Jay argues that Asterisk has not crossed the proverbial chasm for developers and outlines some of the issues he sees.
What is perhaps most interesting about Jay's post is the equally lengthy response by Asterisk creator Mark Spencer. Mark responds to Jay's various points and in doing so provides some good insight into his views on Asterisk's connections to developers, APIs, etc., as well as the differences between the markets that Digium, the company, goes after versus the "market" of Asterisk, the raw telephony platform.
Both Jay's article and Mark's response are definitely worth reading. I'm friends now with both of them and they both bring immense passion and energy to the world of open source telephony. Ultimately they both make the point that we need better tools for developers to create voice applications. This kind of dialogue is great and will only result in better tools in the end. Please do check out the posts.
P.S. Thomas Howe has also weighed in with a post saying this is a clash in world views, which also makes for good reading. I agree with Thomas that we are in a transition into a world of "web-as-a-platform"... basically a transition "into the cloud"... and so we do need "web-centric" interfaces and APIs. But I disagree with Thomas that Asterisk is "tired". To me, Asterisk is just... well... "plumbing". Asterisk is a telephony platform, as is FreeSWITCH... as is Yate... as are all the commercial IP-PBXs. Asterisk is an open source component of the rewiring of our communication infrastructure that we have underway right now. I think anyone, including Mark, would agree that Asterisk has technical challenges it needs to overcome but I, for one, am not ready to write it off yet.