Petter Reinholdtsen wrote: > The discovered issues with insserv 1.12.0-1 has been resolved, and a > sourcefull upload of 1.12.0-2 was done to solve an issue detected on > alpha. This was a problem with insserv expecting deterministic > ordering of entries returned from readdir(), which proved to cause a > problem on alpha. The issue exist in all versions of insserv up to > 1.12.0-2. > > Most of the changes done since the version in Lenny (1.11.0-9) are > updates to the test suite, and the new version pass all the tests > passed by version 1.11.0-9 as well as all the new tests like the one > failing if the readdir() ordering isn't as insserv expect it. The > test suite is very comprehensive. > > In addition, Kel Modderman has done a great job with the upstream > author and got almost all our patches integrated into the new version. > And finally, upstream rewrote the script ordering code to handle start > and stop sequences separately, to solve a long term Debian bug with > "fake" dependenly loops. > > All in all, the new version of insserv is tested and confirmed to do > as it should for all settings where it is expected to work, and also > make the ordering more robust regarding file system features and > script dependencies. > > Please allow it to propagate into Lenny when its resting time in > unstable is over. It will improve the state of the release goal for > dependency based boot sequencing in Lenny.
unblocked Cheers Luk -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
