WoTUG - The place for concurrent processes

BibTeX Proceedings details

BibTex html text dump of PaperDB database.

PaperDB is GPL Software, see: http://paperdb.sourceforge.net.

@InProceedings{MartinJassim97a,
  title = "{A} {T}ool for {P}roving {D}eadlock {F}reedom",
  author= "Martin, Jeremy M. R. and Jassim, S. A.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "1--16",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "We describe a tool, programmed in Java, for the formal
     verification of the absence of deadlock and livelock in
     networks of CSP processes. The innovative techniques used
     scale well to very large networks, unlike the exhaustive
     state checking method employed by existing tools."
}
@InProceedings{Delft97,
  title = "{S}criptic: {P}arallel {P}rogramming in {E}xtended {J}ava",
  author= "Delft, Andr\`{e} van",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "17--33",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "Scriptic is an expression based extension to the Java
     programming language, targeted at user interfaces,
     simulations and parallel computing on shared memory systems.
     The extras are mainly founded on the theory of Process
     Algebra: constructs for non-deterministic choice,
     parallelism and communica-tion. By default, these parallel
     constructs have interleaving seman-tics, rather than
     multi-threading or forking. Specific Java code fragments may
     run in their own threads or handle events from the windowing
     system. This makes interactive applications such as arcade
     games execute as fast as corresponding plain Java versions.
     GUI components such as buttons and menu items are enabled
     and disabled when applicable, without additional
     programming. This paper covers an example application in
     Scriptic, an overview of the language constructs, the
     implementation, originality, previous work and current work."
}
@InProceedings{Demaine97,
  title = "{H}igher-{O}rder {C}oncurrency in {J}ava",
  author= "Demaine, Erik D.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "34--47",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "In this paper we examine an extension to Hoare's
     Communicating Sequential Processes model called higher-order
     concurrency, proposed by Reppy. In this extension,
     communication algorithms (or events) are first-class objects
     and can be created and manipulated dynamically. In addition,
     threads are automatically garbage collected and channels are
     first-class, that is, they can be passed over other
     channels. We describe the design of a Java package that
     implements the main features of higher-order concurrency,
     with similar ease-of-use to Reppy's Concurrent ML system.
     Our implementation can be easily extended to use a
     distributed system, which is a major limitation with
     Concurrent ML. We also hope to bring the idea of
     higher-order concurrency to a wider audience, since it is
     extremely powerful and flexible, but currently only well
     known to the programming-languages community."
}
@InProceedings{HilderinkBroenink97,
  title = "{C}ommunicating {J}ava {T}hreads",
  author= "Hilderink, Gerald H. and Broenink, Jan F. and Vervoort, Wiek and Bakkers, Andr\`{e} W. P.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "48--76",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "The incorporation of multithreading in Java may be
     considered as a significant part of the Java language,
     because it provides rudimentary facilities for concurrent
     programming. However, we belief that the use of channels is
     a fundamental concept for concurrent programming. The
     channel approach as described in this paper is a realization
     of a systematic design method for concurrent programming in
     Java based on the CSP paradigm. CSP requires the
     availability of a Channel class and the addition of
     composition constructs for sequential, parallel and
     alternative processes. The Channel class and the constructs
     have been implemented in Java in compliance with the
     definitions in CSP. As a result, implementing communication
     between processes is facilitated, the programmer can avoid
     deadlock more easily, and the programmer is freed from
     synchronization and scheduling constructs. The use of the
     Channel class and the additional constructs is illustrated
     in a simple application."
}
@InProceedings{Verhulst97,
  title = "{B}eyond transputing : fully distributed semantics in {V}irtuoso's {V}irtual {S}ingle {P}rocessor programming model and it's implementation on of-the-shelf parallel {DSP}s.",
  author= "Verhulst, Eric",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "77--86",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "Virtuoso Classico /VSP is a fully distributed real-time
     operating system originally developed on the INMOS
     transputer. Its generic architecture is based on a small but
     very fast nanokernel and a portable preemptive microkernel.
