C O N T E N T S Introduction Summary International Directory Corporate Concerns Service Providers DSA Survey Contact Points Europe The Americas The Pacific Rim Public Access DUAs Glossary PARADISE Contact Points I N T R O D U C T I O N Piloting X.500 started in the research community in 1988. The end of this beginning is rapidly approaching, as the efforts of implementors and standards makers alike visibly bear fruit. Even with the coming completion of the 1992 standard, many technical issues still need to be resolved along the path to a fully-scaled, commercially-based global network. The success of PARADISE, the COSINE international Directory pilot, has been to show how X.500 can work in an operational environment. Increasingly, interest and awareness in the Directory is coming from large-scale governmental and commercial organisations with real needs to deploy distributed databases. Coupled with the gradual emergence of X.400 (88) conformant products, the decision to migrate to X.500 is growing in popularity. This report, the third in a series of four, shows that many of those who have been sitting on the fence are coming round to tackling the issues involved in interpreting the standard, the integration of X.500 into existing corporate infrastructures and, in particular, X.400 electronic mail systems. Companies, both large and small operating within network communities, now need to anticipate their requirements, understand their motivations, and budget for the expected commitment: the what, the why and the wherefore. And then, once the building blocks of the network Directory have been put in place, the dream of an electronic world map of people can become a reality. S U M M A R Y The current profile of organisations participating in the Directory pilot in Europe is 80% academic/research to 20% industrial/commercial. In North America, the corresponding figures are: 50% academic, 30% government agencies and 20% commercial. The intention of this report is to focus on the needs and requirements of commercial users, with the attempt to answer questions which would make their involvement in the pilot more attractive. To reflect the diversity of activity in X.500, this report contains several feature articles as well as a revised country status report. The first article looks at the need for directories from a corporate point of view, and is followed by an analysis of the position of the European service providers. Finally, following the DUA survey in the last report, there is a DSA survey which looks at 19 X.500 products. Since the last report six new countries are participating in the pilot: Brazil, Croatia, Czechoslovakia, Greece, India and Slovenia. In addition DG XIII of the European Commission recently joined the pilot with a QUIPU DSA. Perhaps more signifcantly, the project has been focussing since the beginning of the year on expanding the pilot into real multi-vendor interworking. In addition to QUIPU, PIZARRO (UCOM.X) and DirWiz which have been in the pilot for the last year, interworking discussions have started with Siemens Nixdorf, who are participating in pilots in Germany, Ireland and the Netherlands, Bull in France, and ICL in the United Kingdom. Expressions of interest for future collaboration have also come from, inter alia, Digital, Marben, Retix and Telia. PARADISE is providing a near- conformance testing service through the PTT Research Laboratories in Groningen, the Netherlands, as well as simple interworking testing against QUIPU. This activity is a major area for expansion and consolidation in the coming year. There have been advances in getting the Directory to serve end-users: at the back of this report there is a list of public access interfaces in 11 countries which are intended to help local users to gain first time access to the Directory and in some cases in their native language. In addition PARADISE provides two public access services centrally from ULCC: o Directory Enquiries to enable enquiries and browsing through the Directory (de); o the Directory Manager to allow users to open and then manage accounts in the Directory (dm). Usage of de has increased from an average of 400 calls per month at the turn of the year to well over 1400 in May - despite the proliferation of nationally-based services. The dm tool will go live at the beginning of July, and it is expected that it will also be run extensively both at the national and local level. The urgent need for user-friendly and practical interfaces is gaining momentum with an increasing recognition that X.500 software has to be portable on to users' desktop platforms before the Directory is going to make any serious headway into corporate thinking, be it a university or an office block. There are already attractive interfaces for the Macintosh, but there is an equal if not greater need for PC- based DUAs. An example of one such interface from Brunel University is shown on page .. . The PARADISE pilot is an umbrella project with a remit to co-ordinate and give focus to pilot X.500 activities in Europe. As such it is totally dependent on the national pilots of its constituent members who in many cases are underfunded (or not funded at all), and contribute their efforts voluntarily. In recognition of this fundamental stumbling block to both growth and quality of service, the European Commission are funding under its VALUE programme an initiative to give support to national X.500 programmes. Although this only applies to the member states of the European Community, it will considerably bolster the profile of X.500 piloting across the whole continent. The thrust of the initiative is on improving or creating infrastructure with a view to establishing a sound base for good quality services. This tranche of funding follows closely on a similar initiative which was targetted at establishing basic X.25 (IXI) connectivity and X.400 services. The VALUE programme is committing 1 MECU in 1992 to X.500, though more funds may be allocated in 1993. The VALUE programme is also funding a project to look at security services in OSI applications which is creating a lot of interest amongst potential users of the Directory. PASSWORD (Piloting Authentication and Security Services Within OSI for the R&D community) consists of three national consortia (France, Germany and the United Kingdom), each with their own security toolkits and suite of OSI applications (X.400 (88), X.500, ODA and FTAM). The project hopes to pilot these secure applications during the course of 1993 using the X.500 Directory to store public key certificates. It is also expected that the Y-Net project will commit itself to an X.500 pilot in the coming year to run alongside its existing X.400 service. A decision, to be taken by its manufacturer consortium (Siemens Nixdorf, Olivetti and Bull), is expected later this summer. The present PARADISE project is scheduled to finish at the end of 1992. However, there is a recognition that some central co-ordination of the X.500 Directory will be necessary for probably another two to three years, after which time bi-lateral and multi-lateral agreements to share Directory knowledge may replace the need to have a real, rather than a virtual, root of the Directory tree. If PARADISE were to continue into 1993, it would need to really tackle head on the varying interpretations of the standard; but in a way that was realistic both for vendors and service providers, allowing also for an evolutionary consensus based on real operational experience. The research community will also have to recognise the eventual emergence of multiple service providers both at the national and supranational levels, whether it be Y-Net or an association of PTTs organised along the lines of the North America Directory Forum (NADF). However this comes about, there will clearly be long term mutual benefit from both dialogue and technical co- ordination. I N T E R N A T I O N A L D I R E C T O R Y EUROPE DSAs Orgs Entries Europe 3 4 2,000 Austria 3 18 1,516 Belgium 2 3 64 Croatia 1 1 10 Czechoslovakia 1 1 10 Denmark 3 340 1,491 Finland 18 19 75,500 France 20 18 5,000 Germany 30 135 7,269 Greece 1 15 100 Iceland 1 30 720 Ireland 3 7 2,100 Israel 1 1 10 Italy 3 5 2,529 Netherlands 5 101 2,128 Norway 11 738 23,346 Portugal 3 2 1,295 Slovenia 1 1 10 Spain 6 59 4,538 Sweden 8 43 20,118 Switzerland 16 15 32,226 United Kingdom 48 53 57,991 --------------------------- TOTAL 188 1608 239,071 THE WORLD DSAs Orgs Entries Australia 25 107 44,351 Brazil 1 1 10 Canada 17 11 44,286 India 10 8 695 Japan 15 15 2,300 New Zealand 1 2 1,510 United States 108 106 468,634 --------------------------- TOTAL 177 247 561,786 GRAND TOTAL 365 1856 801,857 C O R P O R A T E C O N C E R N S There are no global telephone directories. There are no global postal directories. There are no generally available organisational directories. What the X.500 technology provides is the potential to create the infrastructure for all those things to happen and more. As the next few pages illustrate, the X.500 Directory has the multi-media potential which would allow a businessman in Europe to look up a supplier in Japan, identifiable through their company logo, to visually identify the product, listen to a 30 second audio description, and then find the name and telephone number of the appropriate salesman, who would be recognisable at any future meeting from his Directory photograph. If the businessman preferred to contact the salesman via electronic mail, he could transparently transfer the mail address into the header box of his mail system, and the message could be on the other side of the world within minutes. Fantasy? The X.500 technology exists and is being used, albeit not as extensively or as creatively as described above. The network infrastructure is in place, though predominantly across research networks. But, as the demand for commercial non-R&D traffic increases, there will be a breaking down of the barriers between the research and commercial communities together with a significant increase in the capacity of existing lines. Despite this it is widely predicted that X.500 will not take off in the corporate world for another two to three years. In examining the reasons for the sluggishness of this potentially enormous market, it is necessary to look more closely at some of the non-technical issues raised by X.500 - which could be described as strategic, cultural and political - and examine how they are being dealt with. WHY BOTHER? The first set of questions any businessman is going to raise are along the lines of "why should I bother?": o what does the technology involve in terms of resources: how much is this exercise going to cost me? o what are other people doing about this: are there better, cheaper solutions you're not telling me about? o how will this improve productivity, efficiency, market visibility: what's in it for me? o what kind of backing exists in terms of product support, network support and training: does anybody else know about this stuff? o why are you telling me about this: what's in it for you? STANDARDS The informed opinion is that X.500 technology is still too naive for the corporate market. The reasons given for this are that both users and vendors alike are still grappling with X.400, and coming to terms with the 1988 standard. X.500 was originally conceived as an extension to X.400, and as a tool for other OSI applications such as electronic mail and file transfer. The repercussion of the significant changes between the 1984 and 1988 standards in X.400 can be seen in the number of products that still have not migrated to the later standard. With X.500, the standards' makers recognised that there was insufficient time to incorporate