On Tue, Jul 27, 2010 at 10:36 AM, Michael <[email protected]> wrote: > I just think that the format is the obstacle that makes developing some level > of automation too tedious.
I disagree. Suppose we combined a sufficient combination of: - Knowing which files contain configuration for a given package; - Knowing which directories contain configuration for the package; One might set up an SCM repository for that package's data, and automatically track: - Versions tied to package releases - Versions captured from what administrators had set up There are a number of ways to capture such, including: - Tag version sets in an SCM like git/monotone/hg/... - Capture all the configuration in tarballs that get stowed, marked with names based on some naming convention, consistently in some place. - Capture all the configuration embedded in XML files that get stowed, marked with names based on some naming convention, consistently in some place. Actually, I rather like the tarball idea. SCM's cool, but probably overkill. -- http://linuxfinances.info/info/linuxdistributions.html -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected] Archive: http://lists.debian.org/[email protected]
