> I seem to recall that Joey Hess once filed a bug against one of my > packages (xprint-common probably, though I can't confirm it) addressing > this kind of question. I think there's a debconf "best practice" which > he wanted me to adhere to, something along the lines of not explicitly > mentioning "yes" or "no" at all, since you never know which kind of > front-end the debconf question will be presented in. Just ask the > question in the text, and let the buttons provide the answer, something > like that. Yep, that's in the developer's reference. I often happen to file BR about this. > > Might be an idea to check the debconf docs or ask Joey to be sure. > > I'm cc:ing Joey anyway since I'm being a bit vague, but I'm pretty sure > there's a best-practice to provide guidance. "Accept" or "Decline" has been sometimes recommended but I actually prefer recommending something like "Choosing this option will blahblah". This bug report indeed drives the intent for a general rewrite of the debconf templates to have them comply with what is written in the developer's reference (FWIW, *I* am the one who wrote that part of the DRE, with the help of Joey with the earlier versions of the document) I have already mentioned I volunteer for doing this. Of course, before committing such changes, I'll probably ask for feedback to debian-x. However, David mentioned plans to rework the debconf stuff and we'd probably better have this to happen before major rewrite work is done.
Attachment:
signature.asc
Description: Digital signature