On Tue, Sep 12, 2006, Robert Millan wrote: > > My opinion is that whoever creates the environment variable which lists > > the socket should make sure the socket is available before spreading > > the news, > It did, and at that time the socket was present. It's just that the socket > is not there anymore.
If everyone would have done so, the current code wouldn't have to special case the absence of the socket as a "try again" error. My suggestion is to return a hard error instead and fix the applications to wait for the socket to exist before they publish the environement variable. -- Loïc Minier <[EMAIL PROTECTED]>
