Doug Turner wrote: > OnStart is called now without any status because nothing has happened > yet. It is an notification that tells the observers "get ready". One > can expect an OnStop to tell us if things went sucessful or if the > transaction failed. Today, if we fail to connect to the remote server > (because of a DNS problem) the OnStop (not the OnStart) is called to > tell us what happened. > What would passing a status in OnStart buy us? What kind of errors do > we expect to throw here? if we break the current semantics of coupling onstart|onstop (which is what's currently proposed), there's no need to push a result code into onstart. Jud
- nsIStreamObserver changes? Judson Valeski
- Re: nsIStreamObserver changes? Darin Fisher
- Re: nsIStreamObserver changes? Doug Turner
- Re: nsIStreamObserver changes? Darin Fisher
- Re: nsIStreamObserver changes? Judson Valeski
- Re: nsIStreamObserver changes? Darin Fisher
- Re: nsIStreamObserver changes? Judson Valeski
- Re: nsIStreamObserver changes? Doug Turner
- Re: nsIStreamObserver changes? Judson Valeski
- Re: nsIStreamObserver changes? Darin Fisher
- Re: nsIStreamObserver changes? Scott MacGregor
- Re: nsIStreamObserver changes? Darin Fisher
- Re: nsIStreamObserver changes? Judson Valeski
- Re: nsIStreamObserver changes? Judson Valeski
- Re: nsIStreamObserver changes? Darin Fisher