[asterisk-users] subject: 1.4 vs 1.6

[Tilghman Lesher]

On Wednesday 24 February 2010 10:16:25 Miguel Molina wrote:
> Gergo Csibra escribió:
> > Wednesday, February 24, 2010, 3:56:50 PM, David wrote:
> >> On Wed, Feb 24, 2010 at 9:20 AM, Juan Sandro <juan.sandro at hotmail.com> 
wrote:
> >>> Hi Guys
> >>>
> >>>
> >>> We are using asterisk 1.4 on all of our platforms for a while now.
> >>> Some of our partners recommended to use asterisk 1.6 in order to
> >>> improve overall stability and performance.
> >>>
> >>> Can someone please let me know if you have a such experience?
> >>> Also, do you have any other negative or positive comments on 1.6
> >>
> >> If it isn't broke, don't 'fix' it.
> >>
> >>
> >>
> >> There are benefits to 1.6, like dramatically enhanced SIP support,
> >> much faster dialplan processing, easier faxing, changes to dialplan
> >> syntax, and lots of other features. I would say the improvement of
> >> going to 1.6 is only if you are trying to expect more from the same
> >> gear, or want the new features. If you're not actually having
> >> problems, don't change anything.
> >
> > Yes, and check this page:
> >
> > http://www.asterisk.org/asterisk-versions
> >
> > as you can see, the 1.4 version is LTS, and the 1.6 isn't, but the
> > upcoming 1.8 will be LTS too. So don't change to 1.6 :)
>
> That sounds reasonable, but as I have seen through several years
> following the asterisk project, when 1.8.0 will be released it will be
> far less stable than the more used and mature 1.6.0.X, for example. I
> would prefer to do a middle step for upgrading, that would be 1.4.X -
> 1.6.0.X - 1.8.X when it becomes really stable. Asterisk history has
> shown us that a newly released branch, no matter if it's LTS on the new
> release schema, will need time and a large user base that adopts it to
> report bugs and help stabilize it. I would not underestimate the actual
> 1.6.X branches.

Additionally, it's worth noting that the dates above are meant to be the
EARLIEST dates that development, security fixes, etc. will end.  It is quite
possible that we will elect to extend some of them.  The whole idea is to
give companies advance notice of at least six months before we stop
supporting a release.

The end is coming; but it might be delayed.  :-)

-- 
Tilghman