|
Country House, CityWest Conference & Golf Resort, Saggart, Dublin, Ireland, 13-16 April 2004
This event was a good opportunity to learn about different aspects of
the Microsoft .NET framework and future technologies and to meet other Academics as well as people from Microsoft Research. Speakers included people from
Microsoft Research, Microsoft Corporation and Researchers who are working with Microsoft technologies.
Details about the presentations and slides
Machine learning research at Microsoft Research Cambridge
Christopher Bishop, Microsoft Research
[Download the presentation
or the
videos]  Biography: Christopher Bishop obtained a B.A. in Physics with First Class Honours from Oxford in 1980, and a PhD in Theoretical Physics from the University of Edinburgh in 1983 with a thesis on quantum field theory under the supervision of David Wallace and Peter Higgs. He then joined Culham Laboratory where he worked on the theory of magnetically confined plasmas as part of the European controlled fusion programme.
He subsequently developed an interest in pattern recognition, and became Head of the Applied Neurocomputing Centre at AEA Technology. In 1993 he was elected to a Chair in the Department of Computer Science and Applied Mathematics at Aston University, where he was a member of the Neural Computing Research Group. He then took a sabbatical during which time he was principal organiser of the six month international research programme on Neural Networks and Machine Learning at the Isaac Newton Institute for Mathematical Sciences in Cambridge, which ran from July to December 1997. After completion of the Newton Institute programme he joined Microsoft Research in Cambridge. He was also elected to a Chair of Computer Science at the University of Edinburgh where he is a member of the Institute for Adaptive and Neural Computation in the Division of Informatics. His research is concerned with the development and application of probabilistic methods for inference and learning.
Accessibility
Gary Moulton, Microsoft Corp.
[Download the presentation]  Biography: Gary Moulton is responsible for coordinating
Microsoft's efforts with assistive technology manufacturers. Prior to working for Microsoft he was manager of Disability Solutions at Apple Computer, Inc. Gary was trained as a clinician and has worked for children and adults with disabilities. He taught child growth and development at the undergraduate and graduate levels. Gary has his doctorates in education and business administration. He is one of the authors of Accessible Technology in
Today's Business.
A lap around Longhorn, the next version of Windows
Stephen Turner, Microsoft UK
[Download the presentation and
code samples] Abstract: The next version of the Windows operating system, codenamed Longhorn, will offer a leap forward in both functionality and usability, as well as introduce radical advances in system security and software development / deployment. This session will provide a code-centric high level walkthrough of the three main pillars of Longhorn: - The presentation pillar focuses on the new graphics framework, codenamed Avalon, a unified model for creating UI, documents and media content.
- The data pillar explores the new storage system of Longhorn, codenamed WinFS, a common schematised store to unify definition and access to application types.
- The communications pillar looks at the SOAP messaging substrate, codenamed Indigo, a service-oriented programming model for cross-program communications over a wide variety of network topologies.
 Biography: Stephen Turner got stuck into computers at the blush age of 13 when his high school teachers, no longer able suffer his constant abuse, conspired to dump him on an unsuspecting physics lab at the local university. There he explored the magic of lasers and logic gates, and suffered frequent syntactic disruption as he navigated a treacherous course from card punch to reader. This was a very long time ago. Today Stephen still cant get enough of the really cool stuff, and so now fills his days at Microsoft exploring and evangelizing Windows Longhorn and Visual Studio Whidbey. Amazingly, they pay him to do it. When Stephen just cant bear reading yet another email (and Microsofties get a lot), he spends most his time riding his horse. Though only five (the equivalent to a 15 year old with an attitude), Harvey is already 6 feet at the shoulders and still growing! As this means that falling off can really hurt, Stephen really should stop doing it.
Cω
Claudio Russo, Microsoft Research
[Download the presentation]
Abstract: In the last decade, strongly-typed, garbage-collected
object-oriented languages have left the laboratory and become mainstream
industrial tools. This has led to improvements in programmer productivity
and software reliability, but some common tasks are still harder than they
should be. Two of the most critical for the development of the next
generation of loosely-coupled, networked applications and web services are
concurrent programming and the processing of relational and semi-structured
data. Cω is an experimental language designed at Microsoft Research that
makes programming with concurrency and external data simpler and less
error-prone.
The Cω approach to both the data and control issues is to extend C# with
new, first-class, types and language constructs, rather than relying on
external libraries, mappings and tools. The concurrency extensions, based on
the join-calculus, provide a simple and powerful asynchronous programming
model, which is applicable in both the local (multiple threads on a single
machine) and distributed (asynchronous message over a LAN/WAN) settings. The
data extensions add new type constructors giving first-class support of both
relational (database tables) and semi-structured (XML trees) data.
Generalized member access allows XPath-like processing to be expressed
directly within the language, and checked by the compiler.
