Software testing is a very time-consuming and tedious activity and accounts for over 25% of the cost of software development. In addition to its high cost, manual testing is unpopular and often inconsistently executed. Software Testing Environments (STEs) overcome the deficiencies of manual testing through automating the test process and integrating testing tools to support a wide range of test capabilities. Most prior work on testing is in single-thread applications. This thesis is a contribution to testing of distributed applications, which has not been well explored. To address two crucial issues in testing, when to stop testing and how good the software is after testing, a statistics-based integrated test environment which is an extension of the testing concept in Quality Programming for distributed applications is presented. It provides automatic support for test execution by the Test Driver, test development by the SMAD Tree Editor and the Test Data Generator, test failure analysis by the Test Results Validator and the Test Paths Tracer, test measurement by the Quality Analyst, test management by the Test Manager and test planning by the Modeller. These tools are integrated around a public, shared data model describing the data entities and relationships which are manipulable by these tools. It enables early entry of the test process into the life cycle due to the definition of the quality planning and message-flow routings in the modelling. After well-prepared modelling and requirements specification are undertaken, the test process and the software design and implementation can proceed concurrently. A simple banking application written using Java Remote Method Invocation (RMI) and Java DataBase Connectivity (JDBC) shows the testing process of fitting it into the integrated test environment. The concept of the automated test execution through mobile agents across multiple platforms is also illustrated on this 3-tier client/server application.

The primary purpose of this thesis is to explore what issues may arise during development of a framework for handling and display of streamed real-time data. In addition to this, it also investigates how the display of different types of data, along with a change of execution platform, impacts execution time. Through the undertaking of two case studies, each split into developmental and an experimental phases, the thesis goes through the development of such a real-time data handling framework. The framework was developed in both stationary and mobile forms, and the developmental issues encountered along each of these paths are highlighted. Afterwards, the results gathered from performance tests run on each framework version were compared, in order to ascertain whether the handling and display of different data types, along with a change in execution platform, had had an impact upon the frameworks execution time. The results from the developmental observations revealed that the most commonly encountered issues were those relating to program latency, commonly due to sub-optimal program architecture along with connectivity issues encountered during data streaming. The second most encountered issue regarded the choice of an appropriate display method, in order to communicate changes in the displayed data along with correlation between several tracked data points. The results from the experimental comparisons revealed that while the impact on execution time caused by the use of calculated data, as opposed to raw data values, was marginal at most, a change of execution platform impacted said time drastically. By porting the framework to the mobile platform, the different processes whose execution time were measured during the tests experienced an increase in execution time ranging from 2405% all the way to 15860%. The authors recommend that the framework be developed towards gaining the ability to connect to any given relational database, and to handle and display the data therein, in order for it to have application areas other than as a test instrument. Further, the authors also recommend that additional tests be run on the framework using a wider variety of stationary and mobile devices, in order to determine whether the conclusions drawn from the results in the thesis hold up in the face of greater hardware variety. / Denna studies primära mål är att utforska vilka problem som kan uppstå under utveckling av ett ramverk för hantering och visande av streamad realtidsdata. Utöver det undersöks även hur visande av olika datatyper, ihop med ett byte av exekveringsplattform, påverkar exekveringstiden. Genom utförandet av två fallstudier, båda uppdelade i utvecklingsoch experimenteringsfaser, går denna studie igenom utvecklingen av ett sådant ramverk för hantering av realtidsdata. Ramverket utvecklades i både stationär och mobil form, och de utvecklingsrelaterade problem som påträffades i vardera fall belyses. Efteråt jämfördes resultaten framtagna genom prestandatester, som kördes på samtliga ramverksversioner, för att upptäcka om hantering och visning av olika datatyper, samt ett skifte av exekveringsplattform, hade påverkat ramverkets exekveringstid. Resultaten från de utvecklingsrelaterade observationerna visade att det mest påträffade problemet hade att göra med programlatens, vanligtvis p.g.a. ickeoptimal programarkitektur kombinerat med konnektivitetsproblem. Det näst mest påträffade problemet hade att göra med valet av en passande visningsmetod, för att kunna förmedla förändringar i den visade datan, samt korrelation mellan flera följda datapunkter. Resultaten från de experimentella jämförelserna visade att medan påverkan av exekveringstiden som uppstått genom användandet av kalkylerad data, till skillnad från rådatavärden, endast var marginell som bäst, påverkade förändringen av exekveringsplattform denna tid drastiskt. Genom att porta ramverket till den mobila plattformen upplevde de processer vars exekveringstid mättes under testerna en ökning från 2405% hela vägen upp till 15860%. Författarna rekommenderar att ramverket utvecklas mot förmågan att koppla till godtycklig databas, och att kunna hantera och visa datan från denna, för att ha ett användningsområde bortom användandet som testinstrument. Vidare rekommenderar även författarna att ytterliggare test utförs på ramverket med en större variation av stationära och mobila enheter, för att kunna bekräfta om slutsatserna som dragits utifrån resultaten av denna studie kvarstår efter att de utsatts för mer varierande hårdvara.