[asterisk-users] [asterisk-dev] AstriDevCon 2014: Agenda itemDeprecate AMI/AGI(Ben Klang)

[Matthew Jordan]

On Wed, Oct 22, 2014 at 1:55 PM, Paul Albrecht <palbrecht at glccom.com> wrote:

>
> On Oct 22, 2014, at 11:31 AM, Matthew Jordan <mjordan at digium.com> wrote:
>
>
> On Wed, Oct 22, 2014 at 11:14 AM, Paul Albrecht <palbrecht at glccom.com>
> wrote:
>
>>
>> On Oct 22, 2014, at 10:33 AM, Joshua Colp <jcolp at digium.com> wrote:
>>
>> > Paul Albrecht wrote:
>> >> Really? Shouldn’t something this major affecting the entire Asterisk
>> >> community get discussed on the lists? Any idea what Leif is talking
>> >> about when he says the community is in transition, moving from dial
>> >> plan model to external control.
>> >
>> > It was something Ben Klang brought up and wanted to talk about - it's
>> > not something that has been decided 'nor does anyone know what the
>> > future entails. Any further discussions will naturally occur on the
>> > mailing list and in fact some things have explicit action items to bring
>> > them up on here.
>> >
>>
>> The suggestion that Asterisk should consider deprecating AMI/AGI is
>> “crazy talk.” It doesn’t merit discussion and shouldn’t be on the agenda in
>> the first place. It’s completely impractical and can never happen.
>> Moreover, Leif seems to think we (the asterisk community) are in
>> transition. What does that mean? Are we abandoning the dial plan?
>> Seriously? That’s never gonna happen either. ARI isn’t easier to use than
>> dial plan scripting. I guess one could hope that "what happens in Vegas
>> stays in Vegas”, but I don’t think the Asterisk community has that kind of
>> luck.
>>
>>
> Just because someone decided to bring up a radical idea does not mean we
> refuse to discuss it.
>
>
> So you agree that deprecating AMI/AGI is “crazy talk” but you’ll discuss
> it because of your open-mindedness?
>

I didn't say that the idea of deprecating AMI/AGI is "crazy talk". I did
say that radical ideas - and even ones that some folks think are crazy -
are all fine to discuss at AstriDevCon.

The whole point of AstriDevCon is to have a large, free, and open
conversation about Asterisk Development. I fundamentally disagree with the
notion that that should be discouraged.

This is an open source project. Communication is done in an open,
> transparent manner. People should feel like they can bring up interesting,
> radical, and yes - even crazy - ideas.
>
>
> By the same token, when you propose ideas, you must be prepared for honest
> criticism and accept it in graciously rather than simply resorting to
> argument ad hominem.
>

You didn't have honest criticism. You labelled a discussion point as "crazy
talk" and said we shouldn't have even discussed it.

There was no ad hominem attack. I never attacked you. I never even attacked
your statements. I simply defended the free exchange of ideas in
AstriDevCon. I have no problem doing that.

On the other hand, you did callously label an Asterisk Developer's
admittedly ambitious idea as "crazy talk". In the future, you may want to
choose your language more carefully if you wish for others to have a more
open discussion with you.



> If you don't like that, you don't have to participate in the discussion.
>
>
> You haven’t really responded to the substance of my post, that is, is
> asterisk abandoning the dial plan?
>

There are Asterisk users (who also happen to develop) who would like to
minimize the dialplan necessary in their systems, to the point where they
may no longer even need the dialplan. This is a fundamentally sound idea
for some systems, particularly those that require scaling Asterisk out to
many machines.

There are also some Asterisk users who build complex applications on top of
Asterisk, and who find having to use multiple interfaces cumbersome. They
like ARI, and would like to see it able to do more than what it currently
does today.

Fully deprecating a feature in Asterisk is non-trivial. You must have:
(1) A logical and full replacement for the feature
(2) Buy-off from the developer community
(3) Several major versions of the project in which the deprecated feature
must remain

Even in the case of point #3, deprecated features have often lasted in
*many* versions of Asterisk. We are enormously conservative in what we
choose to remove from the project.

I would imagine that things as important as traditional dialplan, AMI, or
AGI would be very difficult to ever deprecate.

Finally, as I've noted to you before [1] [2], please don't cross post
across lists. As this discussion is about AstriDevCon, it should be on the
asterisk-dev mailing list.

[1] http://lists.digium.com/pipermail/asterisk-dev/2013-October/063075.html
[2]
http://lists.digium.com/pipermail/asterisk-app-dev/2013-October/000113.html

-- 
Matthew Jordan
Digium, Inc. | Engineering Manager
445 Jan Davis Drive NW - Huntsville, AL 35806 - USA
Check us out at: http://digium.com & http://asterisk.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.digium.com/pipermail/asterisk-users/attachments/20141022/70a9155a/attachment.html>