certain vital functions, including access control and replication, into the 1988 standard and so postponed their inclusion until the 1992 revision. This to some extent has inhibited implementors from over- commitment to product development or pushing operational deployment. Taken logically, and on the basis of the X.400 model, this would suggest that there would be no real X.500 penetration of the market until 1998 or beyond. However, there is a real need and keen interest in a distributed directory that has prompted the research community to examine whether X.500 will work in a large-scale operational environment. The answer that PARADISE gives is that it does. COMMERCIAL REQUIREMENTS The primary role of the X.500 Directory, as perceived by many non-initiates, is to provide a repository for electronic mail addresses, and possibly telephone and fax numbers as well. For a large corporate manager, the initial requirement is internal: to map the topology of the organisation on to a database structure, and to ensure that the right information is accessible to the right people within the organisation. An SME, on the other hand, has a much greater interest in publicising itself to as wide an external target audience as possible. For a corporation in which the majority of electronic telecommunication is done internally, making company information available externally is a secondary issue; indeed it raises the unattractive spectre of breaches of security and commercial confidentiality. A lot of companies get very concerned at the prospect of revealing their organisational structure to rivals, identifying their experts for fear of poaching, or even admitting the existence of certain activities, such as an X.500 Development Team, which could reveal company strategy. Most X.500 implementations currently have simple authentication and access control; access control allows authenticated access to objects and attributes in the Directory so that departments, individuals and specific data about individuals can have restricted visibility. A practical starting model for an organisation with such apprehensions would be to restrict public access to information concerning sales, marketing and accounts - though these departments are not likely to contain the electronic mail addresses sought after by a research worker! HETEROGENEITY In a large heterogeneous working environment, there is likely to be a plethora of both hardware platforms and software products for a multitude of different tasks. The two that affect decisions about X.500 are electronic mail and corporate/enterprise directories. Given that procurement executives start off trying to make life easier rather than more complicated, the questions raised by X.500 might be: o can our four in-house mail systems (All-in-One, cc:Mail etc) use the same X.500 system to share each others' directory information? o will X.500 make it easier to send electronic mail? o what will happen to the Oracle database we installed two years ago if we adopt X.500? Most electronic mail products have their own "native" directory system which enables users to point to other users on the same local network for addressing. As these directories are not yet X.500 they do not allow users from one system to interrogate the directory belonging to another, thus inhibiting communication. However, a number of vendors have adopted a directory clearing house approach as an interim solution to this problem. This involves a client-server protocol to flag changes to remote databases and then update a central directory. A user on one system can then interrogate this central utility about data on a different system. This is not an X.500 approach, but is effective in the short term. The eventual X.500 solution will be to involve all the electronic mail systems to have their own "native" X.500 directories which then can talk X.500 protocol to each other as well as to the outside world. This integration of X.400 and X.500 is seen as crucial in getting both standards, and the products which represent them, off the ground. Most organisations already feel plagued with too many electronic databases, and recoil in fear at the thought of another one. Over the last five years, large relational databases such as Oracle, Ingres, Informix etc have attracted great attention from those who have sought to integrate everything there is to know about their company into a single piece of software. For them, there is a requirement not to abandon the work they've already carried out, but to find a suitable X.500 product which will extract subsets of data from the corporate database for the X.500 Directory. SMEs For small-to-medium size enterprises (SMEs), particularly those which do not have the resources or the networks staff to run an X.500 Directory, other solutions need to be found. The options rely upon having access to a remote DSA to hold organisational data which is registered in the appropriate place in the Directory tree. The data can then be managed remotely by: o the installation of a local DUA; o access to a remote DUA; o distributing and updating the data by non-network protocols ie post, fax etc To facilitate the participation of SMEs, PARADISE has developed a remote management tool which is accessible over any terminal type on public X.25 and Internet lines. The interface enables a user to add a new organisation into the Directory, and then to modify individual entries. For further details of this service, known as "dm", see page 23. CONCLUSIONS For the corporate user the advantages of the electronic OSI Directory are threefold: o it can solve the problems associated with finding MHS addresses, both within an organisation or in the wider world; o eventually it can replace both the internal telephone book, and the external telephone book for white and yellow page services; o it can be used as a tool for marketing and sales, which in conjunction with the rapidly growing interest in EDI (Electronic Data Interchange), could transform the way that business transactions are carried out before the end of the century. The following features in this report seek to address two of the questions the businessman was asking earlier: o who is going to be providing services? o what products can I buy? S E R V I C E P R O V I D E R S A critical requirement in the original tender for the PARADISE contract was that there should be at least one major service provider organisation involved. The rationale for this was that, however well the international pilot proceeded, eventually X.500-based Directory services would be offered by the PTOs. The R&D community which had been responsible for developing and piloting X.500 technology is also one of the largest potential user groups. To ensure that as smooth a transition as possible should be effected, it was hoped that service provider participation in such a pilot liaison role would provide the catalyst to bilateral recognition of perspectives and requirements between the two communities. There are three companies involved: o PTT Telecom, the Netherlands o PTT Telecom, Switzerland o Telecom Finland All three are involved in the ETSI (European Telecommunications Standards Institute) group on the Directory, which, together with a similar EWOS (European Workshop on Open Systems) group have worked on specifications of the protocol and distributed operation capabilities of the Directory so that interworking between systems can occur. These documents have helped define the Directory A-profiles and Q-profiles for Europe. EXISTING DIRECTORIES At present many of the PTOs in Europe offer an electronic telephone book for national phone numbers, as well as electronic access to the directories of other PTO telecommunications services. Most use the TPH28 protocol to electronically connect directory assistance operators to the directory systems of other national PTOs (within Europe only) in order to obtain telephone numbers. TPH28 is fairly simple by comparison with X.500, and is well-established for operator-based enquiries. However, in September 1991, it was decided by the CEPT 51 Group (Conference Europeene des administrations des Postes et des Telecommunications) that enhancements to the protocol would be discontinued in favour of concentrating on applying X.500 for electronic telephone directory enquiry systems. Few disagree that there is an urgent need for directory services, and that in principle, X.500 offers the standards and the technology to make these services available. And yet, from the perspective of very large scale services, the present X.500 standards are inadequate. THE SURVEY In the context of this background, the troika of PTOs in PARADISE sought to evaluate the position of European service providers with regard to X.500. This investigation was carried out by means of a large questionnaire with some follow-up interviews. During the course of 1991, over 20 service providers were approached, most of whom responded. The questionnaire addressed a wide range of X.500 issues which included: o technical aspects o commercial feasibility o acceptance of the standards o experience with products o the relation of X.500 to TPH28 The processing of the responses was directed to clarify the image of X.500, and to analyse what should be done in order to accelerate the introduction and the acceptance of X.500. In general, the conclusions were: o although European service providers are very interested in X.500 developments, only some had developed some degree of technical competence: the majority had not. Consequently, there was a lot of confusion about the quality and feasibility of the provision of X.500 services; o X.500 was accepted as the standard for directories. The 1992 version was considered to offer more than the 1988, particularly for access control and distributed entries. The stability of the standard was considered to be a major problem. Many wanted to delay action until the standard is more stable; however, there was a common belief that an X.500 service will need to be started in the future; o experience with implementations was very limited: only five products were mentioned more than once in response to a question on awareness; o most did not see a market demand for X.500 services on their own, but only as a support for other services; nor did they believe that potential users are educated about X.500; o X.500 was primarily seen as an added extra to a base service like X.400. There appeared to be little faith that X.500 could be operated as a stand-alone commercial service. The lack of charging and accounting mechanisms was mentioned several times as a problem preventing the potential operation of a commercial service. o many suggested that a service group of future operators was necessary for Europe along similar lines to the NADF in North America; The overall picture was of a concensus of slowly but steadily increasing interest concerned about commercial feasibility, raising user awareness, the stability of the standard as well as the need for a broader product base. What is being done and what can be done in the future to take this position a few steps further? THE WAY AHEAD The much-sought after transition from the TPH28 protocol is proceeding with the specification of an experimental protocol, TPH500, which is intended to bridge the gap to full X.500 migration for directory access. However, the telephone enquiry system is just one small part, albeit the first, in PTO plans to deploy X.500 in order to provide users of their other telecommunications services, for example Videotex and X.400, with access to a global information base. In addition, quite apart from the survey which was directed towards the value-added service departments of