The Secure Open Desktop Project

(This page uses style sheets.)

The success of open software on servers has been well documented. Indeed much of the Internet depends on it eg GNU/Linux, Apache and Bind. However, many people feel that, amongst other things, much open software has a user interface too complex to allow its use on the desktop and therefore the Microsoft environment reigns almost unchallenged.

The public awareness of the need for secure computing systems is growing rapidly, Current security efforts, such as anti-virus software, suffer from the flawed assumption that adequate security can be provided in applications. In reality, the need for secure operating systems is essential. The USA National Security Agency's white paper 'The Inevitability of Failure: The Flawed Assumption of Security in Modern Computing Environments' should be read by all who are concernd with achieving secure systems that enable e-business. This is at www.nsa.gov/selinux/inevit-abs.html.

Also the USA National Security Agency has released to the public domain security updates to Linux to allow its use for secure USA government computing.

netproject is addressing these issues in undertaking consultancy and project development work for UK security organisations. The results of this work will be disseminated to netproject members in briefings, demonstrations and white papers.

A number of developments, particularly the recent Gnome and KDE desktops, OpenOffice from Sun (derived from StarOffice) and Gnome Evolution (an Outlook competitor), have addressed these criticisms. This now means that replacing the traditional Microsoft PC desktop with open software has become a reality.

Reacting to these developments, netproject has started the The Secure Open Desktop Project which is designed to:

1. Inform members of the latest advances in the use of open software on the desktop.
2. Gather members views and requirements and, on their behalf, lobby the software providers.
3. Provide a forum for discussion and an interchange of ideas.
4. Organise seminars at which speakers, pre-eminent in their field, will brief members.
5. Build and show demonstrators of how to deploy Open Sourcein organisations.
6. Address the issues of maintaining compatibility with existing Microsoft Windows applications.

The needs of members are likely to be diverse. However, we believe that there are many common problems and therefore it is very important to provide a forum for the exchange of ideas and comments. To this end we will maintain a web site which will not only contain information, case studies and links but also the facility to join in discussion groups. The scope of this project will be very much dictated by the members.

The word Desktop in the title is intended to included user access to systems in the broadest sense. This means that the project will not be limited to the traditional Microsoft PC desktop example but will include, for instance, PDA's and thin client replacement for dumb green screens. netproject realises that the range of user interfaces is very broad. Indeed, this is why open software can provide a compelling solution.

The software needed to provide a user interface can be divided into:

1. The operating system environment

The focus is likely to be on GNU/Linux, but there are other possibilities eg FreeBSD, OpenBSD and NetBSD. The commitment of companies like IBM to GNU/Linux is likely to mean that it will be the major recipient of development effort in the future.

2. The control of the interface itself eg a window manager

The user interface will depend very much on the intended use. A simple vt220 emulation may be sufficient or a full blown Gnome environment may be needed. In either case open software can be tailored appropriately.

3. A set of applications

This is perhaps the weakest area at the moment. Replacements for Word and Outlook are underway but standard applications such as payroll and general ledger are also needed. These would enable the use of open software in many organisations particularly SMEs.

The word Open is used to avoid difficulties with the semantic wars being waged over the distinction between Open Source and free software. Our intention is to consider software which provides at the very least:

• Access to all source code.
• The freedom to change the code and amalgamate it into the member's own code without significant constraint.
• The freedom to use the software without significant constraint.

Note that payment for software is allowed. These provisions, in our opinion, are the chief attraction of open software. Any possible savings on the initial cost of software would be an added benefit. However support costs of any software will, in general, soon dwarf the initial purchase price and open software is also likely to require this expensive support. On the other hand, the issue of repeating license fees or the constant need to upgrade, with the levy of additional fees, probably means that proprietary software will be significantly more expensive than open software.

Other reasons for choosing open software include:

• Security. Most of the software, particularly the operating system is built with security in mind. Because the source is freely available, security has to be achieved through proper design rather than obfuscation and secrecy. The use of a properly configured operating system will significantly reduce the vulnerability to viruses and other attacks.
• Because the source is available, code can be tailored to meet the specific needs of an organisation. It is this ability to tailor the GNU/Linux environment that allows it to be used in such diverse settings as onboard car computers, PDAs, embedded control devices, thin clients and full blown desktops.
• Choice. The flexibility provided by open software provides the discerning IT director with opportunities not available with proprietary software.
• Increasingly the ways users access systems are becoming more diverse and designing access through an open environment will mean that members are not being tied to the whims of proprietary systems.
• Many sites have a heterogeneous mix of servers and the need to use open rather than proprietary standards becomes more important. These standards are an integral part of open software.

The Open Desktop Project aims to provide an objective assessment of the pros and the cons of using open software. We believe that in some circumstances the case for use can already be easily made. For those organisations where this is not possible at the moment, we feel that they need to keep track of this rapidly changing area to enable them to make informed decisions. In all cases the Open Desktop Project at http://www.netproject.com/opendesktop/ will provide valuable information and discussion.