docs: write down the composition pattern, not just the one case
Some checks failed
CI / backend (push) Failing after 7s
CI / naming (push) Successful in 2s
CI / frontend (push) Successful in 9s
CI / migrations-mysql (push) Failing after 7s

The MECHANISM was already a documented platform contract - ADR-001 defines
partof as composition, makes controls propagate through it, and walks it first
for map-position inheritance. The part-marker work used that rail rather than
inventing one.

What was undocumented is the PATTERN built on it: several devices answering to
one identifier, each becoming its own asset filed under a parent. It existed
only as a collector behaviour for part markers plus a docstring in the device
map, so nothing told anyone how to apply it to another device type, or when not
to.

ASSET-COMPOSITION.md covers when to reach for it and when the shared identifier
is a numbering fault instead, what propagation buys, how to declare a device
type through the map or a per-site setting, what a backup kind must do to
follow the device rather than the parent, how to find the next case with
check-shared-machines, and why the parent is not disposable once devices hang
off it - deactivating it breaks filing, and a hard delete cascades through
backuprevisions.
This commit is contained in:
cproudlock
2026-08-11 12:04:35 -04:00
parent c90ebcbc7c
commit db2b9280e7
3 changed files with 665 additions and 0 deletions

154
docs/ASSET-COMPOSITION.md Normal file
View File

@@ -0,0 +1,154 @@
# Composition: several devices under one parent
When more than one physical device answers to a single identifier, model each
device as its own asset and file it under a parent. This page is the pattern:
when to reach for it, what you get for free, and how to make the collector do
it without writing code.
The mechanism is not new - `partof` and its propagation rails are the ADR-001
platform contract. What is new is the recipe.
## The problem it solves
A machine number is supposed to identify one thing. Sometimes it does not.
At West Jefferson, several Telesis part markers serve one operation number:
0613, 0615 and WJPRT each have more than one. Their configurations differ, most
often by COM port. Treating the operation as the device collapsed them into a
single record, and the damage was quiet:
- Their backups overwrote each other, so the history read as one device
flip-flopping between configurations that were really two devices.
- No question about an individual device could be asked at all - how many there
are, which port one is on, which one failed.
- Every PC driving one contested the single `controls` link to the operation,
so whichever reported last appeared to own it.
None of that announced itself. It surfaced as duplicate backup rows, weeks
later.
## When to use it
Reach for composition when ALL of these hold:
- Several physical devices share one identifier, and that is CORRECT - not a
numbering mistake. Two PCs carrying the same machine number by accident is a
fault to fix on the PC, not a model to build. `flask relationships
check-shared-machines` tells the two apart.
- The devices are individually interesting: they fail, get replaced, carry
their own configuration or calibration.
- Something already identifies each device. One device per PC is the easy case,
because the PC names it.
Do NOT reach for it when the parent is simply a location. A room holding six
printers wants a Location, not a parent asset.
## What you get
Filing a device `partof` a parent buys behaviour already built:
- **Control propagates.** `controls` propagates through `partof` (ADR-001, and
seeded by `flask seed reference-data`). A PC controlling a device therefore
controls its parent by inheritance, so the PC does NOT need - and must not
have - a direct link to the parent. That is what stops several devices
contesting a link only one can hold.
- **Position inherits.** Map-position resolution walks `partof` first, so a
device with no coordinates of its own shows at its parent's position.
- **History separates.** Anything keyed on the asset - backup revisions,
relationships, notes, audit - is per device instead of merged.
## Making the collector do it
A PC that drives a subordinate device declares it in
`plugins/computers/pctypemap.py`:
```python
SUBORDINATE_DEVICE_MAP = {
'gea-shopfloor-partmarker': {
'assettype': 'machine', # core AssetType for the device
'typename': 'Part Marker', # type within that vocabulary
'description': 'Telesis part marker',
'suffix': 'PARTMARKER', # device asset number = <PC>-<suffix>
'partof': True, # file under the reported machine number
'label': 'collector:partmarker',
},
}
```
`partof: True` is the switch that matters. It files the device under the
operation the PC reports AND stops that PC claiming the operation directly.
