Dundee: Helen, Kenny, Will, Mark, June, Dom, Petr, Simon, Roger, Riad, Emil, Balaji,
Remote: Seb, Jean-Marie, Josh, Wilma, David, Mark, Andreas, Chris, Melissa
Start: 2:00 pm
(2-3 minutes each)
(pabot). New library (0.4)
workflow</u>](https://trello.com/c/GJiuntTK/92-workflow-vi-facility-managerlight-admin) for next training (end of January)
Kubernetes presentation. Simon, 6th Feb.
Assume knowledge of Docker & Docker-compose
Set aside time to consolidate Docker & other training
ROI/HDF5 (PRs & Grants)
Update on 5.4.2 timeline
Targeting release for Tuesday
Any final changes to get in?
Last PRs on Thursday
Public user testing
Enable public user temporarily on web-dev-merge
Need a way to test other server configs
Read-only work - to mainline OR IDR branch
Likely 5.4.3 (maybe 5.4.4) by Users’ meeting.
Roger - HDF data from ROIs in sample files.
Changing process wrt Bio-Formats in particular
Adding some formats in the near future
Try same approach with OMERO too.
Features related to scripts and scripts UI
More new hires involved
Run scripts in parallel, show progress/feedback to user etc.
Use as much real test data as possible
Microservices- rendering regions. Improving scalability
IDR Next Release (Seb)
0.4.5 planned for February - 2 new Datasets.
Create new production server and start import soon
Need to update/fix all notebooks
Mid-term future - moving to OMERO 5.4.x
Need to improve tracking of data submission requests.
Looking at redmine as an option to manage these
(5 min. max; tech. Discussion should be highlighted to relevant people and rescheduled)
Breaking jobs & devspaces
Ci cleanup. Remove breaking jobs (been disabled for a while now)
devspace for testing
Mark: will anything test DB upgrade paths?
Seb: no - this will still need to be a manual setup
JM: ported Training testing to devspace - except Matlab
Kenny: taking DB dumps etc from old nightshade2 and rsync to squig.
June: Should be able to confirm user meeting dates next week
(20-25 minutes plus 15 minutes questions max)
Recent discussions on team priorities…
Current state (of the art):
OME-TIFF - tiff files with ome.xml file for metadata etc
BigDataViewer (BDV) - HDF5 file with bdv.xml
BDV is efficient for pixel storage, multi-dimensional etc
But, BDV is not versioned, no upgrade path, limited metadata support
Combine best features from OME-TIFF and BDV.
Target Users’ meeting 2018
Containers: TIFF, H5, KLB
Reading BDV v1 filesets
Reading writing OME-BDV v2 filesets
Clients (ImageJ/BDV stack, OMERO?)
Implementation languages - Java, C++?
OME Data Model / spec
Add BDV xml elements
OME Pixels elements - deprecate vs upgrade
Subresolutions handling [design
Binary data access: plates -> blocks
OME-XML code generation (Metadatastore API)
What about Bio-Formats?
Moving towards B-F 5.8.0
New format readers (Glencoe software)
Releases as needed
Open-source team - no planned work, maintenance/support
Mark: What about ROI storage?
Seb: masks stored as for images