[asterisk-dev] [Code Review] 3343: res_pjsip: Enable DNS support.

[Paul Belanger]

On Fri, Mar 14, 2014 at 1:19 PM, Matthew Jordan <mjordan at digium.com> wrote:
> On Fri, Mar 14, 2014 at 9:02 AM, Olle E. Johansson <oej at edvina.net> wrote:
>>
>> It would mean continuing to maintain Asterisk's pjproject fork until those
>> changes were (hopefully) accepted upstream, released, and then waiting for
>> the rpm/deb packages to catch up.  Not to mention that someone would
>> actually have to _do_ all of this work.  Could all volunteers please raise
>> their hands? ;-)
>>
>>
>> If this is how we are going to manage our product then I'm getting really
>> worried. Are we controlling our own software?
>>
>
> Of course we are controlling our own software! The problem of "I need
> something in this library, and adding it brings me out of step with
> the upstream until they accept it" is certainly not new to Asterisk's
> usage of PJSIP. It's the same problem every open source project has
> faced when they depend on another open source project. And there is,
> of course, an easy solution: write a patch to the open source project
> you depend on, submit it upstream, and in the meantime deal with the
> differences.
>
> Is that a bit harder than embedding a library or writing the entire
> library yourself? Of course. Does it mean that some changes that you
> might make if you "owned the code" you instead choose not to make
> because you have to submit it upstream? Sure. That's part of the cost
> of depending on someone else's code. But the benefits of using a
> library far outweigh the disadvantages: the fact that in a year or so
> of time we were able to produce a functioning SIP channel driver with
> 95% of the features in chan_sip and additional features besides should
> be proof enough of the benefit.
>
+1 external lib for SIP support please. I still believe it was the
right choice to make, just means new battles to be had getting new
code into upstream.  What I don't want to see us do is fork said
library and start maintaining it, because of difficulties getting code
merged.

I cannot speak to pjsip or the DNS patch, since I have not used them
yet. But, if pjproject supports it, why not use it?  If we can expose
DNS for other channel drivers, great, but that should be a discussion
point.  Either way, I'm happy Sean and Josh figure out how not to
directly ready from /etc/resolv.conf for nameservers, that was some
ugly code :D

-- 
Paul Belanger | PolyBeacon, Inc.
Jabber: paul.belanger at polybeacon.com | IRC: pabelanger (Freenode)
Github: https://github.com/pabelanger | Twitter: https://twitter.com/pabelanger