Re: /media for sarge?
sean finney <email@example.com> schrieb:
> On Wed, Apr 28, 2004 at 05:41:40PM +0200, Frank Küster wrote:
>> > i think the issue was more of annoyance than impact... if a local
>> > admin removes said dirs, he/she'd probably like them to stay gone.
>> How can he have removed them if he's upgrading from woody? The postinst
>> script could easily check the version from which it is upgraded.
> i addressed that in my next paragraph:
>> > would be good for? that is, in the maintainer scripts, if the version
>> > change is from pre-fhs to post-fhs, ask the user "upgrade to latest
>> > fhs spec?" this way, i think everybody would get what they wanted.
That doesn't address my point: I think that Santiago'a argument ("we
cannot create it, because the local admin might have removed it on
purpose") isn't particularly strong, because we can easily check whether
we are upgrading from a pre-fhs version (as you name it), where he
cannot have removed it.
>> Yes, and everybody would have to read the latest fhs document before
>> deciding. Or you explain in long detail why it might be clever to create
>> those dirs, and under which circumstances it might be disadvantageous
>> (are there any?).
>> This is debconf abuse, I'd say - there's a safe default for everybody.
> i'd argue that this is exactly the kind of question you'd want to have
> asked in debconf. the fact that there's a safe/reasonable default just
> means that there's no reason to set it at a priority of mid (or better low)
If there's a reasonable default, you shouldn't bother the admin at
all. Nobody has, so far, given any reason _why_ an admin should want to
remove them. There will be some people who do, for sure - but then let
them do it, and give the directories to all those who might finally find
> the people who are annoyed by too many debconf questions already have
> their minimum priority in debconf above this, people who want to have
> a better idea of what their packages are doing or have more control
> get the question,
I do want to have a better idea, and control over what packages are
doing. But I don't want to be asked questions where 99% of the users
will give the same answer, and there's no risk at all for the remaining
Frank Küster, Biozentrum der Univ. Basel
Abt. Biophysikalische Chemie