Forgive me, Nick, but if that problem is known to Microsoft then why have they not been able to tell me so in plain language (including reason(s) and solution(s))?
Maybe because there is no solution yet ? As I said before, third party software is suspected. I have IE8, and my CD/DVD auto-run works fine, but I may not have software resident on my PC that you have on yours.
I did mention reason as well as solution. All I have had from Microsoft is advice to ring a premium rate number. They have not said anything about knowing about the problem.
I'm interested though. Does Autorun work with all programs (including IE8) or just some (such as Acrobat Reader)?
Unfortunately, Roger, it's typical of too many software companies these days. They just let a number of volunteers use their partly developed software for a while in the hopes that the most common bugs can be sorted out before the software is released to the world at large.
You've just been moaning about auto-run not working on your computer, but it works fine on mine. How do you suggest that Microsoft tests its software out on millions of different PC configurations, and on millions of different permutations of hardware, without doing beta-testing with volunteers ? Even then some bugs get through.
I have to take issue there. So called "beta testing" is not testing. Proper testing requires that the input data, action of the software on the data and expected result are all known.
Hardware and software permutations are irrelevant because the individual software instructions are acting on a definable number of variables and it matters not how those variables are generated. A common example is validating a date. It matters not one jot whether the date on a particular computer is correct, all that matters is whether it is a valid date or not.
A proper test dataset will be built to include the appropriate set of variables - but that takes money and effort which too many companies are loath to expend.
There is absolutely no way in which the input data is known in "beta testing". Therefore, the expected result cannot be known.
Companies which use "beta testing" have no way of knowing whether (e.g.) one factor is tested a thousand times or a thousand factors are tested once. As I said previously, all they can do is hope that the most common bugs can be sorted out before the software is released to the world at large.
Graham
