gea-program 0.6.0__tar.gz → 0.8.0__tar.gz

{gea_program-0.6.0/gea_program.egg-info → gea_program-0.8.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: gea-program
-Version: 0.6.0
+Version: 0.8.0
 Summary: GEA-Program: downhole gauge monitoring - canonical measurement records, quality rules, drift evaluation with a re-fit change log, well-test validation, alarm management, model cards, store-and-forward, monthly SLA measurement, FAT/SAT protocols and a dashboard, generated from live objects in the client's reporting outline.
 Author: Daniel T. Murphy
 License: MPL-2.0
@@ -110,7 +110,10 @@ reaches a client document. Nothing unmeasured is ever reported as met.
   leg lives on the dashboard: a Seismic page where a station or an array is
   added by upload with its rigs list and station file, refreshed as a job,
   and read - spectrum, lines, verdicts, bearing - with its Seismic Station
-  Report under Reports. See *The second leg*.
+  Report under Reports; and the SAR panel on that page, a control panel and a
+  screen that plays the leg's arithmetic forward in time on the labelled
+  synthetic scene (`seismic_film.py`), every frame stamped SIMULATION_SELF_TEST.
+  See *The second leg*.
 
 ## Quick start
 
@@ -122,8 +125,8 @@ gea dashboard --catalog-well volve_f12_f14_production_excerpt:15/9-F-12:10000 --
 gea client-report --report accuracy --out client_report
 gea model-cards --out model_cards
 gea sbom --out sbom
-gea accept                     # the product gate (258 checks)
-gea help drift                 # the help library, by the job (15 pages; the same text is on every dashboard page)
+gea accept                     # the product gate (302 checks)
+gea help drift                 # the help library, by the job (16 pages; the same text is on every dashboard page)
 gea guide                      # the click-by-click tester guide (docs/TESTER_GUIDE.md)
 gea gui                        # the desktop window (pip install "gea-program[desktop]")
 gea workspace --path C:\site --action init --name "Pad 3"     # a site folder
@@ -149,12 +152,12 @@ site session is reproduced offline and how the mapping logic is tested.
 ```
 pip install "gea-program[live]"                       # asyncua, paho-mqtt, pymodbus, pyserial (or one extra at a time)
 gea wits0  --write-example-config wits0.json          # drill floor: transport tcp | listen | serial; item codes -> tags, units
-gea wits0-sim --port 5001                             # a WITS0 sender to rehearse with (no rig needed)
-gea wits0  --config wits0.json --seconds 60 --record floor/session.jsonl --out floor --stream-csv floor/stream.csv
+gea wits0-sim --port 5001                             # a WITS0 sender to rehearse with (no rig needed) - run it in a SECOND window and leave it running
+gea wits0  --config wits0.json --seconds 60 --record floor/session.jsonl --out floor --stream-csv floor/stream.csv   # in the first window, while the sender runs
 gea witsml --write-example-config witsml.json         # a WITSML 1.4.1 store: url, uids, mnemonics -> tags; credentials by env name
-gea witsml --config witsml.json --seconds 60 --out store
+gea witsml --config witsml.json --seconds 60 --out store                 # needs a real store: the example url is a placeholder until you edit it
 gea opcua  --write-example-config opcua.json          # edit: endpoint, security, credential env names, nodes
-gea opcua  --config opcua.json --seconds 60 --record opc/session.jsonl --out opc --stream-csv opc/stream.csv
+gea opcua  --config opcua.json --seconds 60 --record opc/session.jsonl --out opc --stream-csv opc/stream.csv   # needs a real OPC UA server, likewise
 gea mqtt   --write-example-config mqtt.json           # edit: broker, TLS, credential env names, topics
 gea mqtt   --config mqtt.json --replay mq/session.jsonl --out mq_replay     # no broker needed
 gea ingest --file floor/stream.csv                    # a stream CSV feeds the historian port like any other
@@ -251,10 +254,10 @@ text: JSON, JSON lines, CSV, Markdown, HTML.
