Continuity rule: the bar runs thirty days, ten of them worked

An activity records an actual start and an actual finish, and the drawing fills every date between them. Type the dated observations you hold and the filled span comes apart, 30 days of bar over 10 worked days, with a 20 day gap the record does not cover and a probe date that reads as worked on the drawing and not worked in the observations. The same output over those two denominators gives 10 units a day and 30, and nothing here reconstructs what happened inside the gap.

Live result

The dated record leaves a 20 day gap inside a bar the drawing fills end to end.

days in the filled bar30 days
days the record supports10 days

At the probe date the drawing reads worked and the dated record does not.

The typed record

Try a scenario

Same scope and information cut

Both records need to describe the same activity and the same information cut before their dates can be compared.

Actual start

Actual finish

Probe date

Output quantityunits

Dated worked segments

Segment 1

Worked from

Worked through

Segment 2

Worked from

Worked through

At least one dated worked segment is required.

The rule and its breaker

The rule is commonly stated as follows: a recorded actual start and actual finish mean the activity was worked throughout every date between them.

The smallest breaker spans 1 March through 3 March. Work is observed on 1 March and 3 March, so the drawing fills 3 days while the dated record supports 2 and leaves a 1 day gap.

The rule is restored when dated observations cover every civil date in the span, either as one continuous segment or as adjacent segments with no missing date.

The computable witness

Days in the filled bar: 30 days
Observed worked days: 10 days
Days not supported as worked: 20 days
First date not supported as worked: 2026-01-06
Interruptions: 1
Drawing at the probe: worked
Dated record at the probe: not worked
Output per bar day: 10 units a day
Output per observed worked day: 30 units a day

What does not break the rule

One continuous thirty day observation covers the whole bar, so the gap is zero.

Two adjacent observations from 1 February through 5 February and 6 February through 10 February still cover all ten days.

Boundary

It compares a filled two endpoint bar with dated observations typed for the same activity, without reconstructing history, verifying work, finding concurrence, assigning cause, or deciding entitlement.

Here is the maths

The formula, in words

Count civil dates inclusively from actual start through actual finish. Count every date inside the nonoverlapping worked segments. The difference is the unsupported gap. Divide the same output once by bar days and once by observed worked days.

G=B-W

where $G$ is gap days, $B$ inclusive bar days, and $W$ the union of observed worked days.

A worked example

From 1 January through 30 January there are 30 inclusive days. The two five day observations support 10 days, so the gap is 20 days. Output of 300 gives 10 units per bar day and 30 units per observed worked day.

30 - 10 = 20\qquad 300 \div 30 = 10\qquad 300 \div 10 = 30

\begin{gathered}30 - 10 = 20 \\ 300 \div 30 = 10 \\ 300 \div 10 = 30\end{gathered}

What this does not model

The arithmetic does not infer work between observations, verify any observation, test concurrence, assign responsibility, establish cause, or decide entitlement.

Source basis

The witness uses public civil date subtraction, interval union and division. It reproduces no definition or wording from sold guidance.

The whole record prints. Your browser can save it as a PDF instead of printing it.

A two endpoint bar has no supported interior, and this witness will not guess one.

Filled bar witness record