On 26 October 2016 at 02:14, Edward Burns <edward.burns_at_oracle.com> wrote:
> Be that as it may, I have been unable to persuade Vincent that an API
> modification is needed.
>
That is most disappointing.
IS there a test implementation that uses the new APIs to implement ALPN
that disproved the criticisms that we have made? If so - great we were
wrong. If not, then I really think the API needs more testing before it
is finalized!
The SslEngine change we suggested is small - almost trivial and would
significantly simplify implementation and efficiency. Without it, I
suspect that most implementations will ultimately not be compliant with the
specifications.
> At this point I recommend we accept the present
> API and proceed with the next phase: having a conference call with the
> interested stakeholders to help define the API for the backport.
>
I'm available for all the times listed and I have forwarded the invitation
to Simone Bordet
regards
--
Greg Wilkins <gregw@webtide.com> CTO http://webtide.com