[Petter Reinholdtsen] > If the new version do not get the correct ordering, i would like to > try to extract a test case to see if we can figure out why.
On second thought, I believe insserv is correct with the provided dependency information, and that the problem is the maintainers expectations for the stated dependencies, not with insserv interpretation of the dependencies. The dependency header state that splashy should start after mountkernfs and stop before everything, and insserv insert the script to start after mountkernfs (actually after everything) and stop it before everything. Thus, the ordering is correct according to the provided dependencies. If I understand the maintainer correctly splashy should start as soon as possible after mountkernfs, but this is not stated in the dependency information. To get that effect, the X-start-before header can be used to document that it should not just start after mountkernfs, but it should start before some script too. Perhaps mountdevsubfs is a good choice? Happy hacking, -- Petter Reinholdtsen -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
