Change of release and governance model for Thunderbird

Ludovic Hirlimann lhirlimann at mozilla.com
Tue Aug 7 08:18:04 UTC 2012


On 7/11/12 5:27 PM, Wayne Mery wrote:
> (original posted on tb-enterprise)
>
> Quoting Ben Bucksch <ben.bucksch at beonex.com>:
>
>> On 10.07.2012 16:35, Jb Piacentino wrote:
> Many, many Thunderbird crashes and top crashers aren't identified 
> until release. Most of these crashes are core related, and a good 
> percentage are not topcrashes for Firefox. The end result is they take 
> longer to identify and take longer to fix - sometimes a few 
> development cycles.  This presents two problems. 1. Between TB17 and 
> TB24 there will probably be an accumulated "pile up" of unfixed core 
> crashes affecting TB.
>
We could also backport these ourselves and deal with Gecko drivers to 
make sure that they get included on the ESR branch.

> Effect on ESR 24
>
> Partly because of the above, I have serious doubts that the outline we 
> have now provides adequate testing to make ESR 24 as stable as it must 
> be for public release. It may need 2-3 cycles of beta for the quality 
> to be adequate.
>
>
That's doable and That's how I envision things to be.
> Hopefully Mark's update of the process will have addressed these items.
>
> Lastly, are we reasonably sure that Thunderbird releases can wait a 
> full year for features/changes delivered in the "next" Gecko? Example 
> considerations: crash reporting, session recovery, toolkit featues, 
> addons, editor, spell checker.
We can't have both world at the same time. Can't get the cow and the 
milk. I think our users are ok with that.


If I recall correctly the cycle's length for Gecko branches will tend to 
be shorter so it should fix a few of the aforementioned issues.

-- 
@lhirlimann on twitter
https://wiki.mozilla.org/Thunderbird:Testing




More information about the tb-planning mailing list