Scor­pio News

This is entirely superfluous as the file is opened automatically by the call to create. This occurs twice (very inefficient programming).

The bug in PCB is also easily fixed if you have a Pascal compiler, which I don’t. I did get round this as I shall explain later. The bug here is far more obscure. In track delete mode, mark one end of a track twice and you will find that the track i deleted (although it stays on the screen). If several tracks meet at the point you mark, you may expect them all to go, but actually only the first one defined is removed. The can be fixed simply by rearranging the list of conditions in a very long “if” statement, which I have no intention of copying it here.

I reported the HiSoft C bug to HiSoft, who replied quickly explaining that it is CP/M which is actually at fault. It relates to the keyhit() function which does not work correctly with rawin() due to the lack of direct console status checking in CP/M 2.2. Other than that I have found this to be a very good product despite the fact that I don’t possess Dr Dark’s library. In fact I have no books on C at all. This is my normal modus operandi having learned BASIC from some example programs (and the Nascom manual), Z80 machine code from the Mostek technical manual (an excellent document, but they would never claim it to be tutorial) and Pascal from nothing (we didn’t even have the manual (Oops!)). I must admit I failed to learn ALGOL this way but I blame that on not having the runtime package. I am learning C by the process of translating the PCB program from Pascal. This is not simply to learn C, nor is it to fix the bug (though I have done that). I find it entirely inadequate that it has no hard-copy options, so I am writing those into my C version as well as catering for larger boards. When I have finished it I may well make it available for a small sum.

While on the subject of learning to use things, I must disagree with Dr Dark regarding Gempen and Wordstar. I have used Gempen a bit but found it most inconvenient. Wordstar, however, seems fairly self-explanatory by comparison, and the main reason for the two Key commands such as ^KQ is to allow you to type text. As I recall it necessary to exit from typing in (With ^Z)before Q will work on Gempen thus making that a two key sequence also. Besides which, on machines like recent Galaxy’s there are function keys winch can be set to do that. And as regards 32 Megabyte files, I have frequently exceeded 50K or so and more than once I have gone over 64K with both documents and source file. I agee WordStar costs s lot, but it is a top quality product and you get what you pay for.

Regarding the Nascom anniversary “happening”, I would like to see something “proper” (i.e., a bit more than a few people in a pub) but I would not wish to see the cost get beyond the less wealthy among us. Equipment demos would be nice.

As to I.B.M.s i Scorpio News, it’s probably a good idea providing they are treated differently from the “colourful comes” (i.e. like you already do with 80-BUS stuff). The I.B.M. is rather less of a “black box” than most but generally still rather more of one than 80-BUS equipment, it would be nice to see it “exposed”

Yours faithfully, R. Pearce, Chadlington, Oxon,

RAM-disk makes life bearable

Dear Sir,

I totally disagree with Dave Hunt’s comments in Issue 1, p35 about the general uselessness of extended memory – either as pageable executable RAM, or a virtual disk (RAM-disc). I happen use two MAP 256K RAM cards (= 512K) as virtual disk with a Nascom-2 (32K paging), and find it almost indispensable.

I have never written – or intend to write (for the Z80) single programs which exceed the 62K TPA, but its primary function IS as a RAM-disk. It’s not a case of being something “bigger than you” (I don’t know any other Nascom users around here), it’s the fact that when editing large text files e.g. source code or data lists, the whole file can be held in RAM at once, allowing editing to proceed without disk waits. Wordstar’s swapping overlays in and out is unnoticable. I find waiting for disk access tiresome nowadays – I suppose it wouldn’t be such a problem if I was running a winchester like DH.

The speed of assembly and compilation is also greatly improved (with the output to RAM-disk, naturally), and making multiple copies of a disk is a soap when the whole disk can be held in