I founded the AskoziaPBX Open Source project while studying in Germany in January of 2007. It was a response to the numerous phone system solutions being developed around the Asterisk toolkit which were, quite honestly, only for telephony wonks.
I remained the project lead and primary developer until April of 2011.
Internet telephony was quickly becoming a thing and it was amazing! I wanted something simple. Everyone should be able to use this stuff.
The software was originally named AfroPBX (another free open public branch exchange). At least two months of commits went into a repository with that name. I had a logo and everything. However, I started to want something more distinct. After bouncing names around for weeks I settled on a domain name I already owned. Askozia. It was originally bought to host subdomains for various projects I was working on. Problem solved.
After six months of development, the first public beta was announced on the asterisk-users list and released on the 1st of June, 2007. It shipped as an embedded firmware which ran on systems with as little as 32MB of storage, 64MB of RAM and a 200MHz CPU.
Over the next 14 months I released 24 public betas, ultimately bringing out 1.0.0 in September of 2008.
It was exactly what I wanted. As simple as it could get, immediately useful to a wide audience and rock solid.
The community grew and grew. Mailing lists, forums, thousands of downloads for each release. Translations for 12 different languages were submitted and we saw use of the software in over 170 countries. Good feelings were had.
PBX manufacturers started to take notice and we landed a development contract to port the firmware to a major vendor's hardware! This triggered a massive undertaking because their appliance was based on a Blackfin processor.
At the time, Askozia only ran on the x86 architecture. And, more critically, it was based on FreeBSD which did not support Blackfin at all. I first needed to port the entire firmware from FreeBSD to Linux, then add support for Blackfin (an architecture that doesn't have an MMU). Another manufacturer came aboard after the move to Linux and we added support for Power PC.
By the end of 2009, the standard litany of features were all there for PBXs and we were running our own custom Linux distro.
How do you make money with free software? An excellent question for which there isn't a simple answer. We tried support contracts, licensing to hardware platforms, charging for the handbook, inserting ads on the presence and asking for donations all with varying success.
I was never willing to implement a licensing mechanism in the finished product. Over time, I learned that unless you force someone to pay for a product, they will simply refuse. No matter how useful, no matter how much money their business has made off of it, no matter the quality.
I did come away with a realization: not everything is suited for Open Source. If it's basically a finished product already, there's not much left for the community to contribute. A physics library, for example, is an excellent candidate for Open Source.
If the market weren't so saturated with telephony solutions, I'd do the whole thing over again!