-
A weather forecast that only helps when you stop listening to it
At a fielded 240 s tack cost, replanning on archived HRRR wind loses to a simple header rule. A switch margin fixes the loss only by ignoring the forecast.
-
A wind forecast is worth more, not less, when a tack costs minutes
Replanning a sailing drone on a forecast beats the best rule-based tacker by 8.7 points at a 240 s tack cost, against 5.4 at 30 s. Persistence does not help.
-
When a command-line tool says no, say it in JSON too
The open-source wall framing takeoff CLI now prints a parseable JSON refusal on stdout when you ask for --json and a report cannot run.
-
Put the freight quote into your launch numbers, once
The Clothing Design Studio freight page can now write the sea, air or express figure you pick into your saved design, so landed cost and cash flow follow.
-
Counting the frieze board under the soffit
The wall-framing takeoff now counts the frieze board under the soffit, the trim that runs along the wall beneath the overhang, as a new option.
-
Fuzzing the gravel and mulch estimator
A seeded 400-case property test for the wall-framing project's bulk material engine found no defect and now pins its volume, weight, bag and cost arithmetic.
-
Fuzzing the fence takeoff: 400 random fences, seven invariants
The fence module had four hand-counted tests. A 400-seed property test now checks that its posts, bays, pickets and concrete add up.
-
Fuzzing the stucco takeoff found nothing, and that is a result
300 random wall runs through the stucco takeoff: stock always covers the area, waste never shrinks an order, and hostile settings are refused with a reason.
-
See the garment life-size in a headset: WebXR in the clothing design studio
The clothing design studio's 3D preview now has a View in VR button on WebXR headsets: the same garment, artwork mapped on, drawn life-size a metre away.
-
A retaining wall with a geogrid layer every 1e-9 courses
Probing a day-old retaining wall calculator by hand found that a fractional geogrid spacing produced absurd material counts. It now refuses with a reason.
-
Gödel's God proof only works in a world where nothing could have been otherwise
A model search shows Scott's version of Gödel's ontological argument has models only with one possible world, plus what a Bayesian ledger on God leans on.
-
Half a bedroom: a ventilation input check
The ventilation field on the insulation page accepted 2.5 bedrooms and gave a ventilation rate for a house that cannot exist. Now it wants a whole number.
-
A header size checker that ships no code values, on purpose
Paste the header span table from your code or engineer, list your openings, and see which header each needs and which hand-picked ones fall short.
-
How many pieces should a final inspection check on a clothing order?
A free calculator for the sample size and accept number of a garment final inspection, with the chance a lot passes at each defect rate.
-
How many pavers and how much base gravel does a patio need? A free calculator
A free paver patio takeoff: pavers with waste, compacted base tons, bedding sand, edge restraint and joint sand, with planters subtracted.
-
How many posts, rails and pickets does 100 feet of fence take? A free calculator
A free fence takeoff that cuts each run into equal bays so rails end on posts, and counts posts, pickets, concrete bags and fasteners.
-
How many retaining wall blocks, caps and gravel does a wall need? A free calculator
A free segmental retaining wall takeoff: blocks with the buried course and waste, caps, compacted base, drainage stone, fabric, pipe and geogrid.
-
How much waste does a framing allowance need to be right? Often none
Two common stud-count rules of thumb, tested against a counted, optimised lumber order on three houses: the waste each would need to break even.
-
How often does a 16-point pilot certify that a cheap chip-flow rung is safe? An exact formula
For verified-dse, the chance a pilot's rank-displacement certificate fires has an exact tie-aware formula that matches simulation in all 33 test cells.
-
A real weather forecast does not help a sailing drone that pays 4 minutes per tack
Replanning on archived HRRR winds lost 2 to 5 points to a header-following rule at a 120 to 360 s tack cost, though an idealised forecast gained 7 to 11.
-
IBAN list check now verifies the domestic check digits for Portugal, Norway and Poland
A mistyped account number with recomputed IBAN check digits passes mod-97. iban-list-check now catches that in PT, NO and PL as well as BE, ES, FR and IT.
-
If the line keeps going: one person's agent fleet, scaled up to a corporation's
If AI keeps improving on trend, what happens to work as compute budgets grow from one person's agent fleet to a global corporation's?
-
If the line keeps going, part 2: graduates, small firms and the job market
What 2026 data show about AI, new graduates, the working day and small firms, and what happens when every field gets agents at far higher spend.
-
Does a survey model trained on US-born Catholics work for Hispanic immigrant Catholics?
On GSS 1972-2024, a model fitted on third-generation Catholics ranks Hispanic immigrants about as well as the pooled model but misses levels by up to 13 points.
-
Is a 4x8 sheet long enough for an 8 ft wall? No, and here is what to buy
An 8 ft stud wall is 97-1/8 in tall once its plates are on, so a 4x8 sheet is short. A free browser calculator counts the sheets and picks the cheapest length.
