Post-TC1 strategizin'

[Richard Ryniker]

On Thu, 04 Aug 2011 14:11:43 +0530, Rahul Sundaram wrote:
>On 08/04/2011 06:11 AM, Adam Williamson wrote:
>> I thought it'd be a good idea to take a step back and plan a strategy
>> for post-Alpha TC1.
>>
>> We're in trouble for Alpha; we're a long way behind schedule, and we
>> have a lot of blocker bugs remaining. TC1 is in pretty bad shape.
>
>Why did this happen?  What went wrong?  The strategy should be based on
>that. 

I doubt anything "went wrong".  The F16 branch occurred when Fedora's
semiannual release objective said it had to occur, in order to meet a
target release date, not when Rawhide satisfied some stability criterion
associated with "not too many bugs."  It is likely this F16 branch simply
captured more problems than usual.

Therefore, the present need is a short-term strategy to handle the
F16 problems efficiently.

In the longer term, experience may lead to better techniques to handle
problems after a branch - schedule options, pre-branch tests, pipelined
schedule with two branches simultaneously in development... whatever
works.