@heading[Full Screen Services]

The work of the 'Full Screen Services' group is still at a very early stage,
consisting of exchanges by letter.
 A meeting of the group will be 
held during the Networkshop in Copenhagen.

There were responses from six countries (including the UK) to the original 
request for information and a number of points have arisen from the responses.

Currently there is very little full screen activity in use
 across national networks. There is some use across local networks 
although this
tends to be very specific, ie just for Dec VT100 type terminals or IBM 3270
terminals. Both the DFN and the UK Academic Community use Triple X 
(X.28, X.29 and X.3) to provide such services although it is recognised that 
Triple X is not ideally suited to provide such services both from the lack
of facilities in the protocol to provide support and from the significant 
cost and performance penalties from running in modes which require single
character packets to be transmitted through the network with character
echo being provided from the host end.

There are at least two separate areas that need to be considered. The first 
is the type of service that handles 'editing-type' screen services, extended
to a more general handling of a 'text' screen for host programs and secondly
the handling of a graphic or a mixed text and graphic screen in a 
consistent manner.

In both cases there are two distinct sides to the problem, firstly there
is a need for a device independant method for a host to drive a screen,
secondly, and currently less well understood, is the method of handling 
input from the user so as to give maximum flexibility to the user
interface, eg for form filling applications, while minimising the quantity
of traffic across the network.

There is a widespread belief that ISO will provide a VTP that will address
the problems, there is however, widespread pessimism that such a protocol
will be produced in the near future. The early '90s being seen as a probable
timescale. The current draft
has virtual terminal facilities for handling output to screens but
 does not have facilities for managing
 the input from terminals so would suffer from the same performance cost
as Triple X.

Two of the communities, DFN and the UK, are doing work on providing facilities 
across national networks. The DFN work is a pilot study to provide a service
to three types of terminal in the first instance and to investigate the mapping
of different specific terminal functionalities to and from a virtual terminal
protocol. This project is still in progress.

The UK Academic Community has been interested in a solution to the problem for
a number of years and a decision was taken to produce a specific protocol,
able to run transparently over Triple X, to provide full screen services.
This protocol, called 'The Simple Screen Management Protocol' was published
in July 1985 and the UK community now has a number of trial implementations both
for specific terminals, micros and host editors. The micros run the protocol
directly while the terminals are handled by a small add-on box which is 
placed between the terminal and the PAD.


@heading[summary]

There is a need across the Academic Communities for full screen support,
this is not seen as necessary for inter community use but mainly for use within
the national networks.  ISO VTP will not be available in the desired 
timescale so it would appear that unless action is taken each community
will produce it own ad hoc solution. Although this will not cause
interworking problems to any extent it does mean a large amount of
unnecessary effort
and conflicting requests to manufacturers for individual solutions.
  The task of the group is to see if there is any
agreement on the production of a common solution.

@flushright[B.Gilmore
May 1986]