On Fri, Jul 08, 2011 at 10:42:21AM +0100, Roger Leigh wrote: > It was flagged up here: > http://people.debian.org/~biebl/lib-init-rw/archive-scan.2011-07-08.log
Aha! > If neither of these are true, and it's just for a testsuite/examples, > then simply replacing all instances of /lib/init/rw with /run will > be sufficient. That's right, it's just test data, used to test backup/retrieval. In fact it's not even that: t/sampledata/etc -> /etc So the test is exposing issues with the installed packages on the test system! I suspect the upstream author(s) thought that this is a neater way of having a large-ish test suite of mostly textual data; it does mean that lots of analysis will vary entirely depending on the system on which the test is being run. At some point in the past, they had a copy of a subset of one of the author's /etc instead: this was problematic because they had init scripts for non-free stuff copied in (vmware etc.) Anyway, the merits of this symlink not withstanding, I guess bup is not impacted by the /run transition. > Thanks for looking at this! You are welcome. -- Jon Dowland -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
