Discuss:Concept of Stable Release - Future

Discussion in 'General' started by briankelly63, Dec 13, 2009.

  1. briankelly63

    Joined:
    Oct 28, 2009
    Messages:
    19
    Likes Received:
    0
    Dear Forum-mates,

    I like to stimulate some thought and discussion on the idea of the "Stable Release". I come from the days of big iron proprietary telephone systems. They were big, expensive, high performance and lasted a long time. They were also rather limited in terms of flexibility and upgrade-ability those characteristics were controlled by the corporations that produced them, innovation was rather slow at the feature level.

    Society in general these days likes to have the best and newest. Last years must have electronic device is this years cast off or at least its old firmware is. Create a new software or firmware release and right away many of us will load it up right away.

    As you look at the three major distros (PIAF, TRIXBOX & ELASTIX) in no particular order. You can see different philosophies in terms of how each distro is implemented. Rather than going into each one I'll try to stay with the big picture and the point I'd like to see input on...

    It would seem that in some cases we have completely lost sight of the whole "Stable Release" concept. For the younger among us this concept used to mean that the distribution and ALL its component parts were functional, TESTED and locked down as a group. On that basis all the features, no matter how many were integrated, were all guaranteed to function. Yes, that didn't always happen but then there have always been bugs!

    If you change even one component you'd break the concept of a "Stable Release" and you'd be on your own. I'm not advocating the return of the dark ages but I do think there are elements from the past that should be re-examined and carried forward.

    I'd like to see us, as industry professionals, promote a more stable VOIP / Asterisk platform that employs extensive testing within a release. Patches as needed would be incoporated in and become part of the next "Stable Release". As needed there might be hot-fixes that are cautiously applied to critical situations that arise ant particular sites but these would be very limited and site specific.

    Why do I think that might be a good idea? As I read this and other forums I say many many people applying fixes, module replacements, yum updates, OS updates, RPM's and feature scripts constantly with very mixed results. It becomes very difficult for others to help resolve issues when the Computer, Telephony hardware, OS version and all the software application pieces are slightly different than the next persons.

    Its clear the forum leaders and older timers in the business are emphatic about production vs. sandbox machines often to no avail. Even if you go with that well tested concept how would you insure that your installation is the same as someone else s installation of the same "Stable Release".

    So my intial thread here is meant to stimulate some thought an hopefully an outpouring of comments on distro philosophy, updates, support and the future of VOIP stability and reputation within our present and future client base. Think about your current distro and your own practices, how stable and standard is your clients release or across your own customer base?

    I hope the results of tread will also help guide the developers in the techniques they use to deploy future releases.
     

Share This Page