     It was further on ported in single and virtual single
     processor implementations to a wide range of processors. As
     the basis of any real-time application is a scheduler that
     allows the developer to use the minimum schedule to satisfy
     the real-time requirements, a number of derived Virtuoso
     tools have been developed with complementary
     functionalities. This paper describes the rationale for
     developing the distributed semantics of Virtuoso's
     microkernel and describes some of the implementation issues.
     The analysis is based on the parallel DSP implementations as
     these push the performance limits most for hard real-time
     applications. The Virtuoso tools are being ported and
     further developed in the DIPSAP-II and EURICO OMI/Esprit
     projects."
}
@InProceedings{Gook97,
  title = "{R}econfigurable {C}omputing",
  author= "Gook, Roger",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "87--87",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "The name may be familiar of old to the WoTUG community, but
     it has now been adopted by one of the fastest growing
     sectors of the silicon industry. Reconfigurable Computers
     are computing systems whose hardware architecture can be
     modified by software to suit the application at hand. The
     core component is the FPGA. Remarkable performance gains are
     achieved by placing an algorithm in an FPGA for embedded
     applications, compared with using a microprocessor or DSP.
     This is because an FPGA takes advantage of hardware
     parallelism while reducing the timing overheads needed for
     general-purpose microprocessor applications. For example the
     time taken by load/store operations and instruction decoding
     can be eliminated. Reconfiguration enables the FPGA to
     provide a problem specific computer for highly optimised
     application performance. Just as high level programming
     languages liberated the first microprocessors programming
     languages will liberate the FPGA. The first of these
     languages to become commercially available is Handel-C.
     Handel-C is based on the CSP Model; it was developed by the
     Hardware Compilation Group at the University of Oxford and
     is to be marketed by ESL. The Handel-C tools enable a
     software engineer to target directly FPGAs in a similar
     fashion to classical microprocessor cross-compiler
     development tools, without recourse to a Hardware
     Description Language. Thereby allowing the software engineer
     to directly realise the raw real-time processing capability
     of the FPGA. The skills and expertise gained by the WoTUG,
     provide the group with a competitive advantage to develop
     the innovative algorithms, applications and products in this
     domain."
}
@InProceedings{SingletonCook97,
  title = "{A}n {O}pen {S}ystems {S}trategy for {D}istributed occam {E}xecution",
  author= "Singleton, Paul and Cook, Barry M.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "88--103",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "We demonstrate the feasibility of distributed execution of
     occam programs within computer networks which support open
     systems standards for inter-process communication, remote
     execution and program hosting (e.g. hardware-independent
     programming languages and operating-system-independent
     APIs). We first propose a general architecture, and then
     describe a constructive proof of its viability (i.e. a
     prototype). which uses a novel multiplexed virtual channel
     protocol over UNIX sockets. Finally we summarise some of the
     opportunities for further development which this project has
     created."
