<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
  <head>
    <meta content="text/html; charset=ISO-8859-1"
      http-equiv="Content-Type">
  </head>
  <body text="#000000" bgcolor="#ffffff">
    On 08/03/14 01:14, Matthew Jordan wrote:
    <blockquote
cite="mid:CAN2PU+5_L5ZzFQQvbT8JPkwepYDxs81UzEC9pFP+xF_B2OJFmw@mail.gmail.com"
      type="cite">
      <div dir="ltr">On Thu, Mar 6, 2014 at 4:32 PM, Damien Wedhorn <span
          dir="ltr"><<a moz-do-not-send="true"
            href="mailto:voip@facts.com.au" target="_blank">voip@facts.com.au</a>></span>
        wrote:<br>
        <div class="gmail_extra">
          <div class="gmail_quote">
            <blockquote class="gmail_quote" style="margin: 0pt 0pt 0pt
              0.8ex; border-left: 1px solid rgb(204, 204, 204);
              padding-left: 1ex;">
              <div text="#000000" bgcolor="#ffffff">
                <div>
                  <div class="h5"> On 07/03/14 08:21, Matthew Jordan
                    wrote:
                    <blockquote type="cite">
                      <div dir="ltr">On Thu, Mar 6, 2014 at 3:42 PM,
                        Damien Wedhorn <span dir="ltr"><<a
                            moz-do-not-send="true"
                            href="mailto:voip@facts.com.au"
                            target="_blank">voip@facts.com.au</a>></span>
                        wrote:<br>
                        <div class="gmail_extra">
                          <div class="gmail_quote">
                            <blockquote class="gmail_quote"
                              style="margin: 0pt 0pt 0pt 0.8ex;
                              border-left: 1px solid rgb(204, 204, 204);
                              padding-left: 1ex;">
                              <div text="#000000" bgcolor="#ffffff">
                                <div>
                                  <div> On 07/03/14 07:29, Matthew
                                    Jordan wrote:
                                    <blockquote type="cite">
                                      <div dir="ltr"> On Thu, Mar 6,
                                        2014 at 3:22 PM, Paul Belanger <span
                                          dir="ltr"><<a
                                            moz-do-not-send="true"
                                            href="mailto:paul.belanger@polybeacon.com"
                                            target="_blank">paul.belanger@polybeacon.com</a>></span>
                                        wrote:<br>
                                        <div class="gmail_extra">
                                          <div class="gmail_quote">
                                            <blockquote
                                              class="gmail_quote"
                                              style="margin: 0pt 0pt 0pt
                                              0.8ex; border-left: 1px
                                              solid rgb(204, 204, 204);
                                              padding-left: 1ex;">
                                              <div>
                                                <div>On Thu, Mar 6, 2014
                                                  at 3:31 PM, George
                                                  Joseph<br>
                                                  <<a
                                                    moz-do-not-send="true"
href="mailto:george.joseph@fairview5.com" target="_blank">george.joseph@fairview5.com</a>>


                                                  wrote:<br>
                                                  ><br>
                                                </div>
                                              </div>
                                              For me to be on-board with
                                              the change, we'd have to
                                              apply it to all<br>
                                              channel drives that
                                              implement said codecs
                                              allow / disallow logic, so<br>
                                              sip.conf,
                                              chan_ooh323.conf,
                                              gtalk.conf, h323.conf,
                                              iax.conf,<br>
                                              jingle.conf.<br>
                                              <br>
                                              That way all our
                                              documentation /
                                              functionality is
                                              consistent among<br>
                                              channel drivers.<br>
                                              <span></span><br>
                                            </blockquote>
                                          </div>
                                          <br>
                                        </div>
                                        <div class="gmail_extra">Yeah...
                                          that will never happen.<br>
                                        </div>
                                        <div class="gmail_extra"><br>
                                        </div>
                                      </div>
                                    </blockquote>
                                  </div>
                                </div>
                                I assume this is about the codecs
                                option. If so, why couldn't it be
                                implemented in all the channel drivers.
