I assume the goal here is to develop a carrier grade soft-switch. I understand features are great but features would cost us. As we add features the stability would degrade. I agree that adding C# support is an intresting idea but do we really need it adding mono into FreeSwitch I understand its a great thing for C# developers but mono is not super-stable. I personal opinion is to work towards defining a minimum set of features that are absolutely critical and then making those features stable and relatively bug free. In Freeswitch much of the features or protocol support is based on external libraries this can be a good thing and a bad thing too. Since each of those libraries would probably have bugs and that means making sure that those libraries are also actively maintained. -- Vikram Rangnekar
The key here is "modular" softswitch, which means, that while you make sure the features you need are up to par, you can not load the ones you don't need. So if you don't need any functionality provided by mono (likely in the early stages to be mostly advanced IVR type of functionality) then you don't need to use it. In regards to using others libraries, I agree that we have to make sure they are maintained and stable, but from a practical standpoint, using other peoples libraries means that there are other people using them outside of freeswitch as well, which should be a plus for finding bugs and making sure you have stability, not a minus. As for defining a minimum set of features, that too will very much depend on the application it is being used for. For a basic TDM to sip softswitch, you need a stable core switching platform, as well as a handful of modules for routing, billing, and TDM and SIP. Our primary focus right now is to make those sorts of core functionalities bulletproof. Mike
-----Original Message----- From: Vikram Rangnekar [mailto:vr@udel.edu] Sent: Thursday, August 31, 2006 1:34 PM To: freeswitch-users@lists.freeswitch.org Subject: [Freeswitch-users] Stability vs Features
I assume the goal here is to develop a carrier grade soft-switch. I understand features are great but features would cost us. As we add features the stability would degrade. I agree that adding C# support is an intresting idea but do we really need it adding mono into FreeSwitch I understand its a great thing for C# developers but mono is not super-stable.
I personal opinion is to work towards defining a minimum set of features that are absolutely critical and then making those features stable and relatively bug free.
In Freeswitch much of the features or protocol support is based on external libraries this can be a good thing and a bad thing too. Since each of those libraries would probably have bugs and that means making sure that those libraries are also actively maintained.
-- Vikram Rangnekar
_______________________________________________ Freeswitch-users mailing list Freeswitch-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users
UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users
On Aug 31, 2006, at 12:33 PM, Vikram Rangnekar wrote:
I assume the goal here is to develop a carrier grade soft-switch. I understand features are great but features would cost us. As we add features the stability would degrade. I agree that adding C# support is an intresting idea but do we really need it adding mono into FreeSwitch I understand its a great thing for C# developers but mono is not super-stable.
Since mod_mono is a module you can leave it out and not chance any stability issues if thats your choice. We don't strong arm you into a way of doing things. (ie we don't glue the lego bricks together)
I personal opinion is to work towards defining a minimum set of features that are absolutely critical and then making those features stable and relatively bug free.
That is our goal also. The core MUST be stable thats the bottom line. http://video.google.com/videoplay?docid=-4613750174577358330&q=cluecon That video is Anthony speaking about freeswitch at cluecon. Just pay attention to the audio.. the video is blurry.
In Freeswitch much of the features or protocol support is based on external libraries this can be a good thing and a bad thing too. Since each of those libraries would probably have bugs and that means making sure that those libraries are also actively maintained.
Yes we do depend on external libs. But we also push fixes and changes back to those projects so that stays a two way street. If one becomes un-maintained we'll just take over on that and continue with maintaining it. The bottom line is STABLE CORE. The core will extend itself to building features with ease once its complete. /b
I agree that the modular nature of freeswitch allows for greater flexibility and probably stability. I put my thoughts to the list just so that everyone understands the great need for a super stable switch core and hopefully get more developers excited on that core part of Freeswitch. On 8/31/06, Brian West <brian.west@mac.com> wrote:
On Aug 31, 2006, at 12:33 PM, Vikram Rangnekar wrote:
I assume the goal here is to develop a carrier grade soft-switch. I understand features are great but features would cost us. As we add features the stability would degrade. I agree that adding C# support is an intresting idea but do we really need it adding mono into FreeSwitch I understand its a great thing for C# developers but mono is not super-stable.
