Omissions:

Areas which did not achieve much comment:

• User identification, authentication and authorisation. Activities within this area are ongoing within the UK AC community as a whole, and are therefore not discussed here. It is likely that a backbone model for EaStMAN might be required to support (unspecified) security functions at the network level - through packet filtering and so on, but the deep issues of security are reserved for institutional policy and differentiation.

• A related function – that of eavesdropping, disruption, and so on, which are network issues did not solicit much discussion either. It is assumed that protection against these activities is ongoing and part of the environment.

• The deployment of multicast applications did not achieve much discussion. This is an interesting observation since the use of IP multicast is being actively worked on in several standards bodies, and indeed being tested within the SuperJANET and University of Edinburgh environments, at least.

2.1 Expansion on topics

1. The locations for learning are becoming more diverse, and are not totally located on campus, MAN or SuperJANET.

There is a growing requirement for access from areas such as the home, business, international, schools, E-University. Modularisation of courses adopted by students may require virtual / modular learning environments where resource must be brought to the student (the JISC term "joined up computing" may be apt). This is important to projects such as SCHWEIMS and to Further Education models.

In the future, students will probably be much more mobile and we will need to ensure that all University VLEs/MLEs are able to speak to each other

A requirement is therefore to address the subjective experience of network delivered learning. Where possible, the campus experience should be maintained independent of the client location i.e. geographical independence. This implies that the material being developed for distribution may require to be created with a view to delivery by multiple mechanisms, some of which may not be bandwidth rich.

R1: The external reach beyond the EaStMAN backbone requires consideration, both self provided (institutional, or across MAN) provided and SuperJANET provided. This important point is emphasised: any move to make EaStMAN more accessible to non-traditional learners in non-campus based locations would be welcome.

This latter point reflects some requirement to integrate the local educational or information community – an aim of Digital Scotland

2. Networked resource offers the opportunity to slip synchrony – that is, pace and self-scheduled activity.

One concrete application was singled out: in the context of undergraduate teaching, the configuration and layout of open student laboratories may require some alteration. In particular, the links should be attached at high capacity in order to deliver the various media requiring support. In some sense this high capacity requirement may seem to conflict with the aims of A above, but it should be viewed as indicating that there may be a need to address the distribution of "quality" material in multiple ways tuned to and determined by the capabilities of the receivers. High capacity linkages may be used on campus, but adaptation for support for slower external access must be considered.

Where access to tutors via conferencing facilities is supported, then one needs to consider the video and audio support from open laboratory areas: should these be from particular bookable systems only, or one way video tutor to student?

R2: Consider the provisioning of areas on-campus where material requiring high capacity is delivered, and also consider the multiplicity of potential receivers outwith those areas.
3. Not all applications which are being envisaged and developed for network deployment are or will be "media-rich", but they still present quality of service requirements.

The types of application which are predicted as being first to achieve wide deployment are not generally those involving multiple media such as video and or audio. Particular example applications are mostly web-based: simulations, assessment exercises, delivery of text-based interactive material, and so on. It is considered that video and audio will soon be built into this delivery.

The important characteristics of these applications are that:

• there is likely to be mass usage in considerable numbers;
• response time is particularly important.

Response is a function of network support combined with the capability of the responding application and the platforms involved in its support, including the client. The combination of mass usage and response implies that there may be bursts of high activity involving many small data transactions. To deliver good response within a background of high activity implies one or both of two general solutions:

• bandwidth headroom to support the data distribution and or
• recognition that traffic engineering and prioritisation will become an essential feature of new network systems.

It is also the case that the servers and the means of distributing and or balancing their network and CPU load also require attention.

When presented with a hypothetical choice, on-campus delivery of high capacity bandwidth rather than scheduling use of the appropriate bandwidth was selected. With the growing base of Gigabit Ethernet capability, this may be indeed already be available on-campus, but when services are accessed externally, where the bandwidth may not be in super-abundance at a reasonable cost, then policy-based or scheduling schemes may need to be adopted.

R3: Applications will require high capacity to deliver good response characteristics to large numbers of users.

Servers need to be suitably scaled and load-balancing paradigms may be required.

Policy and prioritisation need support on the EaStMAN network equipment and eventually on end-systems to support responsive access to services and also to access external applications

Any service that is offered over the MAN should exhibit certain minimum service level characteristics thus leading to user confidence in the system and, probably, greater uptake of services. For example, the view has been expressed that effort to ensure that these service levels are met would be welcome, rather than attempting to extend the range of services in such a way that the requisite service levels may not be met.

4. In the short term, the delivery of streaming video and audio, with synchronised text, may have a more obvious application to T&L than conferencing based methods.

One difference between conferencing and streamed delivery based systems is that the usability of the latter is less dependent on encode/decode delays and network induced latencies - this follows from the "TV channel" nature of many of these implementations. That is, the subjective experience of interactions between real people in verbal or visual communication is considerably affected by delays as low as fractions of a second. Latencies of 0.1 - 0.2sec are perceptible but considered to be less disruptive to communication. Given this feature, it is still important to recognise that such streams require guarantees with respect to packet loss and jitter (the variation in latency).

