**1. Introduction**

One of the factors on which the financial results of the business company depend is the qual‐ ity of software which company is using. Scientific software plays even more special role. On its quality depend the reliability of the scientific conclusions and the speed of scientific prog‐ ress. However, the ratio of successful scientific software projects is close to average: some part of the projects fails, some exceeds the budget, and some makes inadequate product.

The misunderstandings between scientists as end users and software engineers are even more frequent as usual. Software engineers have a lack of deep knowledge of user's domain (e.g., high energy physics, chemistry, and life sciences). In order to avoid possible problems,

scientists sometimes try to develop "home‐made" software. However, the probability of fail‐ ure in such projects is even higher, because of the lack of the knowledge of software engineer‐ ing domain. For example, scientists in common cases do not know good software engineering practices, processes, and so on. They even can have a lack of knowledge about good practices or good artefacts of the software, made by its colleagues.

We stand among the believers that this problem can be solved using the Wikinomics. The idea of Wikinomics (or Wiki economics) is introduced by Tapscott and Williams [15]. Wikinomics is the spatial activity, which helps to achieve the result, having available resources only. Wiki technologies are laid on very simple procedures: the project leaders collect critical mass of volunteers, who have a willing and possibilities to contribute in small scale. The sum of such small contributions gives huge contribution to the project result. The Wikipedia or Wikitravel portals can be presented as a success story of mass collaboration [17].

In other hand, we believe that the mass collaboration can help to improve only part of the scientific development process. We need a software developing solutions, oriented to services and clouds in order to use all available computational power of the distributed infrastructures.

Service‐oriented computing (SOC) is extremely powerful in terms of the help for developer. The key point of modern scientific applications is a very quick transition from hypothesis generation stage to evaluating mathematical experiment, which is important for evidence and optimization of the result and its possible practical use. SOC technologies provide an important platform to make the resource‐intensive scientific and engineering applications more significant [1–4]. So any community, regardless of its working area, should be supplied with the technological approach to build their own distributed compute‐intensive multidisci‐ plinary applications rapidly.

Service‐oriented software developers work either as application builders (or services clients), service brokers, or service providers. Usually, the service repository is created which contains platform environment supporting services and application supporting services. The environ‐ ment supporting services offer the standard operations for service management and hosting (e.g., cloud hosting, event processing and management, mediation and data services, service composition and workflow, security, connectivity, messaging, storage, and so on). They are correlated with generic services, provided by other producers (e.g., EGI (http://www.egi.eu/), Flatworld (http://www.flatworldsolutions. com/), FI‐WARE (http://catalogue.fi‐ware.org/ enablers), SAP (http://www.sap.com/pc/tech/enterprise‐information‐management/), ESRC (http://ukdataservice. ac.uk/), and so on). Two dimensions of service interoperability, namely horizontal (communication protocol and data flow between services) and vertical matching (correspondence between an abstract user task and concrete service capabilities), should be supported in the composition process.

Modern scientific and engineering applications are built as a complex network of services offered by different providers, based on heterogeneous resources of different organizational structures. The services can be composed using orchestration or using choreography. If the orchestration is used, all corresponding Web services are controlled by one central web ser‐ vice. On the other hand, if the choreography is used and central orchestrator is absent, the services are independent in some extent. The choreography is based on collaboration and is mainly used to exchange messages in public business processes. As SOC developed, a number of languages for service orchestration and choreography have been introduced: BPEL4WS, BPML, WSFL, XLANG, BPSS, WSCI, and WSCL [5].

Our proposal has the following innovative features:

scientists sometimes try to develop "home‐made" software. However, the probability of fail‐ ure in such projects is even higher, because of the lack of the knowledge of software engineer‐ ing domain. For example, scientists in common cases do not know good software engineering practices, processes, and so on. They even can have a lack of knowledge about good practices

We stand among the believers that this problem can be solved using the Wikinomics. The idea of Wikinomics (or Wiki economics) is introduced by Tapscott and Williams [15]. Wikinomics is the spatial activity, which helps to achieve the result, having available resources only. Wiki technologies are laid on very simple procedures: the project leaders collect critical mass of volunteers, who have a willing and possibilities to contribute in small scale. The sum of such small contributions gives huge contribution to the project result. The Wikipedia or Wikitravel

In other hand, we believe that the mass collaboration can help to improve only part of the scientific development process. We need a software developing solutions, oriented to services and clouds in order to use all available computational power of the distributed infrastructures. Service‐oriented computing (SOC) is extremely powerful in terms of the help for developer. The key point of modern scientific applications is a very quick transition from hypothesis generation stage to evaluating mathematical experiment, which is important for evidence and optimization of the result and its possible practical use. SOC technologies provide an important platform to make the resource‐intensive scientific and engineering applications more significant [1–4]. So any community, regardless of its working area, should be supplied with the technological approach to build their own distributed compute‐intensive multidisci‐

Service‐oriented software developers work either as application builders (or services clients), service brokers, or service providers. Usually, the service repository is created which contains platform environment supporting services and application supporting services. The environ‐ ment supporting services offer the standard operations for service management and hosting (e.g., cloud hosting, event processing and management, mediation and data services, service composition and workflow, security, connectivity, messaging, storage, and so on). They are correlated with generic services, provided by other producers (e.g., EGI (http://www.egi.eu/), Flatworld (http://www.flatworldsolutions. com/), FI‐WARE (http://catalogue.fi‐ware.org/ enablers), SAP (http://www.sap.com/pc/tech/enterprise‐information‐management/), ESRC (http://ukdataservice. ac.uk/), and so on). Two dimensions of service interoperability, namely horizontal (communication protocol and data flow between services) and vertical matching (correspondence between an abstract user task and concrete service capabilities), should be

Modern scientific and engineering applications are built as a complex network of services offered by different providers, based on heterogeneous resources of different organizational structures. The services can be composed using orchestration or using choreography. If the orchestration is used, all corresponding Web services are controlled by one central web ser‐ vice. On the other hand, if the choreography is used and central orchestrator is absent, the

or good artefacts of the software, made by its colleagues.

70 Recent Progress in Parallel and Distributed Computing

plinary applications rapidly.

supported in the composition process.

portals can be presented as a success story of mass collaboration [17].


The rest of the paper is organized as follows: Section 2 presents overall idea of the platform for research collaborative computing (PRCC). Section 3 presents Web‐enabled engineering design platform as one of the possible implementations of PRCC. Section 4 outlines the architecture, and main components of our other systems based on Wiki technologies. Section 5 describes the comparison of similar systems. Finally, the conclusions are made and future work discussed.
