Skip to content

Flywheel Core 20.1

20.1.0

DEPRECATIONS

Support Extension for BrainBrowser and EEG viewers

  • The integrated EEG and BrainBrowser viewers will be removed in the 20.3.0 release
  • The removal was originally planned for 20.2.0.
  • The removal of BrainBrowser is only temporary, until it can be updated for compatibility and reintroduced no later than Flywheel version 20.5.0.
  • The latest details of this change can be found here.

FEATURES

Bulk Import

Local Import via CLI
  • Data can be imported directly from a local machine using the new (BETA) CLI.
    • Simply specify a local directory path with the --storage option instead of a storage ID.
      • E.g., fw-beta import run ... --storage ~/data/2025-02-01
    • If interrupted during the upload step, local imports can be resumed using the new --resume-local <import-job-id> option.
      • Once the upload step is complete, local imports continue processing on the server-side just like any other import job.
    • System requirements are extremely minimal, because very little processing occurs locally.
      • The CLI first uploads the data off of the local computer and into a temporary server-side storage, then starts a standard servers-side import from that temporary storage location.
      • This way, all of the heavy processing (mapping, grouping, zipping, etc.) is performed entirely server-side and can be retried without requiring a re-upload.
    • De-identification is NOT (yet) available for local imports.
    • For more details, refer to the new (BETA) CLI documentation.
Other Import Enhancements
  • Bulk Import is accessible from the global left navigation pane before selecting a group or project.
  • The Bulk Import jobs list can be filtered by a combination of multiple statuses instead of just one status at a time.

Bulk Export

  • The Bulk Export jobs list page has been redesigned for improved ease of use.
    • The Export job listing is now paginated instead of having an infinite-scroll behavior.
    • The "created" and "modified" timestamps can now be used for filtering data for Bulk Export.
    • The Bulk Export jobs list can be filtered by a combination of multiple statuses instead of just one status at a time.
  • Auto-suggestion and auto-completion are available when specifying which metadata field to use in a container filter rule for Bulk Export.

Data Views

Secure CSV and TSV files

  • Starting with Flywheel version 20.1.0, CSV and TSV files generated by Flywheel Core services will provide greater data integrity and security via content escaping, to prevent Microsoft Excel from interpreting the content as formulas.
  • Values that start with a =, -, +, or @ will have a \ prefixed to it.
  • Escaping only applies to string (text) values. (eg. negative numbers are not effected)
  • More detail regarding this change is available here

Info

You may need to update how you process impacted files if your processing is sensitive to this new escaping.

FIXES

Bulk Import & Export

  • Fixed an issue with Bulk Import and Bulk Export where the audit report for a cancelled job would contain multiple entries for the same source file with slightly different information.

Bulk Import

  • Fixed an issue with Bulk Import where users needed the admin Project Role to upload data. The read-write Project Role should be sufficient.
  • Fixed an issue with the "Resolve Conflicts" page for Bulk Import where the currently-selected conflict item could not be resolved without selecting the checkbox on its row first. This step was unnecessary since the conflict item was already selected (the checkboxes are only necessary when selecting multiple conflict items to resolve en masse).
  • Improved the description text used in the "Add data" dialog for Bulk Import to remove ambiguity and provide clear confirmation whether de-ID will be performed or not.
  • Fixed an issue with the Local Upload dialog for Bulk Import where selecting any area outside of the upload dialog would cause the upload to be cancelled and require the user to start the upload over from the beginning. Now, a confirmation window will be displayed in this case, and the dialog will not close without confirmation.

Bulk Export

  • Fixed an issue with Bulk Export where the list of export jobs could not be retrieved for locked projects.

Collections

  • Improved the usability of the Collections Management Web UI -- making the tab naming and their content consistent with Project Management.
  • Resolved issue where text type collection attachments could not be created or updated with the standard web text viewer.

Core Web App

  • Fixed an issue where the Flywheel application would become unresponsive if left open for a long period of time.

Data Management

  • Resolved an issue where Move Session to Project feature could incorrectly identify a conflicting subject label when a conflict did not exist, due to the value of subject code.

