The FR80 passed its acceptance tests with a few minor reservations in the middle of May. Rather than wait until all the necessary software changes have been made on the 1906A, 360/195 and FR80, it is proposed to start a service on 2 June 1975 with what is currently available. Additional software will be released as soon as possible.
All the FR80 hardware described in FR80 User Note 1 has arrived and is available to users. The precise positioning of camera images depends quite crucially on the final setting up of the system. Consequently, the camera images specified initially are no longer correct. The new values are given in new User Notes.
There is a LOADGO software system available on the FR80 which allows a number of jobs on a single magnetic tape to be processed automatically without operator intervention. This greatly speeds up the processing of short jobs and, for this reason, users are urged to run jobs via the spool system that will be available on both the 1906A and 360/195. Preferential treatment will be given to work arriving via the spool systems. The LOADGO system is currently being modified to allow automatic logging of work through the FR80. This should be available within the next few months. All the packages generated at ACL will provide the relevant LOADGO records automatically.
Initial tests on the FR80 indicate that excellent colour output can be obtained. However, a great deal of experimentation needs to be done before a recommended colour chart will be available. To obtain a particular colour, it is necessary to specify the filters to be used, the intensity of the lines drawing, the number of times the line is overstruck and possibly the vector drawing speed. The aim will be to provide routines in the main packages which hide most of this complexity fom the user.
For the moment it is suggested that users keep to the primary and secondary colours (red, blue, green, cyan, yellow, magenta). Of these, the secondary colours (cyan, yellow, magenta) are preferred as these only require a single filter in the light path. Colour film processing is a new field for the Laboratory. It is possible that results in this area may not be completely perfect in the first few months.
The main set of software packages for the FR80 are already available on the 1906A. A considerable amount of work in rewriting existing packages has had to be done over the last few months. Due to limitations in the number of staff available, some rationalisation of the existing graphics software has had to be done. The main packages available on the 1906A are described below.
The major addition to the set of software available is the SMOG package which is a small basic graphic package allowing the user to access most of the facilities available on the FR80. This is considerably more efficient than the existing packages for the SD4020 and should be of great use to those programmers needing to incorporate graphics facilities in their own packages or large programs.
A complete manual for the system is currently being typed and should be available shortly.
Users of SPROGS should be able to run their existing programs with very little change. The main addition to SPROGS has been a set of routines which make it upward compatible with SMOG. It should be possible for all SMOG programs to run as SPROGS programs if the user finds he requires the facilities available in SPROGS. It is possible for files of SPROGS graphical output to be passed to a SMOG program.
Minimal changes have been made to GROATS so that existing programs can run on the FR80. There are unlikely to be major extensions to this package in the future. In particular, no facilities are provided for accessing the hardware characters of the FR80.
A complete implementation of the old SD4020 FORTRAN package has not been provided. This is oriented towards the SD4020 and has been superseded by SPROGS and SMOG. SCSIM is a set of routines that simulate the important facilities in this package. It should allow users to run existing programs but some changes are essential. It is recommended that, in the long term, users should convert existing programs to SMOG or SPROGS.
The system available on the 1906A will be extended and improved so that the user gets a number of subsidiary benefits for generating output for the FR80 via the spooling system. Already, it is possible to view particular frames before they are plotted on the FR80.
The basic hardware characters in the FR80 can be generated at a size between 12 and 277 raster positions. this is too small for titling on hardcopy or cine film. In the past, packages on the 1906A and 360/195 have contained their own character fonts and this is likely to continue in the future. However, characters generated in this way are quite expensive in computing time.
The basic software on the FR80 has been modified so that the packages may request a hardware character greater than the maximum possible size available. In this case, software within the FR80 simulates the hardware characters using vectors. The only character fonts available in SMOG are produced in this way. It is also possible to generate these pseudo-hardware characters using SPROGS.
Unfortunately, some of the character forms available in the FR80 fonts look rather bizarre when expanded greatly. This is particularly true of the characters A and W and is due to the limited number of vector directions available in the hardware character generator. Eventually, these characters will be redefined so that an acceptable font is available.
The main packages to be made available on the 360/195 are SMOG, SPROGS, SCSIM and MUGWUMP. Implementations of SMOG and MUGWUMP already exist and versions of the other two packages should be available shortly.
There is some systems work still to be done on the 360/195 before the spooling system is complete. It is not proposed to start any service on the 360/195 before this becomes available. Full details of the software will then be given.
All the graphical packages available on the 1906A and 360/195 can output to the microfiche camera on the FR80. However the main use of the microfiche camera will be for large volumes of textual information. Special programs will be available on both the 360/195 and 1906A to allow fiche with relevant titles to be produced instead of lineprinter listings.
It is possible for the FR80 to produce fiche in a variety of formats with a reduction in size between 42 and 48 times. This is a greater reduction than the usual library fiche standard of 24 times reduction. Consequently, viewers for these fiche are not appropriate for the FR80. Tests are being carried out at the moment to find an acceptable standard which is compatible with software on the 1906A and 360/195.
The current service on the SD4020 will continue until the end of September. Due to differences in camera images, it is inadvisable for a user to switch over to the FR80 if he is in the middle of a project where this is important. For example, a film sequence generated partially on the SD4020 and partially on the FR80 would be unlikely to merge. Line thickness, size and intensity are likely to be different on both machines. However, if this is not a constraint, users are urged to change to the FR80 as soon as possible.
Software support for graphics, as before, is provided in the first instance by the Program Advisory Office. However, since the FR80 is a new machine, it is likely that members of the Basic Software Group will be consulted frequently by PAO in the first few months. It is hoped that users will comment freely on the systems available so that improvements can be made.
