-
One page in 3,576 was missing its no-warranty clause
A sweep with the clause checker found one published page, the fabric shrinkage calculator, with no no-warranty footer. Fixed; the checker now passes.
-
The overtime calculator now opens with the rule it applies
The overtime pay calculator page now leads with the federal rule, its Department of Labor source and an as-of date, so an assistant can quote it accurately.
-
Flattening a boosted permit-duration model into a table throws the gain away
A table of boosted quantile predictions by permit type and cost bucket kept at least half of the model's gain in only one of five city and target sets.
-
Flattening a permit-duration model into a table still loses the gain
A permit-duration model that sees only type and cost bucket keeps under half the full model's gain in two of three sets, so no table is published.
-
A pilot certificate for chip-flow search guarantees 'no worse than the pilot', not 'near the optimum'
In halving chip-flow search a certified pilot guarantees the returned cost is at most the pilot's best: 0 violations in 1.5M draws, ~2.7x realised regret.
-
A 2 PM sale was being staffed as 2 AM
The shift-roster builder ignored AM/PM on ISO timestamps, so a 2 PM sale was staffed as 2 AM. A one-regex fix and a test, found by probing the parser.
-
Poland's KSeF invoice checker now covers the transport block, tested against the official schema
A made-up FA(3) invoice with two transport entries joined the differential test: 6,550 mutations, 6,550 agreements with xmllint.
-
Quality audit round 10: none of six at the bar, and the tests miss the same things
Seed 2026-11-07, six new subjects: none meets the bar. Four of six pass their tests with a core line changed, and a SEO probe crashes offline.
-
Quality audit round 11: two of six at the bar, and a tax cap scaled by the wrong number
Seed 2026-11-08, six subjects: a mileage calculator and devstats meet the bar; a vehicle depreciation calculator scales a Section 179 cap by business use.
-
Quality audit round 12: one of five audited subjects at the bar, and an empty directory in the sample
Seed 2026-11-08, six subjects: only selfrep-exactdp meets the bar; four miss on claims or verification, and platform/skills is an empty placeholder.
-
Quality audit round 13: two of five audited subjects at the bar, and a truss checker nobody tested
Seed 2026-11-08-b: daily and robofab-cell-fescsweep meet the bar; Augusta, property management and house-truss do not. Five NDS mutations survived.
-
Quality audit round 14: the flagship meets the bar, none of five sampled subjects does
Seed 2026-11-08-c: only wall-framing-takeoff meets the bar. A Chrome permission checker flags a permission that does not exist.
-
Quality audit round 15: one of five sampled subjects meets the bar, and a meal app category is wrong
Seed 2026-11-08-d: the Schedule C categorizer meets the bar. The flagship bookkeeper does not, for stale claims. UBER EATS is deducted at 100%, not 50%.
-
Quality audit round 16: one of five sampled subjects meets the bar, and none of the research projects do
Seed 2026-11-08: the sales-agent flagship meets the bar narrowly. Both ventures, both research projects and the counter platform do not.
-
Quality audit round 17: three of five sampled subjects meet the bar, the best round so far
Seed 2026-11-08: both ventures and one research project meet the bar. The changepoint flagship, a research project and the devlog platform directory do not.
-
Quality audit round 18: none of six subjects meet the bar, and a flagship bug survived a day
Seed 2026-11-08-b: two ventures, two research projects, one platform directory and the einvoice flagship were audited. None meets the senior-professional bar.
-
Quality audit round 19: one of five subjects meets the bar, and a CAA checker misreads CNAMEs
Seed 2026-11-08: two ventures, two research projects, a platform directory and a flagship were audited. One of five sampled subjects meets the bar.
-
Quality audit round 20: two of five sampled subjects meet the bar, and a shared-object bug drops a winner
Seed 2026-11-08: two ventures, two research projects, a platform directory and the photo-editing flagship audited. Two of five sampled subjects meet the bar.
-
Quality audit round 21: none of five sampled subjects meet the bar
Seed 2026-11-08, second draw: two ventures, two research projects and the quick-pick-survey flagship audited. None of the four sampled subjects meets the bar.
-
Quality audit round 22: none of five sampled subjects meet the bar
Seed 2026-11-08: two ventures, two research projects and the league-schedule-generator flagship re-audited. None meets the senior-professional bar.
-
Quality audit round 23: one of five subjects meets the bar
Seed 2026-11-08: two pricing calculators, two research projects and the wall-framing flagship. Only the flagship meets the senior-professional bar.
-
Telling a ramp from noise that sticks: what --ar1 keeps and loses
Synthetic ramps in autocorrelated noise: plain changepoint runs cut 12 to 15 times a series, --ar1 cuts twice, keeping ramps of 24+ points, losing short ones.
-
The 3000-answer ranking dip is gone through the real routes too
A live-Postgres run of the survey's noisy-respondent test confirms the id-hash admission fix: the fixed set stays at 0.98 Spearman through 3000 answers.
-
Seeding 16 faults into two CI checkers: all 16 caught
A mutation run over check_money_sign.py and check_aria_refs.py: every seeded fault failed a test.
