Judson Valeski wrote:
> Darin Fisher wrote:
>
>>
>> Okay, then we have it. Let's try to allow OnStopRequest's to
>> precede OnStartRequest's in the event that an error has occured
>> before the channel implementation has called OnStartRequest. It
>> just requires a bit more care on the part of the listener
>> implementations, but that should in general be a good thing anyways.
>
>
> close. You'll *never* get an OnStop() then an OnStart(). If there's
> an error before OnStart() is called, you'll get a single onstop()
> indicating why. If there's an error after onstart() you'll get your
> onstop() indicating why as well.
>
> Jud
>
Right... I think I just misused "precede" ... I meant that OnStopRequest
could occur without a preceding OnStartRequest.
Darin
