[asterisk-dev] Re: [svn-commits] kpfleming: trunk r50538 - /trunk/main/channel.c

[Jared Smith]

On 1/12/07, John Todd <jtodd at loligo.com> wrote:
> While I understand the sentiment here, I'm not sure this is a good
> idea.  This builds in a 500ms post-dial delay issue into every call.
> I've been building systems for three years now, and everywhere there
> is an "Answer" (which, I believe, should be the only method that
> picks up a line and sets up a media channel locally, but that's a
> discussion for another thread) there is a "Wait(.5)" or even a
> "Wait(1)".

Wouldn't this be better served as an argument to the Answer()
application?  We already have one argument for a delay *before*
answering the channel, so why not have one for a delay *after*
answering the channel.  Now the commit log says "when a channel gets
automatically answered by an application", so does this take place
only if you don't call the Answer() application and leave it up to
another application to automagicially answer the channel?

-Jared