<html>
  <head>
    <meta content="text/html; charset=utf-8" http-equiv="Content-Type">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    <p><br>
    </p>
    <br>
    <div class="moz-cite-prefix">On 07/12/16 20:19, Liam Miller-Cushon
      wrote:<br>
    </div>
    <blockquote
cite="mid:CAL4QsgsNNS4E3-GESpJwrgwXjV+=BufPyOz3q3swAONwRiz5vg@mail.gmail.com"
      type="cite">
      <div dir="ltr">What do you think about pursuing this as an interim
        fix? It solves a problem we were seeing trying to analyze entire
        compilation units after flow, and it's harder to observe whether
        by-file is followed correctly after that. Or would you rather
        just wait for the ordering constraint to be removed in 10?</div>
    </blockquote>
    I think it's very possible that the ordering constraint will go away
    early in 10, so perhaps it's not a long wait - after all I think
    this patch would presumably be considered to risky to pursue at this
    stage in 9?<br>
    <br>
    Maurizio<br>
    <blockquote
cite="mid:CAL4QsgsNNS4E3-GESpJwrgwXjV+=BufPyOz3q3swAONwRiz5vg@mail.gmail.com"
      type="cite">
      <div class="gmail_extra"><br>
        <div class="gmail_quote">On Sun, Dec 4, 2016 at 2:11 PM,
          Maurizio Cimadamore <span dir="ltr"><<a
              moz-do-not-send="true"
              href="mailto:maurizio.cimadamore@oracle.com"
              target="_blank"><a class="moz-txt-link-abbreviated" href="mailto:maurizio.cimadamore@oracle.com">maurizio.cimadamore@oracle.com</a></a>></span>
          wrote:<br>
          <blockquote class="gmail_quote" style="margin:0 0 0
            .8ex;border-left:1px #ccc solid;padding-left:1ex">
            <div bgcolor="#FFFFFF" text="#000000">
              <p>Also don't forget that the cycle problem is temporary¬†
                - as discussed in that thread, the fix here is to
                actually get rid of erasure (which is something we did
                already in an experimental patch in Valhalla) which then
                removes the ordering problems with supertypes/subtypes.
                We plan to do further experiments with this in the
                Valhalla repo and move this code to 10 when ready.</p>
              <span class="HOEnZb"><font color="#888888">
                  <p>Maurizio<br>
                  </p>
                </font></span>
              <div>
                <div class="h5"> <br>
                  <div class="m_-981544833932075067moz-cite-prefix">On
                    02/12/16 02:18, Liam Miller-Cushon wrote:<br>
                  </div>
                  <blockquote type="cite">
                    <div dir="ltr">
                      <div class="gmail_extra">
                        <div class="gmail_quote">On Thu, Dec 1, 2016 at
                          6:06 PM, Jonathan Gibbons <span dir="ltr"><<a
                              moz-do-not-send="true"
                              href="mailto:jonathan.gibbons@oracle.com"
                              target="_blank"><a class="moz-txt-link-abbreviated" href="mailto:jonathan.gibbons@oracle.com">jonathan.gibbons@oracle.com</a></a>></span>
                          wrote:<br>
                          <blockquote class="gmail_quote"
                            style="margin:0px 0px 0px
                            0.8ex;border-left:1px solid
                            rgb(204,204,204);padding-left:1ex">
                            <div id="m_-981544833932075067gmail-:wn"
                              class="m_-981544833932075067gmail-a3s
                              m_-981544833932075067gmail-aXjCH
                              m_-981544833932075067gmail-m158bd489f3a3b345">Does
                              this address the potential problem with
                              cycles that you mentioned in<br>
                              the original email?</div>
                          </blockquote>
                        </div>
                        <br>
                        Not fully. It improves the current behaviour by
                        ensuring files always go through attr and flow
                        as a unit, but classes may still be desugared
                        separately from their compilation unit.</div>
                      <div class="gmail_extra"><br>
                      </div>
                      <div class="gmail_extra">One of the test cases has
                        compilation units [One, Two] and [Three, Four],
                        where Two extends Four, and Three extends One.
                        There's no way to linearize that so supertypes
                        are desugared first and files are desugared as a
                        unit.</div>
                      <div class="gmail_extra"><br>
                      </div>
                      <div class="gmail_extra">Are there still plans to
                        remove<span style="color:rgb(0,0,0)"> the
                          ordering constraint in TransTypes?</span></div>
                    </div>
                  </blockquote>
                  <br>
                </div>
              </div>
            </div>
          </blockquote>
        </div>
        <br>
      </div>
    </blockquote>
    <br>
  </body>
</html>