Acorn and Psion - a potential collaboration 1 Background Psion have recently taken the decision to use an ARM processor to power their next generation of hand-held computers, and have commenced the design and implementation of a new portable operating system to provide a suitable environment for application programs; this system is code-named ERA. Although ERA is an evolution of their existing system SIBO (for the Series 3 range), there are a number of major new components which they will either have to build themselves or buy in. RISC OS is a potential source of useful technology, and they are discussing with us possible ways in which we might be able to help them: one area of particular interest is our outline font technology. As part of the process of information exchange they have supplied us with a comprehensive overview of the ERA Software System, and we also gained some more information about their plans at a meeting attended by their Technical Director and three other senior software staff held at Acorn on February 1st. I understand that Acorn is currently looking at micro-kernels as a basis for future operating systems to run on our platforms, and that a number of contenders are under consideration (although the Psion micro-kernel is not currently a front runner). During the meeting it became apparent that Acorn might be interested in more than just Psion's micro-kernel, and that there was an opportunity here for joint development that would be of significant strategic benefit to us; this note summarises my view of this opportunity. 2 Characteristics of ERA Key characteristics of ERA are: - Modern OS designed around a micro-kernel which supports pre-emptive multi-tasking and multiple threads per task - Object-oriented API designed to support applications coded in C++ - Dynamically linked libraries - "Policy-free" window server process - Portable - Designed for ROM - Designed for low RAM requirement - Incorporates features relevant to portable computers (power saving) - Follows industry-standard interfaces (particularly Windows-like ones) wherever this does not conflict with other requirements ERA will be available both as a native OS on their target hardware, and also as a development system running under Windows: all application development will be undertaken on PCs, with the final port to target hardware made by a simple recompilation. Psion intend to make ERA available as an "open" system, with both source and binary licences available, and hope to establish ERA as a standard in the low cost portable market place. 3 Opportunity for Acorn These notes are based on the following assumptions: The lifetime of RISC OS is limited: a replacement is needed within, say, two years. The replacement does not need to provide direct support for running existing applications. (The transitional period can be handled by including RISC OS ROMs in product, or supplying RISC OS compatibility cards as optional extra processors.) We cannot afford to develop a new system on our own. From the information available so far, ERA seems likely to support most of the features that we would wish to provide in a future OS; this is, of course, true of other contenders, but there are a number of advantages which other options do not necessarily provide: Target hardware for ERA is ARM-powered. Psion is a local company with whom we already have a relationship. They appear eager to work with us, and by becoming involved now we will have a real opportunity to influence details of the design to meet our own requirements; indeed, they have stated their willingness to adapt to meet our needs. Psion's target hardware is "lean and mean" - something that is closer to Acorn's culture than, say, state-of-the-art PC workstations. This means that their OS is less likely to lose us our competitive edge (high functionality at low cost) than other PC-based solutions might do. An Acorn "personality" can be implemented as a layer on top of ERA's policy-free window server process. ERA's development system will run on PC workstations. If we add an emulation of the new Acorn GUI, this means that ISVs familiar with the Windows environment will be able to develop applications for Acorn platforms with ease. They will also be able to use all of the standard development tools available in the PC environment (at no cost to Acorn or Psion). Wherever appropriate, ERA's API will be "Windows-like" - again making it easier to encourage ISVs with PC software to port it to Acorn platforms. ERA will be available on both Psion and Acorn platforms (and, perhaps, others) - so there will be a larger market for ISVs producing software for ERA. There are obvious opportunities for Psion and Acorn together to provide a range of compatible solutions to Education (and other markets) starting with hand-held "pocket books" and rising through to networks of desktop machines. 4 The way forward Apart from commercial and contractual considerations, there must remain technical questions relating to Psion's detailed plans and how they hope to achieve them. The best way to obtain answers to these questions is to spend some time with the Psion development team itself, and an opportunity is now available to do just that: Psion have shown a keen interest in our font technology and have welcomed the idea that one or two of our programmers should work with them in London (although the commercial basis of such an arrangement has not been discussed). If Acorn considers that the Psion/ERA route is a serious option for our future, then we should both support the transfer of font technology and start talking to them about the wider issues. I strongly favour this approach and believe that here is a unique opportunity for us to move Acorn's software platform into the Nineties at a cost we can afford. If not, then there seems little to be gained from such an arrangement: we should, instead, be focussing our scarce resource on the options that are considered to be more realistic. xxxx xxxx 3rd February 1995