Leave it False for a device that does not share an identifier - a CMM is a
subordinate device too, but no two CMMs answer to one number, so it links to
its PC and nothing more.
`label` is the relationship origin marker. Only rows carrying it are archived
by a collector push, so links made by hand are never touched. **Do not change
an existing label**: those exact strings are in production databases.
A site adds or retargets an entry without a code change, per ADR-015:
```
Setting: subordinatedevice_<pctype> category: pctypemapping
Value: {"assettype": "machine", "typename": "Marking Laser",
"suffix": "LASER", "partof": true, "label": "collector:partmarker"}
```
A malformed override falls back to the built-in default rather than failing the
collector push - a bad setting must not stop a bay reporting its inventory.
## Making a backup kind follow the device
A backup posted by a PC that drives a device belongs to the DEVICE, not the
parent. `plugins/backups/services/registry.py` resolves this through
`markerforsource`: it finds the PC by the reported `sourcehostname`, follows the
PC's active device link, and files against what it finds. It falls back to the
machine number whenever the device cannot be resolved - no hostname on the
payload, a lean build without the computers plugin, or a PC that has not
reported yet - because filing under the parent beats rejecting a backup.
Two things that matter if you add a kind:
- **`sourcehostname` is load-bearing**, not informational. Without it a backup
files against the parent and merges two devices' histories.
- **A revision chain is (asset, kind, source hostname)**, so two devices on one
parent keep separate chains even before they become separate assets.
## Finding the next one
```
flask relationships check-shared-machines
```
Read-only. Lists every machine number claimed by more than one PC and separates
the two cases by whether child assets exist:
```
0615 11 PCs, 11 child asset(s) - modelled
2026 2 PCs, NO child assets - F31N20R3, F4Z7S7J4
```
The first is composition working. The second is two PCs carrying one number -
fix that on the PC.
## The parent is not disposable
Once devices are filed under it, the parent is doing a real job even though it
may look empty: it is the thing that says those devices belong together.
Deactivating it (`isactive = 0`, which is what "delete" does in the UI) breaks
filing, because the resolver requires an ACTIVE parent - every subsequent
report warns and files nothing.
A hard SQL `DELETE` is worse: `backuprevisions.assetid` is `ON DELETE CASCADE`,
so it destroys history still attached to the parent.
If the parent looks wrong in a list - an operation number appearing among
machines - that is a CLASSIFICATION question, not a deletion one.
## See also
- `docs/adr/ADR-001-asset-as-platform-contract.md` - relationship types,
propagation rails, position inheritance
- `docs/adr/ADR-015-site-specific-configuration.md` - why the device map is
settings-overridable
- `docs/COLLECTOR-INTEGRATION.md` - the collector contract and the part-marker
case end to end

View File

@@ -246,6 +246,11 @@ A PC takes its hostname as its asset number and keeps it.
several PCs on one machine number legitimately, and there the alerts fire on
correct data. Warnings ride in the collector response either way.
The part-marker case below is one instance of a general pattern - several
devices under one parent. `docs/ASSET-COMPOSITION.md` covers when to use it,
what propagation buys you, and how to declare a new device type without a code
change.
