Steve Langasek wrote: > It's entirely possible to address this in the way that Michael describes, at > the cost of added complexity in dh_installinit (more maintainer script > snippet combinations, etc). Considering that there are only a handful of > packages that will ever be affected by this issue (specifically, those few > packages which are covered by the exception in policy to ship upstart jobs > that don't have matching init scripts because they're part of the "core" of > the boot system), I felt that having the complexity in the shell snippet was > the right trade-off. However, if you think it's better to move the > complexity to dh_installinit and have "clean" shell code, I'm happy to do > this.
It's not a whole lot of complexity, right, we're just talking about omitting the update-rc.d call, which AFAICS can be done without splitting anything in autoscripts/, but just avoiding including them. > And there is at least one lintian bug: > > init.d-script-not-marked-as-conffile (warning) > > This lintian warning should not trigger for an init.d script that's not > shipped in the package at all. However the other warning seems legit, at least as a warning. -- see shy jo
signature.asc
Description: Digital signature
