[asterisk-users] Discussion: Are we ready to leave 1.4 behind?

[Michael L. Young]

----- Original Message -----
> From: "Olle E. Johansson" <oej at edvina.net>
> To: "Asterisk Users Mailing List - Non-Commercial Discussion" <asterisk-users at lists.digium.com>
> Sent: Wednesday, April 27, 2011 3:34:03 PM
> Subject: [asterisk-users] Discussion: Are we ready to leave 1.4 behind?
>
> Friends,
>
> We have a discussion on asterisk-dev about the maintenance of the 1.4
> branch. According to the release plans, support for 1.4 was
> scheduled to close in April 2011 - basically now. After that, only
> security patches would be committed. This is already a delay from
> the original plan published by Russell Bryant.
>
> Unfortunately, I think this is way too early. My feeling and
> experience is that 1.8 is not ready for production in the
> environments I work in - large scale installations. Customers are
> not planning migration and all new installs are still 1.4. Tests
> we've been doing with 1.8 has failed within just a short time and so
> badly that customers has not paid me to spend any further time with
> 1.8.
>

Whats the game plan to get 1.8 "ready for production"?  To me, for which I say this with all respect, some are focusing still on 1.4 instead of getting 1.8 to the level that some of the members of the community are wanting to see.  1.4 has been very stable for a while.  To the point that I only pay attention to security releases to be honest.  It has been this way for quite a while now.  I personally have been focused more on using 1.8 when I can, mainly on non-critical servers, yet I will admit that I have enough confidence in it now to use on main servers.  Why?  Because I want to get my production servers off of 1.4 and 1.6.2 due to new features.  But, even if I didn't need or want the new features, the current state of 1.4 is excellent.  If I don't ever make the move beyond 1.4, how can I contribute to a better product?  By experimenting and not giving up at the first sign of trouble with the latest version, I feel that I can help to make 1.8 better which ultimately benefits me and the community.  I would like to hear a game plan before we just say, yes, lets keep focusing on 1.4 and then we will decide a deadline to stop support.  I am afraid that software is programmed by imperfect humans and there will always be a bug or two that crops up from time to time.  Do we want to keep waiting until we feel it is "perfect"?

One thing I have noticed, is that the bug fixes and patches being contributed for 1.8 and trunk are not being taken care of as quickly as it used to back in the early 1.4 days.  My feelings are that it is because there have been too many releases to work on.  Going back to focusing on just 1.8 and trunk, would go a long way to speeding up bug fixes to 1.8.  Again, just my opinion.

> Last time we went through this process with a LTS release (which we
> did not know then) it took over one year before we had a stable
> product to migrate away from 1.2 and jump on the 1.4 track.
> Hopefully, with the help of community, we can move up to 1.8 late
> this year or early next year. For me 1.8 is the focus, it's the LTS
> release.
>
> Not having a supported 1.4 version from the Digium-hosted
> repositories will mean that we will have to move to separate
> repositories or branch off from the main track. I already maintain a
> ton of subversion branches with various patches to 1.4 It takes a
> lot of time to manage this version that is a fork from the main 1.4
> branch. I will soon have to start working with subversion branches
> for 1.8 to create a compatible version for my customers to test,
> since most of the patches is not part of 1.8. After a few years of
> doing this, I know the work involved with managing code myself.
>
> The Digium team wants to go ahead and not support 1.4 any more, I
> want to keep 1.4 open for normal bug fixes. What do you think?

Was this really Digium's decision?  You keep mentioning Digium and implying them as the evil one in all of this (perhaps I am just misunderstanding your tone in your emails and if I am, I sincerely apologize for this) when I seem to recall plenty of discussion around these time lines and it was the community who set the deadlines, not Digium.  Digium is just trying to abide by the time lines outlined for them by the community.  They have already been nice enough to extend the deadline in order to finish working on outstanding bug reports and patches.  They have bills to pay too and have really tried to extend an olive branch to everyone in the community.  There has been a lot of activity on the 1.4 branch lately.  If I am wrong, I will gladly retract my comments.

>
> Kevin proposed that the community maintains the 1.4 branch without
> support from the Digium team. I don't think that's a good solution,
> but it may be the only solution.  I haven't got the resources to
> manage the 1.4 code myself, so I won't step forward as a maintainer
> if I can't get proper funding. Anyone else out there that has the
> time and resources to manage the code?

Again, I don't feel Digium really should be brought up in this discussion.  It seems that the main motivation for bringing this up is to get people to speak up as if a commercial entity is trying to hurt the community somehow by no longer being able to offer their resources on an older code base.

There are many people and projects who financially are depending on 1.4.  If this was so important to them, why was it not brought up sooner?  Why don't these ones then help with the maintenance of bugs?  So far I keep hearing, "Yes, lets keep supporting 1.4.  Oh but I can't help out with maintaining it".  All companies involved with or who use Asterisk are benefiting from the community.  Why are they not helping to get 1.8 stable, or "production ready"?  Instead, they don't seem to balance between benefiting from the community and contributing back so that the latest version with the new features that others may need, gets to a production worthy state.  The asterisk project needs to move on in order to stay where it is as one of the leaders in the industry.  I am looking forward to 1.10 and would love to see the community not being distracted with older releases.  This in the end will benefit the whole community.

So, where is the game plan to make 1.8 the community's priority since it is LTS?  That should be established first.  Then, how do we go about continuing to support major bug fixes found in 1.4 as a community with the left over resources (again the community's resources since we are the ones who established the deadlines to begin with and it falls on us, not Digium, if we want to change it now).  That should be the order of priorities in my mind.

I hope my comments are taken the right way.  My goal is to offer another viewpoint and contribute to a positive discussion.

Michael
(elguero)