1. Welcome (John Ketchell, CEN/ISSS)
2. Introduction to COPRAS (Bart Brusse, COPRAS)
2.1. What COPRAS is
2.2. Process
2.3. Criteria for selecting projects
3. Different paths towards standardisation (Scott Hansen, Open Group)
3.1. The challenge of specifications
3.2. How COPRAS can help
4. Keynote Address (Peter Wintlev-Jensen, European Commission)
5. Agenda and targeted objectives (Tatiana Kovacikova, ETSI)
6. Break-out cluster 3: semantic-based systems and languages
6.1. Companies and projects in the semantic cluster
6.1.1. W3C - World Wide Web Consortium
6.1.2. Uni-Verse (Gert Svensson, Royal Institute of Sweden)
6.1.3. SATINE (Asuman Dogac)
6.1.4. Artemis (Asuman Dogac again)
6.1.5. Alvis (Mike Taylor, Index Data)
6.1.6. SIMILAR NoE (Jean Vanderdonckt, University of Louvain)
6.1.7. Digital Business Ecosystem (Peter Stanbridge, Korora)
6.2. Ongoing standardisation activity
7. Plenary review of breakout sessions (Bart Brusse)
7.1. Cluster 1: Broadband access
7.2. Cluster 3: Semantic-based systems and languages (not in order)
7.3. Cluster 2: Security issues
7.4. Cluster 5: eLearning issues (not in order)
7.5. Cluster 4: Smart houses and home networking
8. Close (John Ketchell)
[Brief description of how today is going to work.]
About 35 people came to this meeting.
An overarching 6th Framework project that is not part of any one of the objectives, but addresses all of them. It is supposed to improve interfacing between FP6 IST projects and standards bodies.
COPRAS intends to guide other EC projects in their efforts to have their work formally standardised through whichever is the appropriate one of the plethora of standards bodies. It is mandatory for EEC projects to do this where appropriate, but the process is often difficult. COPRAS aims to smooth the path to make standards available more quickly, and also to reduce duplication of effort.
COPRAS is not itself a standards body.
It runs from 1st February 2004 to 31st January 2007.
### This phase is supposed to have been completed, but we have not even seen this survey, let alone filled it in. Wray?
[We have not been through the selection process, but seem to have got in anyway through the side door.]
Primary selection criteria included:
Secondary criteria:
164 projects were invited to fill in the questionaire. Of these, 92 responded by June 2004 (56%). From these COPRAS identified 40 ``tier 1'' and ``tier 2'' projects most likely to benefit from participating. Of those 40, 36 were invited to participate in the kick-off meeting.
The goal of this kick-off meeting is to define about 14-18 projects for which COPRAS will work on Standardisation Action Plans.
There are many prerequisites to having a standard not only published but widely adopted - which is of course the real goal. First of all comes the quality of the specification itself. Is it ambiguous? Inconsistent? Incomplete?
Assuming it's good, then two paths are often possible:
IST Calls in 2003-04: about 400 projects supported out of 2500 applications.
The IST strategic objective for 2005-06 is: ``anywhere, anytime, natural access to ICT-based services and applications''.
Two big changes from FP5 to FP6:
Technology chains are becoming more complex. Whereas before it was possible to standardise a small part of a technological chain in isolation, the focus now is more on intergrated systems.
Unwarranted editorialising: I think this is a terrible idea, running contrary to everything that we know of how best to build complex systems by decomposition of a complex whole into simpler parts.
Challenges today include:
Despite the difficulties, the EC strongly encourages IST projects to contribute to standards (de jure and de facto) as an important European added-value of project funding. COPRAS support will facilitate this task.
The goals for this afternoon are:
The outcome of all this should be tailor-made Standardisation Action Plans (SAPs) that suit the individual projects involved.
For the afternoon, the break-out sessions will be as follows.
Goals for this session:
Method of work:
Works in four domains covering 20+ activities, and has 50+ groups. Has produced 80+ web standards including HTML, XML, Web Services and the Semantic Web.
[Long description of W3C's internal processes snipped.]
Open source distributed system for graphics, sound and acoustics.
Based on the Verse protocol - a UDP-based real-time, low-latency, low granularity protocol for geometrical 3D information, replacing file transfers with real-time communication. It is this protocol that they intend to standardise.
Verse does not use an existing representation, such as VRML or OpenGL, for its 3D information: it provides its own representation optimised for real-time transmission. So Uni-Verse could be construed as a project to put Quake on a sound theoretical footing.
The Verse protocol was originally designed in 1999. The Uni-Verse project is improving, extending, clarifying and documenting the protocol.
W3C will not touch this protocol because, although is it an Internet protocol, it is not a Web protocol in the strict sense of being layer over HTTP. The W3C member present suggests that this would an appropriate area of work for the IETF.
Exploiting semantically rich web services in the travel industry.
GDSs (Global Distribution Systems), acting as travel agents, integrate information from many aggregators. e.g.
Existing GDSs are legacy systems that must be driven by highly trained human users and which do not interoperate with other systems. They run over dedicated network cables rather than the Internet.
Travel agents are keen to change all this. They have formed a consortium called the Open Travel Alliance (OTA) which has come up with detailed XML schemas for every kind of message that can be exchanged in the travel industry. A few early adopters have started to use OTA-defined web services. They claim large efficiency gains.
