Re: Comm API notes

From: Alex Rousskov <rousskov_at_measurement-factory.com>
Date: Wed, 10 Sep 2008 21:21:07 -0600

On Thu, 2008-09-11 at 14:05 +1200, Amos Jeffries wrote:
> >
> >> > COMM_ERR_CLOSING interface will be removed. The read, write, and
> >> > accept notifications will not be scheduled after comm_close is
> >> > called. New user code should register close handlers instead.
> >>
> >> +1 on that.
> >
> > The question is whether to do it now (v3.1) or later (v3.2).
>
> No question is, can the current code be used in production without it?
> If yes, then its 3.2 issue, otherwise its blocking 3.1.

The code can still be used with COMM_ERR_CLOSING interface in place.
Otherwise, there would not be a question :-). It would be relatively
easy to remove this API though and doing so would simplify
troubleshooting and code maintenance. However, if you think we should
not remove it now, I am fine with leaving it "as is" for v3.1.

Thank you,

Alex.
Received on Thu Sep 11 2008 - 03:22:08 MDT

This archive was generated by hypermail 2.2.0 : Thu Sep 11 2008 - 12:00:12 MDT