<html>
  <head>
    <meta http-equiv="Content-Type" content="text/html; charset=utf-8">
  </head>
  <body text="#000000" bgcolor="#FFFFFF">
    <p>Patrick,</p>
    <p>I did not attend the IVOA interops at Capetown and Victoria.</p>
    <p>I was not aware of the discussions and decision.</p>
    <p><br>
    </p>
    <p>François-Xavier<br>
    </p>
    <br>
    <div class="moz-cite-prefix">Le 13/06/2018 à 20:19, Patrick Dowler a
      écrit :<br>
    </div>
    <blockquote type="cite"
cite="mid:CAFK8nroPe_MMXv3uoN3hPn4J6gqdoNp=hnuFNtkDikXE8Y8JOw@mail.gmail.com">
      <div dir="ltr">
        <div class="gmail_default" style="font-size:small">The intent of
          DALI was that the butterfly polygon is not valid. We didn't
          say it explicitly and maybe should have. <br>
        </div>
        <div class="gmail_default" style="font-size:small"><br>
        </div>
        <div class="gmail_default" style="font-size:small">More
          generally, the DALI xtype="polygon" is a simple polygon
          (convex or concave) with no crossing segments. We had to
          decide between picking a winding direction or for the polygon
          to be limited to less that half the sphere. At the Capetown
          interop the decision was to pick CCW winding direction and
          that definition went into DALI-1.1. <br>
        </div>
        <div class="gmail_default" style="font-size:small"><br>
        </div>
        <div class="gmail_default" style="font-size:small">Pat<br>
        </div>
        <div class="gmail_default" style="font-size:small"><br>
        </div>
        <div class="gmail_default" style="font-size:small">PS-In my
          experience, that construct was always a bug in the polygon
          generation s/w</div>
        <div class="gmail_default" style="font-size:small"><br>
        </div>
        <div class="gmail_default" style="font-size:small">PPS-Anyone
          using postgresql+pgsphere should note that winding direction
          doesn't matter for spoly because that library chose the "less
          than half the sphere" definition<br>
        </div>
      </div>
      <div class="gmail_extra"><br>
        <div class="gmail_quote">On 13 June 2018 at 01:05,
          Francois-Xavier PINEAU <span dir="ltr">&lt;<a
              href="mailto:francois-xavier.pineau@astro.unistra.fr"
              target="_blank" moz-do-not-send="true">francois-xavier.pineau@astro.unistra.fr</a>&gt;</span>
          wrote:<br>
          <blockquote class="gmail_quote" style="margin:0 0 0
            .8ex;border-left:1px #ccc solid;padding-left:1ex">
            <div text="#000000" bgcolor="#FFFFFF">
              <p>Dear Marco, DAL,</p>
              <p>I may be wrong, but I think that the STC definition of
                the inside of a polygon is not compatible with "complex"
                shapes.</p>
              <p>Example: if we take the case of a simple 4 vertices's
                polygon having a butterfly shape (i.e. having two
                crossing great-circle arcs), then the inside of one
                "wing" is in the counter-clockwise sense while the
                inside of the other "wing" is in the clockwise sense.<br>
              </p>
              <p>How to deal easily with such a case while remaining
                compatible with the STC definition?</p>
              <p><br>
              </p>
              <p>A polygon is just a sequence of vertices with
                great-circle arcs connecting the consecutive vertices
                (plus the last vertex connection the first one).</p>
              <p>The great-circle arc connecting two vertices's is the
                smallest of the two complementary great-arcs (and hence
                is always &lt;= pi (we have a degeneracy with vertices
                having an angular separation of pi)).</p>
              <p>There is no ambiguity on the inside and we have fast
                implementations for polygons having possibly very
                complex shapes (ray-tracing method).</p>
              <p>It is true that then the definition does not allow to
                describe the "exterior" (the complement on the sphere)
                of a polygon as being itself a polygon. Is it a problem
                in practice?</p>
              <p>From my (biased) point-of-view, it is more important to
                support complex polygons (with a simple and fast
                algorithm), and to possibly ask for an extra boolean
                parameter in use cases requiring the complement of a
                polygon.</p>
              <p>Do you agree / disagree?<br>
              </p>
              <p><br>
              </p>
              <p>Kind regards,</p>
              <p><br>
              </p>
              <p>François-Xavier<br>
              </p>
              <div>
                <div class="h5">
                  <p><br>
                  </p>
                  <br>
                  <div class="m_7475597954916928517moz-cite-prefix">Le
                    12/06/2018 à 17:45, Marco Molinaro a écrit :<br>
                  </div>
                  <blockquote type="cite">
                    <div dir="ltr"><span
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial;float:none;display:inline">Dear
