Release Notes — v1.5¶

Release date: March 2026

This release adds thread flow visualization, execution timeline SVGs, data waterfall charts, and enhanced result reports — delivering rich inline diagrams and improved notebook rendering across all execution modes.

Thread Flow Diagrams¶

`config.flow()` — Processing Pipeline SVG¶

Thread configs now expose a flow() method that returns a FlowDiagram object showing the thread's processing pipeline: sources, joins, transforms, and target.

config = ctx.load("stg_customers.thread")
flow = config.flow()
flow.save("pipeline.svg")   # export to file
flow                         # renders inline in a notebook

Join sources are positioned above their target join node to show the data flow clearly. Primary sources appear on the left with horizontal arrows through the transform chain.

flow() is only available for thread configs — weave and loom configs return None.

Thread flow diagram showing sources, joins,
transforms, and target — Example flow diagram for a `fact_orders` thread with two join sources, filters, derives, and a merge target.

Dark Mode Support¶

All SVG renderers accept a dark keyword argument to force a specific color palette:

config.flow(dark=True)    # force dark palette
result.dag(dark=False)    # force light palette
result.dag()              # auto-detect via prefers-color-scheme (default)

Execution Timeline SVG¶

Execute mode results for weave and loom runs now include a phased execution timeline Gantt chart. The timeline shows lookups, pre-hooks, thread execution groups, and post-hooks as sequential phases with wall-clock duration bars.

Execution timeline Gantt chart showing
phased execution — Timeline for a weave with a lookup materialization, quality gate pre-hook, two parallel execution groups, and a post-hook.

Data Waterfall SVG¶

Per-thread results include a data waterfall chart showing row counts at each processing stage: rows read, after transforms, quarantined, and rows written. Preview mode shows the first two stages only.

Data waterfall chart showing row counts
through processing stages — Waterfall for `fact_orders`: 15,000 rows read, 14,200 after transforms, 85 quarantined, 14,115 written.

Annotated DAG SVG¶

Execute mode results include an annotated DAG with colored status badges (success/failure/skipped) and duration annotations on each thread node.

Annotated DAG with status badges and
duration labels — Post-execution DAG for a loom with six threads across three execution groups, showing success badges and per-thread durations.

Enhanced Execute Mode Reports¶

Execute mode notebook rendering now includes:

Executive summary — status, duration, thread/row counts
Per-thread detail — expandable sections with flow diagrams, data waterfalls, output schema, and data samples
Execution timeline — phased Gantt chart (weave/loom runs)
Annotated DAG — status-badged dependency graph (weave/loom runs)

Enhanced Preview Mode Reports¶

Preview mode notebook rendering now includes per-thread expandable sections with output schema tables and data sample previews with row numbers.

Telemetry Enhancements¶

ThreadResult now carries output_schema and samples fields
ThreadTelemetry includes rows_after_transforms for waterfall charts
resolved_params is available on all telemetry levels (thread, weave, loom) for the outermost execution scope

New Public Exports¶

FlowDiagram is now exported from weevr.engine

Compatibility¶

No breaking changes. All existing configurations and API usage continue to work without modification.

Component	Version
Python	3.11
PySpark	3.5.x
Delta Lake	3.2.x
Microsoft Fabric Runtime	1.3