discrepancy in ringback doc/code

• Messages sorted by:[ date ][ thread ][ subject ][ author ]

Hi,

Peter T. Breuer wrote:
> > > >   again in the next <t> seconds for mgetty to pick
> > > >   up. If no call comes, mgetty will exit.
> > > 
> > > in addition to the timeout discrepancy (30 vs 17 in the code) i just noted,
> > > i'm wondering why there is an extra timeout value at all? why not just
> > > wait <t> seconds after the last RING is noted? 
> > 
> > How do you know that it was the *last* RING? This is what the 17 seconds
> > are for -- make sure that you know that the other end has hung up.
> 
> You know when the other end has hung up after only 6s without a
> ring after at least the first ring 

This is exactly the 17 seconds timeout. You *need* more than just 6
seconds to handle countries where RINGs can sometimes be up to 
12-14 seconds apart.

> and anyway you could just tell mgetty
> to go into ringback mode when it receives say 3 rings (RINGS = 3) 

How do you know that it has rung *three* times on the receiving end? With
modern digitized telephone systems, RING on the receiving side and
the "ringing" sound on the dialing side are completely asynchronous -- so
when you hear three sounds, it's well possible that mgetty has only seen
two RINGs, or already four.

[..]
> The present system is uncomfortable for the configurer without any
> reason for being so. I don't like it at all. It is particularly galling
> to not be able to call back within 15s because that is about how long
> it takes me to hangup and press redial and have the dial tones emitted.

The present system is *foolproof*. The main design goal behind all
mgetty coding is "do it in a way that will break as seldom as possible".
If you mess with the timing constants too much, other people will complain
that it doesn't work reliably for them anymore.

But you have the source -- change all constants the way you like it best
for yourself -- that's *why* you have the source.

> > This is done on purpose, to avoid false alerts if you get "normal" calls
> > on that line as well.
> 
> There are much better ways to achieve this - that was the point of the
> message quoted, I believe! 

No, there are *no* better ways. Telephone systems are way too complex
and way too different in countries round the globe to count on certain
behaviours in your local phone systems.

gert

-- 
USENET is *not* the non-clickable part of WWW!
           //www.muc.de/~gert/
Gert Doering - Munich, Germany      gert@greenie.muc.de
fax: +49-89-3545980     gert.doering@physik.tu-muenchen.de
.