the PTOs, many of the research departments of these same companies have become actively involved in national pilots participating in PARADISE with a view to gaining operational X.500 experience: o British Telecom (Martlesham) are connecting their QUIPU DSA to the UK pilot and are also involved in the NADF pilot; o Telecom Finland is running a QUIPU DSA allowing access to its 60,000 entry mailbox database, and have also been field-testing HP X.500; o France Telecom, who have collaborated with CAP SESA on a TPH28/X.500 gateway, are active partners in the mixed academic and industrial pilot, OPAX; o PTT Research, the Netherlands are providing diagnostic and near-conformance test services as part of the PARADISE project; o the Swiss PTT is running a QUIPU DSA and providing its internal telephone directory comprising 470 departments and 9727 entries; o Telecom Eireann (Ireland), Jutland Telephone (Denmark), Belgacom (Belgium) and Postes & Telecommunications (Luxembourg) all hope to have some involvement in the forthcoming CEC-funded VALUE national X.500 projects. Outside the EC, the Croatian and Slovenian PTOs are working together with their respective research pilots; o three companies, Telecom Finland, Televerket (Sweden) and British Telecom have all carried out extensive development work on proprietary X.500 (or X.500-like) implementations, whereas the Norwegian Telecom, Telecom Iceland and two public service providers in Portugal, Sevatel and Marconi, are experimenting with the ICL Carrier product. THE FUTURE? The formation of the NADF (North America Directory Forum) in 1990 has generated a lot of interest in Europe. The NADF is an open collection of organisations which offer, or plan to offer, public directory services based on X.500 in North America. The group has defined service and schema definitions, as well as software tools, with the goal of implementing inter-connected public directory services, starting in 1992. It is increasingly felt that there should be an EDF (European Directory Forum), which would be a peer association to the NADF, and would seek to achieve similar goals through the PTOs in Europe. Such a group would rely on technical input from ETSI/EWOS and the experience of the PARADISE pilot, and would initially need to reach agreements on technical, operational and commercial interest among its members. There could also be a close relationship with any Eurescom (European Institute for Research and Strategy in Communications) projects. Based in Heidelberg, Eurescom is just over a year old, and at the moment is owned by 23 public netwrok organisations, though the number is increasing. An EDF is seen as an ideal building block in establishing public X.500-based directory services by the mid-1990s. As a follow-up to their PTO survey, the three companies are preparing a similar study on how X.500 is viewed by several large, multi-national companies, the results of which will be available in the next report. Austria Radio Austria Telecommunications Denmark Telecom Denmark Finland Telecom Finland Telematics France France Telecom Transpac Germany Deutsche Bundespost Telekom Ireland Telecom Eireann Italy Italcable SIP Teleo Netherlands PTT Telecom Norway Norwegian Telecom Portugal Sevatel Spain Telefonica Services Sweden Televerket Switzerland PTT Telecom UK BT The European service providers who participated in the PARADISE PTT Survey D S A S U R V E Y One of the stated goals of the COSINE project is "to create a market-pull for OSI products". By implication one of the goals of the PARADISE project is to create a market-pull for X.500 products. To date, this task has been severely hampered by the lack of available implementations, even at the stage of pre-production or field testing. As 1992 progresses the situation is markedly improving, and in this overview of current product releases we will be outlining either what is currently available or what is expected to be on the market before the end of the year. In addition we will be discussing the problems of interoperability, and what solutions are currently available, including those provided through PARADISE. The survey of X.500 products is not exhaustive, nor is it meant to be: the intention is to give some indication of vendor activity. It also does not attempt to provide product evaluations: the market is too new for an evaluatory report to make sense at this stage. Four categories of X.500 software vendors can be identified through the survey: o public domain software (QUIPU); o software houses who specialise in communications software and OSI-based consultancy and related services (E3X, Retix, SoftSwitch, X-Tel); o equipment manufacturers who supply office automation systems, and include an X.500 component as part of an integrated OSI solution for proprietary platforms (Digital, HP, ICL, SNI); o service providers, who fall into two groups: o those who have developed X.500 software for their own internal use with possible plans for commercialisation (Telia, VTT) ; o those who have targetted specific markets and have developed products around a marketing strategy (BT, GEIS) Currently in the PARADISE pilot, the profile of implementations in use is roughly as follows: o QUIPU (Public Domain) 350 o UCOM.X 500 (E3X) 20 o DIR.X (SNI) 1 o OfficeTeam (Bull) 1 o DirWiz (Sytems Wizards) 1 The PARADISE pilot has been labelled QUIPUcentric: it is up to vendors and users to come forward with non- QUIPU implementations to rectify this imbalance. The survey will be kept up to date and will be available from the PARADISE info-server. In addition there are other on-going reports on X.500 products, most notably from: o DISI, the Ruth Lang/Russ Wright "Catalog of Available X.500 Implementations" which has progressed to RFC status (RFC 1292); o Technology Appraisals, who produce a regular report "OSI Products" which includes a summary table of X.500 products as well as profiles of most leading software vendors. 'TWIXT THE CAMEL AND A HARD PLACE According to General H Norman Schwarzkopf (Ret), who gave the keynote address at the Spring Interop in Washington DC this year, interoperability in computer networking is solely "a management issue." Unfortunately that doesn't seem to be as true in peace time as it was during Desert Storm. It is often believed that if a product passes standards' conformance tests, it will necessarily interoperate in a network environment with any other product that passes similar tests. Unfortunately, this is rarely the case: comparing conformance test results with true interoperability is like comparing chalk and cheese. Testing can be carried out in a number of different ways and interpretations of standards vary considerably. Using products in a real, multi-vendor, multi-protocol network - running heavy traffic and subject to many random events - will be very different to running the same product in a test laboratory. And, critically, the interoperability requirements that really matter to users may not be those checked in conformance tests. To overcome these discrepancies requires both vendors and users to devote considerable time and resources to understanding the issues. The slight incompatibilities in the implementations of the X.500 standard are well-documented. However, irrespective of whether these differences are large or small, the result is some breakdown in the transmission of information. To alleviate these problems, conformance testing and promotion organisations have appeared. Notably, these are: o COS (Corporation for Open Systems International) in the United States; o SPAG (Standards Promotion and Application Group) in Europe; o POSI (Promoting Conference for OSI) in Japan. These bodies were formed to develop uniform interpretations of standards and, most importantly, to test and certify vendor products for compliance with these interpretations. However, no two conformance-test organisations take the same approach. There are no standards for testing standards; each has adopted a set of protocol profiles and testing methodologies that reflect local biases. However, COS, SPAG and POSI are being attracted to the idea of collabotrating on standard profiles and testing procedures for OSI products. But even if profiles from conformance test houses do match, users still can not be certain that products will truly interoperate because these tests will not reflect the real-world environment in which a product will operate. There are three levels of interoperability testing: o theoretical conformance, when the vendor's technical design document is compared to the definitions of the X.500 standard; o test sequences that check the product operates in accordance with the standard. Each function is tested in isolation using specialised equipment. Obviously these tools and the tests will vary between one test house and another; o "bilateral" interoperability. This checks that a product will interoperate with another vendor's product. These tests may simply evaluate isolated protocols, but often products that operate adequately in a single protocol environment fail when multiple protocols are used. "Full-suite" testing ains to overcome this, and is now provided over major conformance test networks such as EurOSInet in Europe, OSIcom in Australia and OSInet in the United States. EurOSInet is the present forum within Europe for vendors to establish interworking testing for their products. It meets every month to agree testing procedures and methodologies. The principle of establishing test DSAs with on-going availability for limited interworking has the value of allowing vendors to test latest releases against other vendors' upgrades, and is one which is being considered for future PARADISE interoperability planning. The way the PARADISE project is trying currently to address these subtleties of interpretation are by looking at the problem across the three levels described above: o near-conformance testing through the PTT Research Laboratory in Groningen who were involved in the CTS2 project on equipment developed by Danet. The tests are aimed mainly and in-depth at testing for the correct use of DAP and Directory Service. DSP has not been covered yet, due to the complex matter of configuring DSAs for testing; o practical interworking testing using a simple script, which has been the basis of the first interworking report and has focussed on simple testing of non-QUIPU DSAs from a QUIPU DUA. There is clearly a need to diversify this activity, and it is hoped that future reports will cover multi-lateral testing. o metrics' proformas which establish a broad, operationally-biased set of criteria in which products can be assessed. METRICS PARADISE has produced a set of metrics for the pilot to define some guidelines by which DSA (and DUA) products can be measured. Although an X.500 DSA may conform to specifications in the standard, protocol conformance is not in itself the hallmark of a usable implementation. A DSA ought to be able to: o perform operations within a reasonable time; o offer good throughput of queries; o handle a reasonable volume of data; o provide some sort of access control if modification operations are provided; o finally, a DSA and its data must be manageable. Whilst it is difficult, and not particularly meaningful, to compare one DSA with another, it is possible to pose a series of questions which will allow a user to determine whether a particular X.500 implementation is suitable for their requirements. One of the final sections contains questions relating to replication and access control, which are to be described in the 1992 standard, but have already been implemented in many 1988-based products. The metrics documents for both DSAs and DUAs will evolve when there is