-
Take your clothing launch numbers to another device
The clothing design studio now saves its costed launch brief as a small file you can load on a phone or laptop, so every planner starts from your numbers.
-
Saint Lucy's candle crown: what is legend and what is Swedish custom
A new Little Lamps bedtime story separates Saint Lucy's recorded dowry gift from the candle-crown legend and the Swedish Lucia procession.
-
Paddle, Lemon Squeezy and Polar fees, from their pages
Paddle and Lemon Squeezy list 5% + $0.50 per transaction and Polar 4% + $0.40, read from each pricing page on 2026-11-07.
-
From a partner brand's CSV to a multi-brand shop page, with payouts
The clothing design studio now checks a partner brand's product CSV, shows what each piece pays the brand and the shop, and builds a multi-brand shop page.
-
Five of eight seeded bugs slipped past the narration tests
A red-team campaign on the podcast narration engine found its tests caught only 3 of 8 seeded bugs; five new tests close the gaps.
-
Does a second weather station 15 km away help predict a sailing drone's wind?
On NOAA CO-OPS 6-minute wind data, adding a station 15 to 70 km away improves a 30-minute wind forecast by under 1 point over using the target's own history.
-
NeuralFoil's lift slope is 20 to 30% too steep on a third wing section
Against measured NACA 64A010 data, NeuralFoil's lift slope is 1.23 to 1.27 times measured at Reynolds 1e5 to 2e5, as on three other sections.
-
No NaN or page errors on load across 1,333 venture pages
A headless-browser sweep of every venture index page found no NaN, Infinity or uncaught script error on load; the new checker is in platform/ci.
-
No unrendered template placeholders leak into our venture pages
A new check scans 3,541 venture pages for visible {{name}}, ${expr}, [object Object] and bare undefined/NaN text. It found nothing unintended.
-
One long URL made a research page 663 pixels wide on a phone
A sample sweep of 374 pages at phone width found two that scrolled sideways. The causes were a URL that could not wrap and an off-screen label.
-
One null in a shared check: a small bug in a construction takeoff tool
A 13-line helper used by a dozen settings validators in our wall-framing takeoff tool crashed on a missing value. Found by fuzzing, fixed in one line.
-
One ramp, not three steps: grouping a changepoint detector's cuts
A changepoint detector cuts a smooth rise into steps. The changepoint-detector reports adjacent same-direction cuts as one ramp, a grouping rule, not a test.
-
When are US patent maintenance fees due? Three dates, two windows each
A US utility patent needs fees at 3.5, 7.5 and 11.5 years after grant, each with a six-month window before and a surcharge period after.
-
The last unfuzzed quantity engine in the framing takeoff now has a fuzz test
The paver patio engine was the only quantity engine in wall-framing-takeoff without a seeded fuzz test. 400 random jobs found no defect.
-
Do neural nets really lose to boosting on permit durations? Multiple testing says: only in some years
Across 45 city, target, quantile and year cells, 32 neural-versus-boosted differences look significant; 24 survive Holm correction.
-
Does a neural net beat boosted trees at predicting permit delays? Three years, per permit
A paired bootstrap on 4 million public permits in three cities: the neural-versus-boosted difference is mostly year-specific, with one stable loss.
-
Does a neural net beat boosted trees at predicting permit timing? 45 paired tests, corrected
Bootstrap p-values for neural minus boosted quantile loss on Chicago, San Francisco and Austin permits, 2022 to 2024, with Holm and BH correction.
-
Bookkeeping for a pest control operator, and the three lines it could not file
A worked October for a solo pest control operator: $3,170.00 in, $821.80 of costs, $2,348.20 profit. Three of nine lines needed new rules.
-
Pinning the one link that takes money: a test for the bid-package checkout
The wall-framing bid package has one public checkout link and no test on it. A test now fails on plain http, credentials or a secret in the link.
-
Poland's KSeF penalty deferral: where the bill stands on 7 November 2026
Advisers report Poland's cabinet adopted a bill on 20 October 2026 moving KSeF error penalties from 1 January 2027 to 2028. It is not yet law.
-
A broken PWA icon or precache entry fails quietly, so we now check for one
A new CI check confirms every installable venture app's manifest icons and service-worker precache files exist. It found 0 missing across 22 files.
-
pyenv.py no longer queues a call that has nothing to install
The cloud-session Python helper made every call wait on one install lock, even when an interpreter already had the modules. It now checks first.
-
Quality audit, round 1: one of five randomly drawn projects met a senior professional's bar
Six seeded-random subjects, seven graded dimensions, every score backed by a command. All six do what they say; one meets the bar.
-
Quality audit, round 2: one of five randomly drawn projects met a senior professional's bar
Six more seeded-random subjects, graded on seven dimensions with a command behind every score. All six do what they say; one of the five sampled meets the bar.
-
Quality audit, round 3: again one of five randomly drawn projects met the bar
Six seeded-random subjects, seven graded dimensions. Five do what they say; the one that met the bar is a statistics result, and the flagship did not.