WatDis makes large networks easier to browse without losing diagnostic context

When a model gets large, the pain does not always start in the solver. Often it starts earlier: the canvas is heavy to navigate, inspecting a finding interrupts the flow, and it becomes harder to tell whether the slowdown comes from drawing, diagnostics, or another runtime layer. This week, WatDis made meaningful progress on exactly that problem.

Less friction while moving through dense networks

The clearest shipped work was a concentrated canvas pass for large-network browsing. Mainline WatDis picked up lighter pan and zoom preview paths, an overview mode with adaptive detail thresholds, viewport culling, and more atomic viewport updates.

In practical terms, the app is now trying to spend less work while someone is actively moving through the network and save the full redraw for more reasonable moments. That is not the same as a finished production benchmark story, but it is a concrete improvement in how WatDis handles the cost of dense-model exploration.

Diagnostics that help without yanking the workflow around

The second strong improvement was about inspection flow, not just rendering. On large networks, WatDis stopped auto-running QA during import or browsing and kept those checks available manually and before hydraulic runs.

That matters because useful inspection should not accidentally become an expensive operation. In the same spirit, Diagnostics now cleanly separates two actions that used to blur together: Zoom is for navigation to a finding, while Open Props is for inspection without moving the viewport. View-only actions also stopped entering undo history, which removes avoidable overhead when the model is already large.

Better signals for validation, not just intuition

Another valuable part of the week is that the canvas now exposes local draw and visibility stats. That is not a flashy user-facing feature, but it is a good engineering move: follow-up testing can check visible counts and draw timing directly instead of relying only on gut feel when a demanding model such as Barranquilla is re-tested.

That distinction matters. In technical software, it helps to leave a better trail for measurement instead of treating responsiveness as something purely subjective.

Hardening with the right amount of restraint

WatDis also tightened script loading in watdis-next with a more restrictive origin policy. That is real security and operational hardening, with an important caveat: the current runtime still needs unsafe-eval, so the honest read is “tighter control,” not “fully locked down.”

The same restraint matters for the rest of the week. A large solver, QA, and operational-diagnostics planning drop also landed on main, but it would be misleading to market that as confirmed hydraulic-runtime improvement. The stronger public story is still canvas responsiveness and clearer diagnostics flow.

Why it matters

The practical takeaway is simple. WatDis is better positioned to let engineers move through large models with less friction, inspect findings without losing context, and validate canvas behavior with better signals.

This is not a week for claiming a solver miracle. It is a week for making complex models easier to work with and for giving the next round of technical validation a stronger base. In a serious engineering product, that is meaningful progress.