Recently I have been seeing more references to FreeSwitch on the web and I thought I would take a look to see if this is something good or just another project with its 15 minutes in the spot light.

I've looked at many Open Source products but there are only 3 products that I have put in clients sites with confidence. These products include Asterisk, pfsense (great software based router), and untagle (great software router with all kinds of filtering/spam/antivirus options). All these products when configured correctly with the right hardware can work reliably (no reboots required) for long periods of time.

For the most part I am happy user of Asterisk and I would like to see the product improved. Unfortunately the only thing that stands in the way is Digium. Its a great open source company but it will never give us what we actually desire for free. We all want a free open source “stable” telephony system. Digium does provide such products except they are not completely open source or free. You can purchase switchvox and pay Digium to support you.

I don't think there is anything wrong with paying for good hardware/software or support. The main problem is conflict. You can't have a truly open source product if you also have a commercial version right beside it. It really doesn't allow a company to focus on providing good “general” support to the open source version if there are competing products from the same company.

Let me be clear that there is a difference in having one product with some commercial options. Some companies provide commercial options or paid support (for specific or unique setups) not available in the open source versions. An example of this is Untagle, it uses many open source products, wraps them in a great interface – but also provides commercial plugins for additional spam and antivirus protection. Its your choice if you want to purchase these or opt for paid support. The product is solid and there is no confusion

There are numerous paid and non paid developers working on the open source version of Asterisk. There are new features being added all the time. Unfortunately with each new feature that is introduced it causes a cascade of new bugs to appear. The bugs are ultimately fixed but the fixes cause new problems to appear. You say that this is normal for any open source software. This is not true, Asterisk has created a parallel release system for 2 source branches, the 1.4 and 1.6 releases.

The 1.4 is meant to introduce small incremental changes to existing code and bug fixes. The 1.6 is meant to introduce new features and change the architecture of how the software actually works. This all sounds like a great idea but it creates more problems than it solves because we get into different scenarious but ultimately we have 2 bug ridden source branches to look out for. The reason Digium did this was because they wanted the adoption rate of the new versions of Asterisk to happen a lot quicker.

Why would Digium want this? Well you just have to understand with open source software we are all beta testers. The quicker Digium can get code out and tested the quicker that code can make it in their commercial offerings and consulting projects. Again not that there is anything wrong with this model, You should be a aware that there is a cost to open source. You may not pay up front with money but you ultimately will pay in testing, failed installations, and unhappy clients, and ultimately less money for you if you are installing the open source product.

It took a while so why some of you may be wondering how can FreeSwitch solve this? Here is a list of reasons:

1. FreeSwitch is not a hardware company – their focus is not getting dahdi/zaptel drivers fixed.

   (I gave up on the zaptel/dadhi architecture early on as it was a major problem with reliability and quality – switched to external gateways)

2. Single code base – no 1.4 or 1.6 nonsense – design a good architecture from the start and fix the issues as quickly as possible.

3. A lot of security and just plain nasty bugs pop up in the Asterisk sip stack. FreeSwitch uses an Open Source SIP stack provided by Nokia called sofia which is more compliant than the home grown version provided in Asterisk. It also lets FreeSwitch developers work more on FreeSwitch core functionality if they don't have to fully maintain their own SIP stack.

4. Telephony Application development – Asterisk is just a PBX – you really have to strugle to create apps using the dial plan and limited scripting capabilities. FreeSwitch provides you with a number of clean interfaces to scripting languages and allowing greater control of what you can do in those scripting languages. The real money for Open Source Telephony will be in the application development. This will be unified communications and what open source works well in. Taking several different apps/code to create a solution for a client at a much lower cost than commercial offerings.

5. Works in Windows – As much as some of us try to avoid Windows its tough and sometimes impossible for some solutions. There is more hardware support for windows than Linux/*bsd. There are also more commercial developers in Windows. This means that a product like FreeSwitch has a better chance in being part of a voip/telephony solution than Asterisk which is non existent on Windows. This allows developers to embrace a product like FreeSwitch because they don't need to learn a new new development platform ir give up any of their code.

6. Asterisk has trained a lot of us to be VOIP experts. What this means is that our knowledge or time has not been wasted. We can transfer this to FreeSwitch without the huge learning curve it took to learn Asterisk. So while it might have took Asterisk a long time to achieve its user base it won't take anywhere as long for FreeSwitch.

7. Fresh Start – Unfortunately Asterisk wide usage has also created marketing power for competing commercial PBX companies. There are thousands of Asterisk users with failing systems (mostly the fault is in their setup) that have posted messages online creating this tarnished appearance of Asterisk) This is easy ammo for competing companies to say just look online and you will see all the problems Asterisk has. Hopefully with FreeSwitch a more positive experience can be created.

I will continue to use Asterisk but I am probably one of many now looking at alternate open source products such as FreeSwitch. What I am hoping for in the long run is that more Asterisk competition will create better products from Asterisk and FreeSwitch.