Ben Finney writes ("Bug#331532: [PROPOSAL] change §10.4 "set -e OR check return status" to AND or be rewritten"): > The final question is: Why is this necessary in policy? If this were > not implemented, would any of these programming errors *not* warrant > an important bug? If it's already grounds for a bug report, what good > is done by proscribing it in policy?
The lack of `set -e' is the single most common mistake in maintainer scripts. Or at least, at the time I added that sentence, it was. I agree that the policy manual shouldn't be a general repository for listing bugs not to have. But surely it's sensible to mention the one or two obvious bugs that so many people get wrong ? Like any bit of the policy manual, it's there to tell you how things probably ought to be done. If you have a good reason not to do as it says then you should do something else ! Ian.