                        DAL members,</span><br
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">
                      <span
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial;float:none;display:inline">at
                        the recent IVOA Interop in Victoria it was
                        pointed out</span><br
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">
                      <span
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial;float:none;display:inline">by
                        Alberto</span><br
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">
                      <br
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">
                      <a
href="http://wiki.ivoa.net/internal/IVOA/InterOpMayy2018DAL/ivoa_201805_micol_polygons.pdf"
                        style="color:rgb(17,85,204);font-size:12.8px"
                        target="_blank" moz-do-not-send="true">http://wiki.ivoa.net/internal/<wbr>IVOA/InterOpMayy2018DAL/ivoa_2<wbr>01805_micol_polygons.pdf</a><br
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">
                      <br
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">
                      <span
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial;float:none;display:inline">that
                        not all the data providers follow the STC
                        specification</span><br
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">
                      <span
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial;float:none;display:inline">about
                        the winding direction of polygons stored in
                        their</span>
                      <div
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">archives.<br>
                        <br>
                        STC states that "The inside of the region is
                        defined as 
                        <div>that part of coordinate space that is
                          encircled by the </div>
                        <div>polygon in a counter-clockwise sense".</div>
                        <div>And this is to be considered when looking
                          at the</div>
                        <div>celestial sphere from the inside.</div>
                        <div><br>
                        </div>
                        <div>The different behaviour of the polygons
                          stored at </div>
                        <div>different sites creates an interoperability
                          issue.</div>
                        <div><br>
                        </div>
                        <div>This email is a request on data providers
                          to check</div>
                        <div>on their data and implementations (when
                          dealing</div>
                      </div>
                      <div
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">with
                        polygons) to solve the presented issue.</div>
                      <div
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"><br>
                      </div>
                      <div
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">James
                        also put together a python code snippet </div>
                      <div
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">implementing
                        a simple test, it is available here:</div>
                      <div
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"><br>
                      </div>
                      <div
                        style="text-decoration-style:initial;text-decoration-color:initial"><span
                          style="font-size:12.8px"><a
                            href="https://gist.github.com/jd-au/45d1cdc0c6e2a7bc848a4be8f46c8958"
                            target="_blank" moz-do-not-send="true">https://gist.github.com/jd-au/<wbr>45d1cdc0c6e2a7bc848a4be8f46c89<wbr>58</a></span><br>
                      </div>
                      <div
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"><br>
                      </div>
                      <div
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Thank
                        you in advance!</div>
                      <div
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial">Cheers,</div>
                      <div
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"> 
                             Marco &amp; James</div>
                      <div
style="font-size:12.8px;text-decoration-style:initial;text-decoration-color:initial"> 
                            (your DAL chair &amp; vice)</div>
                      <br>
                    </div>
                  </blockquote>
                  <br>
                </div>
              </div>
            </div>
          </blockquote>
        </div>
        <br>
        <br clear="all">
        <br>
        -- <br>
        <div class="gmail_signature" data-smartmail="gmail_signature">
          <div dir="ltr">
            <div>
              <div>Patrick Dowler<br>
              </div>
              Canadian Astronomy Data Centre<br>
            </div>
            Victoria, BC, Canada<br>
          </div>
        </div>
      </div>
    </blockquote>
    <br>
  </body>
</html>