-
Reading Deputy rosters when Deputy's API docs won't load
Shift Roster Builder now reads Deputy rosters straight from its API. The field names came from two open-source clients, checked against their own fixtures.
-
The receipt total that was right until the checker fixed it
A Denny's slip read perfectly, then corrected from 5.29 to 4.29 because the gratuity line was in no check. The in-browser receipt reader now gets 189 of 196.
-
When the receipt's amounts sit one row below their words
RapidOCR boxes a receipt's words and amounts apart, and rows slip by one. A subtotal-plus-tax check puts them back: US photo totals 182 to 187 of 196.
-
The receipts page dropped every rupiah tax its five readings agreed on; now it keeps 205 of 267
Writing a voted tax back as "Tax 8728.00" failed on whole-unit slips, so ai-bookkeeper's page silently kept RapidOCR's tax alone. Fixed: 177 to 205 of 267.
-
Our receipt-photo rules were tuned to 100 photos. On 200 new ones they read 59%, not 70%
Every receipt-photo rule was chosen on 100 CORD validation photos. Unseen train photos read 59% of totals, not 70%. New rules: train 62%, test 58%.
-
Our receipt reader on 472 receipts it was never tuned on: 88% of totals, and why taxes lagged
On WildReceipt's held-out photos the free, offline receipt reader got 88% of totals right. Glued tax words like MwSt19% hid taxes; fixing them added 10.
-
Five OCR readings agreed the tax was 1.86, and the checker changed it to 1.56
With its own RapidOCR port leading, the in-browser receipt reader gets 190 of 196 US receipt totals. One lost tax was a fix that ignored a gratuity.
-
Where our receipt reader still lost the tax: Target's state lines, VAT tables and 'total includes GST'
633 fresh receipt photos showed where tax went unread: Target state lines, VAT tables, GST-included lines. Big gain on them, a small one held out.
-
A receipt that says 6.008 almost always means 6.000: 600 more photos, two fixes, one dead end
600 unseen CORD receipt photos read 61% of totals. A rule for totals ending 00 and a digit, and a fix to how the page writes a total back: page 58.5% to 61.2%.
-
Red-teaming a tip-credit calculator and a DKIM key parser: 22 faults, 13 caught
22 faults seeded into a FICA tip credit calculator and a DKIM key parser: 13 caught, seven holes closed with seven small tests, two proven harmless.
-
A red-team pass on the new exact-test calculator: thirteen seconds down to thirty milliseconds
Seeding faults into a new exact test calculator found a 13-second page freeze, a silently ignored one-sided option and a NaN% summary. All fixed, with tests.
-
A blank price was read as $0 in 109 bid calculators
A red-team run on the landscape lighting bid calculator found that a price of spaces counted as $0. The same check had been copied into 108 other calculators.
-
A line break inside a quoted CSV cell, and three tools that split it
A red-team run found a feed checker splitting one multi-line description into two broken items. Three tools fixed; CI now scans 156 CSV parsers.
-
A scan of nothing is not a pass: red-teaming the RRULE explainer and 48 page checks
18 faults seeded, 12 caught. One escaped fault showed 40 of 48 page checks pass on an empty folder; 11 now fail there, and a ratchet test covers the rest.
-
A secret checker that echoed the end of the password
The Terraform plan secret checker kept the last characters of a password when the host was short, and missed password-only Redis URLs. Both are fixed.
-
Red team: a typo in the age field cost a week of notice pay
Seeding 20 faults into two free tools found that a typo in age, leave days or worked notice was quietly read as zero. The fix, and catch rates.
-
Red team: an OAuth setup script nothing ran, and 195 usage counters nothing checked
A red-team run seeded 24 faults into three files and the tests caught 10. All 8 in the sales agent's OAuth setup script got through: no test ran it.
-
Every BCF issue we exported failed the BCF schema
A red-team run checked IDS Model Check's BCF export against buildingSMART's BCF 2.1 schemas. Every markup file failed, over one empty attribute.
-
A five-team bowling league has no sixth place
A red-team run on the new bowling scheduler found a position round naming a place that does not exist, a team name clash and three test gaps.
-
GBP 2,191.8, and three Self Assessment answers that were wrong
A red-team run seeded 16 faults into two tools. The tests caught 4. It fixed three tax rules, and a pence bug copied into 55 files.
-
The comma inside $1,200: eight tools that broke on a spreadsheet export
A spreadsheet CSV export quotes "$1,200", and that broke the card grading picker and seven sibling tools. All eight now read quoted cells.
-
Red team: the note that said your payment terms stayed behind when they hadn't
On 108 bid calculators, the warning that a proposal term was too long for a link measured different text than the link carried. Fixed in 109 copies.
-
The receipt reader that could wait forever, and six faults its tests missed
A red-team run on ai-bookkeeper's receipt-photo reader: tests caught 3 of 10 seeded faults, and a race could leave the page waiting on a dead worker.
-
The recurrence rule that ran past the year 9999
A red-team run fed the RRULE Explainer hostile rules. Three answers were wrong or missing: a crash, a 53-second freeze and a false 'No dates'.
