backlog: plan GUI buttons — wiring fixes (020a), Browse/Save/Load (020b), Generate-scenario (020c) #31
Loading…
Reference in a new issue
No description provided.
Delete branch "docs/gui-buttons-plan"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
What
Work-order plan (docs-only) for the four GUI gaps found while inspecting the
GUI: three new backlog files + overview index rows.
Key finding (P0): the "missing generate button" is really a broken
generate path.
app.pyreadsself._power_sat_var/_power_db_varin_on_model_change/_on_set_power, but neither attribute is evercreated (verified by AST scan). Toggling any signal checkbox raises
AttributeErrorinside the Tk callback, the selection never lands in themodel,
can_generatestays False — the Generate button stays disabledforever. Additionally, the Settings-tab entries, the Channel-tab Doppler/
delay spans and the fmt/placeholder comboboxes have no trace/command
wiring (
_on_setting,_on_spanare dead code) — only thesignal-selection tab is wired.
The three work orders
power panel the step-013 code assumed (fixes the crash), wire every
Settings entry, the span entries and the comboboxes to the model.
extension), Save scenario… / Load scenario… over the existing
Scenario.to_json/from_jsonsidecar format; load maps the scenario backthrough
update_manyand rebuilds the selection checkboxes via theexisting
sat_spec_for_systemhelper; buttons lock while a job runs.build → validate →
draw_channels), then write the sidecar-only JSON(resolved seed + drawn channels included) — hand the result to
gengnss scenario FILE -o out.i16for byte-identical generation. CLIbyte-parity is an acceptance test.
All three are model/test-driven per repo conventions (no display needed for
the core tests; gui-marked smokes for the dialogs).
Docs-only change — no code touched.