Le jeu 12 octobre 2006 07:45, Cameron Dale a écrit : > On 10/9/06, Pierre HABOUZIT <[EMAIL PROTECTED]> wrote: > > I'm not sure, IMHO that's important, but as python-support > > depends upon python himself, the correct chain of depends exists. > > so that's not a major problem, merely an important one. > > I've done some more investigating, and I've reconsidered the severity > of this bug. I now think that my initial idea to make it an RC bug > with severity serious was correct. This problem, though it will in > fact build, install, and run fine, is still a violation of the python > policy, section 2.4 and 3.1, which specify:
that's not true. > Section 2.4: Packaged modules available for the default Python > version (or many versions including the default) as described in > Module Package Names, Section 2.2 must depend on "python (>= X.Y)" you must depend upon python >= X.Y iff there is a constraint on python, and that case just works in python-support. > Section 3.1: Programs that can run with any version of Python must > begin with #!/usr/bin/python or #!/usr/bin/env python (the former is > preferred). They must also specify a dependency on python, with a > versioned dependency if necessary. that dependency is implied by the python-support dependency, so it's ugly, but exists, hence the important severity, not the serious one. > As a result of this problem, I am considering resetting the severity > to serious on this bug, or opening a new one with this severity. I'd > rather not do this though, so please feel free to present some more > reasoning on why this is not a serious bug, and I will consider it. I'll try to bug joss again so that he does not forget that bug again, but no there is nothing broken, only an annoyance. lintian is not the Truth, it's a helper to be near it. What you look at is not even an Error but a Warning. -- ·O· Pierre Habouzit ··O [EMAIL PROTECTED] OOO http://www.madism.org
pgpVCWbYQTVfM.pgp
Description: PGP signature
