2023-04-25

Attending: Will, Frances, Jean-Marie, Andeas, David, Jason, Maadhav, Khaled, Dominik, Melissa, Kyle, Seb, Susanne Kunis, Dom, Petr, Chris, Erin,

Start: 2:00 pm UK

Intros: Maadhav Kothari - Glencoe, formerly at Zeiss Susanne - Osnabruk, I3D:bio

Accepting minutes from last meeting

Project Status

(2-3 minutes each)

• IDR (Frances):
  • testing data for next submissions, working with authors
  • David looking at series ordering in ZarrReader
  • converting idr0033 idr0036 and others to NGFF
  • Khaled: will present search later (see below)
  • GCDR submission in preparation - see https://drive.google.com/drive/u/0/folders/1bf4xAR7IK4RZCCMRowzWpylDIdlT_V8s
• OMERO (JM):
  • migrating repos to remove python 3.7 support
  • replace custom portalocker with dependency - failing on Windows
  • Andreas: reviewing old PRs
  • Will: will need to upgrade node.js on merge-ci at some point, needed for some js upgrades (e.g. omero-parade)
  • Seb: Kevin’s histogram PR on big images, needs to support old data
  • JM: release process should allow anybody to release as needed
• Bio-Formats (David)
  • ZarrReader needs few days work to support Fileset swapping
  • 6.13 milestone nearly done. Release on Thursday. Lots of dependency updates
• NGFF
  • ngff-histo paper “last call”
  • Will: new script to create symlinks after in-place import to support import without chunks.
• GerBI
  • JM: watched presentation on OMERO at TIM2023 - was success - likely to be adopted elsewhere
• Glencoe (Chris)
  • Maadhav started - see above
  • Preparing for update to Django 4.2
  • Changes are compatible with current Django 3.2
  • ABRF meeting - will have booth
• Community (Petr)
  • EastBio workshop yesterday - good feedback
  • NEUBIAS - 8th May Porto
  • Ongoing planning Japan & US trips
  • Richard Acton (Developer from HDBR) - preparing workshop setup with Renku
  • Jason: UK Digital Reasearch Infrastructure meetings

AOB

(5 min. max; tech. Discussion should be highlighted to relevant people and rescheduled)

Main Topic

(20-25 minutes plus 15 minutes questions max)

Khaled search-engine access permissions:

• Elasticsearch indexes key/value pairs
• Currently no authentication - suitable for public data eg. IDR
• Public user Access
  • configure public user before indexing
  • other data is private
• Auth App
  • JSON Web Token (JWT) used for auth
  • Token is digitally signed
  • JSON payload specifies user_id, user_groups etc.
  • Auth App logs into OMERO with user/password
  • Creates the JWT with user/groups info
• SearchEngine with Auth App
  • Uses the JWT created by Auth App
  • Client gets JWT from Auth App and sends it with query to Search Engine
• Implementation
  • prototype of Auth app with Flask framework

Discussions:

Chris: with microservices we integrated with OMERO.web to allow single login etc. Implementation with JWT is hard to maintain security.