|★ LITTÉRATURE ★ ENGLISH ★ THE ANATOMY OF THE CPC'S|Amstrad Action) ★|
|The Anatomy of the CPC's||Littérature English|
Like many of First's books these days, this one hails from German computer-book publishers Data Becker. Unfortunately it has lost a certain something on the way from Düsseldorf to Pangbourne. At one point for example, it assures us that 'The SCREEN PACK is slaved to the TEXT and GRAPHICS PACKS. It practically serves as their executive.' Later on in the same section it refers to the action of control code 07 as 'Klingel'rather than the 'Bell'you might have expected. Minor niggles perhaps, but clarity's important in technical books. There's no sense putting extra obstacles in the way of understanding, when most authors can create plenty of their own.
The book breaks down into three main sections: hardware, operating system and Basic. The hardware section is not particularly noteworthy, taking for the most part a chip-by-chip analysis of the CPC's innards. While this can be quite interesting, it contains very little that Amstrad's firmware guide doesn't handle better. A fair amount of effort is wasted describing features that Arnold's design prevents the user from getting at. Do you really need a book about what Arnold might have been? To make things worse, there are quite a few omissions and inaccuracies. The authors claim that they could only find information on half of the possible RAM configurations on the 6128, even though the missing configurations are at least partially documented and can in any case be worked out with a little effort. Why they didn't make that effort I'm not sure, but galling to pay £15 for a book and then have to work it out yourself.
They also misunderstand the conditions under which key-clashes occur, and state boldly that such clashes have no adverse effects. In fact key-clashes can cause serious problems in games and make twin joysticks almost useless. It is important that programs requiring key combinations are written with this in mind, and the key combinations chosen accordingly.
INTO THE ROMS
The sections on the operating system and Basic are where the book starts to earn its price. These are built round near-disassemblies of the lower and upper ROMs respectively. I say 'near-disassemblies'because there are no details of the actual source code itself. Rather there are extensive notes on the precise function and structure of ROM routines.
The idea is that you print out your own disassembly of each ROM, and then use the book as a set of notes explaining each section. If you don't have a disassembler, fear not: there's one in the appendix for you to type in. If you don't have a printer, on the other hand, you're probably wondering why the book doesn't contain a source-code listing as well as the admittedly comprehensive notes.
I'm not entirely convinced that lack of space is the reason: most of the time the notes occupy only the left side of each page, and frequently the hex or source code would explain itself perfectly well anyhow. I don't know whether legal worries intervened, but the ROMs are copyright and this may have something to do with it. Or maybe Schneider owners all have printers.
One thing even printer owners should ask themselves about is the actual usefulness of these listings. For starters, they only cover the 6128 ROMs. While the authors are at pains to point out the similarities between the different CPCs, this is bound to cut down the book's appeal to 664 and more particularly 464 owners.
More importantly, rummaging around in Arnold's ROMs has little or no practical value. Routines in the lower ROM can and should be called via the firmware jumpblock, as the authors themselves state at one point. Calling ROM routines directly makes for compatibility problems, and prevents interception of operating system calls.