Re: [patch] Use correct finish function

[David Zeuthen <david fubar dk>]

On Fri, 2008-07-04 at 00:29 -0400, Mike Rooney wrote:
> Yes, that does appear to be an error, doesn't it. What are the
> consequences of it? In other words, what is a use case that would
> provide the wrong effect before the patch and right effect after the
> patch?

I guess that the only reason it works right now is because the
implementations of eject_finish() and mount_finish() is similar in the
volume monitors currently in use (e.g. 'hal' and 'gvfsd'). Remember that
more than one volume monitor can be active via the (hidden)
GUnionVolumeMonitor in libgio.

FWIW, I discovered this when working on creating infrastructure for
easily moving volume monitors out of process (patch set to follow in a
few days).

      David