 
 ### The help library
 
-`gea help` lists fifteen pages indexed by the job a reader arrives with -
+`gea help` lists sixteen pages indexed by the job a reader arrives with -
 start, bring data in, quality rules, drift, well tests, alarms and the month,
 the site, instruments, shut-ins and build-ups, seismic (the second leg),
-patches, files, notifications, which code is running, upkeep. Each page carries four lines and stops: the
+patches, files, notifications, which code is running, audit / update, upkeep. Each page carries four lines and stops: the
 command, what it writes, the one number to check, and what the page will not
 call a measurement. The pages ship inside the package (`gea/help/`), and the
 same text is what `/api/help` serves and what the dashboard shows under every
@@ -394,6 +397,155 @@ a rig not on the list, and any quantity in physical units until a station's
 response has been applied and named. `--action selftest` runs the test on a
 synthetic scene and labels its output SIMULATION_SELF_TEST.
 
+The SAR panel at the foot of the Seismic page is the control panel and the
+screen. "Run the scene" is `gea workspace --action sar-film` as a job: the
+labelled synthetic array scene, one frame per window, written under
+`reports/seismic/SIMULATION/`. The screen plays it - the trace, the
+spectrogram column arriving, the beam power map on the slowness grid swinging
+onto the working rig, the map with the bearing and the array's tolerance
+wedge against the rigs where the scene put them, the tally per rig, and at
+the end the array detectability test's verdict on the whole scene, the same
+function a real record goes through. Every frame carries SIMULATION_SELF_TEST
+and the line "not a measurement of any ground"; a real record is refreshed on
+its station and the page shows what it earned, never a film. `gea seismic
+--action sar-film --out film.json` writes the same film from the command line.
+
+### Tracks - the time dimension
+
+The array step gives one bearing for one window. `seismic_track.py` runs it
+window after window: a bearing history per array (a bearing with the array's
+own tolerance where the beam was coherent, a gap where it was not, change
+points where the bearing moved beyond the tolerance), a position history from
+two or more arrays crossed window by window (a position only where two or
+more arrays were coherent, the bearings crossed at 15 degrees or more and the
+point lies in front of every array - every other window a gap with its
+reason), and the track: the principal line through the longest continuous
+segment of positions, with its heading, length and rate, the ellipses beside
+every point, and segments listed apart where positions jump farther than
+their ellipses allow (a coherent background crossing in a quiet hour is not
+the rig an hour later). The verdict against ground truth - the lateral's
+surveyed points or a permit's surface hole with its date, interpolated to
+each position's time - is TRACKED, PARTIAL, NOT_TRACKED or INSUFFICIENT, with
+the fraction and the heading difference printed. The labelled scene behind
+`--action track-selftest` is a bit advancing along a straight lateral past
+two arrays. On a site a track is two or more array stations and an optional
+truth CSV, refreshed as a job into the Seismic Track Report; the track view
+shows the map with every position's ellipse, a slider through time and the
+bearings over time, and the SAR panel's second scene plays the same thing
+forward with the track drawn as it is earned.
+
+Ground truth from a permit export: `gea permits --file export.csv --out
+rigs.csv --within LAT LON KM` finds the columns of a regulator's query (an
+identifier, a surface position, a spud or approval date) by name, with a
+mapping file when the guesses are wrong, and writes the rigs CSV the tests
+read beside an import note that says which columns were used, how many rows
+were dropped and why, and how many end dates it assumed (a permit rarely says
+when drilling stopped). The Seismic page's add form takes a permit export in
+place of the rigs CSV and keeps both.
+
+### Several rigs at once
+
+One bearing is one direction for every machine in the band, so when two rigs
+work together the crossing of whole-band bearings means nothing. That is the
+wall between "a rig" and "a field", and `seismic_signature.py` is the step
+through it. While a rig works alone, the bins that stand above the local
+floor in its windows and not in the quiet ones are kept as its signature -
+the same lines the detectability test already prints. In the windows where
+several rigs work together, the cross-spectral matrix is sliced to one rig's
+lines and beamed, which gives a bearing per rig in the same window; a
+machinery line arriving as one plane wave is coherent in its own bin even
+when the band as a whole is noise, which is why it works.