This talk will provide a general overview of the features of Cω and
include a number of demos.  Briography:
Claudio Russo is Research Software Development Engineer at Microsoft
Research, Cambridge. He has a PhD in Computer Science from the University of
Edinburgh. Before joining Microsoft, Claudio designed and implemented the
extended module system of Moscow ML, a popular bytecode compiler for
Standard ML. At Microsoft, he has been busy developing SML.NET, a Standard
ML compiler with object oriented extensions, that targets the Common
Language Runtime and is integrated with Visual Studio .NET. Working with
Andrew Kennedy and Don Syme, Claudio is also contributing to the design and
implementation of Generics on the Common Languate Runtime. Claudio's
research interests include programming language theory, module systems and
types.
Introduction to the Shared Source CLI 1.0 (a.k.a. Rotor)
Peter Drayton, Microsoft Corp.
[Download the presentation]
SML.NET
Claudio Russo, Microsoft Research
[Download the presentation] Abstract: The .NET Common Language Runtime presents a unique opportunity for programming language researchers: the combination of runtime infrastructure, language interoperability support, and class libraries make niche languages which target the platform more plausible as alternatives to mainstream commercial languages. In this talk Ill describe SML.NET, our compiler for the functional programming language Standard ML. Ill focus on our novel language extensions that support interop with other .NET languages and on ongoing work integrating the compiler with Visual Studio,
Microsoft's development environment.
Biography: see above
Effective Web Browsing and Searching on Mobile Devices
Ralph Sommerer, Microsoft
Research Abstract: Web pages with complex layout do not display well on mobile devices, because their small displays cannot accommodate the space requirements of such pages. Therefore, they require extensive amounts of scrolling, both horizontally and vertically. This quickly leads to disorientation within a page. In order to make optimum use of the small displays on mobile devices for Web browsing and searching, it is necessary to enforce both overview and detail concerns of page viewing. However, to provide a consistent Web browsing experience across a variety of devices it is also necessary to preserve as much as possible of the original layout of Web pages and hence their familiar look and feel even when constrained by small displays. SmartView is a browser feature that improves the browsing experience on mobile devices in two ways: First, it provides an overview in the form of a zoomed out image of the page, with outlines segmenting it into detail regions; the user can select any of these regions in order to view it separately from the rest of the page. Second, it rearranges Web page layouts in such a way that they fit the display, reducing any scrolling to a single direction. It thus connects the user with the familiar content design and presentation as experienced on the desktop. SearchMobil extends SmartView with the aim of providing support for searching: pages retrieved by a search engine are displayed in an overview that is annotated to show the locations of search terms. Users can therefore quickly direct their attention to the most promising region of a page.  Biography: Ralph Sommerer is a researcher at Microsoft Research. Ralph obtained his diploma in Computer Science from the Federal Institute of Technology (ETH) in ZricSwitzerland, in 1989. He then joined the Institute of Computer Systems at ETH, where he worked as a research and teaching assistant in the fields of software component architecture and online documents. He concluded his project with a Ph.D. in 1996. After leading a project group in a company and developing software tools for embedded systems, he joined Microsoft Research in Cambridge in January 1998. Togewith Chuck Thacker, he built Microsofts first experimental Electronic Book (eBook) prototype, a conceptual study that eventually led to Microsofts eBook Reader. Hes been working for a while now on aspects of
Smart Reading and the Mobile Web, such as layout, Web browsing and search on small devices.
The Storage User Experience in Windows Longhorn
Walter Smith, Microsoft Corp.
[Download the presentation]
Abstract: As hard drives grow ever larger, and new data types like
digital music and photos proliferate, traditional file systems don't do
enough to help users deal with their data. Windows Longhorn provides a new
user experience for organizing, searching, and sharing data, based on a new
storage system derived from both file system and database technologies. Learn
about Microsoft's approach and the philosophy that motivated these exciting
enhancements.
 Biography: Walter Smith
is a Senior Architect in the Windows User Experience team, currently working
on Windows Longhorn storage and communications features. In his eight
years at Microsoft he has contributed to Internet Explorer, the Windows
Shell, and MSN Explorer. Previously he was at Apple Computer where he was a
major contributor on the Newton project. Walter has a B.S. in Applied
Mathematics from Carnegie Mellon University.
Why trust is bad for security?
Dieter Gollmann, Technical University Hamburg-Harburg
[Download the presentation] Abstract: Security goes through fashion cycles and
today's fashion is trust. This talk will explore some of the many meanings this word has been used in, and is currently being used in. I will argue that trust is not a unifying concept bringing together different branches of information security, but a source of great confusion and superficiality. I will also point to interesting research directions hidden under the cloak of trust.  Biography: Prof. Dieter Gollmann received his Dipl.-Ing. in Engineering Mathematics (1979) and Dr.tech. (1984) from the University of Linz, Austria, where he was a research assistant in the Department for System Science.