}
@InProceedings{WelchWood97,
  title = "{H}igher {L}evels of {P}rocess {S}ynchronisation",
  author= "Welch, Peter H. and Wood, David C.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "104--129",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "Four new synchronisation primitives (SEMAPHOREs, RESOURCEs,
     EVENTs and BUCKETs) were introduced in the KRoC 0.8beta
     release of occam for SPARC (SunOS/Solaris) and Alpha (OSF/1)
     UNIX workstations. This paper reports on the rationale,
     application and implementation of two of these (SEMAPHOREs
     and EVENTs). Details on the other two may be found on the
     web. The new primitives are designed to support higher-level
     mechanisms of SHARING between parallel processes and give us
     greater powers of expression. They will also let greater
     levels of concurrency be safely exploited from future
     parallel architectures, such as those providing (virtual)
     shared-memory. They demonstrate that occam is neutral in any
     debate between the merits of message-passing versus
     shared-memory parallelism, enabling applications to take
     advantage of whichever paradigm (or mixture of paradigms) is
     the most appropriate. The new primitives could be (but are
     not) implemented in terms of traditional channels, but only
     at the expense of increased complexity and computational
     overhead. The primitives are immediately useful even for
     uni-processors -- for example, the cost of a fair ALT can be
     reduced from O(n) to O(1). In fact, all the operations
     associated with new primitives have constant space and time
     complexities; and the constants are very low. The KRoC
     release provides an Abstract Data Type interface to the
     primitives. However, direct use of such mechanisms still
     allows the user to misuse them. They must be used in the
     ways prescribed below else else their semantics become
     unpredictable. No tool is provided to check correct usage at
     this level. The intention is to bind those primitives found
     to be useful into higher level versions of occam. Some of
     the primitives (e.g. SEMAPHOREs) may never themselves be
     made visible in the language, but may be used to implement
     bindings of higher-level paradigms (such as SHARED channels
     and BLACKBOARDs). The compiler will perform the relevant
     usage checking on all new language bindings, closing the
     security loopholes opened by raw use of the primitives. The
     paper closes by relating this work with the notions of
     virtual transputers, microcoded schedulers, object
     orientation and Java threads."
}
@InProceedings{TidmusMiles97,
  title = "{P}refetch {D}ata {M}anagement for {P}arallel {P}article {T}racing",
  author= "Tidmus, Jonathan and Miles, Roger and Chalmers, Alan G.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "130--137",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "The particle tracing method uses a stochastic approach for
     global illumination computation of three-dimensional
     environments. As with many graphics techniques the
     computation associated with the image generation is complex.
     Parallel processing offers the potential of solving the
     computational complex particle tracing more rapidly.
     Distributed memory parallel systems are scalable and readily
     available. However, large environmental models are often
     bigger than individual node storage capabilities requiring
     data management to distribute and control the movement of
     environmental data as computation proceeds. Prefetch data
     management attempts to reduce idle time associated with
     remote data fetches by anticipating the latency and
     requesting required data items prior to their actual use
     during computation. This paper demonstrates how attention to
     work division and supply coupled with prefetch data
     management can be utilised to minimise overheads associated
     with a parallel implementation and reduce overall image
     synthesis time."
}
@InProceedings{HarrisonBrown97,
  title = "{WEAVE}: {A} {S}ystem for {D}ynamic {C}onfiguration of {V}irtual {L}inks",
  author= "Harrison, S. R. and Brown, Chris R.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "138--151",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "This paper describes Weave, a system which has been
     developed to support the use of a DS Link (IEEE 1355) based
     parallel computer architecture. Weave is an extension layer
     on IIPC (a simple transputer and UNIX parallel processing
     environment which provides dynamic process management,
     hardware transparency and message passing.) Although Weave
     is suitable for any T9000 Transputer network, it has been
     specifically designed to support the use of the AIVRU DS
     Link Vision Engine. A brief description is given of this
     machine, which processes live digital video using a mixture
     of hardware modules and software processes, all
     interconnected by DS Links. Weave provides the ability to
     make virtual link connections between processes on demand at
     run-time. These connections may be disconnected when no
     longer required, and hence the whole hardware architecture
     is dynamically reconfigured automatically to suit the
     requirements of the software application. A small functional
     interface provides processes with the ability to alter their
     own connectivity, and that of other processes. A temporal
     locking mechanism for each virtual link controls when it may
     be disconnected, and when pending connection requests can be
     fulfilled. This locking mechanism is driven by the action of
     communication over the virtual link. The Weave system
     supports transparently, the creation and destruction of
     connections between software processes and the image
     processing hardware modules (and between hardware modules
     directly). Also provided transparently by Weave is support
     for the use of hardware message replicator module(s) that
     multicast virtual link data to any number of DS Link
     recipients."
