> We also have a long term problem with the OnError event. It skips raising
an exception and
> simply calls the event without the actual error information being
available, and in a few
> places continues program flow, not sure how useful this is, it probably
confuses users.
The OnError event
> IMO the correct way to handle this enhancement is to extend
> ESocketException to add more fields with more details about the
> exception, without changing the current error message.
In principle I agree, although I'd add an option for the friendly error
message to be the default instead of
