On Mon, Oct 30, 2006 at 01:06:08PM +0100, Marc Haber wrote: > On Mon, 18 Sep 2006 04:33:49 -0700, Steve Langasek <[EMAIL PROTECTED]> > wrote: > >The error is, if you don't *need* a specific version of the package, you > >shouldn't depend on it at /all/. Essential means it's always available, so > >there's no reason for you to depend on it. > > I have never understood the reason for this rule, as it is bound to > introduce truckloads of RC bugs whenever a package is moved out of > essential.
The idea of essential packages is to avoid truckloads of dependencies on them. Packages aren't moved out of essential. For that reason, libraries (such as libc) cannot be essential, as the new package would move in and the old one move out on a soname change. Explicitly naming these dependencies doesn't help either, because hardly anyone will ever test things on a system where an essential package is missing. That is, even things like pbuilder and piuparts need a "minimal" system for testing, and that is defined as "essential packages only" (plus build-essential in the case of package building). Thanks, Bas -- I encourage people to send encrypted e-mail (see http://www.gnupg.org). If you have problems reading my e-mail, use a better reader. Please send the central message of e-mails as plain text in the message body, not as HTML and definitely not as MS Word. Please do not use the MS Word format for attachments either. For more information, see http://pcbcn10.phys.rug.nl/e-mail.html
signature.asc
Description: Digital signature