greater familiarity with more products and more cross-fertilisation of ideas between developers. All the PARADISE documentation described here form part of the PARADISE Interoperability Pack (PIP), together with the following: o RFC 1274: The COSINE and Internet X.500 Schema (Paul Barker & Steve Kille) o RFC 1275: Replication Requirements to provide an Internet Directory using X.500 (Steve Kille) o RFC 1276: Replication and Distributed Operations extensions to provide an Internet Directory using X.500 (Steve Kille) o RFC 1277: Encoding Network Addresses to support operation over non-OSI lower layers (Steve Kille) These documents are all available from the info- server. For further information, please contact the Helpdesk. ALCATEL The TWICE X.500 product comprises the fully conformant U-DSA and U-DUA modules, both of which use the U-ASN1 compiler. U-DSA communicates using three communication modes: referral, chaining and multicasting. U-DUA is a full screen user interface on a standard terminal, and offers user-friendly access to the Directory query functions through the use of pull-down menus and dialogue windows. The global results of a read, list or search request can be saved in a file as can the X.400 attributes of an object in the notebook used by U.UA, the X.400 messaging interface. U-DUA can also save/restore the parameters for a request (and use pre-stored requests). Three specific U-DSA attributes are used to manage information relative to the naming context (knowledge information for each DSA): o Know-mastered: list of the subtrees for which the DSA is directly responsible o Know-copy: list of the subtrees for which the DSA manages a local copy o Know-shared: list of the subtrees the responsibility of which is shared with other DSAs In its current version, the DSA provides a simple authentication service without protection, but with encryption of passwords in the Directory. A powerful authentication service will be included in the next release with the following features: o generation of both public and private key pairs through the certification authority o use of the RSA algorithm o software service (with no use of specific hardware). The public and private keys are kept by the certification authority in a file encrypted in accordance with the RSA algorithm. Alcatel plan to enable interaction between the DSA/DUA and the MTA/MTS of TWICE X.400 Product: TWICE X.500 Contact: Philippe Andrieu Alcatel TITN Answare 1, Rue Galvani - BP 110 91301 Massy Cedex, France Telephone: +33 1 69 81 16 40 Fax: +33 1 69 20 78 62 BT Cohort 500 was planned from the outset as an X.500 Corporate Directory System based on BT's analysis of the the requirements of the corporate market. At the time it was considered that having full X.500 functionality was not the most significant priority. BT intends that future releases of Cohort will conform to X.500. The requirement identified was for a system which could supply a single consistent source of directory information for various applications throughout an organisation, such as PABX operators/call reception and management information systems. Cohort can be used to produce high-quality, easily customised printed directories and reports: output is via the Xerox Ventura DTP package. A major component of many Cohort systems is the Cohort POD - a dedicated screen-based directory for switchboard operators. An IBM (or compatible) PC-based application, the POD gives exceptionally fast access to essential contact details, helping to improve efficiency and speed up call handling. A number of PODs can be linked together so that several operators can share the same information. Depending on system configuration, data can be downloaded from Cohort for local management. To achieve these objectives, Cohort 500 contains a family of inter-related products and services, the main component of which is the Corporate system, based on the ORACLE relational database. The Corporate system has an X.500-conformant data structure and is capable of linking via an SQL interface to other relational databases capable of accepting SQL. Eventual X.500 support for X.400 and FTAM is hoped for. BT's Directory Products Unit has also developed a CD/ROM-based product, Phonedisc, which contains 17 million telephone directory entries with a search time of seven seconds. Product: Cohort 500 Contact: BT, Columbia Centre Market Street, Bracknell RG12 1JG, UK Telephone: +44 344 861961 Fax: +44 344 860872 BULL Bull adopted the Siemens Nixdorf DIR.X implementation in order to incorporate it into their OfficeTeam product with their proprietary mail system. Bull participate in the OPAX pilot group in France, and are performing interworking tests with U.COM X. There has been interest from Bull UK to do the same with QUIPU. As with SNI, Bull are expected to wish to promote their product through the Y-Net project. Product: OfficeTeam Contact: Gilles Bogo 7 rue Ampere 91343 Massy Cedex, France Telephone: +33 1 69 93 76 39 CONTROL DATA Since early 1991 Control Data have been shipping their X.500 product which is based on QUIPU. It is an integral part of the Mail*Hub, a product suite which ties together a variety of email protocols, allowing users to exchange mail freely between diverse mail environments such as DECnet Mail 11 and All-in-1, IBM PROFS/SNA, X.400, SMTP, UUCP et al. X.500 Directory Services are utilised throughout the design wherever a directory service is needed eg for configuration, routing and address mapping information. The Control Data client base covers the spectrum of government agencies, universities and private industry, and is international in scope. Product: Mail*Hub Contact: Kevin Jordan Control Data Systems 4201 Lexington Avenue North Arden Hills, MN, USA Telephone: +1 612 482 6835 Email: kej@vesta.udev.cdc.com DIGITAL Digital have been field testing their X.500 product recently in Switzerland and elsewhere, and hope to formally announce a release date towards the end of the summer. It will be sold both as a stand alone product, and as part of Digital's messaging package, to facilitate directory look- up and routing. Product: Digital X500 Directory Services Contact: Nick Tatham, Development Manager DEC Ltd, PO Box 121 Reading RG2 0TU Telephone: +44 734 868711 Email: tatham@forty2.enet.dec.com E3X This product originated as a development implementation called PIZARRO which came out of work done at INRIA under the ESPRIT THORN project. E3X, a software company with an exploitation arrangement with INRIA, integrated PIZARRO into their product range, along with M.Plus, the INRIA X.400 MHS, at the beginning of 1990. As a member of ARISTOTE, the Department of Industry research network, E3X participates in OPAX, the French branch of PARADISE, which extensively uses UCOM.X 500 DSAs. UCOM.X 500 is made up of four parts: o U-DSA o U-DUA o API500, the X.500 API built upon DAP, which may be used to develop applications modules that make directory requests o ADM500, administrative tools that offer local base management. The DSA which offers both DAP and DSP, makes no use of an external DBMS, though E3X plan to allow such interactions by making SQL queries. The database is implemented through a unique file containing all the entries. When the DSA is started, it reads this file and builds the DIT in memory directly from the contents of the database. The DIB is kept on disk, so that when an incoming request has to be resolved, disk access is made only after the name resolution which takes place completely in memory. However, as disk access is expensive in terms of performance for search operations, the system is configurable through inverted tables to allow direct access to entries. At the beginning of 1991, the CHR de Lille, one of the major French hospitals started an unusual X.400/X.500 project using UCOM.X in order to provide a formulary exchange in a health care environment based on electronic mail where the recipients of a message are deduced from: o the identity of the originator o the kind of formulary The goal of the project is to eliminate wastage through the creation of several directory maps providing contiguous schemas of the hospital. Another innovative project using X.500 was undertaken by La Poste in March 1991 to store information about its new X.25 network. The aim of the project was to be able to retrieve network addresses and network types associated with any specific plug in a building owned by La Poste Product: UCOM.X 500 Contact: Philippe Brun E3X, Immeuble Les Algorithmes Batiment Pythagore Entree A Route des Lucioles Sophia Antipolis 06560 Valbonne, France Telephone: +33 93 65 34 65 Email: phb@osi.e3x.fr HEWLETT PACKARD HP X.500 can be used for accessing names and email addresses for multi-vendor messaging backbone networks. It can also be used for the development of networked applications requiring distributed directory functionality. HP OpenMail users will be able to access the enterprise- wide HP X.500 distributed directory directly from the terminal interface. X.500 addresses can be selected and automatically returned to OpenMail, simplifying the process of mailing multi-vendor electronic mail. X/Open X.500 APIs (XDS and XOM) are included with the product as well as high level APIs on top of XDS which simplify and speed application development. HP X.500 supports access control and replication as well as all X.400 (88) and X.500 object classes. The product, which will be commercially released in June 1992, has been extensively field tested at Shell (the Netherlands), the Technical University of Berlin and Telecom Finland. Product: HP X.500 Contact: Pascal Boudalier Hewlett-Packard France 5, avenue Raymond Chanas - Eybens 38053 Grenoble Cedex 9, France Telephone: +33 7662 5738 Fax: +33 7662 5323 X.400: c=FR,a=ATLAS,p=HP,o=HP,OU1=HP6330, s=Boudalier,g=Pascal ICL ICL developed some of their expertise in distributed directories through the ESPRIT THORN project. Their X.500 product is an integral part of the Carrier400 suite, although it can be sold as a separate entity running on UNIX 5.4. From the 1992 recommendations, both improved authentication and shadowing are being added. In addition, access to an external database is being incorporated allowing information to be transferred to and from the DSA using SQL protocols. CarrierUA (the systems user agent) is a UNIX-based package which can support up to 50 terminals simultaneously, and when using dial-up connection many times that number of low traffic users. At Cebit two years ago, ICl interworked with products from Siemens, Nixdorf, IBM and Retix, and at Open Systems in November 1990 with OSIWARE in Vancouver. ICL has CarrierDSC deployed with a number of large users such as the South African, Norwegian and Icelandic Telecoms, as well as Sevatel and Marconi who are service providers in Portugal. They have a commitment to provide MCI with a DSA capable of a database holding at least 120,000 entries. It is hoped that AT&T will participate in the NADF pilot using CarrierDSC. The Department of Social Security in the UK have also installed a version, along with QUIPU and other implementations for evaluation. In 1990, Carrier was part of an evaluation carried out by Logica on behalf of the European Commission's DG IX (the Informatics Directorate) who have subsequently decided to go ahead with an ICL-based pilot for document distribution. Product: CarrierDSC Contact: Glyn Wild ICL, Six Hills House London Road, Stevenage SG1 1YB, UK Telephone: +44 438 726161 x8063 X.400: c=GB,a=GOLD 400,p=ICL,o=ICL,ou=STE0902, s=Wild,i=GA ISODE CONSORTIUM The recently-formed ISODE Consortium, which includes representation from North American and European government agencies, industry and research networks, will be making