He was a Lecturer in Computer Science at Royal Holloway, University of London, and later a scientific assistant at the University of Karlsruhe,
Germany, where he was awarded the venia legendi for Computer Science in 1991. He rejoined Royal Holloway in 1990, where he was the first Course Director of the MSc in Information Security. He was a Visiting Professor at the Technical University of Graz in 1991, an Adjunct Professor at the Information Security Research Centre, QUT, Brisbane, in 1995, and has acted as a consultant for HP Laboratories (Bristol). He joined Microsoft Research in Cambridge in 1998. In 2003, he took the chair for Security in Distributed Applications at the Technical University Hamburg-Harburg, Germany. He is a Visiting Professor with the Information Security Group at Royal Holloway. Dieter Gollmann is one of the editors-in-chief of the International Journal of Information Security and an associate editor of the IEEE Security & Privacy Magazine. He has authored a textbook on
Computer Security.
Domain Specific Modelling
Alan Cameron Wills, Microsoft Corp. Abstract: A domain-specific language is a language designed to express the requirements and solutions of a particular business domain. The insurance industry, telephony, ice-cream manufacture, etc, all have their own vocabulary, constraints and patterns which can be most conveniently expressed in a purpose-made language. With clever engineering, we can design a compiler and runtime architecture that can input such definitions and generate an insurance or telephony or ice-cream system. A domain specific language can be seen as a coherent kit of components with an explicit grammar. Within certain domains, such tools already exist, and substantial improvements in development time are reported. This talk will provide an overview of the use of domain-specific languages, together with the associated development processes and tools.  Biography: Alan Cameron Wills works in
Microsoft's Whitehorse team, developing the next generation of tools for Visual Studio. Until 2003, Alan was a methods consultant (and founder of Trireme International Ltd). He is co-author of Objects, Components and Frameworks with UML: the Catalysis Approach [1998], which advocated the use of precise abstractions to model component based systems. During the 80s, he worked as a researcher in the Formal Methods group at Manchester University. He started programming in 1972, and can still read punched paper tape and edit it with a ball-point pen and sticky tape.
.NET: Road to Pervasive Integration
Rafal Luckawiecki, Project Boticelli Ltd
[Download the presentation] Abstract: XML, Web Services, Loose Coupling and Standards are keys to the way software architectures have evolved over the past few years. One of the goals of this evolution, however, is something much more difficult to achieve: a situation where all software, devices and systems can interact with each other without any need to integrate them. While we are quite a few years or a decade away from reaching that ideal the first steps have already been made. Microsoft's .NET and the related Web Services Architecture coordinated by WS-I.org are one of the most important recent architectural developments in IT. Not only will this session will provide a high-level overview of WSA protocols, but it will also explain the more long-term evolution of software towards the ideal of pervasive integration.  Biography: Rafal Luckawiecki, Project Boticelli Ltd: In his role as eVisioneer and Director at Project Botticelli Ltd Rafal is responsible for analysing, planning and forecasting the changes in the field of Information Technology. Through direct association with his clients, Rafal is working closely with teams of up to 150 software developers, as well as with investors and their boards of directors. This work allows him to practice the best principles of Microsoft Solutions Framework, and Microsoft Operations Framework - the secrets behind
Microsoft's and many other IT companies success. Amongst his past experiences, Rafal was also one of the founders of Bot Inc., a Seattle software development company. In his other previous roles at Oxford Computer Group Ltd and at Aris Corporation Rafal gained extensive practical experience in consulting and professional technical speaking on a variety of subjects, which he continues to perform at Project Botticelli Ltd. His frequent conference keynotes and other speeches have been sought after by many company directors, CEOs and business leaders. These included many Microsoft TechEd
conferences, Butler Group events and very prestigious addresses in a
number of European countries upon their governments invitations.
Rafal specialises in three fields: study of programming models, security and cryptography, and project and operational management of IT. His work focuses mainly on: team building, management of software development ventures, designing successful IT support and operations, current and future software design patterns, e-Business, Microsoft .NET strategy, IT security, OO development, and messaging. In his spare time, Rafal has been writing a new book on Microsoft Solutions Framework aimed at its practitioners. Last but not least, amongst over 50 of his professional Microsoft qualifications he has previously obtained the MCSE, MCSD, Internet, MCT and unique MSF Master Trainer endorsements. Rafal also holds an MSc in the Foundations of Advanced Information Technology from Imperial College, University of London.
