Mike Jagdis wrote: > On Sun, May 06, 2007 at 02:14:47PM +0800, Steve Underwood wrote: > >> Be aware that the medium term goal is to throw out chan_zap completely. >> This is why I have only done minimal things to areas like caller ID. For >> example, old_callerid.c is the gather bits and pieces of caller ID >> handling that need throwing away with chan_zap. >> > > I'm all for replacing chan_zap in the longer term (and all the other > inherited yuk). Is there any work done on it? > Yep. Quite a lot has been done, including fully functional CLI for analogue lines :-) Unicall is the replacement for chan_zap. People have been using it for ages as an MFC/R2 platform, but I have pretty complete protocol modules for some other protocols, and the goal is to grow that, and make it more cross-hardware. chan_zap is no more a pig's ear than zaptel, so a reliance on zaptel is a pretty bad thing. >> I am extracting the extra features you need from your patches, and >> adding equivalent functionality to spandsp. Why didn't you do that? Its >> where this still belongs, since all the rest of the message construction >> and analysis occurs there. >> > > I did it that way because I was trying to contain the changes. Step 2 > is to move the callerid stuff to spandsp and/or chan_zap (its only > other use is res_adsi). Step 3 onwards is to clean up chan_zap so it's > comprehensible and then rewrite it so it's sane. But that's not > something I'd wish on one big mongo patch! >
Regards, Steve _______________________________________________ Callweaver-dev mailing list [email protected] http://lists.callweaver.org/mailman/listinfo/callweaver-dev