+
+What it refuses is as much of the method as what it claims. A rig that never
+worked alone has no signature and gets no bearing. Two rigs whose lines fall
+within the spectral resolution of each other are NOT_SEPARABLE, both of them,
+rather than handed a bearing each. A rig whose own lines all sit above the
+array's spatial Nyquist is ALIASED: the measured beam has another peak within
+a tenth of the height of the one found, the geometry cannot say which is the
+source, and the peaks it cannot separate are listed instead of a direction.
+Run through two arrays, every rig that both pointed at in the same window
+gets its own positions, ellipses and track, and a track whose extent is
+inside its own ellipses says `NOT_RESOLVED` instead of reporting the heading
+of a scatter as a direction of travel. `gea seismic --action signatures |
+multi-beam | multi-bearings | multi-track | signature-selftest |
+field-selftest`; on a site, an array station with a rigs list learns the
+signatures at every refresh and a track carries a track per rig beside the
+whole-band one.
+
+### The machine behind the lines
+
+A signature as a list of frequencies is a list of facts about bins. A pump at
+1.7 strokes per second is one fact about a machine, and it puts energy at
+1.7, 3.4, 5.1 and 8.5 Hz. `seismic_harmonic.py` searches every line divided
+by every order for the comb that explains the most of them, normalises it to
+the largest spacing the orders found allow - a comb of every second tooth is
+that comb at twice the spacing, and the larger one claims less - and reports
+the fundamental as a rate per minute. That is a far stronger claim than the
+same lines listed separately, and the module says how strong: `by chance` is
+how often this record's line density puts that many lines on a comb by
+accident, and a family above the ceiling is set aside rather than reported.
+Fewer than three lines is not a family, because any two lines define a comb;
+two spacings that explain the same lines equally well are AMBIGUOUS; and a
+fundamental whose half would fall below the analysed band is flagged, because
+the machine may be running at half that rate with only its even harmonics in
+view.
+
+The other half is time. A pump's rate follows the work: a line at 1.40 Hz
+walks to 1.85 Hz over a tour, no single bin stands above its floor in enough
+windows, and the rig all but disappears from its own signature - which is the
+defect this band exists to fix. Every line is followed window to window by
+nearest neighbour within a drift ceiling, each peak refined inside its bin by
+a parabola so a walk smaller than the bin width can be read, and each track
+carries STEADY, DRIFTING or INTERMITTENT with its drift in Hz per hour. A
+drift no larger than the bin width is not called a drift. The walk is also an
+observable in its own right: `rate_history` reads the rate over the record,
+and a rate that moves is the machine's load changing, not a different
+machine.
+
+Both halves meet in attribution. A track belongs to a source when it passes
+within tolerance of a line that source was learned on - or when it stands in
+a small whole-number ratio, in a window they share, to a track that source
+already claims, which is how a harmonic the fixed bins missed comes back to
+its machine wherever it has walked to. A line two sources could both claim is
+given to neither. A line no listed source claims is listed as exactly that:
+something is making it and the program was not told what.
+
+And it closes the loop on ground truth. Every verdict in the leg rested on
+the rigs CSV's working window - the detectability test, the array test and the
+track all believed it. `activity_from_signature` measures it instead: the
+fraction of a source's own lines standing above their floor, window by
+window, each line looked for within a fraction of its own frequency of where
+it was learned, because a harmonic walks as far as its order. The spells it
+finds sit beside the declared ones with AGREES or DIFFERS and the counts both
+ways. A permit date is a permission, not a drilling log, so a disagreement is
+a finding and not an error - and which of the two is right is not decided
+here.
+
+`gea seismic --action harmonics | harmonic-selftest`; on a site every station
+refresh does all of it - a single sensor as well as an array, using its own
+detectability test's lines per source when there is no array to learn from,
+so one geophone can still say when each rig was working - and the Seismic
+Station Report and the station page carry the section, with the tracks drawn
+against time.