their first release based on the existing public domain ISODE software available under licence at the end of 1992. Future plans for the development of QUIPU include; o a new database API to allow multiple databases within a QUIPU DSA o accommodation of the X.500 (92) recommendations; o inclusion of lightweight protocols, suitable for implementation on smaller machines; o facilitiation of multi-ADDMD operation; o addition of X.509 (public key) based authentications; It is expected to have full integration with PP, the ISODE X.400 application, towards the end of 1992. Product: QUIPU Contact: Steve Hardcastle-Kille ISODE Consortium PO Box 505, London SW11 1DX Telephone: +44 71 223 4062 Email: s.kille@isode.com MARBEN The Marben X.500 product is one of a family of OSI products called Osilogie and is based on the Marben core product OSIAM (OSI Access Method) technology for implementing communications protocols. It is conformant to CCITT X.500 and NIST Agreements and provides distributed Directory Services across OSI environments. It offers full database administration facilities. It supports all X.500 and X.402 object classes and attributes as well as providing X/Open XOM/XDS APIs and an API for database management. The first pilot release was at the end of 1989 with the subsequent official release in April 1990. The next release (V4) is due to be available by mid-1992 and will include a multiple DIB feature plus replication which will allow information in the database to be copied to a shadow database in part or in whole. Architecturally in line with 1992, it will not initially be fully conformant. A new entity, the Shadow System Agent, will be introduced to handle DIT alignment between master and shadow DSAs. The Marben X.500 implementation offers proprietary access control mechanism, but does not support strong authentication. The product has shown interworking with QUIPU and the Unisys X.500 system. Since its launch, Marben has supplied its X.500 product to a number of clients including GEIS (General Electric Information Services, CAP-SESA (Italy), SUN Microsystems, Control Data, Touch/SAIC and some other major computer manufacturers and software companies; it is commercialised in North Amercica as Alliance OSI X.500. Product: OSIAM_C X.500 Contact: Shaofeng Li Marben Produit 11, rue Curie, 92150 Suresnes, France Telephone: +33 1 45 06 32 31 Email: sli@wtk.suresnes.marben.fr NIST NIST are now deploying a pilot for US governmental use with three sites involved and providing agency data (GSA, NSF and NIST). NASA and HHS are interested in running DSAs, and have requested the software. The Custos pilot was demonstrated at Interop '92 Spring, and consisted of an OSIware DUA connected over the Internet to the Custos DSA at NIST. Work still needs to be on some DSA features, as well as providing a DUA and a management interface. Product: CUSTOS Contact: Richard Colella The National Institute of Standards and Technology Gaithersburg, MD 20899 Telephone: +1 301 975 3627 Email: colella@osi3.ncsl.nist.gov OSIWARE D500 was developed at the University of British Columbia in Vancouver as a general purpose telecommunications directory, supporting all X.500 operations with all X.521 object classes and X.520 attribute types. D500 complies with NIST and EWOS, and can be run as an independent application or be integrated with X.400 (or other) applications such as Messenger 400 from OSIWARE. The DSA supports chaining and referrals, and includes a portable, highly efficient database which assumes very little about the host operating system. The database provides automatic recovery after power interruptions or process failure. A database API is under development that will allow integration with other datbase products. D500 includes facilities to add user-defined object classes without recompiling the code. D500 currently supports simple authentication and extensive access control lists. Product: Directory 500 (D500) Contact: OSIWARE Inc 4370 Dominion Street, Suite 200 Burnaby, B.C. Canada V5G 4L7 QUIPU QUIPU is a full implementation of X.500 and conforms to both the UK and US GOSIPs as well as to the NIST Directory Implementors Agreement. The first version of QUIPU was developed in 1987 to facilitate experimentation with the OSI Directory prior to the provision of large scale services. Started at University College London under the ESPRIT INCA project, developments to make QUIPU operationally viable were funded by the JNT (Joint Network Team), with assistance from many people around the world involved with piloting activities. A key step was to integrate QUIPU into the ISODE package of OSI protocols, and to make it openly available. The ISODE source tree was managed by Marshall Rose until Colin Robbins and Julian Onions took it over in June 1991. The last public domain release of the software (8.0) will be made in June 1992. The DSA comes in the form of a single process static responder using ISODE to provide the OSI stack. The major features of the DSA are: o the local component of the X.500 database is mapped on to UNIX text files, so that the text database can be created and managed easily and flexibly using standard UNIX text processing tools. These text files are read upon DSA start-up, and a set of indices built to provide rapid searching (some configurations have achieved 10,000 entries in under one second); o the database can be accessed, searched and modified using both DAP and DSP; o approximate martch searching is supported using a "soundex" based algorithm; o flexible schema management: object classes and attribtue type definitions are defined in an easy to configure text file; o QUIPU uses de-facto standards and proprietary solutions to overcome the undefined areas in the X.500 (88) standard which include: o replication. QUIPU fully implements RFC 1276 which defines an interim replication protocol for use until the X.500 (92) relication mechanism can be deployed. This necessitates the "data block" simplification to distributed systems; o access control. Each attribute within each entry in the Directory can have an access control policy defined., describibg which users can have read and/or mnodify access to which attributes. QUIPU fully implements "OSI-DS 21" which describes an access control policy for limiting external access to corporate sensitive data; o distributed operations and knowledge management which are achieved by mapping the problem back on to te Directory itself. Each DSA has an entry in the Directory describing the data it holds. o use of attribute inheritance to propagate commonly used attribute values to some or all entries in the local database; o intelligent use of chaining/referrals with several configuration options to overcome the problems of heterogenous network communities; o use of RFC 1277 (NSAP address encoding) which allows non-OSI addresses to be encoded in an OSI NSAP; o can perform limited run time DSA management over the DAP protocol. The QUIPU DUA API provides C language synchronous and asynchronous access to the DAP protocol using C structure definitions and ASN.1 encoders/decoders for all standard attribute syntaxes. It uses OSI-DS 23 notation for presentation and input of distinguished names, or OSI-DS for "user friendly names". It facilitates the addition of attribute handlers for new locally defined attribute syntaxes, as well as the extension of object class and attribute type definition tables. There is optional use of full caching facilities, stub handlers for displaying g3 facsimile images and the option to invoke an external process to handle some attribute syntaxes such as audio. A fully functional example DUA, DISH (DIrectory SHell), uses the API and provides a UNIX-style command for each of the X.500 DAP operations. It can be used from any UNIX shell for writing simple shell scripts. The majority (94.5%) of particpants in the current global pilot are running over 350 QUIPU DSAs in more than 25 countries. Large users in the United States include NASA, the Department of Defense and the University of Michigan. Product: QUIPU Contact: Public Domain (see also: Control Data ISODE Consortium X-Tel Services) Email: quipu@cs.ucl.ac.uk RETIX DS-520 X.500 Distributed Directory Services for UNIX System V is an integral part of the Retix OSI Networking Products family. Designed for OEMs, DS-520 is a complete high- performance implementation of X.500 in source code form, including a DUA, DSA and DSA Manager (DSAM). The DUA is available separately as DS-521 to meet the needs, for example, of software vendors who plan to provide application packages with X.500 Directory interaction capabilities. The DSA includes extended capabilities based on emerging 1992 X.500 standards including access control, replication and shadowing. The DSAM is CMIP-based and provides for local or remote on-line management of the DSA. The product is compliant with NIST, EWOS and UK GOSIP Directory functional profiles. Licences to 25 vendor and system integrators have been issued since the product was made commercially available last year. Product: DS-520 Contact: John Fowler Retix UK Ltd 60 Priestley Road The Surrey Research Park Guildford, Surrey, GU2 5SE Telephone: +44 483 300600 SIEMENS NIXDORF Siemens Nixdorf gained experience in distributed directories through their involvement in the ESPRIT project THORN. DIR.X is fully conformant to the OSI standard as well as the EWOS and OIW implementors' agreement, and can be accessed via: o a program interface (X/Open API to Directory Services - XDS) for calling up directory functions in C application programs; o an operating interface for mask-controlled administration of the directory, implemented by an administrator program. DIR.X supports caching, shadowing, referrals, chaining, simple authentication and access control at object and attribute class level. It has been used with a choice of different data retention systems, such as C-ISAM and the relational database, INFORMIX. Currently DIR.X is used internally by the SNI MAIL.X system to manage administration data like routing information and user data. DIR.X is the adopted X.500 product for the OSF (Open Software Foundation), released as DCE/GDS and DCE/XDS, and through which it has been licensed to a number of OEMs including Bull. The official SNI-product will be available this year as DIR.X V3.0. SNI are collaborating with the national pilots in both the Netherlands and Germany. In Germany they have already deployed a test DSA representing the Siemens Nixdorf organisation. Part of the collaboration is to demonstrate operational interworking with QUIPU, and they will also use the testing facilities at PTT Research in Groningen for near-conformance testing. Siemens Nixdorf participated in the trials carried out by Logica on behalf of DG IX in 1991, which involved interworking trials with ICL's Carrier. It is expected that Siemens Nixdorf will contribute DIR.X to the eventual Y-Net initiative on X.500. Product: DIR.X Contact: Patrick Fantou Siemens Nixdorf International Otto Hahn Ring 6 8000 Munich 83, Germany Telephone: +49 89 636 41203 Fax: +49 89 636 45860 Email: patrick.fantou@sniap.mchp.sni.de SOFT-SWITCH Soft-Switch have a fourfold approach to solving the directory problems of their customers: o EAB: the IBM Enterprise Address Book; o DX: Directory Exchange; o DS: Directory Synchronisation; o X.500 The company believes that most of its clients are not yet ready for X.500 (or rather vice-versa), and yet the problem of co-ordinating existing directories for proprietary messaging systems exists now. The Soft-Switch approach has been to offer their client base a transition through hybrid solutions driven by practical considerations and only propose full X.500 when