[sophacles]

I've been noticing a pattern in the free-software world. Product A becomes a flagship implementation of something. Sendmail for smtp, Apache for http and so on. These are early implementors, lots of people get behind the project, a hundred different paths are taken. In the mean-time there are lots of experimenting, tweaking, and reworking to the entire suite/stack for that something, which get rolled into the original implementation. This results in a giant, lumbering beast of an application. It contains this collective history of the entire movement. All the abortive bits, blind alleys, weird idiosyncracies must be kept, because someone, somewhere depends on it.

All the old timers think this is the bee's knees. Everyone else however gets sick of the giant, and you get exim or nginx, in a hundred varieties that perform better for certain common (or uncommon) subsets of use. These better performing "lightweight" versions shed the cruft and are widely adopted by those who no longer care about say, http servers, and instead just want a simple strut.

Anyway my point is: is it at all surprising that FreeSWITCH has come in to do fill the Asterisk role as a voip strut?

[garyrichardson]

That is often the case, but have you seen the Asterisk code base? It's amazing that it compiles, let alone runs for more than 5 minutes.

[sophacles]

How is this different than what I said? In my example/metaphor Asterisk is the crufty lumbering giant..