}
@InProceedings{Cook97,
  title = "{D}ata-{S}trobe {L}inks and {V}irtual {C}hannel {P}rocessors",
  author= "Cook, Barry M.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "175--188",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "Data-strobe links and message transfer protocols as used by
     the SGS-Thomson T9000 processor are described, as are the
     essential characteristics of the supporting virtual channel
     processor. A method of providing the same functionality
     without the use of a T9000 is suggested and illustrated by a
     T225 processor design using a software virtual channel
     processor and minimal supporting hardware. Finally,
     differences between the international standard, IEEE 1355,
     and the T9000 links from which it was derived are described
     and the implications for virtual channel links highlighted."
}
@InProceedings{WelchPoole97,
  title = "occam for {M}ulti-{P}rocessor {DEC} {A}lphas",
  author= "Welch, Peter H. and Poole, Michael D.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "189--198",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "A multi-processor implementation of occam2.1 for
     interconnected DEC Alpha processors has been derived from
     the Kent Retargetable occam Compiler. Each Alpha processor
     is supported over a PCI bus by a T425 transputer, whose
     links complete the communications fabric. This paper reports
     on the software mechanisms for supporting these platforms
     from occam so that they appear just like any transputer
     system -- a collection of processing nodes connected by
     channels placed on links. Advantage was taken of a
     proprietary multi-threading kernel, supplied as part of 3L
     Parallel C/AXP, to support parallel inter-node
     communication. occam multi- processing is supported by the
     KRoC kernel running within one of the 3L threads. The
     performance of generated code and networked systems has been
     benchmarked, with particular care being taken to measure the
     interaction overheads between the Alpha and its
     communication fabric. An image analysis program was also
     used in the benchmarking as an example of a real
     multi-processor application."
}
@InProceedings{Kalinowsk97,
  title = "{A} tool for optimisation of program execution in dynamic topology systems",
  author= "Kalinowski, Tomasz",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "199--209",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "In this paper, we present a tool for optimisation of
     execution of parallel programs in distributed memory
     multi-processor systems with dynamic interconnection
     networks. The programs are described as Directed Acyclic
     Graphs (DAGs). The tool allows to compare simulated
     execution times for different task scheduling heuristics,
     target system topologies and communication models. A list
     scheduling algorithm, which has been applied, accounts for
     dynamic changes of interconnection structure. We demonstrate
     the efficiency of dynamic networks by comparing schedules
     obtained for dynamic and fixed topology systems. We propose
     a method of validating simulation results in a target system
     composed of T9000 transputers. The method relies on
     comparison of simulation results with execution times of
     synthetic OCCAM applications in the target system. The
     comparison indicates that assumptions taken on program
     execution and system model hold in the system under
     investigation."
}
@InProceedings{SilvaPedroso97,
  title = "{T}he {D}esign of {JET}: {A} {J}ava {L}ibrary for {E}mbarrassingly {P}arallel {A}pplications",
  author= "Silva, Luis M. and Pedroso, Hern\^{a}ni and Silva, Jo\~{a}o Gabriel",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "210--228",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "JET is a parallel virtual machine. It has a dynamic number
     of processors which may run different proprietary operating
     systems. The processors communicate through the slowest
     intercommunication network of the world, do not provide peak
     performance and in the overall the parallel machine can be a
     very unstable computing surface. In other words, JET uses
     the idle CPU cycles of computers that are connected to the
     Internet, being a really inexpensive supercomputer. This
     paper presents the design of a Java parallel library that
     provides support for the execution of embarrassingly
     parallel applications. It inherits the security, robustness
     and portability features of Java and includes support for
     fault-tolerance, scalability and high-performance through
     the use of parallelism."
}
@InProceedings{Tudruj97,
  title = "{F}ine-grained global control constructs for parallel programming environments",
  author= "Tudruj, Marek",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "229--243",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "Problems of evolved control in fine-grained parallel
     programs in distributed memory systems are discussed in the
     paper. Global control constructs are proposed which
     logically bind program modules, assign them to worker
     processors and define the involved flow of control.