they are confident both email services and products support X.500. As its own pragmatic solution to these issues of heterogeneity, the company recently announced EMX 1.0 (the Enterprise Mail Exchange), which is a family of manageable X.400 enterprise mail backbone switches that includes full interoperability with e-mail systems utilising existing protocols. EMX provides a multi-protocol directory with a dual emphasis on strategic participation in X.500 networks and tactical implementation of directory synchronisation protocols, which Soft-Switch introduced in 1989. Each EMX switch includes a relational database management system from Oracle accessed via SQL. The relational database is used for the subscriber directory, the configuration database, and the distribution store. In the third level release of EMX in December 1992, Soft-Switch will incorporate an X.500 Directory based on source code from Retix who also provide the X.400 (88) protocol stack. It is hoped that EMX will be used in the NADF pilot. Product: EMX 1.2 Contact: Greg Loux Soft-Switch Inc 640 Lee Road Wayne, PA 19087, USA Telephone: +1 215 640 9600 Email: gloux@ssw.com SYSTEMS WIZARDS DirWiz was developed under the aegis of the ESPRIT THORN and PROOF projects by this software company from Ivrea. it was the first non-QUIPU implementation to show interoperability with QUIPU in the global pilot, and is available for testing at the System Wizard node. The company are currently discussing with Olivetti the possibility of incorporating DirWiz into Olivetti's IBIsys (Integrated Business and Information sytem) alongside X- Mail for use in the Y-Net project. Product: DirWiz Contact: Franco Sirovich Systems Wizards via Torino Ivrea 10015, Italy Telephone: +39 125 631712 Email: swproof@cs.ucl.ac.uk TELIA RESEARCH The Direct500 project was carried out within Swedish Telecom's research company, Telia. Direct500, containing both DSA and DUA, was used by approximately 100 people over three months within three of Televerket's departments, representing different types of users. The directory service was available on Unix and VMS systems and in office automation systems, All-in-1 and KIS. The service was also made available to PC users connected to a LAN. The field trial used Televerket internal information as well as a small part of the public telephone directory. Response times at the DSA level for a normal look-up (read) was about one second in a database of about 30,000 records. The intention of the project was to experiment with an enterprise directory at providing one logical directory for LAN users and others. The DSA in Direct500 is based on a commercially available relational database (ORACLE), in which the X.500 information model has been implemented with operations using SQL statements. The product thus includes the benefits of a commercially available database such as tools for developing reports, special updating functions as well as support functions for backup and database management. Both All-in-1 and KIS are frequently used with VTxxx asynchronous terminals, so directory access functions were integrated with them to support both a stand alone directory as well as a directory in the mail function, able to cut-and-paste directory addresses into mail headers. DUAs were also developed on both MS-DOS and Apple Macintosh platforms. The MS-DOS DUA supports Windows 3.0; both interfaces make use of windows, graphics, cut-and-paste and are mouse operated. They are implemented as clients, resident in a PC, to a LAN-based server in a UNIX system, connected using a proprietary TCP/IP-based protocol. A DSA-Kernel handles operations between an Operations Dispatcher that interprets the DSP/DAP requests and the ORACLE database through an Embedded SQL which has to pass through a pre-processor before it can be compiled with a C- compiler. FileUpdate is a program used to load information into the database following DAP rules, and is run "on-line" towards the database. FileUpdate loads on average approximately one object per second. This X.500 implementation uses COMIF, a general communication architecture, that maps onto POP (Portable OSI Protocols). Some interworking has been carried out internally with QUIPU. No decision has been made yet on the commercialisation of Direct500. Product: Direct500 Contact: Bjorn Hartsen Telia Research AB Box 85 S-201 20 Malmo, Sweden Telephone: +46 40 10 51 00 Fax: +46 40 10 51 01 Email: bjorn.hartsen@malmo.trab.se UNISYS OSI-DSA is a set of co-operating Unix processes (written mostly in C++ with some C) that provide the defined DSA functionality. The DSA is designed to be a generic repository mechanism, which provides storage for objects required by a variety of applications. In order to accommodate new data types when and as required, a user- definable schema is provided to describe the object classes, their attribute types and syntaxes as well as their inter-relationships. The standard X.500 data types are defined in the schema in the same way that user defined types would be, ensuring that this directory can hold data as diverse as graphic images, genealogical trees or account balances. Unisys first released their X.500 product which includes the OSI-DUA Unix C Program interface library, early in 1991, followed by a second release at the end of 1991. The third release is planned for the end of 1992. The development policy has been to provide features that are in advance of the current standard to ensure that the directory can be used in commercial environments. The DSA includes a forms-based Administration User interface program to allow a human administrator to construct and maintain local Directory information. Development work is in progress to provide support for replication, based on the current ISO draft proposal, for the forthcoming X.500 (92) standard. Other new features include: o caching of local entries and the Directory structure to improve the access time to entries; o tailored search processing for selected attribute types to return search hits over huge numbers of entries within a few seconds; o shadowing of sub-trees, as defined by the 1992 draft; o tailoring the current product to support the Directory Information model as defined by the 1992 draft, which formalises the representation of knowledge reference and administrative control entries. The present OSI-DUA is a proprietary C programming interface. An X/Open conformant DUA supporting XDS, XOM and X.400 packages is currently under verification testing and will be made available in the next release. The product runs on the scaleable Unisys U6000 range of Intel-based processors, and is supported for UNIX System V.3 and V.4. The DSA uses a runtime version of either the Informix or Oracle relational database products to store the DIB. Other database engines will be supported in future releases. Informal interoperability has been achieved against QUIPU 6.0. The product was first demonstrated at Interop '90 and again at Interop in October 1991 as part of the OSI Showcase interoperating with X.500 implementations from Retix, 3Com, CDC and Banyan, some of which are QUIPU-based. Further informal testing against other vendors such as Hewlett Packard has also taken place. The Unisys product was developed in Australia where it is hoped to run an operational pilot within the AARNet Directory. Product: OSI-DSA Contact: Socs Cappas Australian Centre for Unisys Software 115 Wicks Road, North Ryde NSW 2113, Australia Telephone: +61 2 390 1312 Email: socs@syacus.acus.oz.au VTT CVOPS (C-language Virtual Operating System) is a combined environment for design, validation and implementation of communications systems software. The main implementations are Teletex, X.25, MAP and X.500. The CVOPS tool for protocol development is used by several organisations today, and is supported and developed by the Information Networks Group at VTT. The VTT X.500 directory software may be used as a basis for a range of organisational directories: personnel, computer applications, distributed bulletin boards and distribution lists. This Unix-based software has both a DSA and a DUA and supports simple authentication and caching. It allows for 20 simultaneous connections to a DSA, allows the use of wild cards in the search argument snd can handle about 400,00 entries. The developers envisage full support of X.509 and X.520/X.521 in the near future. Product: CVOPS Contact: Asko Vilavaara Technical Research Centre of Finland Telecommunications Laboratory PO Box 34, SF-02151 Espoo, Finland Telephone: +358 0 4561 Email: asko.vilavaara@tel.vtt.fi WOLLONGONG GROUP WIN/DS includes a DSA and a DUA and supports all X.500 operations, object classes and attributes. It contains support for data management and closely follows the NIST OIW Stable Implementors' Agreements. It has interworked with other implementations at Cebit and Interop . Product: WIN/DS Contact: The Wollongong Group Inc. 1129 San Antonio Road Palo Alto, CA 94303, USA Telephone: +1 413 962 7100 Email: support@twg.com X-TEL SERVICES The X-Tel product is a fully-supported, commercially available implementation of X.500 based on the ISODE QUIPU implementation. As well as the DSA, the package includes Motif-based management user interfaces (XT-DUA, XT-OB), fully integrated with their X.400 product, XT-PP. This implementation is being used by a number of sites in the PARADISE pilot. Product: XT-QUIPU Contact: Colin Robbins, Product Manager X-Tel Services Ltd University Park, Nottingham NG7 2RD, UK Telephone: +44 602 412648 Email: x500@xtel.co.uk All logos are registered trademarks of the respective companies. C O N T A C T P R O F I L E A U S T R I A Representative: ACONET Contact: Florian Schnabel Technische Universitat Graz EDV-Zentrum Steyrergasse 30/1 A-8010 Graz Telephone: +43 31 68 736 255 Email: schnabel@edvz.tu-graz.ada.at B E L G I U M Representative: University of Brussels Contact: Nils Meulemans ULB, IIHE - Groupe HELIOS-B CP 230 - Bd. du Triomphe B-1050 Brussels Telephone: +32 2 641 35 53 Email: nils@helios.iihe.rtt.be C R O A T I A Representative: University of Zagreb Contact: Enver Sehovic Elektrotehnicki Fakultet University of Zagreb 41000 Zagreb, Unska 3 Telephone: +38 629 616 Email: enver.sehovic@etf.uni-zg.ac.mail.yu C Z E C H O S L O V A K I A Representative: Slovak Academy of Sciences Contact: Karol Fabian Institute of Automation & Communication Severna 5 974 00 Banska Bystrica Telephone: +42 88 532 07 Email: fabian@uakom.cs D E N M A R K Representative: DKnet Contact: Steen Linden Department of Computer Science University of Copenhagen Universitetsparken 1 DK-2100 Copenhagen Telephone: +45 31 39 64 66 x222 Email: steen.linden@dkuug.dk E U R O P E Representative: PARADISE Contact: PARADISE Helpdesk ULCC 20 Guildford Street London WC1N 1DZ Telephone: +44 71 405 8400 x 432 Email: helpdesk@paradise.ulcc.ac.uk F I N L A N D Representative: FUNET Contact: Manu Mahonen FUNET c/o VTKK PO Box 40 SF-02101 Espoo Telephone: +358 31 343 2210 Email: directory-manager@funet.fi F R A N C E Representative: OPAX Contact: Paul-Andre Pays INRIA, Batiment 15 BP 105 Rocquencourt 78153 Le Chesnay Cedex Telephone: +33 39 63 54 58 Email: pays@faugeres.inria.fr G E R M A N Y Representative: DFN Contact: Panos-Gavriil Tsigaridas GMD Fokus Hardenbergplatz 2 West Berlin 12 Telephone: +49 30 25499 232 Email: dfnds-manager@fokus.berlin.gmd.dbp.de G R E E C E Representative: Network Ariadne Contact: Yannis Corovesis NRC Demokritos 15310 Attiki Telephone: +30 1 65 13 392 Email: yannis.corovesis@isosun.ariadne-t.gr H U N G A R Y Organisation: Technical University of Budapest Contact: Erzsebet Erdei Information Centre Technical University of Budapest Budapest Telephone: +36 1 181 2172/186 8058 Email h5082erd@ella.hu