A PC reporting `pctype = gea-shopfloor-partmarker` is handled differently,
because several markers can serve one machine number and an operation holds any
number of them. Such a PC gets its own Part Marker machine asset

View File

@@ -0,0 +1,506 @@
<!doctype html>
<html lang="en">
<head>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width, initial-scale=1" />
<title>ShopDB printer supplies API - reference and live dashboard example</title>
<style>
:root {
color-scheme: light dark;
--bg: #ffffff;
--card: #f6f7f9;
--text: #16181d;
--muted: #5b6472;
--border: #d8dce3;
--accent: #0d6efd;
--ok: #04b962;
--low: #ff8800;
--critical: #f5365c;
--code: #eef1f5;
}
@media (prefers-color-scheme: dark) {
:root {
--bg: #0d0d1f;
--card: #16162e;
--text: #e8eaf0;
--muted: #9aa3b5;
--border: #2a2a4a;
--code: #10102a;
}
}
* { box-sizing: border-box; }
body {
margin: 0;
padding: 2rem 1.25rem 4rem;
background: var(--bg);
color: var(--text);
font: 16px/1.6 system-ui, -apple-system, "Segoe UI", Roboto, sans-serif;
}
main { max-width: 62rem; margin: 0 auto; }
h1 { font-size: 1.9rem; margin: 0 0 .3rem; }
h2 { font-size: 1.35rem; margin: 2.5rem 0 .75rem; padding-top: .75rem; border-top: 1px solid var(--border); }
h3 { font-size: 1.05rem; margin: 1.5rem 0 .4rem; }
.lede { color: var(--muted); margin: 0 0 1.5rem; }
code, pre { font-family: ui-monospace, SFMono-Regular, Menlo, Consolas, monospace; }
code { background: var(--code); padding: .12em .38em; border-radius: 4px; font-size: .9em; }
pre {
background: var(--code);
border: 1px solid var(--border);
border-radius: 8px;
padding: 1rem;
overflow-x: auto;
font-size: .86rem;
line-height: 1.5;
}
pre code { background: none; padding: 0; font-size: inherit; }
table { border-collapse: collapse; width: 100%; margin: .75rem 0; font-size: .93rem; }
th, td { text-align: left; padding: .5rem .6rem; border-bottom: 1px solid var(--border); vertical-align: top; }
th { color: var(--muted); font-weight: 600; }
td code { white-space: nowrap; }
.note {
background: var(--card);
border-left: 4px solid var(--accent);
border-radius: 0 6px 6px 0;
padding: .8rem 1rem;
margin: 1rem 0;
}
.note.warn { border-left-color: var(--low); }
.pill { display: inline-block; padding: .1rem .5rem; border-radius: 999px; font-size: .78rem; font-weight: 700; }
.pill.ok { background: var(--ok); color: #04220f; }
.pill.low { background: var(--low); color: #2b1600; }
.pill.critical { background: var(--critical); color: #fff; }
/* ---- live demo ---- */
.demo { background: var(--card); border: 1px solid var(--border); border-radius: 10px; padding: 1rem; }
.demo-controls { display: flex; gap: .5rem; flex-wrap: wrap; align-items: center; margin-bottom: 1rem; }
.demo input {
flex: 1 1 22rem; min-width: 14rem; padding: .45rem .6rem;
border: 1px solid var(--border); border-radius: 6px;
background: var(--bg); color: var(--text); font: inherit; font-size: .9rem;
}
.demo button {
padding: .45rem .9rem; border: 0; border-radius: 6px;
background: var(--accent); color: #fff; font: inherit; font-weight: 600; cursor: pointer;
}
.summary { display: flex; gap: 1.25rem; flex-wrap: wrap; margin-bottom: 1rem; color: var(--muted); font-size: .9rem; }
.summary b { color: var(--text); font-size: 1.3rem; display: block; }
.printer { border-top: 1px solid var(--border); padding: .75rem 0; }
.printer-name { font-weight: 600; }
.printer-meta { color: var(--muted); font-size: .85rem; }
.supply { display: grid; grid-template-columns: 11rem 1fr 4rem; gap: .6rem; align-items: center; margin: .35rem 0; font-size: .88rem; }
.bar { background: var(--border); border-radius: 999px; height: .55rem; overflow: hidden; }
.bar span { display: block; height: 100%; border-radius: 999px; }
.printer-rest { color: var(--muted); font-size: .8rem; margin-top: .35rem; }
.status-msg { color: var(--muted); font-style: italic; }
</style>
</head>
<body>
<main>
<h1>Printer supplies API</h1>
<p class="lede">
Reference for the toner and supply endpoints in ShopDB, and a working
single-page dashboard you can lift straight out of this file.
