Acceptance tests on the FR80 were completed on10 March and it arrived at Heathrow on 14 March. It is hoped that the machine will arrive on site within the next week.
Some of the points made in Project Note 10 are not completely accurate:
The progress of work was reviewed at the meeting in relation to the chart in Project Note 9.
The ability to DELAY jobs has been implemented but not tested. JRG requires some delayed jobs in the SPOOL. ACTION: FRAH
It was decided that the main packages should provide routines for delaying and erasing jobs from the SPOOL. A job marked as delayed and ready for erasing would not be erased until the delay period had expired.
It was agreed that a SPOOL utility program was required which could delay a job already in the SPOOL. After some discussion, it was agreed that the Tektronix VIEW program should allow the user to erase the SPOOL job at the same time as it gets read into the filestore. ACTION: JRG,AWB
Most of SMOG for the FR80 has been implemented. The ID frame needs to be generated. Line drawing for the Hewlett Packard has been implemented. Tektronix output is complete.
A major problem has arisen in the defined method of implementation. Permanent pictures which should have been generated as relative pictures sometimes contain absolute line drawing.
In order to make implementation of the Tektronix handler easier and because it was thought that a text output command would leave the beam position undefined, it was decided that routines such as SETXY would not generate any FR80 commands. It was also decided that two beam positions would be used - a current typing position and a current drawing position.
This leads to the following code being produced:
TOXY(X,Y) FRSAV (3.0) TODXY(DX,DY) FISAV
A relative file containing a single line would be defined.
SETXY(X,Y) FRSAV (3.0) TODXY(DX,DY) FISAV
This does not produce a relative file. The file contains commands for setting the beam position to (X,Y) before those to draw the line (X,Y,DX,Y+DY).
TOXY(X,Y) FRSAV(3.0) TODXY (DX1,DY1) UPDXY(DX2,DY2) TODXY (DX3,DY3) FISAV
This produces a file consisting of a relative line followed by commands to set the beam absolutely to (X+DX1+DX2, Y+DY1+DY2) followed by a second relative line.
Neither of the last two results are what a user would expect. The need to have a separate typing position is no longer necessary if use is made of the fact that a text output command leaves the beam position at its initial position.
At a separate meeting, it was decided that the definition of SMOG would have to be changed in order that the obvious commands were produced. The following changes were decided:
HTYPE ('A')which has a single argument, a one-character string, will output the character defined in the same manner as HPLOT.
ADVFLM SETXY(XMN,YMX) HTYPE ('/')The routine HIINES (N) is defined as:
HTEXT('////...//')It is assumed that / is the newline character.
This is now working apart from (possibly) minor bugs.
A version of the macro is available for testing purposes. It is held in file :FSPOOL.VIEW and1will produce a maximum of 30 files.
Several considerations must be taken into account when using this macro:
Examples of the use of VIEW:
RJ TONYMOP1,:FSPOOL.VIEW,PARAM(TONYMOP,FRAMES1-5,FRAMES7) RJ TONYMOP2,:FSPOOL.VIEW,PARAM(1,FRAMES23-END),JD(JT30)
The specification is as in FR80 Discussion Paper 9.
Most of SCFOR is now working on the 1906A. It appears to be considerably smaller and faster than the SD4020 version. A comparison will be made of several programs run on the two systems. ACTION: DVR
Both the FR80 handler and the SMOG routines have been implemented on the 360/195. Due to the large size, about 50 of the routines have been rewritten in ASSEMBLER which resulted in a considerable saving on storage:
SMOG (FORTRAN G) 80 kbyes SMOG (FORTRAN G and ASSEMBLER) 57 kbytes SMOG (FORTRAN H and ASSEMBLER) 47 kbytes
Chris Osland is currently testing the HASP Internal Writer.
It is hoped to have a single Load Module for both SPOOL and tape output.
A program has been written to disassemble FR80 binary files. This currently produces a listing giving the mnemonic correctly but with the address in octal. With each binary file is stored a Symbol Table. A later stage will be to use the Symbol Table information to produce a listing containing symbolic addresses and labels.
A listing of the FR80 Displayer provided for ACL's system has been obtained using the disassembler. Work in the next weeks will be concentrated on defining the necessary alterations required to simulate Large characters by vector drawing. AHF requires the FR80 Displayer commands for outputting large characters as soon as they are available and they may well be needed for other packages.
A basic accounts program has been defined and written by RET. A complete definition of this needs to be produced indicating the name (or names) of the program to be called via the Load and Go Header record. Any operator commands and data for the program must also be defined. This is required by all the package implementers. ACTION: WDS
The Accounts Program is likely to dump output to tape once a week. A 1906A program is required to take this tape and generate cards for WJ's accounts program.
Investigation of the various addressing modes on the FR80 and PDP15 has revealed a number of inconsistencies in the published manuals. A paper will be produced describing all the addressing modes possible on the PDP15. ACTION: RWW
After consultation with PMN, it was decided that the same resolution sub-picture should be used on both 16mm and 35mm. Consequently, only permanent pictures 1, 2 and 3 are now defined. These must be compiled into the copies of the FR80 Displayer in use at ACL.
Some discussion arose as to the sensible default setting to be made by the packages for various cameras. The main problem is with hardcopy. It was finally decided that the default setting would be CINE. If COMIC mode is used with the MANY-UP facility, it will be difficult to work out the ordering of frames in the middle of the roll.
The standard settings are defined as follows:
16mm CINE 35mm CINE Hardcopy CINE Microfiche COMIC
It was decided that the standard Forms Flash to be used in the SD4020 simulator will be blank. As the Forms Flash does not get scaled, this is the only sensible decision. If MANY-UP is being used, the Forms Flash is produced across the complete set of sub-frames in a frame.
In the period before a service is started, it will be necessary for software development on the FR80 to co-exist with operators running test jobs for packages and SPOOL. Due to the small size of the FR80 disc, it is necessary to remove the basic system while doing program development. Consequently, booking periods of less than 1 hour in length will not be very useful. A booking system will be organised by RB.
After testing out various identification frames with the operators, the following set has been defined:
(1) 16mm: Blank Frame Username Frame ID Frame (User's output) ID Frame Blank Frame (2) 35mm: Blank Frame ID Frame (User's output) ID Frame Blank Frame (3) Hardcopy: Blank Frame (complete frame, not sub-frame) ID Frame (User's output) ID Frame (4) Microfiche: New Fiche ID Frame (and title) (User's output) ID Frame
The definition of the Username Frame on 16mn: is as follows:
Each region is specified in absolute FR80 raster positions as (XMIN,YMIN,XMAX,YMAX):
(a) Top Bar (filled in region using vector family command) (5727,9670,10657,10150)
(b) Bottom Bar (filled in region) (5727,6220,10657,6700)
(c) Username Region containing 6 characters (5727,7699,10657,8520)
Each character is output in a 821 x× 821 square. The username is output in CINE mode.
The vector family commands required are:
0010 5727 ) set X,Y 0001 9670 ) 222 29 31 lines at intervals of 16 rasters 1100 4930 DX O111 480 DY 1100 4930 DZ 0010 5727 0001 6220 222 29 1100 4930 0111 480 1100 4930
The format of the microfiche title command given in Project Note 6 has been changed by III. The new format is as follows:
TIT/T1C2LIH1,1/:GSIN00 $MC2L1H13,1/FICHE$3 1$M$C& TIT/C C5L2H6,1/ up to 30 chars title $1 another 30 $T&
Note that second line must not exceed the buffer size in total.
Important points to notice: