In our generated example, 90 minutes of core work fits one continuous two-hour block after adding preparation and switching: the plan takes 115 minutes. The same work cannot all fit into two separate 60-minute blocks under the same rules. The total available time is unchanged; restarting the work changes what must fit inside it.

This is a constructed model, not a reader's observed week or a claim about typical concentration. We selected the durations to test a planning question, executed the allocations and published the complete inputs. No productivity improvement or universal buffer is inferred from the result.

Declare the work, the openings and the overhead

Task A is 60 minutes of drafting a defined comparison section. We allow its core work to be split into 15-minute chunks. Task B is a separate 30-minute check that this example treats as indivisible. The two tasks are independent here; no approval dependency or external deadline is modeled.

Each visit to a task costs 10 minutes of preparation before core work. Returning to A in another opening therefore costs another 10. Changing from one task to another within the same opening costs an additional 5 minutes. There is no separate five-minute charge between days; re-entry preparation is already counted. A task is visited at most once within a block.

Those are explicit fictional rules, not recommended session lengths or measured switching costs. The 120 available minutes are assumed to remain after obligations, travel and protected rest. Actual observations could justify other inputs, including no repeated preparation; that would be a different model.

One continuous opening fits the whole target

For one 120-minute block, an explicit plan is preparation A 10, work A 60, switch 5, preparation B 10, work B 30. It takes 115 minutes and leaves 5. Both defined tasks are complete.

Segment in the continuous blockStart–end offsetMinutes
Prepare A0–1010
Work on A10–7060
Change context70–755
Prepare B75–8510
Work on B85–11530

The core-work sum of 90 was not wrong. It was incomplete as an occupancy estimate: preparation adds 20 and switching adds 5. Do not add those as separate rows if they are already included in the main task estimate.

Why two 60-minute openings do not complete it

A's full 60 minutes plus its preparation would take 70, so A needs at least two visits in the shorter openings. B still requires one visit. At least three visits across two blocks force at least one block to contain a change between tasks. A necessary time bound is therefore 90 core + 30 preparation + 5 switching = 125 minutes, already beyond the 120 available.

The 125 is a lower bound, not a proposed feasible 125-minute schedule. It is enough to refute full completion under these assumptions. The finite program separately examines 25 combinations of declared allocations across the two openings. Twenty-one respect the work limits, and none completes both tasks. In the continuous block, the corresponding search includes two complete orderings.

The search covers the stated 15-minute A choices and indivisible B only. It is not an optimizer for every possible real calendar. The analytical bound explains the failure; a large list of attempted plans is not a substitute for that explanation.

Choose a smaller result and leave the rest visible

Two valid examples use the short openings differently. Neither is presented as the best priority decision or the maximum possible partial progress.

Plan using two 60-minute blocksCore work completedSetup / switchTime occupiedStill unfinished
A30 in each blockA60, B020 / 080B30
A30 first, B30 secondA30, B3020 / 080A30

Both leave 40 minutes in aggregate, as two separate 20-minute gaps. That does not create a new uninterrupted 40-minute opening. The first plan finishes the drafting target and defers B. The second finishes B and only a 30-minute first action on A. Calling that first action “finish the draft” would conceal half the original work.

It may be possible to add a smaller piece under different choices; the two examples do not claim maximal progress. A real deadline may make deferral unacceptable. Then the decision concerns scope, another opening or agreement with the relevant person, rather than a silent reduction of the time estimate.

The total worksheet answers a narrower question

We executed this page's existing worksheet module with the selected totals. Core-only rows show 90 planned and 30 remaining. The continuous plan including overhead shows 115 and 5. Entering the short-block completion lower bound shows 125 and −5. Either 80-minute partial plan leaves 40 in total.

Those are correct total-capacity results. The worksheet does not know block boundaries, divisibility or whether preparation repeats. A positive remainder cannot certify a calendar arrangement. The new allocation code supplies that separate check; the worksheet is not described as broken for doing only its stated job.

For comparison, the program also evaluates a clearly labeled zero-overhead counterfactual. A60 in the first block and B30 in the second then fit in 90 minutes. This result belongs to different assumptions; it is not a method for making real preparation disappear.

Preserve the estimate and the actual record separately

Our previous three capacity cases remain available: carrying 165 minutes into 120 leaves −45; a smaller next action leaves 10; a setup-inclusive example leaves 30. Their inputs and results are fictional too. The new example adds the shape of openings and repeated setup, rather than turning either set into an observed week.

When there is an actual record, separate planned core work, preparation, within-block transitions and work left. A missing input is not the same as an optimistic writing estimate. Preserve the original plan rather than overwriting it with the actual duration, and transfer only the remaining work into the next plan. This article supplies no real actual-time measurements to substitute for that record.

Reproduce the finite cases and inspect the rejected plans

The Python source, segment CSV and full results expose each visit, assumption and remaining task. Run python3 session-check.py if Python 3 is already available. It uses fixed generated inputs, no personal files or calendars, no network and no extra package.

We check four explicit plans, the finite allocation counts and the analytical lower bound. The existing JavaScript worksheet was executed separately for total comparison; that is not an independent scheduling algorithm or an independent editorial review. Seven altered plans are rejected: over capacity, work beyond the target, invalid A chunk, split B, unknown task, negative work and repeated context within a block.

The record distinguishes full completion, a smaller completed action and work deliberately left out. That distinction remains necessary even when all the arithmetic is correct.

Use the worksheet

Enter only what you need. Inputs stay on this page unless you choose browser saving or download a file. No account identifiers are needed. Local saving is off by default; do not use it on a shared device.

Subtract sleep, fixed commitments and time you do not want to allocate first. Blank cells contribute zero to the entered subtotals. The planned/actual difference is shown only after both times are entered for every row; enter 0 only when known.

TaskPlanned minutesRecorded actual minutesNext changeActions

Calculation check: the fictional example has 120 available, 90 planned and 110 actual minutes, leaving 30 planned minutes and a 20-minute difference.