+
+### Audit / Update
+
+The Audit / Update tab holds three things. The whole audit log
+(`records/audit.jsonl`) with filters by who, by action and since a date, and
+a CSV download - Administration keeps the last hundred lines and points here.
+The program card: the running version and code path, the newest release on
+PyPI (`current`, `behind`, or `PyPI not reachable` when the machine is
+offline), which extras are installed here, and links to PyPI, the releases
+with the install kits, and the CHANGELOG; "Update the program from PyPI"
+(admin) is `gea update` as a job - pip --upgrade from the same Python, after
+which the service must be restarted; the program never restarts itself, and
+an offline kit is updated by installing the newer kit. The data card: every
+report against its source - `current` or `stale` when the source changed
+after the report was written - and "Update every report from its source"
+(operator), which is `gea workspace --action refresh-all` as a job: the
+dashboard for every well, then every seismic station, errors collected and
+listed, never hidden.
+
 ### Before a site goes live
 
 `deploy/README.md` is the checklist: keep the service on the loopback
@@ -446,7 +598,8 @@ docs/report_samples/  one rendered example of every client report, from this bui
 docs/           TESTER_GUIDE.md, REQUIREMENTS_MATRIX.md (the scope-of-work mirror that shaped the reports), examples/ (port configs),
                 SESSION_LOG.md (the working record, session by session),
                 commercial/ (pilot proposal, bench readiness, commercial use), HISTORY.md (the ship-by-ship record)
-tools/          standalone_check.py (the self-contained guard; run by ci and ship.ps1), build_installer.py (the kit)
+tools/          standalone_check.py (the self-contained guard; run by ci and ship.ps1), build_installer.py (the kit),
+                render_report_samples.py (docs/report_samples), seismic_reader_check.py (the reader against libmseed)
 CHANGELOG.md    per-release summary (a tag ships only with its section); SHIP_LOG.md is written by ship.ps1
 tests/          pytest wrapper around the acceptance suite and the standalone check
 ```

{gea_program-0.6.0 → gea_program-0.8.0}/README.md

@@ -71,7 +71,10 @@ reaches a client document. Nothing unmeasured is ever reported as met.
   leg lives on the dashboard: a Seismic page where a station or an array is
   added by upload with its rigs list and station file, refreshed as a job,
   and read - spectrum, lines, verdicts, bearing - with its Seismic Station
-  Report under Reports. See *The second leg*.
+  Report under Reports; and the SAR panel on that page, a control panel and a
+  screen that plays the leg's arithmetic forward in time on the labelled
+  synthetic scene (`seismic_film.py`), every frame stamped SIMULATION_SELF_TEST.
+  See *The second leg*.
 
 ## Quick start
 
@@ -83,8 +86,8 @@ gea dashboard --catalog-well volve_f12_f14_production_excerpt:15/9-F-12:10000 --
 gea client-report --report accuracy --out client_report
 gea model-cards --out model_cards
 gea sbom --out sbom
-gea accept                     # the product gate (258 checks)
-gea help drift                 # the help library, by the job (15 pages; the same text is on every dashboard page)
+gea accept                     # the product gate (302 checks)
+gea help drift                 # the help library, by the job (16 pages; the same text is on every dashboard page)
 gea guide                      # the click-by-click tester guide (docs/TESTER_GUIDE.md)
 gea gui                        # the desktop window (pip install "gea-program[desktop]")
 gea workspace --path C:\site --action init --name "Pad 3"     # a site folder
@@ -110,12 +113,12 @@ site session is reproduced offline and how the mapping logic is tested.
