Change of release and governance model for Thunderbird

Ehsan Akhgari ehsan.akhgari at gmail.com
Mon Jul 9 17:44:47 UTC 2012


On 12-07-09 12:57 PM, Kai Engert wrote:
> The one thing I'm worried about is regressions.

Yes, this worries me a lot as well.

> I have a very clear opinion how this should be handled. In my opinion,
> existing functionality in the core engine should never be removed, if
> it's currently being used by important projects such as Thunderbird.
> Instead of removing functionality from the core engine, because it
> allowed Firefox projects to proceed faster (happened in the past),
> developer should take care to implement additional features in addition,
> not as a replacement.

That's not really always an option.  We sometimes clean up parts of 
Gecko, rearchitect other parts and replace yet other parts completely. 
It's not always possible to keep all of the existing code.  And many 
people are cautious of changes which could potentially break 
comm-central (and sometimes people just forget, I've definitely been 
guilty of that myself.)

But the large problem is this is only part of what we should worry 
about.  Sometimes people make changes which doesn't really change any 
functionality, but it breaks something that Thunderbird users experience 
way more than Firefox users.  I've definitely seen a number of great bug 
reports being filed from Thunderbird users in the editor component, 
which we have tried to fix as quickly as possible.  If Thunderbird 
starts to live on something other than Gecko trunk, fixing these issues 
in time would be very hard (especially if Thunderbird decides to live 
off of Gecko ESR branches).

Ehsan



More information about the tb-planning mailing list