Summary:ASTERISK-04810: [patch] [post-1.2] sendDTMF(f) doesn't flash the switch hook
Reporter:tim allen (tallen8840)Labels:
Date Opened:2005-08-10 17:49:15Date Closed:2008-01-15 16:13:46.000-0600
Versions:Frequency of
Environment:Attachments:( 0) app.c.diff
( 1) app.diff
( 2) chan_phone.c.diff
Description:the "f" or "F" gets filtered in acc.c so the channel driver never tells the card to flash the switch hook.


don't the zap channels have a special application for this
Comments:By: Mark Spencer (markster) 2005-08-11 01:40:45

Yes, there is app_flash.  If you want to submit a patch which changes app_flash to do ast_indicate(chan, AST_CONTROL_FLASH); as its own application, and then modify chan_zap and chan_phone to support indication of flash, that would be welcome.

By: Sergey Basmanov (sb) 2005-08-11 10:16:10

Seems that we need to call ast_indicate(chan,AST_CONTROL_FLASH) in ast_dtmf_stream when there is 'f' in digits. chan_zap already have support for AST_CONTROL_FLASH in zt_indicate.
There is no .indicate member in chan_phone. But phone_digit() will handle 'f' correctly. I think there should be at least ph_indicate that takes care about AST_CONTROL_FLASH and returns non-zero on others..
Am I right?

By: Kevin P. Fleming (kpfleming) 2005-08-22 21:30:58

Yes, that sounds like a good plan.

By: Sergey Basmanov (sb) 2005-08-23 02:08:41

Added phone_indicate to chan_phone.
In ast_dtmf_stream, we calling ast_indicate(chan,AST_CONTROL_FLASH) on 'f' or 'F' digits.
Please take a look at this patch is this works correctly. I don't have any hardware to check this.

By: tim allen (tallen8840) 2005-08-23 08:37:29

I have a fairly large re-work of this channel module... added some comments... removed some stuff that didn't seem to need to be there... fixed *0# DTMFs... added progress tone detection... more stuff that I can't remember.

Problem is: I have a Linux box, I have both cards (Line and Phone Jack), I even have a mac... My sun box is in a box and is a Sparc20 running like Solaris 7... so I test w/ Linux, I can maby test compiles with OSX (it's a laptop, no ISA slots), I don't know if it's worth the trouble to un-crate the pizza box (Sun).

Question: Is there an "owner" of this module? Would that person be willing to contact me directly? Clearly, I need a bit of hand holding.  Meanwhile I'll merge sb's changes (which look like they should work) into what I have and test.

Note: Because I was late in understanding this new bug/development system (it's new to me, last time I contributed anything was 2 years ago) I made a bunch of changes (fixed a bunch of bugs). I then found out that one does not just e-mail them any longer so I learned; filed 3 bug reports (4937, 4935, 4936) and attached three patches to those. The changes to app.c should probably be thrown out- we're making the flash app work. The changes to chan_phone.c should be reviewed closely and tested. The changes to phone.conf support the cadence detection and the changes to chan_phone.c.

By: Michael Jerris (mikej) 2005-12-01 13:45:20.000-0600

Can we please get an updated patch for this for current svn trunk.  Thanks.

By: Matthew Fredrickson (mattf) 2006-01-10 14:35:59.000-0600

Applied to trunk

By: Digium Subversion (svnbot) 2008-01-15 16:13:45.000-0600

Repository: asterisk
Revision: 7947

U   trunk/channels/chan_phone.c

r7947 | mattf | 2008-01-15 16:13:45 -0600 (Tue, 15 Jan 2008) | 2 lines

Update chan_phone to indicate FLASH properley (ASTERISK-4810)



By: Digium Subversion (svnbot) 2008-01-15 16:13:46.000-0600

Repository: asterisk
Revision: 7948

U   trunk/app.c

r7948 | mattf | 2008-01-15 16:13:46 -0600 (Tue, 15 Jan 2008) | 2 lines

Allow a digitstream to contain a flash character ('f' or 'F') when sending dtmf string to a channel. (ASTERISK-4810)