 ```
 pip install "gea-program[live]"                       # asyncua, paho-mqtt, pymodbus, pyserial (or one extra at a time)
 gea wits0  --write-example-config wits0.json          # drill floor: transport tcp | listen | serial; item codes -> tags, units
-gea wits0-sim --port 5001                             # a WITS0 sender to rehearse with (no rig needed)
-gea wits0  --config wits0.json --seconds 60 --record floor/session.jsonl --out floor --stream-csv floor/stream.csv
+gea wits0-sim --port 5001                             # a WITS0 sender to rehearse with (no rig needed) - run it in a SECOND window and leave it running
+gea wits0  --config wits0.json --seconds 60 --record floor/session.jsonl --out floor --stream-csv floor/stream.csv   # in the first window, while the sender runs
 gea witsml --write-example-config witsml.json         # a WITSML 1.4.1 store: url, uids, mnemonics -> tags; credentials by env name
-gea witsml --config witsml.json --seconds 60 --out store
+gea witsml --config witsml.json --seconds 60 --out store                 # needs a real store: the example url is a placeholder until you edit it
 gea opcua  --write-example-config opcua.json          # edit: endpoint, security, credential env names, nodes
-gea opcua  --config opcua.json --seconds 60 --record opc/session.jsonl --out opc --stream-csv opc/stream.csv
+gea opcua  --config opcua.json --seconds 60 --record opc/session.jsonl --out opc --stream-csv opc/stream.csv   # needs a real OPC UA server, likewise
 gea mqtt   --write-example-config mqtt.json           # edit: broker, TLS, credential env names, topics
 gea mqtt   --config mqtt.json --replay mq/session.jsonl --out mq_replay     # no broker needed
 gea ingest --file floor/stream.csv                    # a stream CSV feeds the historian port like any other
@@ -212,10 +215,10 @@ text: JSON, JSON lines, CSV, Markdown, HTML.
 
 ### The help library
 
-`gea help` lists fifteen pages indexed by the job a reader arrives with -
+`gea help` lists sixteen pages indexed by the job a reader arrives with -
 start, bring data in, quality rules, drift, well tests, alarms and the month,
 the site, instruments, shut-ins and build-ups, seismic (the second leg),
-patches, files, notifications, which code is running, upkeep. Each page carries four lines and stops: the
+patches, files, notifications, which code is running, audit / update, upkeep. Each page carries four lines and stops: the
 command, what it writes, the one number to check, and what the page will not
 call a measurement. The pages ship inside the package (`gea/help/`), and the
 same text is what `/api/help` serves and what the dashboard shows under every
@@ -355,6 +358,155 @@ a rig not on the list, and any quantity in physical units until a station's
 response has been applied and named. `--action selftest` runs the test on a
 synthetic scene and labels its output SIMULATION_SELF_TEST.
 
+The SAR panel at the foot of the Seismic page is the control panel and the
+screen. "Run the scene" is `gea workspace --action sar-film` as a job: the
+labelled synthetic array scene, one frame per window, written under
+`reports/seismic/SIMULATION/`. The screen plays it - the trace, the
+spectrogram column arriving, the beam power map on the slowness grid swinging
+onto the working rig, the map with the bearing and the array's tolerance
+wedge against the rigs where the scene put them, the tally per rig, and at
+the end the array detectability test's verdict on the whole scene, the same
+function a real record goes through. Every frame carries SIMULATION_SELF_TEST
+and the line "not a measurement of any ground"; a real record is refreshed on
+its station and the page shows what it earned, never a film. `gea seismic
+--action sar-film --out film.json` writes the same film from the command line.
+
+### Tracks - the time dimension
+
+The array step gives one bearing for one window. `seismic_track.py` runs it
+window after window: a bearing history per array (a bearing with the array's
+own tolerance where the beam was coherent, a gap where it was not, change
+points where the bearing moved beyond the tolerance), a position history from
+two or more arrays crossed window by window (a position only where two or
+more arrays were coherent, the bearings crossed at 15 degrees or more and the
+point lies in front of every array - every other window a gap with its
+reason), and the track: the principal line through the longest continuous
+segment of positions, with its heading, length and rate, the ellipses beside
+every point, and segments listed apart where positions jump farther than
+their ellipses allow (a coherent background crossing in a quiet hour is not
+the rig an hour later). The verdict against ground truth - the lateral's
+surveyed points or a permit's surface hole with its date, interpolated to
+each position's time - is TRACKED, PARTIAL, NOT_TRACKED or INSUFFICIENT, with
+the fraction and the heading difference printed. The labelled scene behind
+`--action track-selftest` is a bit advancing along a straight lateral past
+two arrays. On a site a track is two or more array stations and an optional
+truth CSV, refreshed as a job into the Seismic Track Report; the track view
+shows the map with every position's ellipse, a slider through time and the
+bearings over time, and the SAR panel's second scene plays the same thing
+forward with the track drawn as it is earned.