Gears and Compute

  • Fixed an issue where Gear Compute was incorrectly attempting to delete network interfaces on Azure.
  • Resolves an issue when uploading large (>5GB) files to an AWS-hosted site for customers using the self-service HPC Client.
  • Resolves an issue with gear upload failing for dicoms with particular dicom tags.
  • The sidebar in the Search Web UI has received updates, necessary for its ongoing support and future improvements.
  • Resolved an issue where the Session Timestamp filter would automatically populate the range dates.

Smart Copy

  • Resolves issue where Smart Copy filtering settings disappear while they are being typed.

20.1.1

FIXES

Bulk Export

  • Fixed an issue with Bulk Export where some jobs would be flagged as "failed" even though all files were exported successfully.
  • Fixed an issue with the Bulk Export jobs list page where selecting an "Error" chip would open the new export job dialog.

De-Identification

  • Fixed an issue with de-identification when uploading data via a web browser where the remove-private-tags: true option (remove all private DICOM tags) was not applied.
  • Fixed an issue with de-identification when uploading data via a web browser where the replace-with: '' option (replace field with empty string) was not applied.

20.1.2

FIXES

Gears and Compute

  • Resolved an issue with compute build inadvertently missed in patch.

20.1.3

FIXES

Viewers

  • Resolved an issue where .obj files could not be viewed with the integrated BrainBrowser viewer.

20.1.4

FIXES

Bulk Import & Bulk Export

  • Fixed an issue with Bulk Import where long-running uploads from a local computer via the Web App could fail if the user is otherwise inactive in Flywheel for at least 1 hour while the upload is still in progress.
  • Fixed an issue with Bulk Import where some large, long-running import jobs could fail due to running out of memory.
  • Fixed an issue with Bulk Export and Bulk Import where jobs could become stuck in the "cancelling" state in rare cases where the cancellation request is received when the job is in the "finalization" step.

CLI (Legacy)

  • Fixed an issue with fw ingest in the legacy CLI where large files (roughly 5 GB or larger) could not be uploaded to Flywheel.

De-Identification

  • Fixed an issue with de-identification where de-ID profiles referencing DICOM tags using "repeater groups" (e.g., (60xx,3000) or (60xx,4000)) would not pass validation.
  • Fixed an issue with de-identification where de-ID profiles referencing some newer DICOM tags would not pass validation.
  • Fixed an issue with de-identification where de-ID profiles referencing private DICOM tags by PrivateCreator (name) value would not pass validation if their PrivateCreator (name) value contained a comma.

Gears and Compute

  • Resolved an issue causing an intermittent but frequent failure of compute upload for cases where small files are uploaded.

SDK and CLI

  • Resolved an issue where uploading files larger than 5 GiB using the Flywheel CLI or Python SDK (flywheel-sdk) to a Flywheel site hosted in Google Cloud Platform (GCP) can result in an incomplete upload without generating any visible error.
    • Affected Scenarios:
      • Sites hosted on Google Cloud Platform
      • Uploads using the CLI or flywheel-sdk only (data uploaded with the Web UI, Connectors, Viewer Apps, etc. are not affected).
      • Files 5GiB or larger.
    • Affected Versions:
      • flywheel-sdk: Versions 17.5.0 through 20.1.0
      • CLI: Versions 16.19.0 through 20.0.0
    • Identification of impacted files:
      • Impacted files will show up in Flywheel with an incorrect size (smaller than 5 GiB).
    • Resolution and next steps:
      • Update your flywheel-sdk to Version 20.1.4 or later. An upgrade your Flywheel site itself is not necessary.
      • Update your CLI to Version 20.2.0 or later. An upgrade your Flywheel site itself is not necessary.
      • Check if Your Site is Google Cloud Hosted. If you're not sure if your Flywheel site is hosted on GCP, you can use the following flywheel-sdk code to check. 
        1
2
3
4
5
6
        site_url = fw.get_config()['site']['api_url']
site_providers = fw.site_api.get_providers()
if len([p for p in site_providers if p['provider_class'] == flywheel.ProviderClass.STORAGE and p['provider_type'] == flywheel.ProviderType.GC]) > 0:
    print(f"This site ({site_url}) DOES have a Google Cloud Storage Provider")
else:
    print(f"This site ({site_url}) DOES NOT have a Google Cloud Storage Provider")
    • If you believe that you have data impacted by this issue, please reach out to our support team for assistance.