The MONADS Project

SPEEDOS is a successor project to the MONADS Project (see http://www.monads-security.org/). The MONADS Project was initially developed in the late 1970s and early 1980s under Keedy's supervision at Monash University in Melbourne, at that time one of Australia's leading universities for computer science. His three initial research students were John Rosenberg (https://en.wikipedia.org/wiki/John_Rosenberg_(academic) (who designed and implemented the kernel, David Abramson (jointly supervised by Prof Chris Wallace), who was responsible for the hardware (https://en.wikipedia.org/wiki/David_Abramson), and Kotagiri Ramamohanarao, who designed the job management aspects (https://www.wikidata.org/wiki/Q58684235). A key achievement of the Monads research team (after Keedy had taken up a Chair at the Technical University of Darmstadt in Germany in 1982), was the development of a successful capability-based computer system known as the Monads-PC, which incorporated the results of the Monads ideas developed between 1976 and 1982. The Monads-PC was a combination of novel hardware and software built by John Rosenberg and David Abramson.

Following Keedy's departure from Monash University in 1982 the Monads Project and related research continued at various universities.

In Darmstadt, where there was unfortunately no possibility of developing new hardware, Keedy was joined by his former Monash student Mark Evered, whose PhD describes a new programming language "Leibniz – A Language to Support Software Engineering". This was an important forerunner for the later SPEEDOS programming Language TIMOR. Keedy's first German PhD student was Bernd Freisleben, whose PhD "Mechanismen zur Synchronisation paralleler Prozesse" [1, 2] ["Mechanisms for the Synchronisation of Parallel Processes"] laid an excellent basis for new synchronisation mechanisms, which found their application in MONADS and later in SPEEDOS.

After returning to Australia to a Foundation Chair in Computer Science at the University of Newcastle, NSW, in 1985, Keedy was joined there by John Rosenberg. The MONADS work at Newcastle concentrated on two new ideas, the MONADS Massive Main Memory project (with Rosenberg and David Koch) and a successful local area networking project (with Rosenberg and Frans Henskens).

In 1988 Keedy accepted a Chair in Operating Systems at the University of Bremen in Germany, where the MONADS work continued, mainly with the assistance of Karen Vosseberg, whose PhD thesis was entitled "Sichere Ausführungsumgeben für Objekte" ["Secure Execution Environment for Objects"] and Peter Broessler, who wrote a PhD thesis entitled "Flexible und effiziente Unterstützung von Transaktionen auf persistente Objekten" ["Flexible and Efficient Support for Transactions on Persistent Objects"]. However during his time in Bremen Keedy began to concentrate more on the design of a programming language which would be suitable for writing application programs for MONADS (and later for SPEEDOS). He gathered an appropriate group around him for this purpose, including Mark Evered, Axel Schmolitzky and Gisela Menger, who made significant contributions to what was later called the TIMOR programming language. These also moved with him to the University of Ulm in 1993. In 1997 Keedy published a new programming language paper [3] which soon provided the inspiration for his solution to the confinement problem.