[Asterisk-Users] delay before dial on TDM04B

[John Novack]

Rich Adamson wrote:

> <snip>
>
>Don't know, but in most dtmf CO's, a single "w" is more then adequate.
>Once in a great while a second might be needed.
>
>The issue is that some CO's have to attach an internal "receiver" to accept the dtmf (or dialpulse), and some CO's are rather slow to do that. The slowness is very likely to vary by a small amount on each call, therefore inserting a couple of ww's really doesn't hurt.
>
>  
>
Agreed, but it seems, from results by others that it doesn't work all 
the time even in DTMF

>If you really need dial pulse, then open a bug on this as the "w" should be applied to dtmf and dialpulse equally. They are both exactly the same issue and should have been treated the same. But, I'm fairly certain the dialpulse logic was added to asterisk after the "w", therefore whoever did the dialpulse probably forgot all about that piece.
>
>  
>
And of course the problem would seldom even appear if detection of 
dialtone  were present.
See references to REALLY slow dialtone posted in the last day or less.

>Are you attaching asterisk to your x-y switch?
>
I and quite a few  other collectors are using Asterisk as an interface 
to our "legacy" switches in a private network. We may have lost one 
today in New Orleans though! I have both XY, SXS and electronic SXS 
switches, as well as the option to use an incoming selector if necessary.

> If so, opening a bug is likely to be the only way to resolve the linefinder delay,
>
I suppose you are correct, though I have doubts that this will be 
considered very important. It seems there is much more interest in new 
"hot" features, rather than fixing what MOSTLY works.
There is also a problem with DETECTION of dial pulses  through an FXS 
interface. It is too intolorant of slight variations in dial speed. I'll 
see if someone knows how to let that be known. I sure don't

> or, dig through the "w" code and copy the process into the dialpulse code.
>
>  
>
That is not likely to happen, at least by me!
First the problem needs to be defined,

John Novack

>  
>