</p>
<div class="note">
<strong>Where the numbers come from.</strong> ShopDB does not poll printers itself.
Levels are read from Zabbix by IP address, so every endpoint below returns empty
results unless <code>zabbix_enabled</code>, <code>zabbix_url</code> and
<code>zabbix_token</code> are set under Settings &gt; Integrations. That failure is
deliberately soft: you get <code>200</code> with an empty list, never an error, so a
wall dashboard does not blow up when Zabbix is down.
</div>
<h2>Endpoints</h2>
<table>
<thead>
<tr><th>Method and path</th><th>Auth</th><th>What it returns</th></tr>
</thead>
<tbody>
<tr>
<td><code>GET /api/printers/lowsupplies</code></td>
<td>optional</td>
<td>Every printer with at least one supply below threshold, plus a summary. The toner report. Cached 5 minutes.</td>
</tr>
<tr>
<td><code>GET /api/printers/&lt;printerid&gt;/supplies</code></td>
<td>optional</td>
<td>Live levels for one printer, read from Zabbix on request. Not cached.</td>
</tr>
<tr>
<td><code>POST /api/printers/supplies/refresh</code></td>
<td><strong>required</strong> + <code>printers.create</code></td>
<td>Clears the cache so the next read is fresh. Backs the report's Refresh button.</td>
</tr>
<tr>
<td><code>GET /api/printers/supplies/meta</code></td>
<td>optional</td>
<td>Allowed supply types, colors and capacity tiers.</td>
</tr>
<tr>
<td><code>GET /api/printers/models/&lt;modelnumberid&gt;/supplies</code></td>
<td>optional</td>
<td>Cartridge part numbers for a model.</td>
</tr>
<tr>
<td><code>GET /api/printers/lookup?ip=</code> or <code>?fqdn=</code></td>
<td>optional</td>
<td>Resolve an IP or hostname to a printer id.</td>
</tr>
</tbody>
</table>
<p>
"optional" means the endpoint accepts a token but does not need one: a logged-out
page can read it. Only the refresh endpoint requires a token, and the account
behind it needs the <code>printers.create</code> permission.
</p>
<h2>Response shape</h2>
<p>Every ShopDB response is wrapped. Your data is under <code>data</code>:</p>
<pre><code>{
"status": "success",
"data": { ... },
"meta": { "requestid": "6d9d0bb6", "timestamp": "2026-08-11T13:47:41.250953Z" }
}</code></pre>
<h3>GET /api/printers/lowsupplies</h3>
<pre><code>{
"printers": [
{
"printerid": 42,
"printername": "Printer-10-129-22-15",
"assetnumber": "PRN-0042",
"ipaddress": "10.129.22.15",
"vendor": "Xerox",
"model": "AltaLink C8145",
"location": "Building 1 - Cell 4",
"supplies": [
{
"name": "Cyan Toner Cartridge",
"level": 8,
"remaining": 8.0,
"status": "low",
"color": "cyan",
"supplytype": "toner",
"iswaste": false,
"isdrum": false,
"partnumbers": ["006R01737"]
}
]
}
],
"summary": { "total_checked": 37, "low": 4, "critical": 1 }
}</code></pre>
<p>
Only printers with something below threshold appear in <code>printers</code>.
<code>total_checked</code> counts every printer Zabbix answered for, so
"37 checked, 5 listed" means 32 are healthy. A printer is counted once in the
summary, as <code>critical</code> if any of its supplies is critical, otherwise
as <code>low</code>.
</p>
<h3>GET /api/printers/&lt;printerid&gt;/supplies</h3>
<pre><code>{
"ipaddress": "10.129.22.15",
"pingstatus": "1",
"supplies": [ /* same supply objects as above */ ]
}</code></pre>
<p>
<code>pingstatus</code> is Zabbix's reachability value, <code>"-1"</code> when
Zabbix is unconfigured or unreachable. This endpoint hits Zabbix on every call,
so poll it for one printer on a detail view, not for forty on a dashboard - use
<code>lowsupplies</code> for that.
