<html>
  <head>

    <meta http-equiv="content-type" content="text/html; charset=utf-8">
  </head>
  <body bgcolor="#FFFFFF" text="#000000">
    Folks,<br>
    <br>
    Find the minutes below and at<br>
    <br>
     
<a class="moz-txt-link-freetext" href="https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_March_18,_2015#Minutes">https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_March_18,_2015#Minutes</a><br>
    <br>
      -- Mark<br>
    ________________________________________<br>
    <br>
    <div id="globalWrapper">
      <div id="column-content">
        <div id="content" class="mw-body" role="main"><span dir="auto">GlueX
            Offline Meeting, March 18, 2015</span><span
            class="mw-headline" id="Minutes">, Minutes</span><br>
          <div id="bodyContent" class="mw-body-content">
            <div id="mw-content-text" dir="ltr" class="mw-content-ltr"
              lang="en">
              <br>
              Present:
              <br>
              <ul>
                <li> <b>CMU</b>: Curtis Meyer</li>
                <li> <b>JLab</b>: Alex Barnes, Mark Ito (chair), David
                  Lawrence, Paul Mattione, Kei Moriya, Eric Pooser, Mike
                  Staib, Justin Stevens</li>
                <li> <b>NU</b>: Sean Dobbs</li>
              </ul>
              <br>
              <span class="mw-headline" id="Announcements">Announcements</span><br>
              <ul>
                <li> At JLab, the quota on /group/halld was raised from
                  300 GB to 400 GB.</li>
                <li> Paul Mattione has checked in changes to the
                  Analysis Library to add support for histograms and
                  custom ROOT branches. See <a rel="nofollow"
                    class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2015-March/001986.html">his
                    email</a> for details.</li>
                <li> Mark packaged a new release of sim-recon, version
                  1.0.0. See <a rel="nofollow" class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2015-March/001988.html">his
                    email</a> for details. We discussed the new version
                  numbering scheme. Not everyone loves it, but there
                  were no strong objections.</li>
              </ul>
              <br>
              <span class="mw-headline"
                id="Review_of_Minutes_from_March_4">Review of Minutes
                from March 4</span><br>
              <br>
              We looked at the <a
href="https://halldweb.jlab.org/wiki/index.php/GlueX_Offline_Meeting,_March_4,_2015#Minutes"
                title="GlueX Offline Meeting, March 4, 2015">minutes</a>.
              <br>
              <br>
              <span class="mw-headline" id="Intel_Code_Validation_Tools">Intel
                Code Validation Tools</span><br>
              <br>
              We briefly discussed access to the Intel code validation
              tools discussed last time. At present they are not
              available for general use at JLab, though that may change.
              Mike Staib has been using an educational version, Paul
              Mattione is working on a 30-day free trial.
              <br>
              <br>
              <span class="mw-headline"
                id="Commissioning-Branch-to-Trunk_Migration">Commissioning-Branch-to-Trunk
                Migration</span><br>
              <br>
              The copy of the commissioning branch of sim-recon on to
              the trunk was <a rel="nofollow" class="external text"
href="https://mailman.jlab.org/pipermail/halld-offline/2015-March/001984.html">done
                Friday, March 6</a>. Mark reported that he and Simon had
              some problems doing a best-practices-type merge of the
              branch to the trunk. It was unclear how to resolve all of
              the resulting conflicts. Instead the trunk was deleted and
              the branch copied to the trunk location. History on the
              branch was preserved, including history from before the
              branch. So far, no problems with this procedure have been
              reported.
              <br>
              <br>
              <span class="mw-headline" id="Offline_Monitoring_Report">Offline
                Monitoring Report</span><br>
              <br>
              Kei reported that there have been a number of issues
              delaying the launch.
