Attending: Mark, Josh, Seb, Petr, Frances, David, Simon, Andreas, June, (9)
Will, Melissa, J-m, Jason, Kevin, Chris,
Start: 2:00 pm
(2-3 minutes each)
Migration is complete!
Next release: prod84 (idr0071, idr0077, idr0079).
Timeline for S3 storage? End of June
mailman lists with same list name and subscribers. Any preferences as to which lists? Or I’ll probably pick an alias one and a mailman one. Also happy to adjust subscribers according to who does or does not want to be part of the testing. Good to include someone with non-UoD address, could add personal ones.
e.g., NGFF, to see how that goes as a test, or simply some test list that we intend to throw away once done.
My planned default is to create groups where:
Final migration could be done fairly quickly once “go ahead”
Feedback on list names by EOB today - can then start testing
Progress on NGFF - moving to Java
Omero-roi-tool updates yesterday for QuPath guides workflows
OME-TIFF PRs - BioQuant is new ‘commercial partner’ creating
iRODS User Group Meeting (Josh attended)
(Utrecht) - https://utrechtuniversity.github.io/yoda/
links</u>](https://github.com/irods/irods_demo), idrop (uploader)
OME Meeting [Survey
(5 min. max; tech. Discussion should be highlighted to relevant people and rescheduled)
(20-25 minutes plus 15 minutes questions max)
Any feedback from the team?
and prepared. For OMERO 6, permissions become more important. Brings questions about types of operations users would like to do (deleting images…)
oriented question (ROIs, multimodal). See https://docs.google.com/document/d/1eY\_ZajjFLUbylY-KjVqEH3FR-gWqyRImp73HYpMZZIM/edit for preliminary list of buckets
on disk or from object store in current architecture. Easiest place is on micro-service side
thumbnailing micro-service. First candidate.
data access (numpy equivalent in Java-land).
Image-region would quickly follow
perform on-the-fly rendering on small images
way down to 16x16. Seb: Just another resolution
cherry on the cake!
J-M: generation of downsampling will still need to happen
Mark: saving rendering setting on zarr as well?
Currently only the later.
Jason: lots of use cases for significant number of ROIs
dataset (upcoming + historical)
Some work around gallery in addition to specification?
Submitter’s regions? Touches the question of provenance
Some initial discussion about generated results
Jason: also examples generated from training.
output of analysis
from different locations. Effectively decentralized without going through OMERO..
Chris: forces to resolve question of defining the data
sources. In favor of completely separating Image from ROI in NGFF in all cases.
If immutable, makes everything is easier to handle
Josh: brings other problems with it
Decision we will have to make early on. Not made in the
case of OME-TIFF for instance where both representations are possible