80-Bus News

number of spaces, these are contained in a message and displayed in the normal way through SNDTXT. Now to display the word, this is done in a similar fashion to the one already described in LOOP4. The only difference is that each character is tested to see if it is flagged. Each flagged character is ‘visible’ so it is ANDed with 7FH to strip the flag and then displayed directly, each unflagged character is replaced by a ‘.’ which is displayed instead. A space is placed between each character. When 00 is detected, the loop cops out in the normal way, and a call made CRLF2 to send two newlines. After that the program jumps back to LOOP3 to start the whole process again.

Well that just about wraps up that program, with the above comments and the liberal sprinkling of comments all over the program itself, you should be able to follow it. Now to improvements, well obviously if you have an assembler you can reassemble the whole thing into it’s right places for either NAS or CP/M versions. If you’re going to have a go at the NAS version, then might I suggest that you put the ORG at 1000H as you could unwittingly end up with problems if you extended the words table from its current end at 0FFCH. You see, NAS-SYS corrupts the bytes 0FFEH and 0FFFH during its ‘E’xecute procedure as it uses them as intermediate stack space, so if you had data in those two bytes then some funny things could happen. There is no easy way of making the program avoid those bytes so best to reassemble the whole-lot above it. If you’ve got an assembler it implies that you aren’t stuck for RAM space, so what the odds. The words table may be extended to 255 words (the largest random number that can be generated), make sure NOWRDS contains a count of how many words there actually are. There is scope for tidying up the displays although the method used here was chosen so I didn’t have complications with the output routine. In fact it was deliberately made a simple as possible. I remember my handwritten version kept the screen in one place and didn’t scroll. It achieved this by moving the cursor about and overwriting the text for each successive turn round the loop. There is one version in Basic I’ve seen which incorporates some spectacularly gruesome graphics, complete with swinging man and stretchy rope. Lot’s more fun to do in assembler. People reassembling this for a NAS regime could save quite a bit of space by dispensing with the input/​output patches and calling B2HEX direct from NAS-SYS. As should be apparent B2HEX had to be included in the program as there is no CP/M equivalent within the system.

I hope that you’ve noticed that by feeding all input and output through simple user patch areas this program could be made to work with very little modification on almost any Z80 based machine. It could also be modified fairly easily to run an any 8080 based machine simply by substituting absolute jumps for the relative jumps I’ve used (and modifying the one or two DJNZ loops used). This technique of providing user patches is going to play an ever more important part in programs published in this mag due to the diversity of machines we intend to support in future. So if you are thinking of writing a program for inclusion in the library then keep in mind that flexibility is the name of the game, and user patches will become very useful.

Note that the RANDOM routine works completely stand alone, all it needs is its three byte work space and to be fed with the maximum number to pick. I’m told this particular routine is very nicely random, quite a bit better than the ones found in a good many Basics.

So here endeth the lesson, as I said, I haven’t tried it under a NAS regime, but it should work. I know it works under CP/M, and as it only uses two system calls it has no excuse for failing under RP/M. The 3K waste space penalty is ‘a nothing’ when used with disks, for the RP/M tape types, sorry, but it takes about 45 seconds to load when it should only take 4 or 5. I hope you will have a go at it, and that it gives you as much fun getting it to work as it gave me in writing it.