80-Bus News


November-December 1982, Volume 1, Issue 4

Page 23 of 51


or a ‘custom’ version, maybe) the solution is to single-step through TOS until the addresses have been copied to RAM, then modify these addresses to suit, then proceed.

TOS cannot be used with NASSYS running in RAM. This is because the catalogue load address depends on the drive in use (i.e. Drive A catalogue must be loaded into the Drive A calalogue space in RAM) and not on the drive used to create the catalogue. TOS gets round this by ‘throwing away’ the header bytes into ‘ROM’ at location 0. It should be a simple matter (I haven’t tried it) to change the address to, say, £800 which is in the video RAM margin. TOS also uses absolute addresses to copy over the NASSYS subroutine table. Again, this is avoidable. TOS calls STMON, which initialises NASSYS, so just a simple look into $STAB should suffice (unless I’ve missed something). In fact (I know ‘cos I’ve done it) rewriting TOS to have its own PARSE routine and to look in $SSTAB as above releases a bit more space, enough for FOUR whole new commands! It also allows it to run on any version of NASSYS. The bit that initialises the SIO can also be rewritten to take up a fraction of its original length. The lack of a Save ZEAP command in the original version of TOS is to be deplored, especially, as I’ve said, there is enough room to slot one in.

It may also be of interest to note that a software package is available from HS Design in Scotland (from the same guys who designed the HSI1N) for HSIN systems, which greatly extends the commands under Crystal BASIC to include many disk-like commands. The command set looks very comprehensive, and would certainly be worth looking into if you use Crystal BASIC a lot. It will only run on the latest version of TOS (not earlier ones, though they will supply the latest version free with the BASIC Extension if you return the original ROMs), but I doubt very much whether it will run on my ‘custom’ version. I wouldn’t buy it, but only because a) I don’t have Crystal BASIC, b) I’d probably lose the extra facilities of ‘my’ TOS, and c) I don’t use BASIC all that much anyway.


a co eg he mt a

The documentation supplied comes in the form of a small manual which is A5 size. I’ve mentioned the misleading bit about the /NASIO circuitry, and also the appalling circuit diagram (which is so bad it looks more like a wiring diagram). No source listing of the software is supplied either, so what does the manual contain? Well, basically it explains the operation of the various commands fairly well (enough to enable you to save files without getting it wrong) but I found the format a little cramped. An A4 size manual with more help on using TOS under program control (I keep on about this, but since it is advertised as ‘looking like’ a floppy disk system & all floppy disk users use operating system commands to load/save data from their programs, I don’t see why TOS users shouldn’t either) would also be useful (or perhaps a commented sources listing – I’ve done it myself but probably didn’t get it all right, and there’s still one or two bits I still don’t understand). Some information on the drives themselves would also be useful.


I know very little about the ‘Hobbit’ system except that it uses the PIO, so all comments here relate to comparisons with the ‘CFS’ system. CFS also writes in blocks of 2K, also at 6000 bps. CFS commands are ‘menu-driven’ but are less extensive than those of TOS (they do include a ‘save ZEAP’ command, though). CFS data format is a syne. byte, followed by a load address word, followed by a length word, followed by the checksum, followed by the data bytes. This is therefore incompatible with the HSIN format, and CFS systems will be unable to read HSIN tapes, and vice-versa. One day I might ‘con’ a CFS user to help with some routines to overcome this. If and when they are finished I might write it up into an article for the mag. (By the way, can 80-Bus News accept NASPEN files dumped onto these min-cassettes? No, I thought not).

This is an OCR’d version of the scanned page and likely contains recognition errors.

Page 23 of 51