I C E L A N D Organisation: University of Iceland Contact: Marius Olafsson Reiknistofnun Haskolans Dunhaga 5 107 Reykjavik Telephone: +354 1 694747 Email marius@rhi.hi.is I R E L A N D Representative: Trinity College Dublin Contact: Donal O'Mahony Computer Science Department Trinity College Dublin 2 Telephone: +353 1 7021261 Email: omahony@cs.tcd.ie I S R A E L Representative: The Hebrew University of Jerusalem Contact: Juliana Solomon Computation Center Taylor Building, Givat Ram The Hebrew University of Jerusalem Jerusalem Telephone: +972 2 58 5686 Email: il-x500@hunch.cc.huji.ac.il I T A L Y Contact: GARR Representative: Antonio Blasco Bonito CNUCE - Istituto del CNR Reparto Infrastrutture di Rete per la Ricerca Via S. Maria, 36 56126 Pisa Telephone: +39 50 593246 Email: bonito@nis.garr.it L U X E M B O U R G Representative: RESTENA Contact: Theo Duhautpas RESTENA 6, rue Coudenhove-Kalergi L-1359 Luxembourg-Kirchbourg Telephone: +352 42 44 09 Email: duhautpas@restena.pt.lu T H E N E T H E R L A N D S Representative: SURFnet BV Contact: Vincent Berkhout Netwerkontwikkeling Postbus 19035 3501 DA Utrecht Telephone: +31 30 310290 Email: berkhout@surfnet.nl N O R W A Y Representative: UNINETT Contact: UNINETT Directory Project c/o University of Oslo/USIT POB 1059 - Blindern 0316 Oslo 3 Telephone: +47 2 453470 Email: directory-adm@uninett.no P O R T U G A L Representative: Universidade do Minho Contact: Fernando Pinto Departamento de Informatica Grupo de Comunicacoes Universidade do Minho Rua do Pedro V, 88-3 4700 Braga Telephone: +351 53 612234 ext 431 Email: fernando@uminho.pt S L O V E N I A Representative: University of Ljubljana Contact: Marko Bonac University of Ljubljana Jozef Stefan Institute Jamova 39 61000 Ljubljana Telephone: +38 61 159199 Email: bonac@ijs.yu S P A I N Representative: IRIS Programme/FUNDESCO Contact: Celestino Tomas RedIRIS/Fundesco Alcala 61 E-28014 Madrid Telephone: +34 1 4351214 x284 Email: celestino.tomas@iris-dcp.es S W E D E N Representative: SUNET Contact: Roland Hedberg SUNET Umdac, S-90187 Telephone: +46 90 165 204 Email: roland@umu.se S W I T Z E R L A N D Representative: SWITCH Contact: Thomas Lenggenhager SWITCH ETH-Zentrum CH-8092 Zurich Telephone: +41 1 261 8178 Email: lenggenhager@switch.ch T U R K E Y Representative: University of Izmir Contact: Esra Delen Ege Universitesi Bilgisayar Arastirma ve Uygulama Merkezi 35100 Bornova Izmir Telephone: +90 51 187228 Email: esra@ege.edu.tr U N I T E D K I N G D O M Representative: Joint Network Team Contact: Directory Project Manager X-Tel Services Ltd University Park Nottingham NG7 2RD Telephone: +44 602 412648 Email: x500@xtel.co.uk A M E R I C A S B R A Z I L Representative: Federal University of Rio Grande do Sul Contact: Cleber Garcia Weissheimer Federal University of Rio Grande do Sul Institute of Informatics Computer Networks Research Group Porto Alegre - RS Telephone: +55 51 3324241 Email: cleber@vortex.ufrgs.br C A N A D A Organisation: University of Western Ontario Contact: John MacAuley CCS, Natural Sciences Building University of Western Ontario, London, Canada N6A 5B7 Telephone: +1 519 661-2151 6020 Email hacksaw@julian.uwo.ca U N I T E D S T A T E S Representative: White Pages Project Contact: Wengyik Yeong PSI Inc. 5201 Great American Parkway Suite 3106 Santa Clara, CA 95054 Telephone: +1 408 562 6222 Email: wpp-manager@psi.com P A C I F I C R I M A U S T R A L I A Organisation: AARNet Directory Project Contact: Graham Rees The Prentice Centre The University of Queensland St Lucia, Queensland 4072 Telephone: +61 7 365 4143 Email: aarn-ds@cc.uq.oz.au I N D I A Representative: Indian Institute of Technology Contact: Surinder Singh Anand Department of Computer Science & Engineering, Indian Institute of Technology Hauz Khas New Delhi 110 016 Telephone: +91 11 686 7431 x6009 Email: anand@netearth.ernet.in J A P A N Representative: WIDE Project/ISODE Working Group Contact: Hideki Sunahara Dept. of Computer Science University of Electro-Communications 1-5-1 Chofugaoka Chofu-shi, Tokyo 182 Telephone: +81 424 83 2161 ext.4122 or 4172 Email: suna@cs.uec.ac.jp K O R E A Representative: SAIT Contact: Kim Inhwan Samsung Advanced Institute of Technology San #14 Nong Seo-Ri Kihung-Euo, Yongin-Gun Kyung Ki-Do Email: inhwan@silla.sait.co.kr N E W Z E A L A N D Representative: Victoria University Wellington Contact: Andy Linton Department of Computer science Victoria University of Wellington PO Box 600 Wellington Telephone: +64 4 495 5054 Email: andy.linton@comp.vuw.ac.nz E U R O P E The first half of 1992 has seen the consolidation of the piloting activity that flourished in 1991. All national pilots, with the exception of France are running QUIPU DSAs, which since the release of ISODE 7.0 have produced a far greater degree of stability and improved quality of performance. Since PARADISE established its central DUA service in summer 1991, half the countries in Europe have installed national public access DUAs, usually based on "de" or sometimes "fred". Since the last report representatives from Croatia (University of Zagreb), Czecho-slovakia (Institute of Automation and Communication in Banska Bystrica), Greece (Institute of Computing, Heraklio) and Slovenia (Jozef Stefan Institute) have joined the pilot. Interest has also been expressed from Hungary (Technical University of Budapest), Poland and Turkey (University of Izmir). In the cases of Croatia and Slovenia, an interim solution for finding a distinguished name has been sought. Because of the complication that not all installed systems will allow the new ISO codes (c=HR, c=SI), these two republics have been temporarily registered as l=Croatia and l=Slovenia below the root. A second DSA at the University of Leuven (KUL) has joined the pilot in Belgium. In order to promote the use of X.500 services in Belgium, the University of Brussels is providing a DSA to hold a limited number of organisation entries. The new pilot in Czecho-slovakia is running its QUIPU DSA on a SUN SPARCstation-2 with IP connectivity. The purpose of the project is to build a network infrastructure, based on TCP/IP, linking all important academic, research and educational sites together with connections to most important international networks. The X.500 pilot is at the stage of: o defining and building its DIT structure, o checking various kinds of DUAs; o setting-up a DUA for remote anonymous access; o building database files for the DSA; o setting-up DSAs in other sites and testing connectivity. A second university, Aarhus, has joined the Danish pilot, where it is hoped a formal project involving Uni-C, the University of Copenhagen and possibly TeleDenmark (Jutland Telephone) will shortly be announced. OPAX the French Directory pilot is open to members of both the academic and industrial communities and is divided accordingly (Paradis and Eden). The master DSA for France has moved to the UREC/CNRS site at the University of Paris with international X.25, IXI and IP access. Interworking tests are bing carried out between the principal implementation (UCOM.X 500) and both QUIPU and DIR.X (Bull). As the pilot grows, rules for active participation are becoming more stringent as only operational DSAs are registered. The goals of the project over the next few months are: o specify a management tool to improve sub-tree updating; o produce a user guide for newcomers to X.500 o produce a booklet to introduce OIPAX objectives as well as terms of reference. Plans exist to involve the Maison des Sciences de l'Homme and the CNRS. The German VERDI 3 project which finishes in July 1993 is developing supporting tools for its pilot, developing and establishing an in-house GMD telephone information system based on the Directory, as well as investigating the multimedia potential of the Directory. In the last few months the pilot has been interworking with the Siemens Nixdorf test DSA based at Munich. More universities and research institutes (15) in the east of Germany are expected to be running their own DSAs in the pilot in the near future. The Directory in Germany has been accepted as a key application for the future. Several organisations are already basing their in-house telephone and information systems on X.500, though there is a reluctance to offer data to the global Directory because of the data protection laws. DFN have provided considerable documentation and supporting software tools which are available on request. The Irish pilot is hoping to expand considerably when it receives formal backing. Currently based at Trinity College Dublin, the pilot now includes a number of commercial organisations including SNI. Experiments at TCD include a project to use X.500 with EDI, and another to build a directory of mathematicians in Europe (EUROMATH). The second SURFnet X.500 pilot, which hopes to be fully operational by January 1994, is underway in the Netherlands with three major objectives: o maintaining the current infrastructure in conformance with international requirements; o developing a real multi-vendor infrastructure, which has started with the collaboration of SNI; o data management tests. These will be carried out a small number of University sites and will focus on the organisation of data management, privacy/legal issues and integration with existing databases on personnel/students. In the second half of 1992, a central DSA will be installed to serve SMEs that can not afford their own DSA. A central DUA for browsing (a version of de translated into Dutch), and a central DUA for maintaining data will also be installed. UNINETT in Norway are continuing its projects to improve the directory service for its users. A current project is to make the directory fully available within two targetted organisations, by providing software and a documentation package which can then be ported to other organisations who wish to learn from the experience. As part of this pilot UNINETT are using a screen- oriented tool for interactively maintaining white pages information in the Directory (sde - the Simple Directory Editor), in conjunction with a white pages user interface for the Macintosh also developed by UNINETT. However, use of the directory is still predominantly through the mail-responder, which will shortly be updated. The Spanish master DSA has moved from Madrid to Barcelona. In the first part of the RedIRIS pilot, the four backbone DSAs were situated at RedIRIS offices, but since January 1992 have been subcontracted out. These DSAs not only keep the information of the centres at which they are based, but also part of the 110 organisations affiliated to the Spanish R&D network. The services provided include four public access DUAs with the PARADISE DUA (de) translated into Spanish, a mail responder and a helpdesk. The main objectives for 1992 are to involve a greater number of the RedIRIS organisations in providing a white pages service, as well as developing the directory for FTAM applications and X.400. In recent months three new organisations have started operating their own DSAs in Switzerland, and are gradually incorporating their data. In one case, the current policy is to only allow local users to read personal entries. Since April 1992, ETHZ has made publicly available a Hypercard-based DUA on the Macintosh using the DIXIE protocol called ITAXA 1.1. Both the Swiss PTT and the Union Bank have already more than 10,000 entries in the Directory. A DEC X.500 implementation running on Ultrix which is currently in field test will soon to be connected to the pilot for interworking tests. In the near future, there are plans to develop a naming schema to anticipate future pilot growth. Wherever possible, existing registration mechanisms should be used to choose names