---
tags: NGFF
---

# IDR-NGFF 2020-W42

[](https://hackmd.io/EQxsvoKYQK6OGutb3C79Jg)

## 2020-10-16

### Clients
- Josh: additions to the Trevor notes (side problem of a chat solution)
- Will: next client-side thing
- assuming open with is working
- assuming Z/T working
- back to napari? need data, so omero-cli-zarr for one of the specs?
- looked at grid layout. Lays out channels in a grid. Could possibly use the methods.
- best solution may be a pyramid with a plate at the top-level
- or open a new window?
- is the goal sliders for the whole plate?
### Microservices
- nothing new on micro-services
### Datasets
- Dom: start converting one plate from idr0033 using omero-cli-zarr + option 3
- Seb: proposing to convert https://idr.openmicroscopy.org/webclient/?show=plate-5966 ( 41744_illum_corrected of idr0033)
- Josh: registered volumes?

### Format
- Tischi opening a Java version of [n5+cloud issue](https://github.com/zarr-developers/zarr-python/issues/540)
- see https://github.com/BIOP/bigdataviewer-bioformats
- other metadata specs
- voxel size as a priority
- segmentation table
- main issue: impossibility to talk to Zarr AND S3 at the same time.
- Ongoing discussion with Saalfeld et al.
- If working, could be a candidate library
- Could use/test Zarr reader with BIOP aka BDV + BF
- Also possible MOBIE (more complex project)
- David: started writing some code for reader
- test various libraries. Seb: service?

### Summary of next steps
-

## 2020-10-12

Attending: Will, Seb, Dom, J-M, Josh, Frances, Petr, David, Simon

### Clients

- Will: was fixing open with plugin to handle callback to see if something is enabled
- make things work in right-hand-panel
- Josh: blocker on vizarr side?
- Unclear what the status of vizarr/imjoy
- imjoy: have been managing to do ImageJ in browser + pulling data from Zenodo
- main question is: have they handled Bio-Formats? Apparently not yet. Planning to add a few plugins
- Trevor implemented browser-viewing via Zarr, Imjoy
- Also can have a talk with Trevor re JS server
- Simon: propose him to demo at a Tuesday meeting?
- Josh to invite him in for a meeting in the next 3 weeks

### Datasets

- idr0033 original segmentations
- Seb: as soon as prod90 is up, will add the segmentations.

- idr0033 notebooks
- J-M: stripped a lot from existing pipeline. Combines analysis + QC. Removed modules looked for non-existing data
- Generating CSV files rather than saving into the DB
- Running other pipeline to see what's happening

### Format

- HCS specification - https://github.com/ome/omero-ms-zarr/issues/73
- Good reviews
- Possible issues with many items per group
- See new option-3 combining all the suggestions
- 5 layers of folders
- With largest current screens, 500 elements per folder would be max
- Current OME metadata fields: https://github.com/ome/omero-ms-zarr/issues/73#issuecomment-705627752
- J-M: will need HCS-specific client visualization able to handle grid distributed images
- Josh: short term, assuming new JS and/or napari
- J-M: how to link redundant metadata for HCS e.g. rendering settings
- Simon: envisaging "3D" plates? Splitting wells into rows/columns imposes 2D structure
- Josh: C. Tischer et al interested in other layouts. N. Kiaru talking about using SVG to layout things at arbitrary location (see [NGFF-EMBL-HD-2020-10-02.md](./NGFF-EMBL-HD-2020-10-02.md))
- If we need hexagonal plate, that's a separate route
- Seb: assuming we will pick a first layout by end of the week and start hammering it

### Microservices

- https://github.com/manics/docker-example-omero-microservices works!
- problem in configuration file: one port is integer, the other one is a string
- Chris will look at cleaning the errors
- next steps?
- generate thumbnails from Zarr?
- J-M: assuming thumbnails are not stored in Zarr. Reading binary and generating on-the-fly