Rico -mc- Gloeckner wrote:
> Motivation is brought by Goals, aims, guidelines (not rules), etc.
> "Being better than asterisk" is not what i mean, though.  It requires
> some "Where do we want to go" (for example architecture wise), it doesnt
> make sense to develop just to get the code all thrown over a hill
> because it doesnt fit any future architecture.

Architecture-wise, I think we all agree that the Asterisk core is a pig. 
Somehow, it manages to run some fairly big sites, but I'm sure we are 
all familiar with (and tired of) random segfaults and other bizarre 
crap. The Asterisk core is incredibly fragile - it's a wonder any of us 
PBX admins get any sleep at night.

OpenPBX always had an alternative plan to replace the core, and I think 
Freeswitch could be the solution to that. What would be great, is if the 
current OpenPBX dev team stuck together long enough for Freeswitch to 
offer a viable platform, around which we could develop Asterisk-like 
features (excluding the crashing, thanks).

> FWIW, asterisk made some improvements since the fork, but until now i
> still find it a big mess.

This is true - I looked at Asterisk trunk yesterday, and thought to 
myself how familiar some of their new stuff is. More cross-platform 
coding, replacement of db1 with sqlite etc. Maybe OpenPBX was a bit of a 
kick in the pants for Asterisk... ;-)

> Also, its nice that anthm(?) started freeswitch and obviously some
> people are rather moving there, but i felt lost when i first heard of
> freeswitch and iam having thoughts wether freeswitch will make some of
> asterisk's mistakes again.

I'm nowhere near the level of coder that Anthony is, but from looking at 
his code and API documentation, I have to say I like it. It looks clean, 
well-designed, and stable. Not like the inter-dependency hell that 
Asterisk's core is.

> I guess it would be helpful if the commiters to openpbx-svn would say
> something about where they want to go and what they want to do, so that
> the rest of commiters knows where everyone is standing.

For me personally, I feel like OpenPBX (in its current state) may have 
already come to the end of its usefulness. I am trying to fix a bug that 
is affecting two production PBX's I run, where it randomly crashes at a 
CAPI call hangup. This may already be fixed in Asterisk. Even if I find 
the bug in OpenPBX and fix it, how many other bugs are lurking that are 
going to stick their head up in a few weeks? The commercial reality is 
that I need to run with code that is being actively maintained (by more 
people than just myself). Ok, Asterisk is a big yucky mess of code, but 
it has a lot of users, and hence a lot of momentum. I reported a bug in 
Asterisk's Postgres CDR module yesterday (one that I had known about for 
a long time) and was shocked (and pleasantly surprised) when OEJ patched 
it within an hour or two of me reporting it.

My official job role in my company is not to maintain open source 
projects - the give me a bit of leeway since we are developing solutions 
around OpenPBX, but I'm really supposed to be doing other things. At the 
end of the day, we need to know that the software around which our 
solutions are based is stable, actively maintained, and not heading down 
a dead-end. For that reason, I am considering moving back to Asterisk as 
an interim solution, while Freeswitch comes up to speed.

OpenPBX is great - but I just think it was too little, too late. I'd 
love to see an OpenPBX v2.0, based around Freeswitch - I'll definitely 
lend my time to helping out with that.
_______________________________________________
Openpbx-dev mailing list
[email protected]
http://lists.openpbx.org/mailman/listinfo/openpbx-dev

Reply via email to