                                Surely the "codecs list" option could be
                                a simple wrapper for "disallow all,
                                allow list".<span></span><br>
                              </div>
                            </blockquote>
                          </div>
                          <br>
                        </div>
                        <div class="gmail_extra">Damien asked me about
                          this in #asterisk-dev, and I should apologize
                          here - that was a bit of a glib response.<br>
                          <br>
                        </div>
                        <div class="gmail_extra">The reality is that
                          some channel drivers have active maintainers,
                          and core changes that are made (or 'better
                          ways of doing things') do get actively made in
                          those channel drivers. This is the case with
                          chan_skinny, chan_ooh323, and chan_unistim.
                          The channel driver maintainers have done an
                          excellent job working together with the
                          community to keep up with the changes in
                          Asterisk 12.<br>
                          <br>
                          Others, however, have no active maintainer.
                          This doesn't mean they never get a bug fix, or
                          that they are broken in Asterisk 12, but it
                          does mean that there is no one who actively
                          works to keep the channel driver working with
                          all of the latest changes.<br>
                          <br>
                        </div>
                        <div class="gmail_extra">During Asterisk 12, we
                          spent a lot of time working through all of the
                          channel drivers for the changes in the
                          Asterisk core. If we hadn't done that, they
                          would have been broken by the transfer,
                          pickup, and parking changes. I think that's a
                          fair requirement on the project: if you make a
                          change in the core and it breaks someone, it's
                          on you to go fix it.<br>
                        </div>
                        <div class="gmail_extra"><br>
                        </div>
                        <div class="gmail_extra">The question then
                          becomes: do we limit any changes to supported
                          channel drivers if we do not reflect those
                          changes in an unsupported channel driver?<br>
                          <br>
                        </div>
                        <div class="gmail_extra">I don't think that's a
                          fair requirement. It burdens the project: any
                          incremental improvement in chan_pjsip, or
                          chan_sip, or any channel driver really - has
                          to be reflected across all channel drivers.
                          And not all channel drivers are equal: making
                          a configuration change in chan_pjsip is vastly
                          different then making that change in
                          chan_dahdi.<br>
                          <br>
                        </div>
                        <div class="gmail_extra">So: no, I don't think
                          it's correct to require non-breaking changes
                          to be propagated over to all other channel
                          drivers. <br>
                        </div>
                        <div class="gmail_extra"><br>
                        </div>
                        <div class="gmail_extra"> Matt<br>
                          <br>
                        </div>
                      </div>
                    </blockquote>
                    <br>
                  </div>
                </div>
                Thanks Matt<br>
                <br>
                A couple of observations. While I agree with your
                general advice of not restricting changes where
                consistency can't be done, where reasonably trivial (eg,
                setting codecs as an alias for any channel driver using
                allow), it would be nice to try and make such changes
                consistent.<br>
                <br>
                I guess it becomes a matter of where do you draw the
                line in the sand. Basically, if a change is going to be
                made, other drivers should be considered.<br>
                <br>
                The second thing, I only picked this up by luck. A
                couple of words caught my eye as I deleted a PJSIP email
                not relevant to my stuff. It may be worth considering
                how this type of information can be reliably shared to
                interested parties.<span class="HOEnZb"></span><br
                  clear="all">
              </div>
            </blockquote>
          </div>
          <br>
        </div>
        <div class="gmail_extra">Really, the asterisk-dev mailing list
          *is* the appropriate place for this kind of information to be
          disseminated. It's the heartbeat of development in Asterisk -
          we just have to make sure we keep using it :-)<br>
        </div>
      </div>
    </blockquote>
    Sorry, didn't explain myself. My issue was the there's a big PJSIP
    in the subject line when in fact that particular part of the
    discussion was about every other channel other than PJSIP. I think
    it would be reasonable to assume that many not involved/interested
    in PJSIP would just delete the email based on the subject line.<br>
    <br>
    My suggestion was that when issues become broader, they could be
    highlighted, possibly even just a cut and paste of the relevant part
    to a new email with a subject that doesn't include a specific PJSIP
    reference.<br>
    <br>
    Damien<br>
  </body>
</html>