</p>
<h2>The supply object</h2>
<table>
<thead><tr><th>Field</th><th>Meaning</th></tr></thead>
<tbody>
<tr><td><code>name</code></td><td>Raw name from the printer, e.g. <code>"Black Toner Cartridge"</code>.</td></tr>
<tr><td><code>level</code></td><td>Raw value as reported. For most waste cartridges this is percent <em>full</em>.</td></tr>
<tr><td><code>remaining</code></td><td>Normalised percent remaining. <strong>Use this one</strong> for bars and text.</td></tr>
<tr><td><code>status</code></td><td><span class="pill ok">ok</span> <span class="pill low">low</span> <span class="pill critical">critical</span></td></tr>
<tr><td><code>color</code></td><td><code>black</code>, <code>cyan</code>, <code>magenta</code>, <code>yellow</code>, <code>none</code>.</td></tr>
<tr><td><code>supplytype</code></td><td><code>toner</code>, <code>drum</code>, <code>waste</code>, <code>maintenance</code>.</td></tr>
<tr><td><code>iswaste</code>, <code>isdrum</code></td><td>Convenience flags; a waste cartridge is usually worth showing differently.</td></tr>
<tr><td><code>partnumbers</code></td><td>Order codes from the <code>modelsupplies</code> table. Empty until someone fills them in for that model.</td></tr>
</tbody>
</table>
<div class="note warn">
<strong>Do not compute status from <code>level</code> yourself.</strong>
Thresholds are <code>remaining &lt;= 5</code> critical, <code>&lt;= 10</code> low.
The catch is direction: a full waste cartridge is bad, so for waste ShopDB
converts to <code>100 - level</code> - except on Xerox, which already reports
waste as capacity remaining. That vendor rule is why <code>remaining</code> and
<code>status</code> exist. Read them, do not re-derive them.
</div>
<h2>Before you build</h2>
<h3>Base path</h3>
<p>
Instances are served under a subpath, so the API is at
<code>https://tsgwp00525.wjs.geaerospace.net/shopdb/api/...</code> - production on
this server - or <code>/ops/api/...</code> for the dev instance beside it, not at
the domain root. Make the base a variable; do not hardcode <code>/api</code>.
</p>
<h3>Same origin, or CORS</h3>
<p>
The simplest deployment is to drop your HTML file into the instance's web root so
it is served from the same origin - then <code>fetch</code> just works. A page
served from anywhere else is a cross-origin request, and the browser will block it
unless that origin is in the server's <code>CORS_ORIGINS</code> allowlist
(an env var; production refuses to start with a wildcard).
</p>
<h3>Caching and polling</h3>
<p>
<code>lowsupplies</code> is cached for 5 minutes server-side, so polling every
30 seconds gets you the same payload nine times out of ten and buys nothing.
Poll every 2 to 5 minutes. If you need a manual Refresh button, call the refresh
endpoint first - that one needs a token.
</p>
<h2>Minimal example</h2>
<pre><code>const BASE = 'https://tsgwp00525.wjs.geaerospace.net/shopdb'; // no trailing /api
async function lowSupplies() {
const response = await fetch(`${BASE}/api/printers/lowsupplies`);
if (!response.ok) throw new Error(`HTTP ${response.status}`);
const body = await response.json();
return body.data; // { printers, summary }
}
lowSupplies().then(data =&gt; {
console.log(`${data.summary.critical} critical, ${data.summary.low} low`);
for (const printer of data.printers) {
for (const supply of printer.supplies) {
if (supply.status === 'ok') continue;
console.log(`${printer.printername}: ${supply.name} ${supply.remaining}%`);
}
}
});</code></pre>
<h2>Polling, with the failure cases handled</h2>
<pre><code>const BASE = 'https://tsgwp00525.wjs.geaerospace.net/shopdb';
const POLL_MS = 3 * 60 * 1000; // server caches 5 min; faster buys nothing
async function tick() {
try {
const response = await fetch(`${BASE}/api/printers/lowsupplies`, {
headers: { 'Accept': 'application/json' }
});
if (!response.ok) throw new Error(`HTTP ${response.status}`);
const { data } = await response.json();
// Zabbix off or unreachable: 200 with nothing in it. Say so, rather than
// rendering "0 low" as if every printer were healthy.