for the Directory. A M E R I C A S Since the last report, Brazil has established its master DSA at the Federal University (UFRGS) in Porto Alegre. It is a prototype and contains only a few entries. In the short term they are planning to extend the UFRGS database and incorporate other members of the academic community. They will investigate means of directory access, such as mail-responders, though in the long term the directory is intended to be used for network management. Work on the X.500 Directory in the United States progresses on many fronts with work contributed by many groups. Much of this work is co-ordinated through a set of committees and groups with interlocking membership. The key activities are: o the X.500-related working groups of the Internet Engineering Task Force (IETF), especially the Directory Information Services Infrastructure WG (DISI) and the OSI Directory Services WG (OSI-DS); o the North American Directory Forum (NADF); o the White Pages Project (WPP); and o the Fielding Operational X.500 project (FOX). DISI has progressed all its current documents to RFCs. These include: o "A Catalog of Available X.500 Implementations" (RFC 1292); o "An Executive Introduction to Directory Services Using the X.500 Protocol" (RFC 1308); and o "Technical Overview of Directory Servies Using the X.500 Protocol" (RFC 1309). Document ideas currently being considered include a pilot project catalogue, a DSA set-up guide, an advanced usages document, a how-to document on becoming registered in the Directory and a document on the naming philosophy of X.500. OSI-DS are studying the issue of security for competitive products with plans to produce an RFC. A strategy document is planned for Directory Services, which will hinge on the listing versus registration issue. The argument, which centres on where authority for Directory names is derived, will play a key part in the future of X.500. The model favoured by the NADF is the listing model which gives each entity rights to a name based on national intellectual property law. PSI has also designed and tested software to incorporate the Domain Name System (DNS) into the DIT. The NADF, which now has active participation from the United States and Canada, has completed preliminary agreements on naming conventions in the public portions of the North American DIT, as well as agreements on the sharing of naming and knowledge information in those portions of the DIT. Preliminary agreements were also reached on North American service requirements. In response to privacy and security concerns, an NADF sub-group has also begun an initiative to address such issues in the context of the provision of public directory services. This sub-group has developed a User Bill of Rights, outlining the rights of individuals and organisations with respect to their listings in public directory services. An experimental pilot involving all members of the NADF has also begun. This experimental pilot is designed to test the operation of public directory services according to the technical agreements reached by the NADF. The NADF will establish connectivity to other X.500-based directory systems in later stages of the pilot. The WPP co-ordinated by Performance Systems International (PSI) continues to grow at a steady rate and now includes 106 organisations, which has required adding an auxiliary DSA to accommodate its growth. The WPP is moving towards adopting the NADF naming schema outlined in NADF 175. FOX co-ordinates the X.500 directory activities at: o ISI at the University of Southern California which compiles and distributes the Directory Services Activities Report (DSAR) each month as part of the Internet Monthly Report; o Merit Network Inc who are implementing pointer structures for their Network Information Resources; o SRI International where the whois-based Directory interface (x5whois) now accesses a data set of 150,000 entries, though loading the data set on to a DSA still takes an inordiantely long time. x5whois, originally run over a QUIPU DSA, is now being tested on Custos 0.1.1. o PSI co-ordiante the White Pages Pilot. Other activities in the US include the Lightweight Directory Browsing Protocol (LDBP) document, now being revised by PSI and the University of Michigan, and further work on the MaX.500 Macintosh DUA by Merit. P A C I F I C R I M In the last six months, India has joined the pilot with a master DSA at the Indian Institute of Technolgy in New Delhi. Interest in Korea has come from the Samsung Advanced Institute of Technology. There are three main strands to the Japanese WIDE project/ISODE working group: o the problem of handling Japanese characters. A consensus chose to use the T.61 character strings with the JIS (Japanese Industrial Standard) character encoding rule. They finished modification of the DSA and several DUAs to handle Japanese characters. o experimentation with the use of X.500 from an application (MH); o an enhanced collection mechanism for DSA performance information to improve DSA efficiency by means of a new object class "wideDSA" and an attribute type "stat" to store response times between DSAs. WIDE also plans to start a directory service and experiments with X.400 using PP. Since the last report, AARNet have demonstrated at the Australian Networkshop which generated a lot of interest in the Directory especially the storage of images and sound. The demonstration has encouraged a number of new sites to start experimenting with the Directory; this process has been assisted by the provision of a binary distribution of QUIPU for a number of popular hardware platforms. AARNet are currently registering all organisations that are accessible across the Australian network, and will provide basic contact information to replace the AusMap project. At present, those sites running the MHSnet software are being registered, by hand, when a state message arrives from their site at Adelaide University. These messages provide useful contact information although not in an easily- automated processible form. AARNet hope, depending on resources, to provide a service whereby sites that do not have network connectivity or the resources to run their own DSA can provide a small amount of useful information on one of the four AARNet backbone machines (Digital DS3100's). There are plans to conduct some interoperability testing with local X.500 products, particularly that from Unisys. The current economic climate makes it difficult for institutions to embark on projects, such as the Directory, that require significant resources both in terms of equipment and manpower without some obvious benefit for the institution as a whole. At present although there are clearly perceived benefits to running a Directory, future growth will be inhibited due to a lack of implementations of user agents that use the most common computing platform within the institution ie a PC running MS-Windows. While it can be seen as useful to those who have access to Unix workstations and servers, it will not become part of the infrastructure until it can address the needs of the great majority of users. At present there are no useful PC applications: the existing ones either making use of a rare commercial TCP/IP package or are still using an unavailable OSI stack. CSIRO have produced patch kits to provide a look-up facility within the Unix mail systems MH and Elm which allow users to query the Directory to determine electronic mail (rfc822) addresses using a UFN-style format. These kits are available for non-commercial use from ajw@mel.dit.csiro.au. Despite these constraints, AARNet are continuing to assist sites wishing to deploy the Directory. However, the problems facing even sites with technical expertise are in the areas of initial loading data into the Directory, due in part to the diversity of information sources but also to the lack of useful management tools. P U B L I C A C C E S S D I R E C T O R Y I N T E R F A C E S PSS IXI IP Australia 129.127.40.3 130.194.2.68 130.102.128.43 129.78.64.15 (login: fred) Belgium: 222100611 204306500004 134.184.11.4 (login: dua) Denmark: 129.142.96.43 (login: ds) Finland: 128.214.6.100 (login: dua) France: 20800603053201 (login: dua, password: ucom.x) Germany: 26245050230303 Ireland 134.226.32.17 (login: de) Italy: 22225010083212 20432240001212 131.114.2.5 (login: de/fred) Spain: 2142160234013 2043145100103 130.206.1.3 2142540916282 2043145400002 150.214.4.4 2043145300011 147.83.41.13 21452220206402 2043145200002 130.206.73.15 (login: directorio) Sweden 240374810306 130.239.16.15 (login: de) Switzerland: 22847971014540 20432840100540 130.59.1.40 (login: dua) PARADISE Directory Enquiries: 23421920014853 20433450400253 128.86.8.56 (login: dua) PARADISE Directory Manager: 23421920014852 20433450400252 128.86.8.56 (login: dm) PARADISE Dial-up: 1200 bit/s full duplex: +44 71 831 6171 1200/75 bit/s: +44 71 831 6181 (from August 1992: +44 71 405 4222) PARADISE Directory Enquiries Number of Calls IXI IP PSS Janet Dial-up Total 1992 May 103 945 33 357 1438 April 39 719 163 86 26 1033 March 50 287 246 116 699 February 42 191 187 135 555 January 51 137 183 52 423 1991 December 50 112 186 41 389 November 263 103 129 28 523 October 86 198 198 14 496 --- ---- ----- --- ---- ---- 684 2692 1325 829 26 5556 G L O S S A R Y API Applications Programmer's Interface, a set of calling conventions defining how a service is invoked through a software package COSINE Cooperation for Open Systems Interconnection Networking in Europe, a EUREKA project funded by the CEC and 18 European governments and managed by RARE DAP Directory Access Protocol: the protocol used between a DUA and a DSA DCE Distributed Computing Environment DE Directory Enquiries, the PARADISE public access interface DIT the Directory Information Tree DM Directory Manager: the PARADISE public access account manageemnt tool DSA Directory System Agent: the distributed directory database DSP Directory Sysyem Protocol: the protocol used between two DSAs DUA Directory User Agent: the user interface EDI Electronic Data Interchange: the protocol based on the CCITT X.435 standard FTAM File Transfer and Access Management: the OSI standard for file transfer IXI the COSINE X.25 backbone, which carries non-commercial traffic for the research communities ISODE ISO Development Environment: a research tool developed to study the upper layers of OSI. (NB ISO has nothing to do with the International Standards Organisation) JANET Joint Academic Network (UK) MHS Message Handling System: the name of the "electronic mail" set of services as defined by CCITT (X.400) and ISO (MOTIS) OSI Open System Interconnection: the ISO standard architecture for Open Systems; an international effort to facilitate communications among computers of different manufacturer and technology PARADISE the COSINE pilot international X.500 Directory Service PTO a post and telephone operator QUIPU a public domain implementation of the OSI Directory, packaged with the ISODE software and developed at UCL under the ESPRIT project INCA and through the JNT. RFC Request for Comments: the document series describing the Internet suite of protocols and related experiments RFC 822 de facto electronic mail standard for Internet users SME a small-to-medium size enterprise with up to 500 personnel SQL Sequential Query Language TCP/IP Transmission Control Protocol/Internet Protocol: network protocol offering a connectionless mode network service X.25 OSI protocol offering a connection-oriented network service X.400 the CCITT standard defining electronic mail exchange in Open Systems X.500 the series of international standards and recommendations for a distributed directory based on OSI and adopted in 1988; to be revised in 1992