Using Viruses and Bugs to Teach Artificial Intelligence (AI) Programming in C#.NET
Peter Cowling, University of Bradford
[Download the
presentation] Abstract: A good way to get students to push themselves to understand and tool up for difficult new concepts is to make it as much fun as possible. AI is widely regarded as the
next big thing by the computer games industry, with increasingly advanced AI appearing for commercial games. Moreover, games have long been regarded as an important research platform for AI ideas in academia. In this talk we describe how these ideas have been put together in the final year undergraduate/masters module
AI for Games at the University of Bradford. The module is assessed entirely by two
course works. The first coursework is to develop a key AI component for the 2-player board game of
Virus. The second is to develop AI which controls a bug in a simulated environment, an academic version of
Microsoft's
Terrarium application. In each case, a server runs games continuously so that students can access a league table on the Internet, which has maintained a high level of student interest and motivated students to use sophisticated AI and software ideas.
The talk will demonstrate the two applications used in the module and the results server. We will talk about some of the innovations shown by students, both in creating their AI and in forming a thriving
add-on community, and assess the potential of approaches of this type for teaching advanced programming and AI concepts.  Biography: Peter Cowling leads the Modelling Optimisation Scheduling And Intelligent Computing (MOSAIC) research group at the University of Bradford. He has worked extensively in the areas of scheduling, optimisation and modelling, both within academia, where he has published over 50 papers, and as a consultant, having led the team responsible for the development of models and heuristics for the successful SteelPlanner range of steel scheduling software, and having developed systems used for scheduling staff in one of the UK's largest financial institutions. He obtained MA and DPhil degrees from the University of Oxford. Current research interests include flexible evolutionary approaches and
hyper heuristics applied to realistic, dynamic models of real-world personnel and production scheduling, and the development of Artificial Intelligence approaches for game playing agents.
Using the .NET Framework in University Courses
Damien Watkins, Microsoft Research
Abstract: Incorporating new technologies in to a curriculum is an
important task for attracting students, keeping courses relevant and
demonstrating emerging standards but, if not managed correctly, it can also
become an all-consuming occupational hazard. This workshop details
experiences introducing the .NET Framework as the implementation platform in
three university subjects. The first was an undergraduate subject, Software
Component Technologies, which had previously used COM, the second was a
Masters subject, "Distributed Object Technologies", which had used a
combination of CORBA and DCOM and the third was a Windows programming
subject that had previously used the Object Windows Library.  Biography: Damien Watkins is a member of the University Relations team at Microsoft Research Cambridge. His major area of responsibility is the adoption of .NET related technologies in research and teaching within EMEA.
Before joining MSR, Damien founded and managed his own software consulting company called Project 42. Prior to commencing Project 42 Damien lectured at the School of Computer Science and Software Engineering and the Department of Software Development at Monash University. Damien lectured both postgraduate and undergraduate subjects, covering the areas of UNIX System Call Programming, Windows Programming with C++, Component Based Software Development and Distributed Object Oriented Technologies. Damien taught the worlds first .NET Framework course at Monash University. He has also taught at Uppsala University (Sweden) and KMITNB (Thailand). Component architectures that provide language interoperability have always held a keen interest for Damien, and he has written a paper on this topic, titled Handling Language Interoperability with the Microsoft .NET Framework. Since 1998 Damien has been engaged with Microsoft on the development of the .NET Framework. This has required regular visits to Microsoft, Redmond and also included a two-month sabbatical in June/July 2001. Damiens PhD thesis, titled Adding Contracts to Interface Definition Languages, dealt with improving the quality of distributed object oriented architectures, such as DCOM and CORBA. Damien has published several papers on COM/DCOM, CORBA and the .NET Framework. He has presented tutorials, seminars and workshops on COM/DCOM, CORBA and the .NET Framework at SIGCSE 2002 (USA), the Microsoft Research Faculty Summit 2001 (USA), ObjectWorld (Australia), TOOLS Pacific (Australia), TOOLS Europe (France), TOOLS East Europe (Bulgaria) and TOOLS Asia (Beijing). Related publications:
Symphony in C#: Biofeedback and 3D Visualization using the .NET Framework
Robert Burke & Gary McDarby, Media Lab Europe
Abstract: The MindGames Group is an interdisciplinary team at
MIT Media Lab
Europe that integrates intelligent biofeedback, computer gaming and
sensory immersion to develop techniques for positively affecting a
participants state of mind. For over a year, the group has developed
Symphony, an extensive signal processing and visualization
framework that leverages the combined power of the C# language, the .NET
Framework, and associated technologies such as Managed DirectX.
In this talk, we will describe a number of diverse real-time
applications that have been developed using Symphony. One of these
applications, Mind Balance, involves the non-invasive acquisition
and processing of electroencephalogram signals from the surface of the
head, for the purpose of creating a brain-computer interface that
controls a character that is tottering across a tightrope. In another
application, Still Life, conventional video acquisition
technologies are used to generate a surreal magic mirror. Although
originally developed for physiotherapeutic applications, Still Life has
also provided the interactive technical component for live theatre
performances, including a modern dance performance at the United
Nations World Summit on the Information Society. Other applications
being developed with Symphony include Aura Lingua, which uses a
tongue-controlled bi-directional wireless interface; and Peace
Composed, where the depth and complexity of a piece of music are
affected in real-time by a participants level of relaxation, as
measured by their galvanic skin response.
