Paint your week: scheduling around real availability instead of a 40-hour fiction

Underneath every schedule is an assumption nobody examines: how many hours, and which hours, the people actually have. Classic tools assume eight a day, five days a week, every week, for everyone. The assumption is wrong for almost every human who has ever used one. It is wrong for the founder juggling product and sales, the freelancer across three clients, the academic who teaches Tuesdays and Thursdays, the parent whose deep work dies at school pickup, and everyone building something on the side of a day job.

When the availability assumption is fiction, everything downstream is fiction with extra steps. The most honest dependency graph and the most calibrated estimates cannot save a schedule that believes you have forty hours you do not have. Garbage hours in, garbage dates out.

Painting instead of configuring#

Topolog's answer is deliberately tactile: a seven-by-twenty-four grid you paint. Brush the hours that are genuinely available for plan work; leave the rest. Templates (a 9-to-5, an early bird, a night owl) give you a starting wash, and then you correct it to your life: Tuesday evenings, two weekday mornings, the Saturday block. The grid auto-saves and becomes the scheduling constraint for everything.

Painting beats configuring ("35 hours per week") because the shape of availability matters as much as the volume. Twenty hours as four five-hour blocks and twenty hours as ten scattered two-hour slots produce very different realistic finish dates for work that needs continuity. The painted week captures shape natively, and honesty about shape is precisely what the configured number hides. The grid also has the right default the other way: paint nothing on Sundays and the scheduler will never quietly mortgage your Sundays to make a date look good.

What the constraint propagates into#

The painted week is not decoration; it is a hard input to the same engine that computes everything else, and it shows up in three distinct places.

The schedule. The causal-threads view packs the dependency-ordered work into your actual painted hours, day-of-week honest: a 12-hour task starting Thursday on a week with no Friday hours lands Monday, not "tomorrow". This is the same job calendar AI tools perform, with one structural difference: the packing respects the dependency graph and feeds a forecast, not just a tidy week (why that ceiling matters).

The forecast. The Monte Carlo engine converts effort-hours into calendar dates through your painted capacity, per simulated future. The completion spectrum is therefore denominated in your real weeks, not in mythical full-time weeks (how to read one). For a ten-hour-a-week side project, this single correction is routinely the difference between a forecast that says six weeks and a truth that says four months. The four months was always true; the fiction just hid it.

The portfolio. Your painted hours are one pool, and every active plan draws from it. The scheduler allocates the pool across plans and surfaces the at-risk ones when commitments exceed paint (capacity-based workload planning covers the team version, where everyone's grids combine).

Honesty compounds#

The painted week is one of three honesty layers that only work together. Estimates carry their real spread; completions recalibrate the model to your pace; and the painted grid grounds it all in hours that exist. Skip any layer and the others leak: calibrated estimates scheduled into fictional hours still produce fictional dates.

There is also a quieter, personal effect. People who paint an honest week and watch the forecast respond tend to stop making a category of promise they used to make: the one that was only ever achievable in the 40-hour fiction. The grid does not make you slower. It makes you the speed you already were, on paper, in advance, where you can negotiate with it. The plan stops being a hope about a different person's life and starts being a forecast about yours.