vicidial and freewitch = heaven

[bobbymc]

Based on what they wrote freeswitch is the replacement of asterisk.. No hardware needed for timing and looks like it does 10x more then asterisk could ever do..

http://www.freeswitch.org/

matt what do you think?

[mflorell]

Yes, I have heard that before from members of the FREEswitch developers group, but ask them about their native T1 PRI support.

The coding and design of FREEswitch are superior to Asterisk, that's not really even in question. Asterisk is a mess in a lot of ways, but it is tested and in realiable use in millions of installations. It also has a lot of features that FREEswitch does not have. Asterisk also has a much larger company behind it with millions of dollars that will ensure it's future development.

Switching to FREEswitch would involve months of uninterrupted coding. At this time that amount of investment on our part is not justified. We have a reliable package that works reliably on Asterisk.

Would VICIDIAL work better/faster/more-reliably on FREEswitch? possibly, if all of the features can be matched. But before we would begin that, we would need a sponsor for the move that has deep pockets and could afford to pay to have FREEswitch developers fill in the gaps in the FREEswitch featureset.

[bobbymc]

there is a very big possibility that i have a funder for it.. how much are we talking?

[mflorell]

I don't really have any idea. I would guess that a minimum of $100,000 would be needed for our time and the time to have the features that are lacking in FREEswitch programmed by their core developers and it would take a minimum of 6 months.

But at this point in time I would not even take the project on. Mostly because it would halt my development on adding new features to VICIDIAL so I could work on porting it, and I really don't see enough benefit to do that at this time.

[tushkanwn]

Matt, why will it be so costly?

I don't really have any idea. I would guess that a minimum of $100,000 would be needed for our time and the time to have the features that are lacking in FREEswitch programmed by their core developers and it would take a minimum of 6 months.

But at this point in time I would not even take the project on. Mostly because it would halt my development on adding new features to VICIDIAL so I could work on porting it, and I really don't see enough benefit to do that at this time.

[mflorell]

Because I don't think it's worth doing. I don't think FreeSWITCH is mature enough of a platform, with a good enough management and big enough community behind it compared to Asterisk.

[bobbymc]

i gotta disagree.. i pushed freeswitch to limits asterisk could wish they could. plus the community behind freeswitch is larger then you think. There is not 1 single feature that asterisk has that's not in freeswich. On top of all that they support google voice and many other voice platforms that asterisk does not.

=) ALL HAIL FREESWITCH..

[gardo]

Looks like Freeswitch is really superior to Asterisk when it comes to scalability and stability. Not to mention coding. Though not as mature as Asterisk but it's going there. I've read some pretty good and sound arguments about it's advantages and disadvantages over Asterisk.

It's really up to Matt if he wants to port it or not. Though it would really require serious time, effort and money to do it. And as he pointed out, development of Vicidial will be halted when the porting starts. Hopefully there are other takers for this project. Anyone?

[mflorell]

There are a lot of companies using FreeSWITCH(and I personally spoke with several of them at Cluecon last year), but they mostly fork the code and do not maintain upgrade compatibility because of the poor version to version compatibility. Dozens of modules are dropped from release to release and core functionality can be changed as well, sometimes without any notice or updated documentation at all.

We used to use FreeSWITCH in production for our hosted platform as part of the back end, and it worked well when load was low, but once we started to get over 300 channels going through it for the application we were using it for, it started to regularly crash(keep in mind, there was NO media handling). Then we were told that the new version of FreeSWITCH fixed that crashing bug so we upgraded, but that was not as trouble-free as it should have been since some of the modules we were using were deprecated and no longer supported(yet are still in the docs), and they had changed how callerID handling worked as well(again with no documentation of the change). These issues required us to make significant changes to our application, and when we got it back up and running we still had the random crashing issue. In the end we dropped FreeSWITCH and are only using Asterisk and OpenSIPs on our backend without any significant service interruptions for months.

So, this experience did not endear us to FreeSWITCH at all. I admire their goals for a more modular and scalable telephony engine, but for our use we did not find it easy to work with, stable or with sufficient support behind it to continue it's use. This is why we will not be porting ViciDial to work natively on FreeSWITCH at any time in the near future.

[bobbymc]

that's a fair statement and i agree with the troubles matt had as i had them also, but many of these issues have been fixed since, might wanna give it another shot matt =)

[ramindia]

but for a longtime plan

instead of spending time over asterisk

lets start with freeswitch slowly atleast next 2 years we have good application.

this is my view
ram