Do we need feature tracking bugs?

John Poelstra poelstra at redhat.com
Tue Aug 19 22:16:53 UTC 2008


During the feature process review for Fedora 10 it was suggested that 
each new feature have a separate tracking bug which would then block the 
F10 release blocker.

As the Feature Wrangler I agreed to create each feature bug blocker so 
that there was not more overhead on the feature owner.  I confess that I 
have not created any of the blocker bugs yet and am doubting the value 
of doing so... and wishing I hadn't volunteered :)

It is going to take a fair amount of time to go through each feature 
page, determine the BZ component, open a bug, add the bug to the feature 
page, etc. and I'm doubting that these feature blocker BZs will really 
get used.

I'm not seeing the value that would be provided beyond simply adding new 
feature bugs directly to the release blocker.  Yes, we would know at a 
"feature" level if something wasn't looking good for the release, but 
the release blocker is not so big that this wouldn't be readily obvious.

1) Who will be responsible to make sure that each new bug related to a 
new feature blocks its respective tracker? This would be a natural bug 
triage activity, but we do not have enough people participating in bug 
triage right now to take this on.

2) Who is going to walk all the individual feature tracker bugs to make 
sure we are ready to release?

3) What about features that don't have a BZ component--block the distro 
component?

I would like to propose that feature tracking bugs be created on an "as 
needed basis" and not for every single feature listed here: 
https://fedoraproject.org/wiki/Releases/10/FeatureList

John




More information about the fedora-devel-list mailing list