On Tue, Feb 27, 2001 at 09:59:19AM +0000, Julian Gilbey wrote:
> On Tue, Feb 27, 2001 at 10:45:32AM +1000, Anthony Towns wrote:
> > If, after consideration, you think that one or more of the
> > recommendations (SHOULD) or requirements (MUST) in this document
> > don't apply to your package, then you should post to -devel to see
> > if there's some alternative way you can do things in your package
> > or see if policy should be changed, and document what you've done
> > in README.Debian.
> This is also a nice piece of advice, but is orthogonal to the
> suggestion being made.
Uh, reread Sam's message: he was saying that there would be a number of
guidelines that would always have exceptions, I was disagreeing.
> The suggestion is that something like the /usr/doc -> /usr/share/doc
> transition is a MUST,
No. It is not.
The /usr/doc -> /usr/share/doc transition was intended to be a
three-parter:
1) /usr/doc/<package> *must* work, but it *may* (in a
strongly-encouraged sense) be a symlink to
/usr/share/doc/<package>
2) /usr/doc/<package> *must* be a symlink to /usr/share/doc/<package>
3) docs must be in /usr/share/doc/<package>
> but that we will record somewhere, as an
> appendix to policy or with footnotes in situ that as these are
> (relatively) new policy amendments, we should not yet be filing
> (normal|RC) bugs against offending packages, i.e., those which do not
> *yet* follow the new policy.
Yes, this would be possible. It'd also be a bunch of completely pointless
work.
Document what's allowed and recommended. Nothing more. Nothing less.
If what's allowed and recommended changes, well, change policy then. Not
before.
I'm really not seeing what's so hard to understand about this. Policy
should document what we expect of packages right now. If we don't care if
packages haven't been updated, then that's what we should document. We
shouldn't pretend like we care, then contradict ourselves later and say
we don't care ("Packages MUST use /usr/share/doc [0] ... [0] Except they
don't have to if they don't want to").
"Must" directives need to be used *very* sparingly. We're allowed to
distribute buggy packages.
Cheers,
aj
--
Anthony Towns <aj@humbug.org.au> <http://azure.humbug.org.au/~aj/>
I don't speak for anyone save myself. GPG signed mail preferred.
``_Any_ increase in interface difficulty, in exchange for a benefit you
do not understand, cannot perceive, or don't care about, is too much.''
-- John S. Novak, III (The Humblest Man on the Net)
Attachment:
pgpHzaqn9dqR9.pgp
Description: PGP signature