← Guide

Reading the lost-time ledger

The coordinator's and QS's read surface: the two tiers, the three columns, the still-running chase list, the noticeboard lines, the breakdown views, and the marks a record carries.

The lost-time ledger is where the site's Lost time is read: every Blocker record, folded into figures a coordinator chases from and a QS reports from. It lives on the Commercial surface (I) behind the Lost time tile, and needs the lost-time read grant — held by the coordination band and the QS, deliberately not by every role. The tile is a resolving door: holding the grant lands you on this site-wide ledger, and its header says Site-wide so you always know which scope you are reading. If you're a subcontractor without the grant, the same tile lands you on Your firm's lost time instead.

If the counting rules (crew-hours, the send-home test, chains) are new to you, read How lost time is counted first — this page assumes them.

To pull the week's position:

  1. Open the Commercial surface (I) and pick the Lost time tile.
  2. Set the date range; the ledger states the range and the as-at instant beside its figures.
  3. Read the Still running section first (what's still costing time), then the two tiers and their breakdowns.
  4. Export the CSV, or generate the weekly, from the same surface when the meeting needs paper.

Two tiers, and why they never sum

The ledger renders in two stacked tiers:

  • Confirmed — adjudicated records, each carrying who lost the time and who was named responsible (a firm, or nobody at fault). This is the tier the money and the reports stand on.
  • Awaiting triage — raised but undecided, grouped by who lost the time only (blame doesn't exist until confirm). It shows its own hours total and how long the oldest claim has waited, in whole days.

The two totals are never added and never differenced — they are different kinds of number, and the awaiting tier's age line exists precisely to make triage delay visible rather than letting it hide inside a blended figure. Where several records are one multi-day declaration split into stretches, the ledger folds them to one entry per declaration with a segment count ("1 loss · 2 segments") — the fold never invents a calendar span across gaps.

The meeting reads the Blockers tab, not this one

The coordination meeting does not sit on this ledger. It sits on the Blockers tab of the same Completions drill-in, in meeting mode — the day's two bands (every open raise however old, then the report day's decisions), grouped by firm, each row expanding in place into the triage form. Crew-hours and £ stay here; the meeting table carries the send-home count, the "no count · ⟨firm⟩ to say" ask and the chase state, and no money at all.

Two things the meeting writes that read back on this ledger:

  • The attested count. A coordinator may record another firm's count only with a named representative of that firm present at the sitting. The figure is stored as the firm's own wordstood_down_basis stays declared — with the representative's name, their firm and the meeting's date on the count's timeline event, which the basis bundle renders as "spoken by ⟨rep⟩, ⟨firm⟩, at the ⟨date⟩ meeting". There is no attested-by column and no second basis value: a later own-firm correction is simply a later act.
  • The decision. A confirm at the table moves the row's figures out of the awaiting-triage tier into the confirmed one exactly as a confirm anywhere else does. Nothing about the tiers, the axes or the range pill changes because the decision was taken under a projector.

No meeting is stamped anywhere: "since the last meeting" is the daily report's own window, and that window is the only thing band two is cut on.

The three columns

Each entry carries three time figures, deliberately kept apart:

Column What it is Notes
Stopped hours The record's own window — how long the work was held up. Renders whether or not a crew was counted.
Crew-hours Window × people stood down. A dash — never a zero — where no count exists; the dash always says why.
Open for The running age of a record still open. An age, never a loss: never totalled, never priced. A closed record with no declared restart reads no restart recorded.

The gap between the first two columns is meaningful, not missing data: on a counted record it is the crew factor made visible (8.5 stopped hours over a gang of 8 with 3 stood down = 25.5 crew-hours); on an uncounted one it is the undercount, waiting on the firm's answer.

Still running — the chase list

Above the tiers sits Still running: every record whose work hasn't restarted, whichever tier it's in, oldest first. Each row names the firm, its own verb (still stopped / still disrupted / extent not declared), and — where the crew's working day has visibly ended — a sub-line like "still stopped · crew's day ended Tue 16:30", so you can see a record that outlived its crew's shift. Chasing from here is a recorded act: the row shows "chased 3× · last Thu 16:00" (or "not yet chased"), and deliberately mints no automatic notification — every ping is friction paid against the next honest raise, so the chase stays a human call.

The noticeboard (N) carries the same story in one line each for three audiences: the still-running split ("2 still stopped · 1 still disrupted · oldest 49h"), the triage queue line for coordinators ("3 awaiting triage · oldest 8 days"), and a crew-hours KPI line for lost-time readers. Those three segments are independent lenses, not a sum.

