[mythtv] Stable release branch starting post-0.18

[Jan Örnstedt]

Thanks Jarod!!

Best news on this list for a very long time.
I case you need any help testing releases on EPIA & Mac I will happily
assist. I have a multi platform setup and I have no time in keeping
everything up to latest CVS releases but will happily spend time on
"stable" release testings.

Thanks
Janne

On Wed, 23 Mar 2005 01:32:02 -0800, "Jarod Wilson" <jarod at wilsonet.com>
said:
> Hey folks, I've volunteered to be the sucker/punching bag to try to
> maintain a 
> "stable" release branch of Myth. This all seems oddly familiar to a
> recent 
> discussion on the lkml... :-)
> 
> The goal here is to provide an alternative for non-developer users that
> is 
> slightly less scary than using CVS (which honestly isn't scary at all,
> but I 
> understand where people are coming from). Essentially, it'll be
> back-ported 
> bug fixes from the main development branch to the latest major release.
> 
> Anyhow, here's the rough plan to get started:
> 
> - Starting with the 0.18 release, we'll branch a "stable" tree after
> every new 
> release from the main development branch.
> 
> - New releases from the development branch are X.Y.0, subsequent 
> point-releases from the stable tree will be (no surprise here) X.Y.1,
> X.Y.2, 
> etc.
> 
> - The X.Y.Z stable tree will cease to be maintained after the X.[Y+1]
> release 
> is out, only a stable tree based off the latest release from the
> development 
> branch will be supported.
> 
> - There should never be any new functionality added to the stable branch,
> only 
> bug/compatibility fixes.
> 
> - For the most part, patches will be cherry-picked from the main
> development 
> branch. If it hasn't passed muster to be committed there, there must be a 
> very compelling reason to commit it to the stable tree before it goes in 
> there.
> 
> That's all I've got for the moment, I'm sure there's plenty more to
> consider, 
> but I'm too tired to think anymore right now, time for bed. 
> Comments/suggestions welcome.
> 
> -- 
> Jarod Wilson
> jarod at wilsonet.com