I never cease to be amazed by the way computer manufacturers have the nerve to release computers that have not been properly debugged.
Now, I realise how hard it must be to fully debug 32K or so of code, but it should be done. The manufacturers should lend the computer they are about to release to various people, to see if they come across any problems.
Take the BBC for instance. When I bought my BBC about 1.5 years ago it was fitted with the OS 0.1 This operating system didn't allow any use of disks, the RS423, the tube, Econet or sideways ROMs, and the BBC was claimed to be expandable. BBC Basic 1 has bugs in it, and even the humble ZX81 has better mathematical accuracies (hence the development of Basic II). To upgrade my operating system cost £10.
It seems that these computer manufacturers expect the people who went out to buy the machine early (on good faith) to put up with anything.
The QL is another computer that is riddled with bugs. On paper it looks an extremely good machine, but in the flesh? Well, that's another matter. Even if you can have 255 colours on the screen at one time (with 'stipples') as Mr. Lawlor is quick to point out in its defence, that is no excuse for a machine that does not do things as it should do.
What is the point of 128K RAM if the Basic needed to use it is lacking, and the documentation is bad?
It seems little wonder that so many companies are in difficulties - people won't buy their machines because the documentation is so badly written (Take the Dragon for example) or the machine isn't debugged (the QL). Sinclair's Quantum Leap seems to be a very small leap indeed - quantum sized in fact.
22nd September 1984