Satine's role is to provide semantic-based interoperability between OTA-compliant systems and those which are not properly compliant but wish to implement similar services. It also provides a peer-to-peer system for discovering relevant web services.
It is not entirely clear what this project wants to standardise.
There are at least seven candidate standards for healthcare informatics. Among these, one of the most important is HL7 version 2.x, the most widely used standard worldwide. HL7 has recently approved ebXML and Web Service standards for trial use (version 3.x). IHE has also started using Web Services.
As with the travel services, there is a problem with multiple formats describing messages that have similar intent but different details. It is necessary in such cases to perform ontology mappings, which are done using something called ``semantic mediation''.
Artemis's contribution is involved in the mapping, but the details are obscure.
Both SATINE and Artemis make use of OWL = Web Ontology Language, which in turn is built on RDF.
We already know about this :-)
We want to offer two XML formats for standardisation:
I was asked how much we have worked with, or plan to work with, the existing P2P community. I replied that there are two separate P2P communities: a file-sharing one (pragmatic, efficient, unsophisticated) and an academic one (ambitious, perhaps less practical). We have only had contact with the latter, through the WP4 people, primarily EPFL. The file-sharing community might be interested in what we will be offering as it generalises their existing, very specialised, P2P.
(NoE = Network of Excellence.)
The problem: to implement user interfaces that react to changes in the context of use. ``Context of use'' encompasses:
One example is graceful degradation in presenting images when using displays of different resolution, colour-depth, etc. Given a small screen or low bandwidth, approaches include reducing resolution or colour-depth, using a different image-file format or cropping to the most important part of the picture. This is related to the W3C initiative on device independence.
UsiXML = USer Interface XML, a language in which to describe content-aware interfaces.
Various approaches for UI design include:
The project wants to offer the UI specification language for standardisation. How best to position this? As a device independence standard? A web accessibility standard? A ``multimodal Web'' standard?
See www.similar.cc (free registration required).
Peter's role in DBE is dissemination and some semantic aspects.
The idea is to provide an environment for deployment of e-business applications, to ensure for example that small companies are not locked out of tied-up markets.
Two kinds of companies should benefit from this:
DBE has defined a business modelling language. For example, companies need to be able to describe themselves regarding their legal status, the goods and services they provide, etc. The challenge is that the companies they want to serve do not all work in the same domain, so they can't make domain-specific models: instead, they need to create meta-models at a higher level of abstraction. e.g. a printer of business cards has a ``capacity'' (i.e. maximum output rate) while a theatre also has a ``capacity'' (maximum number of concurrent customers) which are related but different concepts.
There is crossover here with the Semantic Web.
The perennial problem of resource discovery raises its head here, as in many of the other projects described today. Given that you want to book a hotel, how do you find hotel-booking services that have hotels in the right area?
Some businesses are already saying that SOAP and WSDL are not powerful enough to handle their requirements.
The difference between this and ebXML is (DBE considers) that ebXML is not looking at the necessary layers of abstraction. For example, ebXML defines an invoice format; but DBE wants to define a way to define invoice formats (if I understand it properly).
[This part of the agenda was handled in parallel with part 6.2]
Eight projects were represented in the session. The general feeling was the COPRAS was going to be helpful. However, the projects do anticipate some problems: for example, the fact that consortia's member organisations are not members of standards bodies.
The short life-cycles of EC-funded projects presents another problem: that in many cases, by the time projects realise that there is scope to standardise the specifications they have produced, the project is close to ending.
Standards bodies should be involved and consulted at an early stage as projects start to put their specifications together.
[Various future meetings considered, and get-togethers of the Broadband cluster planned at larger gatherings.]
The projects involved a broad diversity of themes, including things like 3D graphics as well as business and health-case issues.
Recurring themes included:
Despite widespread interest in P2P (five of the seven projects we discussed want to use it), there does not seem to be any consortium working on standardising it. There may be a role for COPRAS in leading this process.
There were presentations from five projects. Most are quite well advanced in the definition of standardisable outputs, and some have already identified potential partners for the standardisation process.
The question arose of whether there is really a security ``cluster''. There is potential for synergy among the security-related projects, but some of those projects might find that they have more in common with projects in the other clusters.
Four projects were involved, and recognised significant potential for disseminating standardisable work. One project will consider whether it can accelerate part of its work to facilitate more rapid standardisation. If that can't be done, then a two-step process may be required: first, finishing the specification within the project; second, revising it for standardisation (perhaps after the formal end of the project?)
Five projects were represented, and also several standards bodies. There wasn't much time to go through presentations and discuss the projects in detail. It emerged that for standards bodies, timing and momentum are very important. It is a recurring problem that projects start to think seriously about standardisation only near the end of the lifetime.
Different standards organisations sometimes have completely different procedures. Some organisations require that they are involved from the beginning. This raises problems to do with use of resources. Projects need more input from standards bodies on exactly what is required of them in order to submit.
Today is World Standards Day! We have used it well.
The COPRAS steering group will meet the week after next, and now has plenty to discuss. Some recommendations that are likely to be adopted at that meeting are already apparent.
Finally, thanks to the team that put the meeting together and to everyone who came.