Included in our discussion will be some specific examples of how the
.NET technologies have allowed us to develop Symphony's strong
foundation, and also rapidly build and deploy such a wide variety of
applications. We will show how Symphony is currently being used to
build very complex real-time applications by a variety of researchers,
some of whom had limited previous software development experience.
Finally, we will discuss some potential future directions for the
architecture within the Longhorn timeframe and beyond.
More information:
 Biography:
Robert Burke is a Research Associate with the MindGames Group, where
he is the designer and lead architect of the Symphony code base. In
addition to his work on the architecture, he is also the lead software
developer on the projects Still Life, Mind Balance and
Peace Composed. Prior to his research at Media Lab Europe, Rob was a
graduate student with the Synthetic Characters group at the MIT Media
Lab, where his thesis contributed a new ethologically-inspired model for
predictive learning in lifelike virtual characters.
 Biography:
Gary Mc Darby has a background in biomedical signal processing
and is a principal research scientist in Media Lab Europe leading a
multi-disciplinary group called MindGames. The group has a range of skills from
computer science, signal processing, psychology, hardware design, games
programming, graphic design and music composition. The group has produced a
number of widely acclaimed demonstrations notably the Relax to Win
concept where players must relax in a racing game in order to win and
Brain-Child, an immersive video game where gaining control over relaxation
becomes a power in the virtual world. Gary was primarily responsible in
facilitating the construction of an Intel Computer Clubhouse on the Media Lab
Europe campus and he recently featured on RTEs Would You Believe show
which documented his life over the last ten years. He has given numerous talks
on technology and philosophy and has recently been nominated to the European
Academy of Sciences for contributions to human progress.
Keynote speakers
Microsoft Ireland
Terry Landers, Head of Government Affairs,
Microsoft Ireland
Irelands growth
strategy: research the key
Edward Walsh, President,
Irish Council for Science,
Technology and Innovation (ICSTI)
[Download presentation]
Training
An
Introduction to the Common Language Infrastructure
Damien Watkins, Microsoft Research
[Download the
presentation and code samples]
Biography: see above.
Abstract: A number of Component-Based Architectures (CBAs) have
been developed to facilitate the interoperability of software components
that are written in different programming languages. The most widely known
and used of these is OMGs Common Object Request Broker Architecture (CORBA)
and Microsoft's Component Object Model (COM) and Distributed COM (DCOM).
Although not specifically designed to provide language interoperability,
many Java-based technologies, such as Enterprise Java Beans (EJB), also
adhere to the principles of a CBA. The Common Language Infrastructure (CLI)
is the latest CBA to be developed, and it builds on many of the lessons
learned from previous architectures. In this workshop, we will look at the
major components of the CLI. The CLI is an ECMA standard on which Microsoft
has built numerous implementations, including: the .NET Framework, the
Compact Framework and the Shard Source CLI. The CLI is also the
specification on which non-Microsoft implementations are based.
Fusion and .NET Security
Richard Grimes
Abstract: The purpose of these two session is to build upon the
introduction course and to give a deeper understanding of fundamental
aspects of the framework and runtime. These labs will cover two main topics:
Fusion and .NET Security.
Fusion is the technology that .NET uses to locate and load assemblies and
has been designed to prevent the naming, versioning and location issues that
are inherent in Windows Dynamic Link Libraries. Fusion is flexible and
allows users to alter the way it works; it also maintains a log of its
actions so that users can determine the source of problems when the occur.
In this lab we will show how to use the facilities of Fusion and how to
interpret the information that it logs. Well also show how to share
libraries through the Global Assembly Cache (GAC) and how to change
versioning through publisher policy files.
There are two important aspects to .NET Security. The first aspect is
ensuring that the code is legitimate and has not been tampered. The second
aspect is that the code is trusted to perform the actions it attempts. In
this lab we will show how .NET validates that an assembly has not been
tampered when it is loaded and checks to see if it contains dangerous code
through verification. We will also show .NET Code Access Security that uses
evidence obtained from the code, rather than from the user, to determine the
permissions that the code has to perform specific actions. We will show how
to apply code access security to your code and how to configure .NET policy
to change the trust given to an assembly.
