Once all of the data shape issues have been addressed, we need to re-generate the NWB files for the VBO release.
Historically there was a LIMS pipeline for generating these files. Adam thinks we just need to run the "no state-change" queue. It ought to just work, since we are not adding new sessions to the release.
[ ] Once the files are generated, hand them off to Marina's team for validation.
Need to ask Marina's team how to handle this: Should we publish to a staging bucket that they pull from during validation (preferred), or silently publish to the actual bucket and only tell outside users after the science team has validated the data.
The safe move is to publish to a staging bucket and make sure the science team is pulling from that bucket when they validate the data. The staging bucket should be specified using the bucket_name field in the input JSON for the data release tool.
Once all of the data shape issues have been addressed, we need to re-generate the NWB files for the VBO release.
Historically there was a LIMS pipeline for generating these files. Adam thinks we just need to run the "no state-change" queue. It ought to just work, since we are not adding new sessions to the release.
Need to ask Marina's team how to handle this: Should we publish to a staging bucket that they pull from during validation (preferred), or silently publish to the actual bucket and only tell outside users after the science team has validated the data.
The safe move is to publish to a staging bucket and make sure the science team is pulling from that bucket when they validate the data. The staging bucket should be specified using the
bucket_namefield in the input JSON for the data release tool.