if (data.summary.total_checked === 0) {
render.unavailable();
} else {
render.board(data);
}
} catch (err) {
// A wall display must survive a blip: keep the last good render.
console.error('supply poll failed', err);
render.stale(err);
} finally {
setTimeout(tick, POLL_MS); // chained, so a slow response cannot pile up
}
}
tick();</code></pre>
<h2>Refreshing on demand (needs a token)</h2>
<p>
Create a Personal Access Token under Settings &gt; API Tokens for an account with
<code>printers.create</code>. Treat it as a credential: a token embedded in a page
anyone can open is readable by anyone who opens it, so use this on an operator
page behind login, not on a lobby display.
</p>
<pre><code>async function refreshAndReload(token) {
await fetch(`${BASE}/api/printers/supplies/refresh`, {
method: 'POST',
headers: { 'Authorization': `Bearer ${token}` }
});
return lowSupplies(); // next read pulls fresh values from Zabbix
}</code></pre>
<h2>One printer, live</h2>
<pre><code>// Skips the cache and asks Zabbix now. Fine for a detail page, wrong for a loop.
async function printerSupplies(printerid) {
const response = await fetch(`${BASE}/api/printers/${printerid}/supplies`);
const { data } = await response.json();
if (data.pingstatus === '-1') return { offline: true, supplies: [] };
return data;
}
// Have an IP but not an id? Resolve it first.
async function printerIdByIp(ip) {
const response = await fetch(`${BASE}/api/printers/lookup?ip=${encodeURIComponent(ip)}`);
const { data } = await response.json();
return data?.printerid ?? null;
}</code></pre>
<h2>Live dashboard</h2>
<p>
The section below is the real thing, running in this page. Point it at an instance
and it will render. Served from a different origin than the API, it will fail on
CORS - that is the browser doing its job, and the fix is to host the file on the
instance itself.
</p>
<div class="demo">
<div class="demo-controls">
<input id="base" type="text" value="https://tsgwp00525.wjs.geaerospace.net/shopdb" aria-label="ShopDB base URL"
placeholder="https://tsgwp00525.wjs.geaerospace.net/shopdb" />
<button id="load">Load</button>
</div>
<div id="out"><p class="status-msg">Enter a base URL and press Load.</p></div>
</div>
<script>
// ---- the whole dashboard, in one place ----
const COLORS = { ok: '#04b962', low: '#ff8800', critical: '#f5365c' };
const el = id => document.getElementById(id);
const escapeHtml = text => String(text ?? '').replace(/[&<>"']/g,
ch => ({ '&': '&amp;', '<': '&lt;', '>': '&gt;', '"': '&quot;', "'": '&#39;' }[ch]));
async function fetchLowSupplies(base) {
const response = await fetch(`${base.replace(/\/$/, '')}/api/printers/lowsupplies`, {
headers: { 'Accept': 'application/json' }
});
if (!response.ok) throw new Error(`HTTP ${response.status} ${response.statusText}`);
const body = await response.json();
return body.data;
}
function supplyRow(supply) {
// remaining, not level: waste cartridges and Xerox invert differently and the
// server has already sorted that out.
