Debug live displays by mpound · Pull Request #23 · astroumd/slama

mpound · 2026-07-22T14:38:43Z

This is an initial pass at fixing up the JSON files to display monitor points on web displays.

Replaces DSM placeholders with actual redis variables.
For some points, Claude made some initial guesses, some of which were wrong, some were completely invented.
There are more files to do, so I will keep this branch open for when I work on other displays. But there is enough now that it should be merged.
Also minor changes to some python scripts related to investigating the monsys and tests thereof

…cedence With --merge alone, smax.json values win on conflicts (safe default that preserves validated descriptions, units, and thresholds). With --merge --overwrite, the freshly-queried Valkey values win instead, by swapping the deep_merge(base, overlay) argument order. Useful when the live database has updated type or size information. 2 new tests in TestDeepMerge covering overwrite-mode semantics. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Lines like " smax.json canonical names : 4723" from valkeynotsmax.txt contain a colon but are not valid SMAX paths. filter_input now validates the first colon-separated segment against _VALID_SEGMENT regex ([A-Za-z0-9_.][A-Za-z0-9_.\-]*), rejecting any segment with spaces or other non-SMAX characters. 2 new tests: test_skips_summary_stat_lines_with_spaces, test_keeps_paths_with_dots_in_segment. Total: 248 tests passing. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

…licts Generated via smax_from_valkey.py --merge --overwrite, then repaired: - Removed 9 spurious keys (summary stat lines + junk namespaces _test_, pytest_smax, scripts, summitlab leaked from valkeynotsmax.txt) - Removed unfetchable antenna:..:c1dc:1:alarm entry (smax_type=?) - Restored DSM roach2-* __each__ (valkey-wins had replaced it with acc1-9); acc1-acc9 added as explicit singleton nodes with 11 variables each - Fixed antenna:...:if and correlator:swarm:segment __each__+explicit-1 duplicate conflicts by merging explicit entries into __each__ templates - Added DSM singletons from Valkey: colossus, corcon, hal9000, hcn, hcn2, m5, newdds, obscon, obscon1, obsconmk, oldhal, phasemon, tenzing, roach2-19 and acc1-acc9 computers - Updated test_monitorsystem.py: relax RM count to >=8*19 (RM grew from 19 to 642 vars/acc); fix DSM roach2 filter to roach2- prefix (+roach2-19) 248 tests passing. Also includes cleaned intermediate files smax_updated.json and smax_updated-valkeywins.json for reference. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Cross-references all 146 dsm:placeholder:DSM_* occurrences across 19 display JSON files against the refreshed conf/smax.json, proposing a real canonical name for each with a HIGH/MEDIUM/LOW/NO MATCH confidence tag. 74 high-confidence, 33 medium, 12 low, 27 no match. Review only — no display JSON files are modified. Marc reviews and confirms/corrects mappings before they're applied. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Applies the 66 HIGH-confidence mappings from goals/dsm_placeholder_review.md to 14 display config files, replacing dsm:placeholder:DSM_* stand-ins with real smax.json canonical names (UPS, meteorology, GPS8, project info, phasemon vault voltages, YIG/MRG lock+tune, roach2 temps). {ant}/{roach} template variables are preserved. All replaced names verified to resolve via MonitorSystem.get_monitor_point(). 72 MEDIUM/LOW/NO MATCH placeholders are left untouched pending further review. 248 tests passing. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Marc confirmed all 74 HIGH-confidence mappings against the live SMAX database at the observatory. Checked off in the review doc; the 72 MEDIUM/LOW/NO MATCH entries remain open for further review. Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

Updated confidence mapping notes in dsm_placeholder_review.md.

…monitor sanity bounds Catalogs the WACKO_* bound constants and printInvalidFloat/printInvalidDouble call sites in cursesmonitor/src, mapping each to a proposed err_low/err_high pair for the corresponding RM_*/DSM:*/antenna:* field in smax.json. Also documents wacko checks that aren't static hardcoded bounds (computed thresholds, NaN/sign sanity, string corruption) and a schema mismatch found between the H and L receiver polarization nodes while tracing the SIS-bias fields. Proposals only; no changes to smax.json in this pass. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

Two live_redis dumps 5 minutes apart (conf/dump1.out, conf/dump2.out) show the earlier H/L asymmetry was a smax.json authoring problem, not a real receiver hardware difference: L does have lo_plate:pll:* live, the pll fields are actually named BPF/NF (not bandpass_power/notch_power), and H's SIS-bias data is real but lives under sis_bias:mixer:{n}:* rather than the shape currently in smax.json. Also surfaces two live-database irregularities (lo-plate/lo_plate duplication on antenna 8, sis_bias/mixer nesting-order duplication on antennas 1 and 7) that are instrument-side, not smax.json issues. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

…reads Most "DSM ID (monitor C code)" entries in dsm_placeholder_review.csv/md were never actually verified against the cursesmonitor source -- they were placeholder names invented from display labels. Re-derived all 19 file-groups against literal dsmRead/call_dsm_read/dsm_structure_get_element/rmRead call sites, with an exact file:line citation on every row. This surfaced several real mapping errors along the way, not just missing citations: - ~15 entries wrongly marked NO MATCH actually have live fields (RM_POLAR_DUT_SEC_D, DSM_AS_FILE_NAME_C80, coherence fringe amplitude/coherence fields, per-antenna DIPFIT tau fields, SWARM roach temp-age/DDS-delay/loading-factor, etc.) - BDC 8-12/10-12 GHz "ATTEN" cells were pointing at the VGA field instead - The entire yig/mrg group used a fabricated reference:mrg:1:* key namespace that doesn't exist in the live database at all; corrected to the real, live DSM:m5:MRG_CONTROL_X structure that mRG.c actually reads - That fabricated namespace was already live-wired into mrg.json and misc_dsm.json, so those two display files are fixed here too (previously dead cells) Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

