[Sipsak] Re: latest sipsak tarball

[Nils Ohlmeier]

Again cc'ed list for the archive.

On Wednesday 16 November 2005 17:14, peter.3.edwards at bt.com wrote:
> It works!  I ran a test against our main SIP registrar which resolves
> using SRV.  Using the 0.9.2 binary (without SRV support) it says the
> name is unresolvable.  With the new binary, it sends the message and
> returns the response!  :)
>
> I'm ecstatic!  :)
>
> *Many* thanks for your patience and your willingness to implement
> support for a more solaris-friendly dns resolver.  I'm chuffed to bits
> and can't wait to get the other guys in my team using sipsak.
>
> Just to satisfy my own curiousity - I didn't notice any reference to
> libruli anymore.  Have you completely replaced libruli for c-ares or do
> you plan to support either now?

Both is supported now. C-ares is just preferred over ruli because of this 
portability thing.
It is the same for MD5: gnutls is preferred over openssl which is preferred 
over the own MD5 implementation.
The gnutls detection can be turned off with a configure switch. The SRV lib 
detection can not be influenced (yet).

Greetings
  Nils

> Thanks again,
>
> Peter.
>
> P.S.  I also heard back from Everton, author of RULI, who confirmed that
> it can be built on Solaris but is dependent on IPv6.  None of the
> servers I have access to were built with this, so that's why I couldn't
> get it to work.  He said he wouldn't have time to look at the issue
> until today but, now that I have a working binary, I'll let him know
> that it's not a major showstopper for me anymore.