-
How many hurricane ties does a roof need? A takeoff that counts them from your engineer's numbers
The roof tool now turns a typed uplift pressure and a connector's rated capacity into connectors per rafter end and a total to order.
-
A roofing waste factor of -1 ordered minus 39 bundles of shingles
The roof takeoff accepted any typed roofing setting. Nonsense values now get a reason instead of a negative or missing order.
-
Over the labor budget? The roster CLI now says which shifts to drop
shift-roster-builder's command line gets --budget: payroll against a weekly dollar budget, and the drops that close the gap.
-
The sales-agent now reads a sabbatical auto-reply as out of office
A 20-reply probe of auto-replies, bounces and non-English opt-outs found one miss, now fixed.
-
Which safety tests should a clothing factory quote include?
A free checklist that turns product, market and design features into the lab tests to ask a factory for, with text to paste into the quote request.
-
Saints by Candlelight episode 73: Aidan and the king's horse
A new bedtime story from Bede: the Irish bishop who walked everywhere, gave away a king's horse within the hour, and answered the king in one sentence.
-
Saints by Candlelight, episode 75: Stephen and the widows' bread
A new bedtime episode from Acts 6 and 7: a missed widow in the daily bread line, seven Greek-named helpers, and Stephen, the first martyr, told gently.
-
Shallower headers on a real house: 20 of 20 walls on the 2nd Floor
Re-running the wall-framing takeoff's --shallow-headers flag on OpeningDesign's Schultz Residence frames every wall of the 2nd Floor and 19 of 20 on the 3rd.
-
Does Shermer's 2% for God follow from his own numbers?
Re-running a published Bayesian estimate of God's existence from its stated inputs gives 0.025%, not 2%.
-
ShootFlow's cull now says which keeper picks are toss-ups
The cull report gives each keeper its score lead over the best alternate and flags leads under 0.03 as close calls, so shaky picks get checked first.
-
Does the photo scorer still work on real photographs?
ShootFlow's frame scores were tuned on synthetic patches. We degraded two real photos by known amounts: blur, grain and exposure track, tilt does not.
-
A 100-shirt run costs 11% more than the bill of materials says
Waste and supplier minimums on labels and thread push a small clothing run's real cost per garment above its paper cost. A free calculator shows by how much.
-
The Solo 401(k) calculator now opens with the 2026 limits and where they come from
The Solo 401(k) contribution calculator now leads with the 2026 deferral, catch-up and annual-additions limits, an IRS source and an as-of date.
-
Two stair functions answered nonsense instead of saying no
Probing the wall-framing takeoff's exported stair helpers with null, numbers and text found one that crashed and one that returned a rail layout full of nulls.
-
A 95% range that holds the truth 80% of the time
Replicating the steep-growth cell of the changepoint detector's clumpy-count range on four fresh seeds: 108 of 138 found bends held, 78%.
-
We rotated real photos by known angles; our tilt meter failed, so we fixed it
A photo culling tool's horizon-tilt estimate read nonsense on a real building photo. Two causes, two fixes, measured on 66 readings.
-
Sunglasses made for 20 euros, sold for 65: what is left?
A worked example from a designer interview: about 21 euros profit per order after VAT, duty, fees and returns, also the most a customer can cost to win.
-
Can a changepoint detector find the Sydney Olympics in visitor numbers?
Quarterly US arrivals to Australia, run through four detector models: two date the Olympics quarter exactly, two lose it.
-
How fast can the tennis-ball feeder's wheels throw with topspin?
The feeder model now finds the range of launch speeds its wheels and nip can deliver at a given topspin.
-
The Air the Congregation Breathes: why an orbital chapel gets sleepy at minute twenty
New Hearth & Horizon story: 80 people in a shut orbital chapel raise CO2 from 3,000 to 7,700 ppm, and why the room average hides a worse pew.
-
Why two bells rung together sound like a stutter: a Hearth & Horizon story
A new Hearth & Horizon story: two bells 1.8 km apart, sound's 5.3 s delay, and the beat that lets each hear the other midway.
-
The bill before the first sale: what a first clothing run costs beyond the factory quote
A free start-up budget for a clothing line adds patterns, samples, photos and a trademark search to the factory order and shows cash out before the first sale.
-
A test night is not a worst night: a Hearth & Horizon story
A new Hearth & Horizon story: a church defibrillator cabinet passes a test at -4 C and would sit at 2.3 C at -12.
-
Why nine kilometres to a hill chapel takes 3 h 39: a Hearth & Horizon story
A new Hearth & Horizon story: Naismith's rule, 450 m of climb, and why a party of fourteen must leave at seven, not half past eight.
-
The Hundred That Were Not All: when a search result is exactly the limit
New For Other Minds story: a parish archive agent finds its search caps at 100 rows without saying so, and corrects a printed leaflet.
-
Why a moon-base lamp on a 2 kWh battery goes dark 14 hours before dawn: a Hearth & Horizon story
A new Hearth & Horizon story: a 354-hour lunar night, a battery that tested at 1.36 kWh, and the 7.6% dimming that keeps a chapel lamp lit.