Skip to content

Release v0.20.0

Choose a tag to compare

View all tags
@Edwardvaneechoud Edwardvaneechoud released this 24 Sep 16:51
· 15 commits to main since this release
v0.20.0
3835685
This commit was created on GitHub.com and signed with GitHub’s verified signature.
GPG key ID: B5690EEEBB952194
Verified
Learn about vigilant mode.

Changes since v0.19.1.

This release makes dashboards and the catalog easier to use for everyday analysis. Dashboards get KPI tiles for headline numbers, and the SQL editor becomes a proper workspace with tabs. Cloud storage now gets the Delta Lake merge modes and change feed that catalog tables got in 0.19.1. The rest is polish: faster previews and catalog pages, steadier undo/redo, and a batch of fixes. Migration 033 runs automatically at startup.

New features

KPI tiles in dashboards (#765)

You can now add a KPI tile to a dashboard to show one headline number. Pick a column and an aggregation: sum, average, median, min, max, count or distinct count. You can add a prefix or suffix, set the number of decimals and use short numbers like $416.3k. A KPI can also be compared to a target, or to the previous period when a date range filter applies to it. Dashboard filters can now be reordered by dragging.

KPI tiles on a dashboard

Delta Lake on cloud storage (#759)

  • The cloud storage writer adds upsert, update and delete modes with merge keys, and can turn on change tracking.
  • The cloud storage reader can read the change feed since a given version or timestamp.
  • Both are available in the Python API (flowfile_frame).

Catalog SQL editor (#763)

  • You can open several query tabs, and they stay open between sessions.
  • The split between the editor and the results can be resized by dragging.
  • Your queries and results stay put when you leave the SQL tab.
  • You can save a query as a virtual table. A marker shows whether a tab is a draft, saved, or changed since you saved it.

Table export and copy (#768)

You can export tables from the data preview and from notebook output, and copy/paste values straight from a table.

Filter and file format improvements

  • The filter node has a date picker, and date and time values are handled better. (#758)
  • Reads non-UTF-8 gzipped CSVs. Adds setup for IPC streams, with schema detection for Avro and IPC streams. (#756)

Improvements

  • The data preview limits how many columns it shows, so wide tables load faster. (#771)
  • Undo/redo is more stable: rollbacks are reliable and typed descriptions survive an undo. (#767)
  • Catalog pages load faster because they send fewer duplicate requests. (#757)
  • Group By, Pivot and Unpivot have better column selection, clearer output names, and scroll on their own while you drag. (#760)

Fixes

  • The cloud writer no longer crashes when the session token is empty. Cloud storage paths, credentials and Docker-mode access are also tighter. (#766)
  • Dropping a node from the palette onto the action canvas does not add a node. (#769)
  • Node settings stay in sync when you minimize the drawer. (#761)
  • Pressing keys while renaming a notebook tab no longer changes tabs. (#762)
Assets 13
Loading