              <br>
              <ul>
                <li> We upgraded to a new version of ROOT to fix memory
                  leaks in the FFT package that we never call.</li>
                <li> Paul and Nathan Sparks worked on a bug fix that
                  caused crashes at the exit of hd_root jobs when
                  certain combinations of plug-ins are run. They traced
                  it to some global STL containers in one of the static
                  libraries.</li>
                <li> The current set of code is crashing for reasons to
                  be investigated.</li>
              </ul>
              <br>
              <span class="mw-headline"
                id="Summary_of_meeting_with_SciComp">Summary of meeting
                with SciComp</span><br>
              <br>
              Last week SciComp called a meeting to discuss new features
              with us. In attendance were Mark, David, Kei, Justin, and
              Paul.
              <br>
              <ol>
                <li> Chris Larrieu gave a presentation/demo of <a
                    href="https://halldweb.jlab.org/wiki/images/e/e7/Swif.pdf"
                    class="internal" title="Swif.pdf">SWIF</a>. We plan
                  to adopt the farm handling scripts we have been using
                  for data challenges and offline monitoring to take
                  advantage of this system, though work in that
                  direction has not started yet.</li>
                <li> Ying Chen showed slide on the new <a
                    href="https://halldweb.jlab.org/wiki/images/8/84/Halld-WDM.pptx"
                    class="internal" title="Halld-WDM.pptx">Write-Through
                    Cache</a>. Files written to the this disk will be
                  archived to tape after two weeks. When space gets
                  short, files that have been archived will be deleted
                  to make space for new files. There is a minimum size
                  requirement of 3 MB for a file to be archived. 20 TB
                  of space on this device is now available to us.</li>
              </ol>
              <br>
              <span class="mw-headline"
                id="GlueX_Use_of_C.2B.2B11_features_and_GCC_Versions">GlueX
                Use of C++11 features and GCC Versions</span><br>
              <br>
              Dmitry had privately discussed with Mark the idea of using
              C++11 features in our offline code. We have not explicitly
              endorsed or prohibited such use, but Mark thought we
              should discuss it. He showed a <a
                href="https://halldweb.jlab.org/wiki/index.php/GCC_at_JLab,_March_2015"
                title="GCC at JLab, March 2015">table of GCC compilers</a>
              used on systems around the Lab. If possible, it looks for
              now that we do not want to use any language features that
              are not available in the 4.4 series compiler. Dmitry was
              particularly interested in using [<a rel="nofollow"
                class="external free"
href="http://en.wikipedia.org/wiki/Smart_pointer#shared_ptr_and_weak_ptr%7Cshared">http://en.wikipedia.org/wiki/Smart_pointer#shared_ptr_and_weak_ptr%7Cshared</a>
              pointers. He claims that they are available in 4.4.
              <br>
              <br>
              The <a rel="nofollow" class="external text"
                href="https://gcc.gnu.org/projects/cxx0x.html">GCC page
                on these features</a> raises some flags, notably
              <br>
              <br>
              1. "experimental<i>" in GCC, i. e., no guarantee that all
                behavior of these features preserved in future versions.</i><br>
              2. possibility of the need for special compiler flags to
              access these features.
              <br>
              Concern was raised by these items. We did not have an
              expert on hand.
              <br>
              <br>
              One approach would be to simply wait until all of our
              commonly used systems catch up to, say, version 4.8 or
              greater. The only problem with that is that it might be a
              long time coming and meanwhile we lose the benefit of the
              new features which are standard for a lot of developers
              outside of GlueX.
              <br>
              <br>
              David suggested that Dmitry try some of the features he is
              interested and see if behavior is different when using a
              4.4 compiler versus a 4.8 or 4.9 compiler. That would help
              us decide how to proceed.
              <br>
            </div>
            <div class="printfooter">
              <br>
              Retrieved from "<a dir="ltr"
href="https://halldweb.jlab.org/wiki/index.php?title=GlueX_Offline_Meeting,_March_18,_2015&amp;oldid=64291">https://halldweb.jlab.org/wiki/index.php?title=GlueX_Offline_Meeting,_March_18,_2015&amp;oldid=64291</a>"<br>
              <br>
            </div>
          </div>
        </div>
      </div>
    </div>
  </body>
</html>