Thunderbird and Taskcluster (was Re: My personal agenda for TB 52)

R Kent James kent at caspia.com
Mon Jan 25 18:52:14 UTC 2016


On 1/22/2016 3:02 PM, Jörg Knobloch wrote:
> I'm happy to work on other stuff to achieve long term goals if my 
> limited skills permit it. Of course these long term goals would need 
> to be defined, otherwise people like myself who are not visionaries 
> like Joshua and Kent will fix a bug here and there without really 
> securing the future. As far as I can see, TB 45 is looking good, so 
> that should enable us to start looking at the bigger picture.

Joshua and I are both putting most of our effort toward converting key 
pieces of Thunderbird into Javascript. The implicit reasoning there is 
that, long-term, for Thunderbird to survive it will need to move toward 
less dependence on the Mozilla C++/XUL/XPCOM backend.

But the big picture for the next year is that Mozilla is pushing us away 
in various ways, and some of us need to step forward and do things that 
many of us have "limited skills" to do that were formerly done by 
Mozilla. That would particularly include the requirement to move off of 
Mozilla release and build infrastructure.

What you are doing and proposing to do is valuable Jorg. So if that is 
where you feel comfortable working, then by all means do it. But I hope 
that someone can feel the urge to take on aspects of the infrastructure 
problem.

Here's the status of infrastructure issues: Joshua is moving toward a 
combined m-c/c-c, but after that we need to start experimenting with our 
own build management and scripts. Mozilla releng is clearly moving 
toward Taskcluster, and would like us to move there too, so much so that 
if we move that direction we could probably get both staff and 
infrastructure support from them, and possibly even continue to run on 
their infrastructure for an extended period of time. But I am reluctant 
to recommend Taskcluster unless we are convinced that we could also run 
Taskcluster independently of Mozilla. Though Taskcluster is open source, 
making it a public resource that other organizations could run has not 
been a Mozilla priority. Someone needs to investigate Taskcluster and 
see if it possible now or in the reasonably foreseeable future to run 
that independently of Mozilla, should we need to do that in one of our 
possible futures. Otherwise, we need to resist converting to Taskcluster.

Yeah I have "limited skills" at this too.

:rkent


More information about the tb-planning mailing list