 Biography:
Richard Grimes has a BSc in Physics from
Nottingham University and a PhD in Semiconductor Physics also from
Nottingham. He has research experience in GaAs/AlGaAs 2 dimensional
structures and postdoctoral work on semimagnetic semiconductors. After his
science career Richard worked as a software architect working on distributed
object systems in the days before CORBA and DCOM. Richards writing career
started when he wrote the first and most widely used book on DCOM and then
followed that up with books on ATL, MTS and COM+. More recently, Richard
writes and gives talks on .NET with an emphasis on the fundamentals of the
framework. Hes currently writing a comprehensive reference on WinFX, the
framework behind Longhorn.
Richard has written best selling books for Addison Wesley and Microsoft
Press on the .NET framework and Managed C++. He speaks regularly at .NET
conferences in the US and Europe including VSLive, DevConnections, WinDev,
WinSummit, BASTA and DevWeek. Richard writes for a variety of magazines
including MSDN Magazine, VCDJ, Windows Developer Magazine and Developer
Network Journal. Currently, he is a contributing editor of Dr Dobbs Journal
where he writes a monthly column on .NET. Richard has been awarded Microsoft MVP status since 1999 and is an MVP for contributions to the .NET community.
Related publications:
Demos and posters
- Adaptive Performance Monitoring of Component Based Applications, Adrian Mos, Dublin City University
- Agent-Managed
Community Healthcare Transactions, Richard Hill and Simon Polovina,
Sheffield Hallam University
- Dendros, Steven Murdoch, Markus Kuhn and Piotr Zielinski, University of Cambridge
- End-to-End .NET For Field
Marketing, Assad Faizi, University of Reading
- e-Science, the Grid and Microsoft .NET, Mike Jackson, Edinburgh Parallel Computing Centre
-
OpenPMF - Modelling Security for Complex, Heterogeneous, Distributed Systems,
Ulrich Lang, University of Cambridge
- Plugging .NET into academic curriculum, Amala Rajan, Middlesex University
- Sentient Objects for Context-aware Mobile Computing, Vinny Cahill, Trinity College Dublin
-
Supporting Process-Oriented Architectures through Component Composition,
Finbar McGurren, University of Limerick
- Teaching Hardware/Software Co-Design, James Srinivasan, University of Cambridge
-
A web-enabled application and information system for use by GAP year students,
Tom Robinson, University of Reading
Adaptive Performance Monitoring of Component Based Applications
Adrian Mos, Dublin City University Component middleware platforms such as J2EE or .NET address the needs of enterprise projects by providing reusable standardized services (such as persistence and s ecurity) and reliable runtime environments which developers can leverage in a cost-effective manner when building large-scale systems. By concentrating on business logic, less time is spent on the implementation of the application infrastructure, therefore reducing time to market. However, the resulting complexity of the running system is orders of magnitude higher than the complexity of the business logic code alone, as the component containers have a significant computational footprint in order to provide complex lifecycle and systemic services. In such conditions, it is often difficult to precisely understand the causes for performance degradations observed at application level with such metrics as use-case response time. COMPAS is a framework for non-intrusively instrumenting and monitoring the performance of component-based applications at the component level. In addition, it can extract execution models from running systems and present them in UML. By dynamically switching parts of the monitoring infrastructure on and off based on runtime conditions, COMPAS ensures negligible overhead when the system is operating within acceptable limits, and a minimum overhead when some parts exhibit poor performance and need to be inspected. Since the extracted performance data is presented at the conceptual level developers operate with, the time needed to identify the appropriate correctional steps can be minimised. The COMPAS prototype has been implemented for J2EE and the feasibility of porting it to .NET is under evaluation.
Agent-Managed Community Healthcare Transactions
Richard Hill and Simon Polovina, Sheffield Hallam University
Many multi-agent software
systems are being developed with little regard to the economic environment
in which these agents will have to conduct their transactions. Our interest
is in the contractual agreements that exist between community healthcare
agencies and their 'customers' (being elderly patients or their local
government authorities). Economic scarcity is particularly prevalent in such
an area, and not just for its monetary costs as political and social
dimensions have also to be considered. Community healthcare thus highlights
the challenges that must be described in a robust way if agent-managed
services are to be accepted commercially. Typically auction models are
utilised by agent applications as they are widely understood, established,
and cheap to implement, but they are too simple to capture the nuances of
community care payment transactions as described. We therefore consider the
Event Accounting model in relation to the myriad of payment transactions
within the community care environment, and propose a robust
transaction-based framework for the deployment of agent-managed community
care systems in practice.