const width = Math.max(0, Math.min(100, supply.remaining));
const parts = supply.partnumbers?.length
? ` &middot; ${escapeHtml(supply.partnumbers.join(', '))}` : '';
return `
<div class="supply">
<div>${escapeHtml(supply.name)}${parts}</div>
<div class="bar"><span style="width:${width}%;background:${COLORS[supply.status]}"></span></div>
<div><span class="pill ${supply.status}">${supply.remaining}%</span></div>
</div>`;
}
function printerBlock(printer) {
const meta = [printer.location, printer.model, printer.ipaddress]
.filter(Boolean).map(escapeHtml).join(' &middot; ');
// Only what needs action gets a row. A colour laser reports six supplies, so
// rendering all of them buries the empty cartridge among five healthy ones
// and costs the screen space of another printer.
const failing = printer.supplies.filter(s => s.status !== 'ok');
// The healthy ones are still worth a glance - magenta at 12% is not flagged
// today but is the same trip - so they collapse to one muted line rather
// than disappearing.
const healthy = printer.supplies.filter(s => s.status === 'ok');
const rest = healthy.length
? `<div class="printer-rest">also ok: ${healthy
.map(s => `${escapeHtml(shortName(s))} ${Math.round(s.remaining)}%`)
.join(' &middot; ')}</div>`
: '';
return `
<div class="printer">
<div class="printer-name">${escapeHtml(printer.printername || printer.assetnumber)}</div>
<div class="printer-meta">${meta}</div>
${failing.map(supplyRow).join('')}
${rest}
</div>`;
}
// 'Black Toner Cartridge' -> 'Black'. Enough to tell supplies apart on the
// one-line summary without repeating 'Toner Cartridge' four times.
function shortName(supply) {
if (supply.iswaste) return 'Waste';
if (supply.isdrum) return 'Drum';
if (supply.color && supply.color !== 'none') {
return supply.color.charAt(0).toUpperCase() + supply.color.slice(1);
}
return supply.name;
}
function render(data) {
if (data.summary.total_checked === 0) {
// Zabbix off or unreachable. "0 low" here would be a lie, not good news.
el('out').innerHTML =
'<p class="status-msg">No printers could be checked - Zabbix is not configured or not reachable.</p>';
return;
}
if (!data.printers.length) {
el('out').innerHTML =
`<p class="status-msg">All ${data.summary.total_checked} printers healthy.</p>`;
return;
}
el('out').innerHTML = `
<div class="summary">
<div><b>${data.summary.critical}</b> critical</div>
<div><b>${data.summary.low}</b> low</div>
<div><b>${data.summary.total_checked}</b> checked</div>
</div>
${data.printers.map(printerBlock).join('')}`;
}
async function load() {
el('out').innerHTML = '<p class="status-msg">Loading...</p>';
try {
render(await fetchLowSupplies(el('base').value));
} catch (err) {
el('out').innerHTML =
`<p class="status-msg">Could not read supplies: ${escapeHtml(err.message)}. ` +
'If this page is not served from the ShopDB instance, the browser blocked it on CORS.</p>';
}
}
el('load').addEventListener('click', load);
el('base').addEventListener('keydown', e => { if (e.key === 'Enter') load(); });
</script>
<h2>Notes for a wall display</h2>
<ul>
<li>Show the supplies that need action as rows, and collapse the healthy ones
into a single muted line. The endpoint returns every supply for a listed
printer, so this is the client's call: a colour laser reports six, and
rendering all six buries the empty one. Keep the healthy summary rather
than dropping it - a cartridge at 12% is not flagged today but is the same
walk.</li>
<li>Chain the poll with <code>setTimeout</code> in a <code>finally</code>, not
<code>setInterval</code>: a slow or hung request cannot then stack up behind itself.</li>
<li>Keep the last good render when a poll fails. A display that blanks on one
dropped request is worse than one showing five-minute-old numbers.</li>
<li>Distinguish "nothing is low" from "nothing was checked". They both look like
an empty list, and only one of them is good news.</li>
<li>The board reloads itself if you serve it as a ShopDB display; otherwise add
a daily <code>location.reload()</code> so a deployed fix reaches a screen
nobody touches.</li>
</ul>
</main>
</body>
</html>