<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <br>
    <br>
    <div class="moz-cite-prefix">On 2/14/19 5:47 AM, Morgan Fainberg
      wrote:<br>
    </div>
    <blockquote type="cite"
cite="mid:CAGnj6avrRzK-kwc1q0vx9bUR6tCYT16-+4QcNSNooyRaYXKSTQ@mail.gmail.com">
      <meta http-equiv="content-type" content="text/html; charset=UTF-8">
      <div dir="auto">Rethinking my last email... Go with just release
        notes, no need for a bug.</div>
    </blockquote>
    <br>
    <tt>The only thing we lose with this would be a place to see every
      commit that deprecated or removed something in a release (short of
      doing a git blame on the release note). We could still do this
      with bugs and we could drive the tracking with Partial-Bug in each
      commit message. We need to make sure to formally close the bug
      however at the end of the release if we don't close it with a
      commit using Closes-Bug. In my experience, we rarely stage all
      these commits at once. They're usually proposed haphazardly
      throughout the release as people have cycles.</tt><br>
    <br>
    <blockquote type="cite"
cite="mid:CAGnj6avrRzK-kwc1q0vx9bUR6tCYT16-+4QcNSNooyRaYXKSTQ@mail.gmail.com"><br>
      <div class="gmail_quote">
        <div dir="ltr">On Thu, Feb 14, 2019, 06:46 Morgan Fainberg <<a
            href="mailto:morgan.fainberg@gmail.com"
            moz-do-not-send="true">morgan.fainberg@gmail.com</a> wrote:<br>
        </div>
        <blockquote class="gmail_quote" style="margin:0 0 0
          .8ex;border-left:1px #ccc solid;padding-left:1ex">
          <div dir="auto">
            <div>I would go for one tracking bug per cycle or we could
              also just lean on the release notes instead of having a
              direct bug. <br>
              <br>
              <div class="gmail_quote">
                <div dir="ltr">On Thu, Feb 14, 2019, 06:07 Colleen
                  Murphy <<a href="mailto:colleen@gazlene.net"
                    target="_blank" rel="noreferrer"
                    moz-do-not-send="true">colleen@gazlene.net</a>
                  wrote:<br>
                </div>
                <blockquote class="gmail_quote" style="margin:0 0 0
                  .8ex;border-left:1px #ccc solid;padding-left:1ex">On
                  Wed, Feb 13, 2019, at 8:56 PM, Lance Bragstad wrote:<br>
                  > Over the last couple of years, our launchpad
                  blueprints have grown<br>
                  > unruly [0] (~77 blueprints a few days ago). The
                  majority of them were in<br>
                  > "New" status, unmaintained, and several years old
                  (some dating back to<br>
                  > 2013). Even though we've been using
                  specifications [1] for several<br>
                  > years, people still get confused when they see
                  conflicting or inaccurate<br>
                  > blueprints. After another person tripped over a
                  duplicate blueprint this<br>
                  > week, cmurphy, vishakha, and I decided to devote
                  some attention to it.<br>
                  > We tracked the work in an etherpad [2] - so we
                  can still find links to<br>
                  > things.<br>
                  > <br>
                  > First, if you are the owner of a blueprint that
                  was marked as<br>
                  > "Obsolete", you should see a comment on the
                  whiteboard that includes a<br>
                  > reason or justification. If you'd like to
                  continue the discussion about<br>
                  > your feature request, please open a specification
                  against the<br>
                  > openstack/keystone-specs repository instead. For
                  historical context,<br>
                  > when we converted to specifications, we were only
                  supposed to create<br>
                  > blueprints for tracking the work after the
                  specification was merged.<br>
                  > Unfortunately, I don't think this process was
                  ever written down, which<br>
                  > I'm sure attributed to blueprint bloat over the
                  years.<br>
                  > <br>
                  > Second, if you track work regularly using
                  blueprints or plan on<br>
                  > delivering something for Stein, please make sure
                  your blueprint in<br>
                  > Launchpad is approved and tracked to the
                  appropriate release (this<br>
                  > should already be done, but feel free to double
                  check). The team doesn't<br>
                  > plan on switching processes for feature tracking
                  mid-release. Instead,<br>
                  > we're going to continue tracking feature work
                  with launchpad blueprints<br>
                  > for the remainder of Stein. Currently, the team
                  is leaning heavily<br>
                  > towards using RFE bug reports for new feature
                  work, which we can easily<br>
                  > switch to in Train. The main reason for this
                  switch is that bug comments<br>
                  > are immutable with better timestamps while
                  blueprint whiteboards are<br>
                  > editable to anyone and not timestamped very well.