Dendros
Steven Murdoch, Markus Kuhn and Piotr Zielinski, University of Cambridge Imagine the future of the Internet as a single huge document, divided and replicated across thousands of servers, using a new kind of distributed database or file system, based on a next-generation variant of XML. What security mechanisms could encourage users to share control over their data more widely? How far can we detach the question of who controls the servers from who controls the stored data? Project Dendros creates a unifying data-management layer for Internet applications. Its goal is to create, study, and optimize a unifying replacement technology for historically grown standards and technologies such as DNS, file systems, file structuring formats such as SGML/XML, MIME, ASN.1, email headers, X.500, LDAP, and MIME. The core of Dendros is a flexible and simple data model, essentially a tree of byte strings. This is combined with a modular processing model based on access filters, which implement functions such as storage management, remote read/write access, authentication, access control, revision control, and replication. A particular aim of Dendros is the design of new cooperative administration concepts that enable participants with limited mutual trust to cooperate effectively in a single global name space. Instead of a single authoritative central server, the data is managed by a group of peer servers, which independently verify all access conditions for each change request. Therefore, malicious behaviour of some of the servers will not interfere with the integrity and availability of the others. Each server maintains an independent revision history of the data. The right to add a change and the right to mark a new version as
current are orthogonal. This makes it easier for the community to optimistically grant write access to less trusted members and roll back changes if necessary.
End-to-End .NET For Field
Marketing
Assad Faizi, University of Reading
Field marketing largely involves processing considerable amounts of data
into meaningful information such as marketing reports. Gekkos current IT
systems are unable to deliver the level of detailed reporting their clients
(such as Three, Logitech and AMD) are now demanding. Although a basic in
house solution exists, it is very cumbersome to manage and deploy,
compromising the immediacy and quality of reports that clients require.
To develop and implement a .NET solution that will integrate the various
business components into a single IT system which is both scalable and
secure. The .NET platform will enable Gekko Partners Limited to use a
variety of client devices such as PDAs for collecting field data, while
using server based components to process the data into rich content web
reports.
As part of a government sponsored initiative known as the Knowledge
Transfer Partnership (or KTP), the University of Reading, supervised by Dr
Shirley Williams and Dr Andrew Adams, is helping Gekko Partners Limited to
implement such a system. By recruiting a graduate through the KTP programme,
both the company and the university can combine ideas and knowledge
thinking, all contributing to a well developed IT system.
Gekko Partners Limited is a BTL agency entering their 3rd year of
trading. Gekko provide field marketing solutions to brands within the
consumer electronics, technology and telecoms sectors. Activity is
undertaken across both Consumer and B2B channels which involves data
collection to enable clients to understand, amongst other things, their
channel, stock availability and competitor products. Within the consumer
channel, activity is executed throughout high street retailers such as
Dixons, PC World, Carphone Warehouse etc. All activity requires data to be
understood and presented back in campaign report format.
e-Science, the Grid and Microsoft .NET
Mike Jackson, Edinburgh Parallel Computing Centre The MS.NETGrid project was a year-long project focused upon implementing the Open Grid Services Infrastructure (OGSI) on Microsoft .NET. The goal of the project was to provide a practical demonstration to the UK e-Science community of the applicability of Microsoft .NET technologies to the hosting, development and deployment of Grid Services. A complementary goal was facilitating understanding of the Grid and e-Science within Microsoft. The project was a collaboration between EPCC and Microsoft Research Ltd., funded through the UK National e-Science Centre by the UK Department of Trade and Industry. The achievements of our project were as follows: - Delivery of an implementation of OGSI running under Microsoft .NET - MS.NETGrid-OGSI - based upon the design of the implementation of OGSI provided within the Globus Toolkit 3. This container provides the option to maintain service state within files or an SQL Server database facilitating service persistence and container robustness.
- Delivery of an implementation of a set of demonstrator Grid services running under this OGSI implementation, including an OGSA-DAI service which interacts with SQL Server via ADO.NET technology.
- Delivery of four OGSI and Microsoft .NET training courses at the UK e-Science Institute in Edinburgh and at Imperial College in London to delegates from both the academic and commercial sectors and from a range of UK e-Science institutions across the UK.
In our poster we will present the high-level architecture of MS.NETGrid-OGSI and experiences of attendees at our training courses. We shall also provide a demonstration of MS.NETGrid-OGSI.
OpenPMF - Modelling Security for Complex, Heterogeneous, Distributed Systems
Ulrich Lang, University of Cambridge
The proliferation of different distributed systems platforms and security
technologies complicates the integration of distributed applications and the
migration of existing applications to new technologies. Model driven
software development tries to tackle this problem by modelling the
application logic undistorted by technology and using automated (or
semi-automated) tools to map the model to the particular technology. The
OMG's Model Driven Architecture (MDA) is one predominant framework that
advocates this approach. Distributed systems security faces a similar
problem in that there are many different platforms and security technologies
that need to be integrated.
We present our flexible, model-driven Policy Management Framework (PMF),
which is the result of applying the concepts of model driven software
engineering to security. PMF is available as Open Source software. It is
developed as part of the EU part-funded IST research project COACH
(Component Based Open Source Architecture for Distributed Telecom
Applications).
