consultant, I still encounter customers that do not consider architecture as part of their project. They start projects with vague requirements and are of the opinion that these requirements will become stable during software construction. In our current webbed society, they also forget that the software has to function in an open environment. It is their opinion that interfaces with other partners in an organization network will become stable at the end.

What they forget is that partners also have to adjust their software to be able to function in the new environment. A software component can only be developed, in case all interfaces are fully specified. IT architecture has several views, e.g. a business and information view. One of the views is: the set of components and their interfaces. Clear, concise and complete specification of those interfaces allows all partners in a network to adjust their software simultaneously. The result will be a fully functioning system, operational within planning and hopefully budget.

Business services versus business chains

There has been and still is a growing interest in so-called business chains. These comprise a number of organisations that try to achieve a particular objective. Well known examples are supply chains. Most of these supply chains are centered round one particular organisation, e.g. a large retailer or producing company. We always call them 'supply chain of company X'.

Today, there is a growing interest for service orientation. It starts with technical, webservices. From a business perspective it comprises business services that are supported by IT service, e.g. webservices. Examples are transport services, payment services, etc. Each organisation has its internal policies to offer these services, e.g. by buying parts of the services from suppliers. These policies are transformed to business rules that govern the internal business processes.


This approach has many advantages. Services of organisations can be viewed as bricks for composing business chains. The contractual agreements between organisations for these services are the glue in those chains. In case there is a contract for many business transactions or for a period, the chain is fixed. Otherwise, the chain is per transaction.
Thus one is able to construct chains based on these individual bricks. It allows organisations to dynamically adapt to changes in their environment, which is especially of interest for government agencies controlling laws and regulations.

Choosing standards

Most professionals are of the opinion that only sound technical standards are the best choice. From a technical perspective, this is of course always true.

However, there are more forces that influence the choice of a standard. Take for instance the historic choice between VHS and Betamax (for those that remember these systems) for video, of which the best technical solution did not win. Amongst others, forces that influence choice are substitutes, supplier strategies, government policies, etc. Examples are for instance in the implementation of the webservices standards (WS-*) compared to others like ebXML. Technically, ebXML with respect to data exchange is the most sound. However, many suppliers heavily invest in webservices and offer complete product suites for drag and drop development. Oracle is an example of such an environment. Other reasons for webservices are more in the past. Internet was not that reliable for data exchange. ebXML developed a method for reliable transfer, but many organizations invested in using queuing technology to solve this problem. MQseries was one of the most popular products. Complete EU communication networks are still based on these products.


Thus, there are a number of aspects to consider when adopting a standard. Technical sound is only one of the options.

Folksonomies and social tagging

Part of Wikinomics is the change of consumers into prosumers. They contribute to the production processes of service providers. It seems that policy, government, etc. is a good application area to integrate closely with citizens and companies with respect to future developments and service delivery. However, this leads to a great many questions at this time, e.g. which tools to apply, how to change your interaction with citizens and companies, etc. In this particular case, we just forget about organizational changes and functional profiles of civil servants. We do not yet know what the effects will be, we just have to get things started.

Wikinomics: how to apply the concepts?

With much interest I have been reading the book of Don Tapscott and Williams on Wikinomics. It is very stimulating, but after some chapters it is repeating itself. On the other hand there is the book of Keen stating that such an environment leads to too much information, in which the real information gets lost. I find it challenging to see how the concepts can apply in my profession as consultant in applied research. Do processes of my customers change, what is the impact of the new technology and how can I get these process changes activated? These are some of the questions that are bothering me. Any suggestions are welcome.