     Implementation methods are discussed which assume control
     flow processing decoupled from data processing inside
     executive modules. The proposed constructs are extensions of
     the OCCAM 2 language. They can be incorporated into an
     intermediate code generated by a parallel language compiler
     or can be used by a programmer to define control flow
     between fine-grained program modules assigned to different
     processors. Architectural requirements for efficient
     implementation of the proposed control constructs are
     discussed."
}
@InProceedings{RodriguezSande97,
  title = "{E}xpanding the {M}essage {P}assing {L}ibrary {M}odel with {N}ested {P}arallelism",
  author= "Rodriguez, C. and Sande, F. and Le\`{o}n, C. and Garcia, F.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "244--251",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "A synchronous extension to the library model for message
     passing (Inmos C, PVM, Parmacs, MPI, etc.) is presented.
     This extension, provides a comfortable expression of nested
     parallelism from inside the message passing model.
     Furthermore of being a valuable tool for the presentation
     and teaching of parallel algorithms, the computational
     results prove that an efficiency similar to or even better
     than the one obtained designing and implementing algorithms
     using the native language can be achieved."
}
@InProceedings{Sakellari97,
  title = "{C}ompile-{T}ime {T}echniques for {M}apping {L}oop {P}arallelism",
  author= "Sakellariou, R.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "244--251",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "A synchronous extension to the library model for message
     passing (Inmos C, PVM, Parmacs, MPI, etc.) is presented.
     This extension, provides a comfortable expression of nested
     parallelism from inside the message passing model.
     Furthermode of being a valuable tool for the presentation
     and teaching of parallel algorithms, the computation results
     prove that an efficiency similar to or even bettern tahn the
     one obtained designing and implementing algorithms using the
     native language can be achieved."
}
@InProceedings{SchrettneJelly97,
  title = "{D}ynamic {P}rocess {I}nteraction",
  author= "Schrettner, Lajos and Jelly, Innes",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "261--273",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "This paper concerns with the design of a building block for
     parallel and distributed software systems. We start with a
     very common problem of process interaction and successively
     derive a building block that can be used to construct
     systems that are correct by construction. When copies of
     this building block are connected to each other in an
     arbitrary fashion, the resulting system is deadlock/livelock
     free. An application is outlined where this method was used.
     We also would like to stress that the method of successive
     refinements used in this paper seems to be a fruitful
     approach in the design of protocols."
}
@InProceedings{HaasThornley97,
  title = "{T}he {M}acram\`{e} 1024 {N}ode {S}witching {N}etwork",
  author= "Haas, S. and Thornley, D. A. and Zhu, M. and Dobinson, R. W. and Martin, B.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "274--281",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "To date, practical experience in constructing switching
     networks using IEEE 1355 technology has been confined to
     relatively small systems and there are no experimental
     results on how the performance of such systems will scale up
     to several hundred or even several thousand nodes. Some
     theoretical studies have been carried out for large networks
     of up to one thousand nodes for different topologies. We
     present results obtained on a large modular testbed using
     100 Mbits/s point to point DS links. One thousand nodes will
     be interconnected by a switching fabric based on the 32 way
     STC104 packet switch. The system has been designed and
     constructed in a modular way to allow a variety of different
     network topologies to be investigated (Clos, grid, torus,
     etc.). Network throughput and latency are being studied for
     various traffic conditions as a function of the topology and
     network size. Results obtained with the current 656 node
     setup are presented."
}
@InProceedings{Welch97,
  title = "{J}ava {T}hreads in {L}ight of occam/{CSP} ({T}utorial)",
  author= "Welch, Peter H.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "282--282",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "Java provides support for parallel computing through a model
     that is built into the language itself. However, the
     designers of Java chose to be fairly conservative and
     settled for the contepts of threads and monitors. Monitors
     were developed by Tony Hoare in the early 1970s as a
     structued way of using semaphores to control access to
     shared resources. Hoare moved away from this, in the late
     1970s, to develop the theory of Communicating Processes
     (CSP). One reason for this was that the semantics of
     monitors and threads are not WYSIWIG, so that designing
     robust parallel algorithms at this level is seriously hard.
