NOTE: Abbreviations are Hopgood:FRAH, Brandwood:RB, Burraston:AWB, Chiu:MFC, Francis:AHF, Gallop:JRG, Lewis:JWEL, Nelson:PMN, Ralphs:DVR, Rushby:JMR, Shaw:WDS, Thomas:RET.
FRAH reported that the estimated deliVery date of the FR80 was still early in March.
The operator interface for emptying the SPOOL area had been defined after discussions with RB. A definition of this would be produced. RB commented that the only time that the SPOOL had been lost in the past, it had been impossible to get information from.it at all. Consequently, the only way to find out what programs needed to be re-run was to look at the messages produced on the console log. As the FR80 SPOOL would not output a console message for each job, there was a need for some other method of finding out which programs had been lost. It is possible that the job could output a message to SYSTEM JOURNAL and for these to be retrieved in the event of a disc failure. FRAH will discuss this possibility with CJP.
JRG had started coding the SPOOL read-out program. This work was on schedule.
The vector routines have been tested and work correctly on the Tektronix. One version of LIMITS has been produced and tested. The others should be straightforward. The TEXT routines were being debugged. Currently, these were having to search all the text for newline characters which meant that text generation on the 1906A was quite slow. The routines have been set up in a library which now works. The SPROGS handler for the FR80 has been added to SMOG and appears to be working properly. A magnetic tape that was generated appeared to have the correct orders on it although a detailed study was still needed.
Some parts were ahead of schedule and some slightly behind. A number of problems with the definition of SMOG had arisen and it was decided to have a separate meeting to clear these up.
Output to the SPOOL is working with the correct header records for the LOADGO system. The implementation of the new SMOG command in SPROGS had started although there were some definitional problems.
The SPROGS Supplement has been produced. Version III of SPROGS woulq be released on 7 February. The next release would be the FR80 one.
Testing of the SPROGS simulator for SCFOR was continuing. This had been held up by a number of bugs in SCFOR itself. It was decided that routines for line thickness would not be available in the SMOG simulator unless users felt it was absolutely necessary. RET was currently recoding the SPROGS version of the simulator so that it used its own facilities similar to those available in SMOG. Production of the SMOG simulator would then be much simpler. RET hoped to have the simulator available before he left for the USA.
CAMP had been recoded so that it called SMOG vector drawing routines. It was now held up waiting for a version of SMOG to use .
One problem has arisen with the ID frame. Although a complete square frame can be output on most devices using the same part of the raster, this is not possible on 16mm. As this only uses a much smaller part of the complete raster, outputting the same ID frame would mean that only the centre part would be visible. In any case, it was difficult to read the username even if the complete ID frame was output. It was decided that:
The utility to view SPOOL information on the Tektronix had been defined. This produced a separate file for each frame. Coding was complete but not tested. There was some discussion as to whether the RETAIN or ERASE flag took precedence. It was agreed that information could be put in the SPOOL with both flags set. This would allow the information to be viewed and, if no action was taken, it would eventually be erased.
The output to SPOOL was now working and the definition of the header records had started but no coding has been done. This should be complete in the next two weeks.
Some preliminary work had been done in estimating the problems associated problems. associated with defining a simulator.
RET reported that little work had been done on the SPOOL system. The system had still not been specified. The modifications to allow internal writers in HASP had not been obtained yet from abroad.
MUGWUMP had been defined in terms of SMOG and was now waiting for a 360 version to be implemented.
A version of SPROGS is now working which generates graphical output on the Tektronix or the lineprinter. However, the amount of core required is about 250K bytes. This must be reduced to at least 210K bytes if it is to be usable. The current system is compiled using the G compiler. Compilation using the H compiler is not possible at the moment due to bugs in the compiler. In particular, it cannot handle call-by-reference arguments. SPROGS will be redefined on the 1906A to avoid arguments of this type. The main priority should be to get an FR80 version working. Reducing the core size can be left for the moment.
Due to the size of SPROGS, it seems sensible to produce an FR80 version of POLYGRAPHICS using the SMOG output routines.
The colour film processor had begun to fall to bits after only one week's use. It would be replaced by another one. It is possible that the original machine was a demonstration model.
It now looked as though the FR80 hardcopy camera would only be able to take 12-inch paper. There may be some problems initially in getting a supply of this paper from Kodak. Consequently, an order for paper had also been placed directly with III.
The operators had been on their photographic course. PMN and RB will ask Harry Edwards to give a more advanced course for both operators and programmers.
It was still unclear whether there would be a sufficient head of water for the film processor, As it would take 2 months to purchase a pump, this decision must be made in the near future.
It was:decided that when more than one plot was being output per frame, the individual plots would be called sub-frames. Thus a microfiche frame would be made up of many sub-frames.
The form of the fiche title monitor command described in FR80 Technical Paper No 2, page 6, has been amended to provide automatic numbering of the fiche and to include, in Large type, the username of the job which generated the fiche. Up to 60 characters of user-supplied titling information may be included.
Assuming 42X fiche (16 columns/fiche), the fiche title will appear as:
The form of the commnd is:
TITLE/T1CL1H1,IV1/:GSIN00 $MC2L1H13,IV1,1/FICHE$3 1$M$$& C C5L2H6,1V1,1/ up t 30 chars title $L another 30 chars title $T&
The symbol & is used to represent the character 215 octal.
The package must insert the correct 8-character username in place of the :GSIN00 above.