Replaces dsm:placeholder:DSM_* strings across weather.json, misc_dsm.json, gps.json, coherence.json, project.json, arraymonitor.json, bdc.json, aCmonitor.json, seeing.json, and swarm.json with the real canonical names derived in the previous cursesmonitor C-source audit. Also fixed two pre-existing bugs found while doing this: project.json's "Script PID" cell pointed at a malformed name that doesn't exist live (DSM_AS_PROJECT_SCRIPT_FILENAME_C256_PID instead of the real DSM_AS_PROJECT_SCRIPT_PID_L), and re-applied the mrg.json/misc_dsm.json MRG/YIG/C1DC fixes from the prior commit that had been reverted mid-session. Several of the "real" fields turned out to be multi-element smax.json arrays (2 to 242 elements: mRG/YIG _V2_* fields, BDC _V8_* fields, coherence fringe _V11_V11_V2_* fields) rather than scalars. DataBridge.fetch_cell() has no per-index addressing, so these render as a stringified array rather than a single number -- wired in anyway per Marc's call (real data over a dead placeholder), documented in dsm_placeholder_review.md as a known limitation pending a template/schema extension for array indexing. Remaining placeholders are only cells with no live field at all (confirmed NO MATCH) or croom_iflo.json's IF/LO fields, whose real canonical name needs polarization+mixer indices the {ant}-only template can't express. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

teuben · 2026-07-24T12:28:17Z

clarification: after this merge, will you keep on working here, and merge that back into main with another PR from this branch?

teuben · 2026-07-24T12:37:12Z

The file goals/dsm_placeholder_review.csv has not been written by a person. Should there not be some README reminding reader which files are edited (with or without claude help) and which are generated?

mpound · 2026-07-27T18:03:58Z

clarification: after this merge, will you keep on working here, and merge that back into main with another PR from this branch?

I would merge, continue on this branch, and then do another PR. I suppose I could also create a debug-live-displays-2 if that means less confusion.

teuben

all looks good now. I'd recommend we start a new branch to follow up for next work. I'll also do a special simulation branch.

mpound · 2026-08-05T15:03:55Z

I will delete this branch after merge. I have a new branch for the vectorized display work.

mpound and others added 30 commits June 29, 2026 13:45

HOST not SERVER

b981014

check environment for SMAX_HOST/PORT

b48b939

fix swapped data source points

833d098

add --verbose

9ead54a

debug print

8c14cd5

Merge branch 'main' into debug_live_displays

f137bb1

list of monitor points in SMA database and not in smax.json

8e6c8d7

script to examine a json hierarchy

a445e6f

replace char type with string

c8d1bfc

fix path for json file

8e5de87

Revise mapping notes and remove outdated instructions

0e7074f

Updated confidence mapping notes in dsm_placeholder_review.md.

spreadsheet for keeping track of DSM->SMAX mapping

a40f67f

excel version

6ad6e06

nah

2ed323b

remove debug text

045eb7f

use RACK values for smoke

17165be

fix up DSM variables to match monitor program

2568bc8

add multisite weather

dc28332

points

eb0895a

%

1d5acca

mpound added 13 commits July 16, 2026 16:02

add more points and make sort order same as monitor code

f3a8d75

comma

027d97f

change formatting

9750053

remove invented variables, clean up labels

5c0e711

update formats, use wind server

fab0276

typo

392d4b8

int formats

dcef81e

int formats

d0119a8

fix up formats and taus

97b9000

mark Local

3803ffb

add a bunch of MPs

81e1a37

use the monitor points in BDC_xx.c (BDC Status monitor screen)

c3b2a3e

fix ESTOP mp

f73a7b8

mpound requested a review from teuben July 22, 2026 14:45

Merge branch 'main' into debug_live_displays

1cbca2f

teuben requested changes Jul 24, 2026

View reviewed changes

Comment thread slama/scripts/print_json_hierarchy.py

Comment thread slama/scripts/printmonsys.py

Comment thread slama/scripts/print_json_hierarchy.py

Comment thread slama/scripts/print_json_hierarchy.py

mpound added 2 commits August 4, 2026 15:14

add readme, remove csv

674512d

changes from PR review

ae59a43

mpound requested a review from teuben August 4, 2026 22:16

teuben added 3 commits August 5, 2026 08:35

add routines for completeness

617995a

add debug flag

696872b

add a debug for poiting

aa07591

teuben approved these changes Aug 5, 2026

View reviewed changes

mpound merged commit b16c725 into main Aug 5, 2026
6 checks passed

mpound deleted the debug_live_displays branch August 5, 2026 15:04

Navigation Menu

Uh oh!

Debug live displays - #23

Debug live displays#23
mpound merged 49 commits into
mainfrom
debug_live_displays

mpound commented Jul 22, 2026 •

edited

Loading

Uh oh!

teuben commented Jul 24, 2026

Uh oh!

teuben commented Jul 24, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

mpound commented Jul 27, 2026

Uh oh!

teuben left a comment

Uh oh!

Uh oh!

mpound commented Aug 5, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

Uh oh!

Conversation

mpound commented Jul 22, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

teuben commented Jul 24, 2026

Uh oh!

teuben commented Jul 24, 2026

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

mpound commented Jul 27, 2026

Uh oh!

teuben left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

mpound commented Aug 5, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

2 participants

mpound commented Jul 22, 2026 •

edited

Loading