One row per stored session, titled by its server -- the host the file is named
after and the one CONFIG comes from -- with the group as subtitle. A forget
repopulates rather than removing one row, because the rows carry their index
into the list and dropping one renumbers the rest.
The chooser runs before ScopeModel exists, so declining costs no buffer and no
socket. UDPSCOPE_PROBE_SOURCES=1 prints what the discovery found and exits,
before gtk_init(), so the list is checkable with no display.
collect_layout() now reports the multicast group and interface the app was
launched with, so a stored session says how to reach its source and not just
which host and port it was.
known_sources() reads the session files the layout store already owns, keeping
only those that name a signal: a development machine accumulates one file per
probe run on a throwaway port, and all 140 of them here hold empty slots.
A session is named <host>_<port>.toml, which cannot say whether that source
was multicast or which interface it joined on, so a chooser row could not
reconnect to one. Both go in the top block: a bare key after a [[table]]
header belongs to that table, which is how the enabled key was lost once.
Seven tasks, each ending in a probe assertion rather than a unit test -- this
project's regression net is the headless probes, not a suite. The spec's
multicast round trip is split across three tasks so nothing in the plan needs a
group this machine can join.
Launched with no --server, udpscope silently connects to 127.0.0.1:8000 and
offers no way to pick a source or see which ones it already knows. The design
is a pre-flight dialog: it runs before ScopeModel exists, so a cancel costs no
buffer and no socket and the existing pipeline is untouched.
The list is built from the stored session files, filtered to those naming at
least one signal -- a development machine accumulates one file per probe run on
a throwaway port, 140 of them here -- and the session format gains multicast and
interface keys so a row can reconnect exactly.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
The flicker ("no data" at no constant frequency, worse with the measurements
and cursors panes up) was a torn read of the window's first stamp: acquire()
pins the window to the ring's oldest slot, and reading it at paint time races
the producer lapping it, returning a stamp one ring span too new. x0() is the
origin every trace point is reconstructed from, so one torn read produced a
refused neighbours(), a filter that dropped three quarters of the points, and
a polyline shifted off its axis. It is cached by acquire() under the lock now.
EventSeries::neighbours() answered from the (x - x0)/dx grid, whose drift from
the real arrival stamps arrives as a constant offset in seconds -- a doubling
share of the screen per zoom step, which is the "zooming into a held event
clips it more from the left" report. Both series binary-search the stamps now.
The measurements pane went through the LOD pyramid rather than reading its
window raw (19.7 ms per recompute to 1.2 ms, 0.1% of a 2 M-sample window) and
recomputes on a cadence; the math engine lost a per-slot heap allocation, a
per-sample source lookup by name and a fetch_add that cost more than the
arithmetic it counted.
The cursors pane's mix mode drew a line per *named* slot, so a slot holding a
signal that can never produce a channel -- the stream's timestamp -- got a row
that could only read "--", and such names were re-saved every session. It
counts bound traces now and apply_layout() refuses them.
Four toolbar icons are drawn for this app and compiled in as a GResource: the
stock ones were a record dot for the trigger and three magnifiers for cursors,
measurements and math.
New probes and dump lines for all of it (trace ink, measure cost, math cost,
event zoom, toolbar icons); README and CLAUDE.md record what they cure.
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Martino Ferrari