The breakdowns

The confirmed tier folds four ways, each answering one question:

  • By loser — whose work lost the time (the record's own firm).
  • By responsible — who was named responsible at confirm. Two honest residual buckets: (nobody at fault) — a decision — and (no attribution) — older records that predate the question. They are never merged, because one is an answer and the other a silence.
  • By cause — weather, sickness, plant breakdown, waiting on others, a permit, a clash, another firm's stoppage… and cause not given, the bucket a raise lands in when it never opened the phone sheet's (or the desk modal's) details fold. Cause not given is its own bucket and is never folded into Something else: one says no answer was given, the other names a cause the vocabulary lacks. It is retyped like any other cause, at the meeting table or on the record, and the retype is a recorded correction. Every bucket sub-groups by day, whatever the cause: the trigger is its member records carrying dated instances, never a test on the kind. So "Tuesday's weather" is readable without anyone having declared a weather day — and so is Tuesday's permit hold-up. What the kind decides is only which members may be undated: an id-bearing cause whose record supplied no id has no drill target, so it lands in the bucket's dateless residual rather than claiming a day.
  • By place — where on site the loss landed, derived from the record's attached work. A record whose work spans places reads (multiple places) whole, never split — hours are never apportioned across buckets.

Reading it on the map

The By place breakdown has a map face. Turn the choropleth on and every Named place is tinted by the crew-hours the ledger accounts to it, over whichever tier and axis you are reading — the caption beside it always names the scope, so a tint is never an unlabelled colour. Hovering a feed row highlights the place it was accounted to, and hovering a place highlights its rows; the highlight follows the server's attribution, never where a record's pin happens to sit.

The full-screen Commercial surface covers the map, so the drill-in carries a verb pair for when you want both at once:

  • ⤡ Dock to map on the drill-in's header moves the ledger into a left rail beside the live map — Docked mode. The map is ordinary live ops there: pan it, zoom it, brush against it. Everything you had set comes with you (scope, view, review range, the choropleth and its axis).
  • ⤢ Expand on the docked rail takes you back to the full-screen drill-in, carrying the same state the other way. Escape closes the docked rail like any other panel.

A wash outlives the surface that painted it (Map aid). Close the drill-in — or the docked rail — with the choropleth on and the tint stays on the map, behind a labelled chip reading "Lost time — confirmed tier". From there:

  • the chip's body reopens the ledger as you left it;
  • clears the wash;
  • clicking a tinted place reopens the ledger scoped to that place — the map's own way of asking "what did we lose here?".

Nothing expires the wash on your behalf: the chip's label is what tells you what you are looking at, and dismissing it is your call. Only one wash paints at a time, so turning on the inventory heat map replaces this one (and swaps the chip) rather than blending the two.

The marks a record carries

Records travel with qualifiers — words beside the figure, never deductions from it:

  • disputed — the coordinator doubts the figure (noted, with the note readable).
  • contested — the firm named responsible contests the attribution. The tier total says "…of which 61 contested"; the money never does.
  • changed since confirm — a count, window, evidence or cause changed after adjudication; the confirmer and the named firms are pinged, and the chip expands to before/after.
  • confirmed before close — the decision predates the record's close.
  • raised and triaged by the same person — stated neutrally; on a site with one coordinator this is ordinary, and the answer to it is disclosure, not a rule of conduct.
  • cause withdrawn — the record this one's cause points at was itself decided against. It rides the cause chip only, states the fact and never the reason, and stays silent while the target is undecided, so it can't disclose that an unvetted grievance exists.

None of these six strikes through, discounts or subtracts anything — the ledger's arithmetic has exactly one gate (triage), and everything else is a lens.

Paper and export

Two printed weeklies and one file come off this surface, all gated on report generation or the ledger's own read grant:

  • The full weekly — the principal contractor's working copy: per-firm figures, the marks, the money line, the cause section. Marked not for distribution.
  • The distribution copy — the same sheet minus every £ line, safe to hand around the Friday meeting room.
  • The ledger CSV — the folded entries as columns, dashes carried with their reasons, totals in a trailing block.

What prints is what was vetted. Both weeklies list the confirmed rows and, beneath them, one detail-free line — "N raises awaiting triage, not listed". An unvetted figure never reaches paper and nothing vanishes without a count: the awaiting tier's hours are still stated here, on the surface that may state them, and an evidence pack inherits the same line with the rest of its frozen context. If a figure has to print, confirm it at the meeting table first.

Every one of them prints its date range and its as-at instant beside its figures; a frozen (issued) sheet keeps its issue number, and re-generating mints the next issue with a supersedes line rather than silently replacing ink. The money behaviour on all three is on Standing cost.


Next: Your firm's lost time · Slip & nil returns · Standing cost

Blockers & lost time guide: Overview · Reviewing & triaging · Raising · How lost time is counted · The ledger · Your firm's panel · Slip & nil returns · Standing cost · The register · Scene by scene · Reference

Was this page useful? Report a problem

No cookies and no third-party analytics. A verdict records this page's path — and a search that finds nothing records the words you typed — in this site's own server log.

Generated from source · SimOops living system reference