On Wed, Nov 21 2012, David Paleino wrote: > On Wed, 21 Nov 2012 13:29:59 +0000, David Edmondson wrote: > >> The wicd daemon doesn't currently disconnect any active links when it >> terminates. Doing so results in wpa-supplicant being correctly stopped. >> >> Attached is a proposed patch (I've also uploaded a build of the package >> to mentors.debian.net). > > The patch is not appropriate. It might work, but that's not the right place to > stop a connection. Just FYI, the appropriate place should be at the end of > main(), just before the DaemonClosing signal is emit.
Understood. > Besides that, I don't consider this to be a bug. Rather, I consider it a > feature > -- being able to replace/restart/... wicd while still keeping a connection > alive. Whilst not deeply passionate about it, I disagree. Stopping the daemon should revert the system to the state it was in before starting, else the originally described scenario is difficult to avoid. How does an upgrade of the wicd daemon avoid this problem? And, if you don't believe that this is a bug, shouldn't it be closed as such rather than simply downgraded in severity? > I have thought about doing so since its filing, but I just wanted to think > more > about it, and now it comes: I'm downgrading this bug's severity. I'd agree - it doesn't seem like something to stop the release. > I do understand KiBi's reasons -- even though I *do* *not* consider his > wording > appropriate or fair at all -- and I could consider adding a separate > command-line option to the daemon (to keep the connection alive) and changing > the current default. But this surely is _not_ RC IMHO.
pgpI8kzrMzn0m.pgp
Description: PGP signature
