What’s new in 0.11

trac 0.11 adds data pipelines to and from external storage, user accounts and platform administration in the application, and a new way to organise objects with business tags.

Data import and export

  • Import data jobs: copy data from an external storage location into trac. Each imported data and file object records where it came from — the storage location and path, and the file’s name, size, modification time and content hash — and the original file is kept.

  • Export data jobs: export data from trac to an external storage location. The files written are recorded in the job’s result.

  • Export models in flows: a flow can end with an export model, so a calculation and the publication of its results run as one job. See Jobs & resources.

These features are not available in Community.

Platform administration

  • Manage the platform: platform admins manage licence provisioning, GitHub credentials and user account provisioning from within the application, including a setup wizard which runs the first time a platform admin signs in.

  • GitHub App: model repositories can be accessed through your organisation’s own GitHub App.

These features are not available in Community.

Organising and finding objects

  • Business tags replace business segments. Tags are grouped into domains defined by each tenant, can be restricted to particular object types, and are carried onto every output of a job. Searches match any or all of the selected tags. See Business tags.

  • Object and version names: each object has a name and description shared by all its versions, and each version can have its own name and description, shown in version pickers and version history.

  • Selecting job inputs: a flow can specify custom criteria to use when selecting each input — for example, datasets with a different key, narrowed by name and business tags.

Job logs and errors

  • Logs for failed jobs: job summary pages now show the job log for failed jobs as well.

  • Clearer error messages across the application.

Model development

New modelling tutorials, covering local setup, file inputs and outputs, structured objects, Polars, optional and dynamic inputs and outputs, runtime metadata, and importing and exporting data. See the modelling tutorials.