+
+Ground truth from a permit export: `gea permits --file export.csv --out
+rigs.csv --within LAT LON KM` finds the columns of a regulator's query (an
+identifier, a surface position, a spud or approval date) by name, with a
+mapping file when the guesses are wrong, and writes the rigs CSV the tests
+read beside an import note that says which columns were used, how many rows
+were dropped and why, and how many end dates it assumed (a permit rarely says
+when drilling stopped). The Seismic page's add form takes a permit export in
+place of the rigs CSV and keeps both.
+
+### Several rigs at once
+
+One bearing is one direction for every machine in the band, so when two rigs
+work together the crossing of whole-band bearings means nothing. That is the
+wall between "a rig" and "a field", and `seismic_signature.py` is the step
+through it. While a rig works alone, the bins that stand above the local
+floor in its windows and not in the quiet ones are kept as its signature -
+the same lines the detectability test already prints. In the windows where
+several rigs work together, the cross-spectral matrix is sliced to one rig's
+lines and beamed, which gives a bearing per rig in the same window; a
+machinery line arriving as one plane wave is coherent in its own bin even
+when the band as a whole is noise, which is why it works.
+
+What it refuses is as much of the method as what it claims. A rig that never
+worked alone has no signature and gets no bearing. Two rigs whose lines fall
+within the spectral resolution of each other are NOT_SEPARABLE, both of them,
+rather than handed a bearing each. A rig whose own lines all sit above the
+array's spatial Nyquist is ALIASED: the measured beam has another peak within
+a tenth of the height of the one found, the geometry cannot say which is the
+source, and the peaks it cannot separate are listed instead of a direction.
+Run through two arrays, every rig that both pointed at in the same window
+gets its own positions, ellipses and track, and a track whose extent is
+inside its own ellipses says `NOT_RESOLVED` instead of reporting the heading
+of a scatter as a direction of travel. `gea seismic --action signatures |
+multi-beam | multi-bearings | multi-track | signature-selftest |
+field-selftest`; on a site, an array station with a rigs list learns the
+signatures at every refresh and a track carries a track per rig beside the
+whole-band one.
+
+### The machine behind the lines
+
+A signature as a list of frequencies is a list of facts about bins. A pump at
+1.7 strokes per second is one fact about a machine, and it puts energy at
+1.7, 3.4, 5.1 and 8.5 Hz. `seismic_harmonic.py` searches every line divided
+by every order for the comb that explains the most of them, normalises it to
+the largest spacing the orders found allow - a comb of every second tooth is
+that comb at twice the spacing, and the larger one claims less - and reports
+the fundamental as a rate per minute. That is a far stronger claim than the
+same lines listed separately, and the module says how strong: `by chance` is
+how often this record's line density puts that many lines on a comb by
+accident, and a family above the ceiling is set aside rather than reported.
+Fewer than three lines is not a family, because any two lines define a comb;
+two spacings that explain the same lines equally well are AMBIGUOUS; and a
+fundamental whose half would fall below the analysed band is flagged, because
+the machine may be running at half that rate with only its even harmonics in
+view.
+
+The other half is time. A pump's rate follows the work: a line at 1.40 Hz
+walks to 1.85 Hz over a tour, no single bin stands above its floor in enough
+windows, and the rig all but disappears from its own signature - which is the
+defect this band exists to fix. Every line is followed window to window by
+nearest neighbour within a drift ceiling, each peak refined inside its bin by
+a parabola so a walk smaller than the bin width can be read, and each track
+carries STEADY, DRIFTING or INTERMITTENT with its drift in Hz per hour. A
+drift no larger than the bin width is not called a drift. The walk is also an
+observable in its own right: `rate_history` reads the rate over the record,
+and a rate that moves is the machine's load changing, not a different
+machine.