                  We already have<br>
                  > tooling in place to update bug reports based on
                  commit messages and that<br>
                  > will continue to work for RFE bug reports.<br>
                  > <br>
                  > Third, any existing blueprints that aren't
                  targeted for Stein but are<br>
                  > good ideas, should be converted to RFE bug
                  reports. All context from the<br>
                  > blueprint will need to be ported to the bug
                  report. After a sufficient<br>
                  > RFE bug report is opened, the blueprint should be
                  marked as "Superseded"<br>
                  > or "Obsolete" *with* a link to the newly opened
                  bug. While this is<br>
                  > tedious, there aren't nearly as many blueprints
                  open now as there were a<br>
                  > couple of days ago. If you're interested in
                  assisting with this effort,<br>
                  > let me know.<br>
                  > <br>
                  > Fourth, after moving non-Stein blueprints to RFE
                  bugs, only Stein<br>
                  > related blueprints should be open in launchpad.
                  Once Stein is released,<br>
                  > we'll go ahead disable keystone blueprints.<br>
                  > <br>
                  > Finally, we need to overhaul a portion of our
                  contributor guide to<br>
                  > include information around this process. The goal
                  should be to make that<br>
                  > documentation clear enough that we don't have
                  this issue again. I plan<br>
                  > on getting something up for review soon, but I
                  don't have anything<br>
                  > currently, so if someone is interested in taking
                  a shot at writing this<br>
                  > document, please feel free to do so. Morgan has a
                  patch up to replace<br>
                  > blueprint usage with RFE bugs in the
                  specification template [3].<br>
                  > <br>
                  > We can air out any comments, questions, or
                  concerns here in the thread.<br>
                  <br>
                  What should we do about tracking "deprecated-as-of-*"
                  and "removed-as-of-*" work? I never liked how this was
                  done with blueprints but I'm not sure how we would do
                  it with bugs. One tracking bug for all deprecated
                  things in a cycle? One bug for each? A
                  Trello/Storyboard board or etherpad? Do we even need
                  to track it with an external tool - perhaps we can
                  just keep a running list in a release note that we add
                  to over the cycle?<br>
                </blockquote>
              </div>
            </div>
          </div>
        </blockquote>
      </div>
    </blockquote>
    <br>
    <tt>I agree. The solution that is jumping out at me is to track one
      bug for deprecated things and one for removed things per release,
      so similar to what we do now with blueprints. We would have to
      make sure we tag commits properly, so they are all tracked in the
      bug report. Creating a bug for everything that is deprecated or
      removed would be nice for capturing specific details, but it also
      feels like it will introduce more churn to the process.<br>
      <br>
      I guess I'm assuming there are users that like to read every
      commit that has deprecated something or removed something in a
      release. If we don't need to operate under that assumption, then a
      release note would do just fine and I'm all for simplifying the
      process.<br>
    </tt><br>
    <blockquote type="cite"
cite="mid:CAGnj6avrRzK-kwc1q0vx9bUR6tCYT16-+4QcNSNooyRaYXKSTQ@mail.gmail.com">
      <div class="gmail_quote">
        <blockquote class="gmail_quote" style="margin:0 0 0
          .8ex;border-left:1px #ccc solid;padding-left:1ex">
          <div dir="auto">
            <div>
              <div class="gmail_quote">
                <blockquote class="gmail_quote" style="margin:0 0 0
                  .8ex;border-left:1px #ccc solid;padding-left:1ex">
                  <br>
                  Thanks for tackling this cleanup work.<br>
                  <br>
                  > <br>
                  > Thanks,<br>
                  > <br>
                  > Lance<br>
                  > <br>
                  > [0] <a
                    href="https://blueprints.launchpad.net/keystone"
                    rel="noreferrer noreferrer noreferrer"
                    target="_blank" moz-do-not-send="true">https://blueprints.launchpad.net/keystone</a><br>
                  > [1] <a
                    href="http://specs.openstack.org/openstack/keystone-specs/"
                    rel="noreferrer noreferrer noreferrer"
                    target="_blank" moz-do-not-send="true">http://specs.openstack.org/openstack/keystone-specs/</a><br>
                  > [2] <a
                    href="https://etherpad.openstack.org/p/keystone-blueprint-cleanup"
                    rel="noreferrer noreferrer noreferrer"
                    target="_blank" moz-do-not-send="true">https://etherpad.openstack.org/p/keystone-blueprint-cleanup</a><br>
                  > [3] <a
                    href="https://review.openstack.org/#/c/625282/"
                    rel="noreferrer noreferrer noreferrer"
                    target="_blank" moz-do-not-send="true">https://review.openstack.org/#/c/625282/</a><br>
                  > Email had 1 attachment:<br>
                  > + signature.asc<br>
                  >   1k (application/pgp-signature)<br>
                  <br>
                </blockquote>
              </div>
            </div>
          </div>
        </blockquote>
      </div>
    </blockquote>
    <br>
  </body>
</html>