On Tue, 2006-02-07 at 09:25 -0600, Nathan C. Smith wrote:

> The different people involved all have their own reasons for being involved
> and many have their own projects, all of which will be addressed in time.
> Many of us with the project want something that works now, and others want
> the freedom to add functionality.

I think we all appreciate how much easier it is to get patches committed
in OpenPBX, than battling the internal politics at Asterisk. However,
OpenPBX is still Asterisk at the core, and this is its downfall. While
Asterisk does manage to run some fairly big sites, and presumably
OpenPBX could too, the Asterisk core was never intended to scale that
big. You could say it was largely accidental that it became so popular.
This is the issue that FreeSwitch is addressing - a new core, designed
from the ground up to be stable and scalable.

Asterisk strikes me as the biggest code duplication effort on Earth, due
to their inflexible licensing. I think OpenPBX should be trying to
eliminate some of the duplicated code (and I'm not just referring to
duplicated functions being scattered throughout the code). We should be
aiming to use best-of-breed, off-the-shelf, cross platform libraries for
things like the SIP and IAX stacks. Why should we spend hours debugging
the protocol stacks from Asterisk, if there are solid, widespread
alternative libraries?

If you think this sounds like what FreeSwitch is doing, you'd be right.
The difference is the core. OpenPBX (at least, version 1.x) could be a
transitional product from Asterisk to FreeSwitch, providing the same
configuration environment, core and general behaviour. Ultimately, I
believe OpenPBX needs to drop the Asterisk core, which is the cause of
so much blood, sweat and tears (read: segfaults).

> OpenPBX.org should drop the animosity toward other softPBX systems, and
> gently distance ourselves from the Freeswitch project as they take on their
> own identity and importance.

Somebody summed up the state of OpenPBX quite well in a recent IRC
discussion (referring to OpenPBX tracking the patches added to the
Asterisk tree): it's like the dog that sits patiently under the table,
waiting for scraps to be dropped.

A lot of the OpenPBX developers have apparently moved to the FreeSwitch
camp. I still sort of have a foot in both camps, since I run a few
OpenPBX boxes (two in production use). There's only limited fun in
beating a dead horse though. Don't get me wrong, I support the OpenPBX
concept, but I seriously think that instead of trying to push and pull
the Asterisk code into shape, we should look to a better platform such
as FreeSwitch, and develop Asterisk-like features around it (excluding
the segfaults). Who's to say that OpenPBX 2.0 couldn't use FreeSwitch at
the core?

Daniel "pressureman" Swarbrick

_______________________________________________
Openpbx-dev mailing list
[email protected]
http://lists.openpbx.org/mailman/listinfo/openpbx-dev

Reply via email to