On Wed, Jun 29, 2005 at 12:42:45AM -0700, Steve Langasek wrote: > On Tue, Jun 28, 2005 at 12:34:01PM +0200, Arjan Oosting wrote: > > > > > If it really should not be done, maybe a check could be added to lintian > > > > and/or linda and CDBS and the policy should be updated. > > > > > > There *is* a lintian check for this, and there is no need for policy to be > > > updated -- just for you to follow it. > > Well my package *is* lintian and linda clean, so I guess the check is > > broken or I am doing something awefully wrong ;) I will attach the debug > > output of lintian. > > $ echo 'E: foo source: > depends-on-build-essential-package-without-using-version' | lintian-info > E: foo source: depends-on-build-essential-package-without-using-version > N: > N: The package declares a depends on a build essential package without > N: using a versioned depends. In general a package should not depend on > N: build essential packages but if it must do so, the depends should have > N: a version string. > N: > N: Refer to Policy Manual, section 4.2 for details. > N: > $ > > Yeah, so I don't know why this check isn't being triggered if it applies to > your package. The version of gaim-extendedprefs in the archive doesn't seem > to have any build-deps on build-essential packages, so I can't compare.
The 'build-essential' package itself is not build-essential, it's a helper package depending 'accidently' on all build-essential packages. In the context of build-depends, a dependency on 'build-essential' is essentially a no-op, but this isn't catched by the above-mentioned lintian check. --Jeroen -- Jeroen van Wolffelaar [EMAIL PROTECTED] (also for Jabber & MSN; ICQ: 33944357) http://Jeroen.A-Eskwadraat.nl