What’s new for Polarion by Nextedy – June 2026

Some months are about new capabilities. Others are about making sure nothing breaks when the platform underneath you changes. June was mostly the second kind — and that work is worth more to a validated environment than a feature list.

All Products — Ready for Polarion 2606

Polarion 2606 moves to Tomcat 11, which means javax.servlet becomes jakarta.servlet. For add-ons, that is not a cosmetic change — servlet-based extensions simply stop loading, and Polarion’s own updater refuses to proceed while an incompatible one is present.

We migrated all of them. Risksheet, Powersheet, Gantt, Planningboard and Checklist shipped 2606-compatible builds on 15 June, about two weeks after the platform release, with Nextedy AI following on 19 June.

Two build variants now come out of a single codebase, and both ship inside the same release archive:

  • polarion-2512-and-older/ — the Tomcat 9 / javax build

  • polarion-2606-and-newer/ — the Tomcat 11 / jakarta build

Deploy the one folder that matches your server, never both. Same features and same configuration either way. If you are staying on Polarion 2512 or older, you can still take the newer product versions — you just use the javax variant.

One sequencing detail is worth knowing before you upgrade. The 2606 updater validates extensions for Jakarta compatibility and blocks the update when it finds an incompatible one, so /extensions has to be clear of javax extensions at the moment you run it: either replace each one with its jakarta build, or remove them and reinstall afterwards. This applies to every custom extension on the server, not only ours. Removing them loses nothing — /extensions holds code only, while your data and configuration live in the repository and the database. Siemens covers the platform side in their note on updating custom extensions for Polarion 2606.

Powersheet

A Powersheet requirements traceability matrix in Polarion: user needs with their rationale, linked use steps and validation plan, edited inline in a spreadsheet-style grid.

An Administrator’s Documentation Expert

Configuring a traceability sheet means knowing which property does what. Historically that meant searching documentation, reading someone else’s YAML, or filing a support ticket.

The Admin AI Assistant now sits inside Powersheet administration. You ask a configuration question in plain language and get an answer grounded in the official Nextedy documentation, with citations linking back to learn.nextedy.com so you can verify it yourself.

It handles the questions administrators actually ask: how do I configure this behaviour, which YAML property controls it, show me a working snippet, what does this existing configuration do, why is it behaving unexpectedly, what is the difference between these two concepts.

Two things matter here more than the capability itself:

  • Your configuration stays yours by default. Only the question text leaves your server. Attaching your data model and sheet configuration is an explicit action you take when you want the answer grounded in your actual setup — not something that happens silently.

  • No AI keys, no hard dependency. You do not bring your own AI API key, and Powersheet continues to work normally if the documentation search endpoint is unreachable.

The Assistant ships in the separately-installed, separately-licensed Nextedy AI package. If it is not installed, Powersheet behaves exactly as before.

Cells That Lock Themselves

Powersheet’s isReadOnly option used to be a static true or false. It now accepts an expression evaluated against the row.

That means a cell can lock based on the row’s own state — freezing a description once the row reaches an approved status, for example — without maintaining a separate view for each stage of the process.

One deliberate limitation: a dynamic lock can only add read-only. It never relaxes an existing security or metadata permission. Governance stays where it belongs.

Errors That Tell You What Went Wrong

Three Powersheet changes in June share a theme: the system used to fail quietly, and now it explains itself.

  • Saving a link that violates Polarion’s link role rules used to be accepted, producing invalid data. It now returns an error listing the work item types that are actually allowed.

  • Saving a custom field that does not exist on the row’s work item type used to fail silently. The error now names the field and the type.

  • Default values now work for entities backed by multiple work item types with sparse custom fields — each type uses its own defaults instead of leaking across types. (date-time and time fields do not support defaults yet.)

Silent failure in a regulated environment is worse than a clear error. This is us paying down that debt.

Gantt

Come Back to Where You Left Off

Anyone who works with a large nested Gantt knows the routine after a refresh: re-expand four levels, scroll back to the right quarter, find the item you were editing.

Gantt now restores that context. Expand and collapse state is remembered per user, per chart, and survives reloads. Vertical scroll, the visible date range, and the selected item are all restored after a refresh for the rest of your session.

It sounds small. It is the difference between resuming work and rebuilding your place in it.

A Resource View That Shows What Is Left

Resource bubbles have always shown allocated hours. Useful for reporting, less useful for deciding what to do next.

A configuration property now switches those bubbles to show either remaining available capacity or overallocated hours instead. Same green/red logic, but the number you are looking at is the one you act on.

Related: items deliberately excluded from resource load with task.setLoad(0) no longer inflate the item count above the bubble. The total now reflects only what actually contributes to load.

Small Fixes That Matter

June carried a substantial regression and fix batch, most of it driven directly by customer reports.

Gantt

  • Saving a large number of changed items no longer fails on Tomcat’s 2 MB POST parameter limit

  • PlannedIn saves work for cross-project plans (Project Span)

  • Compare with Date no longer treats every item as new

  • Resources are sorted by name again, not by ID

  • Resource capacity displays correctly in cross-team view when friendly Team IDs are used

  • Milestone marker colours set in the Markers Script are applied again

  • Unscheduled children align with their parent by default again

  • nextedy.Gantt.multiAssignmentSameEffortForAll no longer behaves in reverse

  • addDays is now available in Gantt script utilities for inclusive end-date handling

Powersheet

  • Sheets with several nested expand columns sharing a parent navigation now populate all of them, not just the first

Risksheet

  • The drive dropdown shows “Risksheet” instead of a literal placeholder

What’s Next

July is a heavier month, and we will cover it properly in the next post: formula columns that calculate even when hidden in Risksheet, subqueries and URL-parametrised queries in Powersheet, and first steps towards planning against roles and departments in Gantt.

Until next month — stay efficient, connected, and let’s keep pushing Polarion forward together. 🚀