Release 2026.40

Product Enhancement :

Automatic Date Format Resolution —

Datasets with ambiguous date formats now resolve automatically when a project-level date format is set, so users no longer confirm each one. The Project Settings modal now keeps the default date format checked, shows a compact summary with an Edit button for the full options, and sets new workspaces’ format from the user’s timezone.

Draft Mode Remembered per Dataview —

Mammoth now remembers each dataview’s Draft Mode setting. After you paste steps or apply a template, the pipeline briefly switches to Draft Mode, then returns to your chosen setting once submitted. Draft Mode is off by default.

Source Interpreter Reads Messy Spreadsheets Accurately —

Messy spreadsheets now import correctly, including timesheets, block layouts, multi-row headers, totals and footers, and rows with blank cells. A new Import as is button keeps a file exactly as uploaded and skips review on future uploads to that dataset.

Defect Fixes :

Resolved the following issues:

• Label & Insert Values — Editing or duplicating a Current Date or Current Time step on an existing date column and clicking Apply blanked the whole column; the step now saves correctly.
• Label & Insert Values — Choosing Some cells without adding a condition wrote the value to every row, overwriting existing data; Apply now stays disabled until a condition is added.
• Label & Insert Values — Switching a new column to Date after typing a number produced a 1970 date and left the pipeline stuck on Processing; the typed value now clears and a date picker appears.
• Label & Insert Values — Switching a value back to All cells still allowed adding another value, and Apply failed with a generic error; Add another value is now disabled.
• Data Library — Moving an item into a folder from the More options menu did nothing; the item now moves as expected.
• Data Library — Moving an item to the Data Library root left the move stuck on Processing; the move now completes within seconds.
• Data Library — After a subfolder moved to the root, its former parent still showed a folder count; the count now updates correctly.
• Pipeline — Pasting a copied Bulk Replace step into a different dataset’s dataview selected the wrong column; pasted steps now match columns by name.