Plugging .NET into academic curriculum
Amala Rajan, Middlesex University The software industry has ventured into the integration of windows and the internet with Microsoft .NET which generates Smart Client Software. Microsoft .NET expands the opportunities by using this technology to act on data and communicate anywhere, anytime and on any device. IT industry has already accepted this technology and is searching for qualified .NET professionals for fulfilling its needs. There is a wide gap that exists between this need and the IT professionals emerging from our country. Plugging in .NET can minimise the technological gap between lack of knowledge of IT professionals in .NET and the need and to enhance the technological content within academic programmes. This project involved analyzing the learning outcomes of relevant subject modules within curriculum, and based on this analysis, to identify, design and implement the
plugging in of various elements of .NET architecture into these modules.
Sentient Objects for Context-aware Mobile Computing
Vinny Cahill, Trinity College Dublin This poster/demonstration will present work in progress towards the design of a middleware architecture for the development of proactive context-aware applications from sensor-rich components based on the Sentient Object paradigm. We will describe/demonstrate a visual programming tool for the development of sentient objects and a case studies that are being used to evaluate the approach including autonomous sentient vehicles and a smart couch.
Supporting Process-Oriented Architectures through Component Composition
Finbar McGurren, University of Limerick
As the component based development process progresses into component
composition business processes identified during requirements gathering are
translated into software constructs based upon the guidelines of the
underlying component model. This translation creates a cognitive gap between
the business process models developed during design and the system
implementation. During software evolution considerable time and effort is
expended bridging this gap. It is important therefore that the components of
a system be linked to a particular business process so as to aid the
maintenance task. My research investigates how this gap may be reduced by
adding a process abstraction to the application model using a component
composition language and presents an architectural connector framework that
supports this model.
Organizations face major challenges in the provision of their software
systems and with increased market pressure, they require systems that can
incorporate new requirements and adapt quickly to changing environments.
Despite the requirement for flexible software systems, it is estimated that
approximately 85% of the system lifecycle is spent performing maintenance.
Much of this effort is spent locating the site of evolutionary change. This
problem may be attributed to the ad-hoc techniques used to compose component
based systems and their inability to make business processes explicit
architectural entities. Furthermore these techniques are too low level to
support an evolution process that takes place at a higher level of
abstraction in which business strategy and rules change.
Component based systems have a clear separation between computational and
compositional entities represented in the architecture. This separation of
concerns lends itself particularly well to performing software evolution as
it is possible to change system composition independent of the components
involved. It is important that the techniques used to compose software
systems provide the necessary abstractions for evolving these systems in the
future.
As a solution to these problems the component application model is
extended to include a process abstraction layer. At this layer business
processes are superimposed upon the components found in the business layer.
The application model is now more process orientated, operating at a higher
level of abstraction, and built on top of existing component technology
shielding the developer from infrastructural complexities.
This research makes the following contributions:
- A component composition language for defining business processes at
the process layer.
- An architectural connector framework that supports the process
orientated application model.
Teaching Hardware/Software Co-Design
James Srinivasan, University of Cambridge Thanks to a Microsoft Research Innovation Excellence Award for Embedded Systems we have ported the Windows CE .NET operating system to the Altera EPXA1 development board. This board features an Altera Excalibur FPGA with an integrated ARM9 processor together with a variety of external interfaces. By porting the Windows CE operating system we hope to offer more challenging group and individual projects in our undergraduate Computer Science course as well as providing students with invaluable real-world experience of an entire embedded system.
A web-enabled application and information system for use by GAP year students
Tom Robinson, University of Reading
Using a modular development approach the existing static website of GAP
Activity Projects is being augmented with a number of value added
interactive services. The first is allowing online applications, followed by
a community portal for use by GAP year students and alumni. Later, services
for volunteer and salaried staff will be developed for use in the offices
across the world.
GAP Activity Projects is the largest year-out provider for 17 - 20 year
olds. A not-for-profit organisation providing a diverse range of voluntary
work placements in 33 countries. Placements last between 3 - 12 months and
range from teaching English as a foreign language, schools assistance,
caring, medical and outdoor education to environmental projects. On return,
GAPpers can keep in touch through the Alumni Club and join the Business
Partnership Scheme for help with graduate recruitment.
The project is a Knowledge Transfer Partnership (KTP) scheme with The
University of Reading, supervised by Dr. Rachel McCrindle and Dr. Shirley
Williams of the School of Systems Engineering. These government sponsored
programmes combine the knowledge of the company, graduate and knowledge base
partner. A research based masters degree is a major part of the programme
and will be looking at A web development framework for incorporating
individual identities within corporate websites.
The extensive use of .NET will be key to the success of the project as it
will allow the interactive modules to be expanded upon and customised for
each of the growing number of global offices, each with varying cultures,
resources and business procedures.
|
