This paper is being re-issued as the first of a series of graphics notes on subjects not directly related to SPROGS. Originally issued as SPROGS Paper no. 42.
It has been suggested that the PDP15 be used for a system which reads data from the DMAC digitizer, displays the points on one of the VDUs and sends the data to a fast paper tape punch. This paper outlines the software already available which, if put together, could go towards providing such a system.
There is a routine STYLII in the PDP15 library which reads points from all the graphic input devices including the DMAC. It requires a user provided subroutine, which is called each time a point is read; this user-subroutine could update the display image and add the 2 co-ordinates to an array. STYLII requires - a particular format to be set up on the DMAC itself (,XXXX<space>YYYY<cr><reset1>) and it makes use of certain inputs from the DMAC keyboard to indicate a break in a continuous line (S<cr>) and an exit from the current group of points (E<cr>). STYLII can be made to scale DMAC co-ordinates to VT15 co-ordinates or leave them unchanged; in the latter case, the calling program must scale the point in order to display it, but the original accuracy of the DMAC is retained. STYLII will thin the data by time or distance if required.
If STYLII is thought to be too cumbersome, the DMAC co ordinates can be read using a READ/FORMAT statement in FORTRAN; the FORMAT statement would determine the format that needs to be wired up on the DMAC.
The documentation of STYLII is section 4.6 in the PDP15 Routine Library.
It is not known whether editing is required or simple monitoring only, but this will largely determine the amount of programming that needs to be done. A compromise could be a command to delete the previous track.
If simple monitoring is required, then the existing routines to draw a line and position the beam on the VT15 can be used. Analogous routines for the Tektronix are not yet available.
The intended final medium is paper tape, presumably the numbers are to be ASCII coded, The BSI could be used to send the output to the 1906A and the data can be accumulated in a GEORGE file - the software exists to do this. Paper tape can then be generated using LISTFILE. There seems to be no disadvantage with this method compared with attaching a fast paper tape punch to the PDP15. The program would treat them in a similar way.
The program would have to provide markers in the output in order to divide the user's data into logical subgroups. One of the important parts of the program would be to provide the user with commands to intersperse his data with text.
Documentation of the BSI software mentioned above will shortly be available in the SPROGS Technical Paper series.
(1) If a simple monitoring system is required soon, one could be written using existing software; the additional programming would need to deal with the user's interaction with the system (start reading from DMAC, read user's text from keyboard, output current subgroup to disc or directly to the 1906A, etc) and the grouping of data into subgroups. The user's interaction with the system will initially be quite simple, but if the system proves to be useful, facilities will no doubt have to be added.
(2) An editing system, using the lightpen perhaps, would require more careful planing.
(3) Some decision should be made about the format of the final output. This depends on which other systems will process this data and what format they require.
(4) It does not seem necessary to add a fast paper tape punch to the PDP15 in addition to the existing so-called high speed punch. At present, the BSI software does not transfer binary data to the 1906A, but the project in question should transfer characters because of the additional information included in the data. Any programming that may be necessary (conclusions (1) and (2)) will not be increased by using the BSI.
(5) Depending on how urgently a system is required, it may be worth waiting for further software to be available.
If the system is required soon, then it will be necessary to use DOS. Eventually, though, such a system should be available under RSX, since the use of the DMAC is fairly lengthy and other jobs should run in parallel. In a few months, PIGS, the user interaction system, should be ready (under DOS!), in which case part of the work mentioned in conclusion (1) will already be done; it will also be easier to extend the available facilities.
