[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Linphone-developers] Contributing to linphone

From: Simon Morlat
Subject: Re: [Linphone-developers] Contributing to linphone
Date: Tue, 30 Aug 2011 22:07:46 +0200
User-agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv: Gecko/20110805 Thunderbird/3.1.12

Ok looks much better !

I have a few other remarks:
1) why did call notify_uninit() in linphone_gtk_close() ? Because linphone_gtk_close() is called when the user dismisses the main window, not when linphone leaves. Don't you think the end of main() would be a better place for notify_uninit() ?

2) Regarding the tone to indicate an incoming call, I have found that this kind of tone is actually normalized by ITU-T E180. Here is an extract:

Call waiting tone
The call waiting tone is used to advise a subscriber who is engaged on a call that another subscriber is
attempting to call.
The tone is intended to be sufficiently alerting to succeed in its purpose without interfering with existing
The recommend specification of the tone is one or more cycles defined by a frequency f in the range:
f : 400 to 450 Hz
and with a cadence (frequency sequence) as follows:
a) f on 300 to 500 ms, silence 8 to 10 s (f = 300 ms is preferable to the longer tone since the ongoing conversation
  would be interrupted for a briefer interval); or
b) f on 100 to 200 ms, silence 100 to 200 ms, f on 100 to 200 ms (the total to be no more than 500 ms); 8 to 10 s
Other tones may be acceptable.
The second and subsequent cycles may be at a lower level than the initial one.
Where the tone continues for more than one cycle, it should preferably cease when it is no longer possible to
accept the waiting call.


As you see it is not a dtmf (dual tone) but a single tone that repeats every 8-10 seconds. Could you modify the tone settings to reflect this specification ?

Otherwise I will do it, just tell me if you are lacking time.



On 30/08/2011 20:04, Valentin Beraudo wrote:
Hi Simon,

I don't know why the MS2 patch was wrong, it worked well when I redid it.

I tried to clear things as much as possible and made a linphone_gtk_notify() function that processes the given Call and (error) message to eventually produce a notification, if the GUI is not active. If no call is given, a notification with no actions and the given message as body would be produced.

Thanks for the remarks,


ps : patches attached

Le 29/08/2011 00:46, Simon Morlat a écrit :
Hi Valentin,

Your patch looks good in general, however I think there is a problem in the way you generated it, because it also cancels a few recent commits that were done. See for example in mediastreamer2 the l16.c file that is removed, and the --disable-tests option that is also removed.
Can you check ?

Concerning the notify feature, I have a few remarks:
* Rather than having multiple #ifdef HAVE_NOTIFY, it would be clearer to have all the notification stuff related to call states into a single function. I mean that you could create a single call_state_changed_process_notifications() function to perform all the notifications related to call states changed. * Also I think it could be a more common practice to have notifications popping up _only_ when the GUI is hidden (linphone running in background or main window not visible). Otherwise it is somewhat redundant. What do you think ? I think that should be easy to know with gtk if a window is visible to the user.

Thank you,


On 27/08/2011 16:18, Valentin Beraudo wrote:
Hi Simon,

I could swear it was there.
Hope it'll work this time.


Le 27/08/2011 11:10, Simon Morlat a écrit :
Hi Valentin,

This is very interesting ! Great achievement !
Sorry I can't find the diff in your email, did you forget it ?


On 26/08/2011 17:30, address@hidden wrote:
Hi Simon

I tried putting notification windows via libnotify where they seemed useful : For incoming calls, you can chose to answer or decline the call and, on Ubuntu, it opens a (more intrusive) pop-up window instead of a simple notification bubble. For other notifications (call ending or transferred, errors, resuming, paused by remote), it should work as intended.

I also made a kind of "beep" signal for incoming calls when already in call.
It may need some tweaking, and perhaps doesn't really need two tones.



ps : diffs attached

----- Mail Original -----
De: "Simon Morlat"<address@hidden>
À: address@hidden
Cc: "Valentin Beraudo"<address@hidden>
Envoyé: Mardi 9 Août 2011 15h41:13 GMT +01:00 Amsterdam / Berlin / Berne / Rome / Stockholm / Vienne
Objet: Re: [Linphone-developers] Contributing to linphone

Hi Valentin,

I think it could be smart to have notification windows above the status
icon, when for example an incoming call is arriving.
The current behavior when an incoming call is to show up the linphone
window, but this grabs focus and keyboard input, which is annoying.
We could imagine to have to display instead a smart nofication bubble in
the notication area, saying that "someone is calling", and with two
buttons to decide to accept or decline the call.

From what I understand gnome's libnotify library can be used for that.

Another interesting feature would be that when linphone receives an
incoming call while already in call, it plays some small beeps to signal
the new call to the user.
The current behavior is just to display the new call next to the current one in the tab of calls, but the user might not see it. The playing of
the tones should be done by using the MSDtmfGen object of the
AudioStream that belongs to the active call.

Best regards,


On 03/08/2011 23:50, Valentin Beraudo wrote:

I'm interested in Linphone and have some time this month so I'm
willing to contribute.
I had thought of maybe improving the Gtk interface.

Do you have an idea of features to add (to the gui or elsewhere) ?


Linphone-developers mailing list

Linphone-developers mailing list

reply via email to

[Prev in Thread] Current Thread [Next in Thread]