+
+Both halves meet in attribution. A track belongs to a source when it passes
+within tolerance of a line that source was learned on - or when it stands in
+a small whole-number ratio, in a window they share, to a track that source
+already claims, which is how a harmonic the fixed bins missed comes back to
+its machine wherever it has walked to. A line two sources could both claim is
+given to neither. A line no listed source claims is listed as exactly that:
+something is making it and the program was not told what.
+
+And it closes the loop on ground truth. Every verdict in the leg rested on
+the rigs CSV's working window - the detectability test, the array test and the
+track all believed it. `activity_from_signature` measures it instead: the
+fraction of a source's own lines standing above their floor, window by
+window, each line looked for within a fraction of its own frequency of where
+it was learned, because a harmonic walks as far as its order. The spells it
+finds sit beside the declared ones with AGREES or DIFFERS and the counts both
+ways. A permit date is a permission, not a drilling log, so a disagreement is
+a finding and not an error - and which of the two is right is not decided
+here.
+
+`gea seismic --action harmonics | harmonic-selftest`; on a site every station
+refresh does all of it - a single sensor as well as an array, using its own
+detectability test's lines per source when there is no array to learn from,
+so one geophone can still say when each rig was working - and the Seismic
+Station Report and the station page carry the section, with the tracks drawn
+against time.
+
+### Audit / Update
+
+The Audit / Update tab holds three things. The whole audit log
+(`records/audit.jsonl`) with filters by who, by action and since a date, and
+a CSV download - Administration keeps the last hundred lines and points here.
+The program card: the running version and code path, the newest release on
+PyPI (`current`, `behind`, or `PyPI not reachable` when the machine is
+offline), which extras are installed here, and links to PyPI, the releases
+with the install kits, and the CHANGELOG; "Update the program from PyPI"
+(admin) is `gea update` as a job - pip --upgrade from the same Python, after
+which the service must be restarted; the program never restarts itself, and
+an offline kit is updated by installing the newer kit. The data card: every
+report against its source - `current` or `stale` when the source changed
+after the report was written - and "Update every report from its source"
+(operator), which is `gea workspace --action refresh-all` as a job: the
+dashboard for every well, then every seismic station, errors collected and
+listed, never hidden.
+
 ### Before a site goes live
 
 `deploy/README.md` is the checklist: keep the service on the loopback
@@ -407,7 +559,8 @@ docs/report_samples/  one rendered example of every client report, from this bui
 docs/           TESTER_GUIDE.md, REQUIREMENTS_MATRIX.md (the scope-of-work mirror that shaped the reports), examples/ (port configs),
                 SESSION_LOG.md (the working record, session by session),
                 commercial/ (pilot proposal, bench readiness, commercial use), HISTORY.md (the ship-by-ship record)
-tools/          standalone_check.py (the self-contained guard; run by ci and ship.ps1), build_installer.py (the kit)
+tools/          standalone_check.py (the self-contained guard; run by ci and ship.ps1), build_installer.py (the kit),
+                render_report_samples.py (docs/report_samples), seismic_reader_check.py (the reader against libmseed)
 CHANGELOG.md    per-release summary (a tag ships only with its section); SHIP_LOG.md is written by ship.ps1
 tests/          pytest wrapper around the acceptance suite and the standalone check
 ```

{gea_program-0.6.0 → gea_program-0.8.0}/gea/__init__.py

@@ -122,7 +122,7 @@ from .profile_catalog import (
     read_pangaea_txt,
 )
 
-__version__ = "0.6.0"
+__version__ = "0.8.0"
 __all__ = [
     "aging_rate", "rate_psi_yr", "accuracy_psi",
     "Sensor", "SimulatorConfig", "DownholeEngine",