FreeSWITCH

variable lenght dialplan interdigit timeout (skinny-wait) need to be implemented

Details

  • Type: Improvement Improvement
  • Status: Waiting for reporter Waiting for reporter
  • Priority: Minor Minor
  • Resolution: Unresolved
  • Affects Version/s: None
  • Fix Version/s: None
  • Component/s: mod_skinny
  • Labels:
    None
  • Environment:
    FreeBSD8.0 / FreeSWITCH Version 1.0.head (git-2629a57 2010-06-16 22-42-15 +0200) / SCCP 9.0.3
  • Platform:
    Linux x86/gcc
  • FreeSWITCH GIT Revision:
    FreeSWITCH Version 1.0.head (git-2629a57 2010-06-16 22-42-15 +0200)
  • Reproduced with GIT HEAD?:
    yes

Description

When dialing the skinny-patterns.xml fetches the keys pressend and match them with the regex in the xml file.
However this is working nice, there's no solution for a variable lenght dialplan.

If your regex is ^(\d+)$ wich matches any string with any digits the dialplan match direct on the first press of a digit and doesnt wait for the timer to expire.
I've attached skinny-patterns.xml, default.xml and freeswitch.log.
  1. XML File
    default.xml
    17/Jun/10 5:53 AM
    0.7 kB
    Erik Dekkers (wvds-nl)
  2. Text File
    freeswitch.log
    17/Jun/10 5:53 AM
    24 kB
    Erik Dekkers (wvds-nl)
  3. XML File
    skinny-patterns.xml
    17/Jun/10 5:53 AM
    0.4 kB
    Erik Dekkers (wvds-nl)

Activity

Ascending order - Click to sort in descending order
Hide
Permalink
Mathieu Parent added a comment -
As said on wiki "skinny-wait tells skinny to wait 'data' seconds for more numbers before drop". Currently the timeout parameter is ignored and so it waits forever (means: wait next digit).

To implement what you want, a new application should be added. Here is complete new list:
    - skinny-route tells skinny to route the call directly
    - skinny-drop tells skinny to drop the call directly
    - skinny-wait-drop tells skinny to wait 'data' seconds for more digits before drop
    - skinny-wait-route tells skinny to wait 'data' seconds for more digits before route

Maybe you can propose a better wordings.

In fact, I first need to implement the timeout. For this I will probably need to create a specific thread.
Show
Mathieu Parent added a comment - As said on wiki "skinny-wait tells skinny to wait 'data' seconds for more numbers before drop". Currently the timeout parameter is ignored and so it waits forever (means: wait next digit). To implement what you want, a new application should be added. Here is complete new list:     - skinny-route tells skinny to route the call directly     - skinny-drop tells skinny to drop the call directly     - skinny-wait-drop tells skinny to wait 'data' seconds for more digits before drop     - skinny-wait-route tells skinny to wait 'data' seconds for more digits before route Maybe you can propose a better wordings. In fact, I first need to implement the timeout. For this I will probably need to create a specific thread.
Hide
Permalink
Erik Dekkers (wvds-nl) added a comment -
I think your wordings are okay. One thing I don't understand is why use a drop-method? You match the digits based on a regex.
So when a regex don't match, it will drop anyway cause there's no match from the regex.

Besides that, I think you should wait milliseconds instead of seconds. This way you have more freedrom to configure, e.g. if you like to configure 0.5 seconds before next digit.

Also, i would suggest to implement the timeout value per endpoint/user and/or as global config where the endpoint config overrules the global config. This way you can set a fast timeout for 90% of the endusers and a slow timeout for the other 10% (e.g. older people who can't dial fast). This way you can satisfy every user.

If there's anything I can do please tell.

regards

Erik
Show
Erik Dekkers (wvds-nl) added a comment - I think your wordings are okay. One thing I don't understand is why use a drop-method? You match the digits based on a regex. So when a regex don't match, it will drop anyway cause there's no match from the regex. Besides that, I think you should wait milliseconds instead of seconds. This way you have more freedrom to configure, e.g. if you like to configure 0.5 seconds before next digit. Also, i would suggest to implement the timeout value per endpoint/user and/or as global config where the endpoint config overrules the global config. This way you can set a fast timeout for 90% of the endusers and a slow timeout for the other 10% (e.g. older people who can't dial fast). This way you can satisfy every user. If there's anything I can do please tell. regards Erik
Hide
Permalink
Erik Dekkers (wvds-nl) added a comment -
I've tested this one. It's still waiting forever for more digits.
Show
Erik Dekkers (wvds-nl) added a comment - I've tested this one. It's still waiting forever for more digits.
Hide
Permalink
Mathieu Parent added a comment -
Yes, I have done nothing for this one.
Show
Mathieu Parent added a comment - Yes, I have done nothing for this one.
Hide
Permalink
Marc Olivier Chouinard added a comment -
This is still an issue or not ?
Show
Marc Olivier Chouinard added a comment - This is still an issue or not ?
Hide
Permalink
Mathieu Parent added a comment -
Yes,

(but I don't need it myself).
Show
Mathieu Parent added a comment - Yes, (but I don't need it myself).
Hide
Permalink
Erik Dekkers (wvds-nl) added a comment -
It's still an issue but I stopped using mod_skinny in combination with SIP.
In my setup it just won't work cause some trouble with the codec negotation.
If you need 12 Cisco 7961G's for a nice price please tell, I'm going to sell them anyway
Show
Erik Dekkers (wvds-nl) added a comment - It's still an issue but I stopped using mod_skinny in combination with SIP. In my setup it just won't work cause some trouble with the codec negotation. If you need 12 Cisco 7961G's for a nice price please tell, I'm going to sell them anyway
Hide
Permalink
Ken Rice added a comment -
Hey Nathan, Erik, Math,

can you guys chime in on this bug? is this still an issue?
Show
Ken Rice added a comment - Hey Nathan, Erik, Math, can you guys chime in on this bug? is this still an issue?
Hide
Permalink
Nathan Neulinger added a comment -
I can confirm that I have never had any luck getting the skinny-wait stuff to work either.
Show
Nathan Neulinger added a comment - I can confirm that I have never had any luck getting the skinny-wait stuff to work either.
Hide
Permalink
Mathieu Parent added a comment -
Please, this is _not_ a bug but a feature.

This is not easy to implement as it probably requires a separate thread to handle the wait.

I don't plan to implement this, but patches are welcome.
Show
Mathieu Parent added a comment - Please, this is _not_ a bug but a feature. This is not easy to implement as it probably requires a separate thread to handle the wait. I don't plan to implement this, but patches are welcome.
Hide
Permalink
Nathan Neulinger added a comment -
Only if you remove the documentation/examples that make it look like it is supposed to work.
Show
Nathan Neulinger added a comment - Only if you remove the documentation/examples that make it look like it is supposed to work.
Hide
Permalink
Mathieu Parent added a comment -
Yes, doc and examples updated (Commit 5838050 + wiki)
Show
Mathieu Parent added a comment - Yes, doc and examples updated (Commit 5838050 + wiki)

People

  • Assignee:
    Mathieu Parent
    Reporter:
    Erik Dekkers (wvds-nl)
Vote (0)
Watch (4)

Dates

  • Created:
    Updated: