Adding John Dickinson candidacy for Swift
Change-Id: I06646fd0d2adf4b4ac4e33564bdb88b5dcb06a79
This commit is contained in:
parent
c0f7aade5a
commit
19ce430052
88
candidates/newton/Swift/John_Dickinson.txt
Normal file
88
candidates/newton/Swift/John_Dickinson.txt
Normal file
@ -0,0 +1,88 @@
|
||||
I am submitting my candidacy for Swift PTL for the Newton cycle.
|
||||
|
||||
### Community growth
|
||||
|
||||
During the Mitaka cycle, we've seen the community grow from about 440
|
||||
contributors to 531 today. That's a 20% growth in the last six months.
|
||||
This growth rate is huge, and I'm really happy to see all the new
|
||||
people coming into the community. New contributors bring new ideas,
|
||||
new use cases, and fresh eyes on old problems. This is the hallmark of
|
||||
a healthy community.
|
||||
|
||||
New contributors reflect the broader picture of what's going on with
|
||||
Swift, too. When we see new contributors from companies who are
|
||||
already actively participating in the community, it shows that they
|
||||
are using Swift more. When new companies join the community for the
|
||||
first time, it shows new places where Swift is being used to solve
|
||||
real-world storage problems. Both situations are exciting. Swift is
|
||||
growing. More people are using it to store more data and solve more
|
||||
storage problems.
|
||||
|
||||
As I've said many times before, my vision for Swift is that it will be
|
||||
used by everyone, every day, even if people don't realize it.
|
||||
Community growth is one way we can see that this is happening today.
|
||||
|
||||
### Tracking the community
|
||||
|
||||
One thing I've been working on over the last year is tracking our
|
||||
community and finding ways to understand it and help it continue to
|
||||
grow. You can see some of this work in the graphs below.
|
||||
|
||||
http://d.not.mn/total_contribs.png
|
||||
http://d.not.mn/active_contribs.png
|
||||
|
||||
I've been working on a few more interesting graphs and metrics too.
|
||||
I've shared one of these before: visualizing individual contributor
|
||||
activity over time. You can see the results at
|
||||
http://d.not.mn/contrib_activity.png. This is a rather large graph,
|
||||
but it's fairly simple to read. The x-axis is time (from Swift's
|
||||
initial release to today). Each line on the y-axis is a person's
|
||||
activity in Swift. Blue boxes are for authoring a patch; green boxes
|
||||
are for a review. This graph has been invaluable in helping understand
|
||||
who is contributing to the project and what activity active
|
||||
contributors are engaged in. One thing I've learned is that there is a
|
||||
class of contributor who has been involved for a long time, but is
|
||||
only active infrequently. Most of these people are operators who are
|
||||
normally running production clusters but occasionally need to submit a
|
||||
patch or review upstream.
|
||||
|
||||
Another thing I've worked on is a way to find out what the community
|
||||
as a whole thinks is important. We can start to find a lot of this
|
||||
info from information we already have. For example, we can get a list
|
||||
of every patch that every contributor has starred to see if there is
|
||||
any commonality between them. The patches that are more often starred
|
||||
are likely to be more important, from the community perspective.
|
||||
|
||||
I've taken that basic idea, along with some further parsing of
|
||||
information from gerrit, and created a review dashboard. It's going to
|
||||
keep changing, but you can see its current incarnation at
|
||||
http://not.mn/swift/swift_community_dashboard.html (and linked in
|
||||
the #openstack-swift channel topic). So far, I've seen this dashboard
|
||||
result in a dramatic decrease in the number of unreviewed patches, and
|
||||
as the dashboard improves, I expect it to lead to a similar
|
||||
improvement in review times for important patches.
|
||||
|
||||
### Current struggles
|
||||
|
||||
Tracking the community (both with metrics like above and from simply
|
||||
talking to people) shines a light on common problems in the community.
|
||||
Two of the most-often raised issues are that of long patch review
|
||||
times and review prioritization. We've been improving both of these
|
||||
with several tools in the past, and we're currently in a much better
|
||||
place than we've been in the past. As PTL, I feel it's my duty to
|
||||
enable the community to solve these problems. I will continue to
|
||||
improve the tools we have and create new tools as necessary so that we
|
||||
know what's going on, what to work on, and make every contributor
|
||||
effective.
|
||||
|
||||
### Goals for newton
|
||||
|
||||
Looking forward to the Newton cycle, I want to see three things happen.
|
||||
|
||||
1. Increased contributor growth
|
||||
2. Progress on significant code efforts, including crypto, improved
|
||||
global clusters, and improved performance
|
||||
3. Better tools and info for community prioritization and community tracking
|
||||
|
||||
The community is well-positioned to meet these goals, and I will be
|
||||
honored to continue to serve you as PTL for OpenStack Swift.
|
Loading…
Reference in New Issue
Block a user