A program written to avoid a phone bill ended up as part of global communications infrastructure. Its history contains almost every classic question about commercial open source.
In 1999 Mark Spencer was running a small Linux support company in Alabama. He needed a phone system, and the quotes for commercial PBXs struck him as absurd — a box handling a few dozen lines could cost as much as a house. So he decided to write one.
Early Asterisk was not grand in ambition. It set out to make telephony run on a server. But Mark made a decision with long consequences: he open-sourced it. That decision assembled a group of telecom engineers in a very short time — precisely the people who understood phone networks and were least happy with vendor pricing.
He then founded Digium with a plain model: free software, paid hardware interface cards and support. It worked well through the 2000s, because anyone who wanted a software PBX still needed a card to connect the lines.
Asterisk uses a dual-licence structure: the community edition under GPLv2, with a commercial licence available to vendors who need to embed the code in closed products. That arrangement served two very different audiences at once — developers who wanted freedom to modify, and hardware makers who wanted to hide communications inside their own products.
Many projects have copied the structure since, with mixed results. The difference is that Asterisk's commercialisation did not treat the community as free labour but as a source of requirements: custom work requested by enterprise customers often returned to the mainline as general features.
The hard part of open source is not winning the first users. It is not treating the community as a cost line in year ten.
In 2018 Digium was acquired by Sangoma. For the community it was a moderate shock: the maintainer changed, so would the roadmap? In hindsight the rhythm barely moved — one major a year, even numbers as LTS, a public repository, an open contribution channel.
Sangoma's reasoning was straightforward: Asterisk is the foundation of its entire communications portfolio, from PBX appliances to cloud services. Maintaining the open core means maintaining its own product base. That strategic alignment is the usual precondition for an acquired open-source project to survive.
First, pick a market where pricing has lost touch with reality. When Asterisk started, PBX prices were more than ten times the underlying cost, and that gap was the opening.
Second, make programmability the top priority. Asterisk did not beat comparable projects on feature count; it beat them on 'you can change anything'. Once integrators found they could build solutions nobody else could, word of mouth did the rest.
Third, accept slowness. Twenty-six years on it is still the unattractive system that needs documentation and text configuration files. It is also still the one where you can change whatever you like.