On 10/02/2009 04:53 PM, Rahul Sundaram wrote:
A important oddness of the feature process is that it is not
actually
necessary for the feature to be in the distribution. So you send a nag
mail, feature owners ignores it, you "drop" the feature and the
functionality is still there. So unless the feature owner is actually
convinced it is worth the effort, he or she can let the feature be
"dropped" and continue just working on whatever they are. I don't know
whether this can be fixed or is considered a problem but I think we
should see if we can do it in a different way.
It isn't always a problem at all. Other times it is a huge problem.
NFSv4 by default is an example of a change that could cause problems.
The reason being that it affected numerous other pieces of the distro.
That kind of change needs coordination (even if it's just break the
world at the beginning of the cycle and everyone figures out workarounds
before beta.)
OTOH, Better WebCam Support was self contained among a few developers
who were already talking to each other. This is a feature where
dropping it is only dropping the release notes portion, not the actual work.
So do we need to fix this? I think recognizing that there's different
types of features and some of them can continue even if they are dropped
while others must be (wholly or partially) reverted if they are dropped
is step one. Putting more effort (tracking development, making sure
that the owner knows what things need to be in place for what deadlines,
finding extra manpower if possible if a feature is slipping) towards the
ones that would need to be reverted if they don't complete is step two.
-Toshio