     Fortunately, it is possible to introduce the CSP model into
     Java through sets of classes implemented on top of its
     monitor support. By restricting interaction between active
     Java objects to CSP synchronisation primitives, Jav thread
     semantics become compositional and systems with arbitrary
     levels of complexity become possible. Multi-threaded Web
     applets and distributed applications become simpler to
     design and implement, race hazards never occured,
     difficulties such as starvation, deadlock and livelock are
     easier to confront and overcome, and performance is no worse
     than that obtained from directly using the raw monitor
     primitives. The advantages of teaching parallelism in Java
     purely through the CSP class libraries will be discussed.
     (These libraries were developed jointly at Kent and Oxford
     Universities in the UK and the University of Twente in the
     Netherlands.)"
}
@InProceedings{Hilderink97,
  title = "{C}ommunicating {J}ava {T}hreads {R}eference {M}anual",
  author= "Hilderink, Gerald H.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "283--325",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "This document describes a csp package that contains the
     channel and composition classes in Java according to the CSP
     paradigm. The csp classes form a complete package that
     contains the necessary ingredients for concurrent
     programming with channels in Java. The channel and
     composition concepts are derived from CSp and are developed
     according to the object-oriented paradigm. There is a clear
     pattern of concerns by means of object-oriented techniques
     using inheritance, delegation, genericity , and
     polymorphism. This document is meant as a reference manual
     and gives background information about the realization of
     the csp classes. The use of the CSP channels in Java is
     illustrated by means of using building blocks."
}
@InProceedings{PatrickGreen97,
  title = "{A} {M}ultiprocessor {OCCAM} {D}evelopment {S}ystem for {UNIX} {N}etwork {C}lusters",
  author= "Patrick, D. G. and Green, P. R. and York, T. A.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "289--198",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "This paper describes the OCCNIX multiprocessor environment,
     that enables OCCAM program development and testing, on
     clusters of UNIX workstations. Linked binary level
     interpreters form a virtual Transputer network that uses the
     TCP/IP client-server model and provides hardware independent
     multiple platform access in a similar way to the recently
     released JAVA. Results show a single processor performance
     that is half that of an actual Transputer and a
     four-processor speedup of 0.78. The system also has the
     ability to run development tools such as the ISPY network
     debugger."
}
@InProceedings{MartinJassim97b,
  title = "{H}ow to {D}esign {D}eadlock-{F}ree {N}etworks {U}sing {CSP} and {V}erification {T}ools -- {A} {T}utorial {I}ntroduction",
  author= "Martin, Jeremy M. R. and Jassim, S. A.",
  editor= "Bakkers, Andr\`{e} W. P.",
  pages = "326--338",
  booktitle= "{P}roceedings of {W}o{TUG}-20: {P}arallel {P}rogramming and {J}ava",
  isbn= "90 5199 336 6",
  year= "1997",
  month= "mar",
  abstract= "The CSP language of C.A.R. Hoare originated as a blackboard
     mathematical notation for specifying and reasoning about
     parallel and distributed systems. More recently
     sophisticated tools have emerged which provide automated
     verification of CSP-specified systems. This has led to a
     tightening and standardisation of syntax. This paper
     outlines the syntax and semantics of CSp as it is now used
     and then describes how to design CSP networks, which are
     guaranteed to be free of deadlock, throught a succession of
     increasingly complex worked examples, making use of the
     verification tool Deadlock Checker."
}

If you have any comments on this database, including inaccuracies, requests to remove or add information, or suggestions for improvement, the WoTUG web team are happy to hear of them. We will do our best to resolve problems to everyone's satisfaction.

Copyright for the papers presented in this database normally resides with the authors; please contact them directly for more information. Addresses are normally presented in the full paper.

Pages © WoTUG, or the indicated author. All Rights Reserved.
Comments on these web pages should be addressed to: www at wotug.org

Valid HTML 4.01!