Adrian Chadd wrote:
> Right. This'd make another thing that would need killing in a new
> event io model.
Right.
> How about taking the event io callback code from the eventio branch
> and registering an event if the buffer isn't totally empty after
> the SSL read?
Naa... lets patch this up in the current I/O model first. It is not that
hard to deal with.
From what it looks the eventio model needs to be extended to support
deferred I/O operations and "userspace" events. (note: deferred ==
delayed to a later time when conditions have changed, not like it is
done today)
Regards
Henrik
Received on Fri Oct 19 2001 - 03:30:34 MDT
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:14:33 MST