Thanks, Jared. This has gotten too personal, and I apologize. That was not
my intent. Here's the bottom line, and I'm telling you before I publish it.

You're never going to get Fortune 500 companies or government organizations
to deploy Asterisk when you destroy commands and change syntax in the API
every year without providing backward command/syntax support. However minor
the changes may appear to you, these changes unnecessarily break existing
applications, lots of them. It took me six months to rewrite a dozen 1.2
applications so that they would run reliably under 1.4. I'm not going to do
it again. I'd suggest you go back and reread the references you provided to
the syntax changes. Most of the syntax changes aren't covered at all.

Yes, the app_swift change you noted is minor. But it's still broken. And it
assumes a willing author to rewrite code or give it up to others... assuming
the author can even be located. You easily could have avoided the whole
problem by supporting both syntaxes. And Digium has total control of that
environment! 

What you don't seem to appreciate is that the current approach forces every
organization and application developer to revisit ALL of their application
code with EVERY new Asterisk release: 1.2, 1.4, 1.6, 1.8. It also trashes
the building block model because it destroys all of the components built by
other people. People die. People move on. If their code only works for a
year or two, no one will ever feel comfortable expanding Asterisk to meet
corporate needs for fear that the tools may stop working because some genius
decided that Set(TIMEOUT(digit)=7) looked prettier than DigitTimeout(7).
That one change broke every dialplan on the planet when folks moved to
Asterisk 1.4. It would have taken about two lines of code to support BOTH
syntaxes! Digium doesn't support Asterisk 1.2 today, and how long has it
been since a stable version of 1.2 was first released?? Phone systems are
deployed by organizations to last a decade or more, not a year or two.

Suppose C commands were removed or its syntax was changed as often as you
guys kill off Asterisk verbs? Do you think anyone would write code in C??
Where would Asterisk be? Suppose the COBOL language had been changed and
commands had been eliminated as often as the Asterisk API. That would have
meant that banks would have had to revisit ALL of their application code
annually. Nobody would have used COBOL.

Quite frankly, no organization with any business sense is going to use
Asterisk for the very same reason. It's just too painful to have to rewrite
and revisit tens and hundreds of thousands of dollars worth of application
code so often. Don't take my word for it. Go ask some customers that have
invested in expensive custom code for their phone systems. I worked in that
environment for decades.

Finally, here's a sad footnote. We really wanted to crack some vertical
markets with Asterisk later this year, particularly the medical and hotel
sectors. We finally had all the building blocks in place to do that
including substantial venture capital. So we were ready to begin writing the
applications. But you've made it crystal clear that this would be a huge
waste of time and resources. Sorry to be so negative but trashing command
syntax for absolutely NO reason when both the new and the old syntaxes could
easily be supported is financial suicide in the commercial and government
sectors. Someone at Digium had better start looking at the forest and stop
focusing on the trees before it's too late. Supporting old syntax rather
than trashing it, of course, has absolutely nothing to do with building a
better mousetrap. Just my $.02.
--wm