R4: For streaming applications, probably the most obvious for T&L, the network must be able to minimise both jitter and packet loss characteristics. To support conferencing or interactive modes, these characteristics should be in combination with minimised latency.

In the future, better access to broadcast quality conferencing facilities will encourage and support greater uptake of conference based T&L. It is a fairly safe statement to make that conferencing T&L can provide a useful supplement to streaming based approaches.

5. Conferencing mode is important for developing applications.

Applications for conferencing / interaction mode will form a spectrum range from the live teaching sessions - which may be a mixture of broadcast with question points - through to support functions – for example at the student's laptop or open lab location. It is recognised that the lecture theatre environment poses a particular challenge to both the technology and to practice at the front. Not all of these are network related, but the whole environment needs analysis in order that the network can deliver its part. (This aspect is currently under study within a UKERNA / SHEFC project looking at the development of SMVCN, the Scottish MANs videoconferencing network.)

The cost of using the facilities is important - some will walk from ISDN sites where there is a charge, to a SMVCN site where there is none. Thus, direct cost sensitivity is high in this area.

R5: Conferencing based applications imply that quality of service (packet loss, latency, jitter) must be recognised and supported.

The quality of the current SMVCN is recognised as a desirable target, but needs study with regard to its integration and on-going development.

Zero or low cost at the point of delivery is an important factor to consider in development, that is, the service should be data network based where possible.

The comment has been made that conferencing mode is most effective when used by staff development teams with access to broadcast quality video services. However, as the connection speed is determined by the slowest link, this level of service is often unavailable even to those who have direct access to EaStMAN. Any strategy effort to try and resolve this anomaly would be warmly received.

6. The familiarity and awareness of the teaching staff about the opportunities available is still classed as low (despite great efforts expended by TALiSMAN).

For teaching purposes, the quality and opportunities offered by the SMVCN are recognised, but perhaps not as fully developed as they could be. How can this be improved?

Demonstrator applications are still recommended by some as a means of dissemination. The SHEFC staff development activity should perhaps take this on board, and raise the profile of the institutional and sector opportunities. For example, within the University of Edinburgh, Jim Sheach [MALTS] and John Martin [EUCS] have presented a successful introductory course on the use of videoconferencing studios, it will appear as a standard entry in the staff development handbook.

It is noted that the environment can change and so alter the profile and perception of such services. On the West of Scotland, the Glasgow to Dumfries link has certainly stimulated considerable activity in remote teaching. It is not clear if the Heriot-Watt to the Scottish College of Textiles in Galashiels will result in the same culture change. A corresponding link between the Western General to Royal Infirmary hospitals has yet to achieve significant usage. For success based on usage, the correct mix of degree of requirement and availability of the appropriate technology seems to be required. Need seems to be the one which drives the usage in the above cases.

One factor may be the difficulty of access to high quality services outside of dedicated facilities. Greater emphasis on people/units based within each HEI would, in our opinion, probably have a greater impact. That is, there is a need for staff in each HEI who could take responsibility for 'spreading the word' and arranging for dissemination of information and good practice regarding MAN applications. Although the work done by TALISMAN is excellent it does not seem to have produced much change or greater awareness amongst individual staff within several (perhaps not all) institutions. Unless staff are more aware and have appropriate support, widespread use of the MAN for teaching and learning could be delayed for some time.

7. The best means for supporting the basic infrastructure deployed by tutors and learners needs considerable attention.

Support raises its head in at least two areas:

Firstly, the support of the technical infrastructure, in lecture theatre or studios environments, desktop systems, involving cameras, audio microphones, data sharing and so on. Informed experience in Stirling suggests that students are extremely intolerant of delays and problems caused by the "equipment". Thus to maintain a local-remote audience attention span and to maximise the effectiveness of the experience, the involvement with equipment must be made easy and thereby minimised, procedures for use are important, set-up time allowed for, and so on. It is extremely important that staff and students using the systems are familiar with what is available and the means of best exploiting it. Training and practice are important.

R6: Attention must be given to awareness and the support functions at the technology teaching interface, with training and practice seen to be important. Whether this is institutionally based or centrally, remains for discussion.

Secondly, the creation of media rich material can be both costly and time-consuming. The capture, digitisation and editing of such material may be regarded as presenting more of a specialist technical nature than does pure text, it may also involve a class of high-end equipment which is not generally available to all members of staff. Some are happy with DIY but not all.

It has been commented that compartmentalisation of experience gained by doing one's own thing within (say) a department may not be the most cost-effective way to provide such material. The group further comments that it may be worth considering the centralisation on a MAN basis, of such services. In addition, analogous to the network infrastructure management, there may be a good case for introduction of some facility management (FM) management of servers to support local, EaStMAN (and beyond) distribution. Certainly models which distribute support roles are starting to appear for support of the Further Education sector and its imminent integration within an HE/FE academic network.

R7: Support of the technology and practitioners in T&L is vital to success, details of technology must remain hidden from those who need not know anything about it.

EaStMAN-wide services which go beyond the network infrastructure management (e.g. FM) might be an avenue to explore further.

Support is required for academic staff who have little or no familiarity with this technology particularly regarding effective pedagogy – but this may be going beyond the remit of this group.