Since mod_mono is a module you can leave it out and not chance any stability issues if thats your choice. We don't strong arm you into a way of doing things. (ie we don't glue the lego bricks together)
I personal opinion is to work towards defining a minimum set of features that are absolutely critical and then making those features stable and relatively bug free.
That is our goal also. The core MUST be stable thats the bottom line.
http://video.google.com/videoplay?docid=-4613750174577358330&q=cluecon
That video is Anthony speaking about freeswitch at cluecon. Just pay attention to the audio.. the video is blurry.
In Freeswitch much of the features or protocol support is based on external libraries this can be a good thing and a bad thing too. Since each of those libraries would probably have bugs and that means making sure that those libraries are also actively maintained.
Yes we do depend on external libs. But we also push fixes and changes back to those projects so that stays a two way street. If one becomes un-maintained we'll just take over on that and continue with maintaining it.
The bottom line is STABLE CORE. The core will extend itself to building features with ease once its complete.
/b
_______________________________________________ Freeswitch-users mailing list Freeswitch-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
-- Vikram Rangnekar
Yep Core stability is our main concern features will come once the core has been stamped stable, The API will also be versioned so we can have that be stable also. We don't like moving targets. ;) /b On Aug 31, 2006, at 12:57 PM, Vikram Rangnekar wrote:
I agree that the modular nature of freeswitch allows for greater flexibility and probably stability. I put my thoughts to the list just so that everyone understands the great need for a super stable switch core and hopefully get more developers excited on that core part of Freeswitch.
Stability is a big issue for me too. My background includes time in a military communications corps, and I do not tolerate systems that crash every other day (I'm not mentioning any other projects that we might know that do that...) Yossi Neiman Cartis Solutions, Inc. - http://www.cartissolutions.com Vikram Rangnekar wrote:
I agree that the modular nature of freeswitch allows for greater flexibility and probably stability. I put my thoughts to the list just so that everyone understands the great need for a super stable switch core and hopefully get more developers excited on that core part of Freeswitch.
On 8/31/06, Brian West <brian.west@mac.com> wrote:
On Aug 31, 2006, at 12:33 PM, Vikram Rangnekar wrote:
I assume the goal here is to develop a carrier grade soft-switch. I understand features are great but features would cost us. As we add features the stability would degrade. I agree that adding C# support is an intresting idea but do we really need it adding mono into FreeSwitch I understand its a great thing for C# developers but mono is not super-stable.
Since mod_mono is a module you can leave it out and not chance any stability issues if thats your choice. We don't strong arm you into a way of doing things. (ie we don't glue the lego bricks together)
I personal opinion is to work towards defining a minimum set of features that are absolutely critical and then making those features stable and relatively bug free.
That is our goal also. The core MUST be stable thats the bottom line.
http://video.google.com/videoplay?docid=-4613750174577358330&q=cluecon
That video is Anthony speaking about freeswitch at cluecon. Just pay attention to the audio.. the video is blurry.
In Freeswitch much of the features or protocol support is based on external libraries this can be a good thing and a bad thing too. Since each of those libraries would probably have bugs and that means making sure that those libraries are also actively maintained.
Yes we do depend on external libs. But we also push fixes and changes back to those projects so that stays a two way street. If one becomes un-maintained we'll just take over on that and continue with maintaining it.
The bottom line is STABLE CORE. The core will extend itself to building features with ease once its complete.
/b
_______________________________________________ Freeswitch-users mailing list Freeswitch-users@lists.freeswitch.org http://lists.freeswitch.org/mailman/listinfo/freeswitch-users UNSUBSCRIBE:http://lists.freeswitch.org/mailman/options/freeswitch-users http://www.freeswitch.org
participants (4)
-
Brian West
-
Michael Jerris
-
Vikram Rangnekar
-
Yossi Neiman