Re: PROPOSAL: "Staging Areas"
- To: email@example.com
- Subject: Re: PROPOSAL: "Staging Areas"
- From: Guy Maor <firstname.lastname@example.org>
- Date: 28 Feb 1999 23:09:40 -0800
- Message-id: <email@example.com>
- In-reply-to: Anthony Towns's message of "Fri, 26 Feb 1999 20:04:22 +1000"
- References: <19990225143451.O450@cs.leidenuniv.nl> <Pine.SOL.3.95q.990225142007.18017Gfirstname.lastname@example.org> <19990226103914.A18286@azure.humbug.org.au> <email@example.com> <19990226200422.A21327@azure.humbug.org.au>
Anthony Towns <firstname.lastname@example.org> writes:
> I quote from the developer's reference:
> ] Developers should be very selective in the use of the _experimental_
> ] distribution. Even if a package is highly unstable, it could well
> ] still go into _unstable_; just state a few warnings in the
> ] description. However, if there is a chance that the software could do
> ] grave damage to a system, it might be better to put it into
> ] _experimental_.
> That seems just a little bit strong
There are other reasons to put something in experimental besides a
potential for grave damage. You could certainly put every
incompatible change into experimental before putting it in unstable.
Most people won't even try them, and the ftp administrators will get
irritated at you. Use your judgement.
> People *do* run important servers off unstable
Well, they shouldn't.
These sub-distributions are needed when we'd like to stabilize
something that isn't so stable upstream.