Personal tools
  • We're Hiring!

You are here: Home Community Minutes Conference Calls 2016 2016-11-22 Tuesday Team Meeting

2016-11-22 Tuesday Team Meeting

Dundee: Kenny, Roger, Petr, Dominik, Will, Michael, Tom, June, Simon, Gus, Helen, Mark, David, Jean-Marie, Simone, Chris, Roger

Remote: Sebastien, Wilma, Eleanor, Andreas, Josh, Melissa,

Agenda - 2:00pm Start

  1. Accepting minutes from last meeting

  2. Project Timelines (2-3 minutes each)

    1. Spaces

      1. Files/Formats/OME Files (Sebastien)

        • B-F: cut 5.3.0 m2 release.
        • Architecture changes and JPEG-XR work included
        • Some follow-ups (documentation, testing)
        • Discussing add this to OMERO 5.3 m6
        • OME files: next milestone discussion
        • C++ 11 as a requirement - cf Roger’s blog post
        • MULTIMOT as a driver of the new features (see below)
      2. Metadata (Josh)

        • MULTIMOT
          • 12th, 13th dec multimot devs to visit Dundee
          • Prep for January meeting
          • Similiar to IDR - tabular data with schema
          • Record paths / tracks of cell invasion
          • Will try to provide API for them to implement
          • “Frictionless data” http://frictionlessdata.io/data-packages/
          • Json data
        • IDR paper preparation
          • Outages - E.g. multiple searches
          • 2 more Datasets in preparation for import
      3. OMERO 5.3.0

        • Potential m6 for JPEG-XR
        • Insight to enter maintenance mode soon
        • Dom to move to import / web
        • Josh - can close all Insight trac tickets!
        • Web - discussion about thumbnail caching and using React.js for centre panel
        • Design issue on plate names - any feedback, please add
        • Chris - need to consult community. E.g. blog post etc. Big impact
        • Backport to 5.2 OMERO+ and get feedback from Glencoe customers
      4. Outreach

        • Balaji in Edinburgh on Thursday
        • Petr: good trip to Germany. Everyone asks “How do we get installed/started?”
        • Good workflow & setup. Larger group with smaller time slot but went well.
        • Simon: workshop in London
    2. Other releases/upgrades:

      1. Figure - Balaji re-starting work
      2. ImageJ
      3. Learning
      4. Sysadmin
      5. Website redevelopment/redesign (Liza)
        • Working on design this week - typography, images etc.
        • Helen: working with Gus on 5.3 features, icons
        • Another review soon
      6. Smuggler (Montpellier)
    3. Glencoe Update (Chris)

      • CellProfiler completion
      • Ohsu & harvard scalability improvements, Server & client
      • Thumbnailing caching and invalidation - OMERO.web plugin
      • Elasticsearch to improve OMERO search. JCB data-viewer data.
      • Then port to OMERO 5.2 (end of this week) and test with IDR
      • No work to parallelise searching yet
      • Will replace Lucene searching, using same API initially
      • Evaluate with both searches running alongside, test with real data
  3. Michael Porter presenting on his Map Annotation app and the issues he's encountered

    • New project 6 months ago
    • Assay development - lots of variables to track
    • Antibodies, media, culture duration, scientists
    • Excel spreadsheet not easy to link to data…
    • Using map annotations
    • Projects - different media, Datasets Well #, Images with different Antibodies
    • Matlab tool: dictionary of keys & values. Case sensitivity etc important
    • Can browse P/D/I and check existing Key-value pairs.
    • Adding to “All images in Dataset” is tricky to undo
    • Searching: case sensitive for keys but not for values.
    • Whitespace is a problem
    • Chris: not tried to index key-value pairs in new search work
    • New search should be able to provide existing keys & values to use for auto-complete
    • Michael: useful to use colon in search to target map-anns
    • Chris: but other bugs with E.g. colons in keys/values, handling other special chars etc.
    • Currently no ‘types’ - everything is just a string. Can’t do ‘more than’ etc.
    • Michael: working across different user’s data is tricky
    • Ola’s mapr might help, but don’t know yet
    • Chris: infrastructure can’t support ‘context’ E.g. find image in Dataset with ann
    • Petr/Will: what about using tags? Easier to add & query. Organise by tag-set
    • Gus: convert map-annotations to tags?
    • Chris: yes we’ve done this. Tags don’t work so well with many different values
    • Roger: use hql? Chris: works OK but not with larger image counts
Document Actions