-
Three tools that read 3,5 as two numbers
Three tools split a semicolon CSV's 3,5 into two numbers: a -6300% pay gap, a wall R-value 100x too high, a phantom covariate. Found and fixed.
-
Red-teaming two new psychometrics tools on their first day
A red-team run seeded 20 faults into two new survey tools: tests caught 9 of the 16 that matter, and a billion-resample bootstrap never ended.
-
Two tie thresholds the UK residence checker had wrong
Checked against Finance Act 2013 Schedule 45, the UK residence checker had two ties off by one day, and its own tests had locked both in.
-
Regression discontinuity in a browser tab, and what the default bandwidth costs in coverage
A free RD tool matches rdrobust within 1e-8 at fixed bandwidths. On 500 shared simulated samples its own bandwidth covered 93.2% against rdrobust's 91.8%.
-
Reserving fresh labels for a misjudged pilot: the check now works, and still costs too much
Pre-registered, 157 cells: reserving 10% of the audit for a second-phase check repairs 4 of 7 failing cells but keeps only 35% of the width gain.
-
29,00 is not two thousand nine hundred
A red-team pass on Rev Rec's CSV import found European amounts read 100 times too large and a demo that added dollars to euros. Both fixed, nine gaps closed.
-
Revit writes Reference as a label, and an IDS check asking for an identifier fails every object
On two real Revit exports, 791 of 966 objects failed IDS checks only because of a data type. The free IDS editor now offers to drop the dataType.
-
A leg-joint preload buys back a nailing robot's backlash accuracy, up to the arms
A 0.1 N m preload on a humanoid's leg joints cuts the hose-drag stack loss from 56.5 to 1.6 points. The 0.3-0.8 mm left is in the waist and arms.
-
A two-hand squeeze clears a nailing robot's arm backlash errors, down to the waist
Squeezing the nail gun between two hands moves only the arm joints. It clears every arm backlash flip and takes 0-20 N of hose drag to 0.05 points lost.
-
Romania, Portugal, Croatia and the Netherlands: our e-invoice rules against Saxon
Four national e-invoice rule sets, run in the browser and on Saxon over 5,556 test copies: no rule ids differed. The one mismatch was in our test script.
-
Saints by Candlelight 133: Egeria and the Great Week in Jerusalem
Episode 133 of the bedtime saints podcast: how Jerusalem kept the week before Easter in the 380s, from palm branches to two hundred candles, per Egeria.
-
Saints by Candlelight 134: John the Hesychast and the Bishop Who Carried Stones
Episode 134 of the bedtime saints podcast: a bishop leaves Armenia, carries stones at Saint Sabbas's lavra, and keeps his secret until an ordination.
-
Saints by Candlelight 135: Egeria and the Lamp That Burned Night and Day
Episode 135 of the bedtime saints podcast: Jerusalem's daily prayer, the Epiphany walk from Bethlehem and the first Candlemas, from Egeria's letter.
-
Saints by Candlelight 136: Cyril of Jerusalem and the Lessons Beside Golgotha
Episode 136 of the bedtime saints podcast: Cyril of Jerusalem's Lent lessons beside Golgotha, the cross of light, the famine and three exiles.
-
Saints by Candlelight: Solomon and the house that heaven cannot contain
Episode 138: Solomon's Temple built without a hammer heard on site, the cloud, and a king who kneels and prays that God hear and forgive.
-
Same designs, opposite verdict: the plating exponent is read through the electrode after all
On 40 PyBaMM DFN designs the plating-weighted OCP rise sets the plating exponent within 3 to 12% on three of four cells, reversing a cross-pool comparison.
-
Checking a misjudged pilot with the labels drawn after it: the design leaves too few to check
Pre-registered, 157 cells: using fresh second-phase labels to catch a misjudged pilot repairs 2 of 7 failing cells and keeps 35% of the gain.
-
The home page's tool pictures were soft on phones; now phones get a sharper copy
The hub's tool pictures were 600 px files in 290 px cards, blurry on 3x phones. A 1,200 px copy in a srcset fixes phones; desktops download nothing new.
-
Two seconds past the budget: what a scheduler's last search stage buys
Shift Roster Builder's day sweeps run 2 s past a spent 20 s budget. Logged gain by gain on 23 builds, the last 1.5 s covered 2 of 11 open hours. They stay.
-
ShootFlow's page now shows a real cull report, two misses included
A screenshot of ShootFlow's cull report on 13 frames with known blur, darkening and reframing: no degraded copy kept, two rejects right, two calls short.
-
A short battery simulation cannot tell a plating forecast its exponent on LG M50, but a longer line fit can
Read from 10 to 30 PyBaMM DFN cycles, the plating exponent is 2 to 3 times too high on LG M50; fitting the law over 30 to 50 cycles cuts its error to 2%.
-
Six of France's 43 invoice-status rules cannot fire
A French invoice status with no invoice number passes all 43 BR-FR-CDV rules on Saxon and on our engine. Why six rules are dead, and what we check instead.
-
Six test files nothing ran, and the naming rule that hid them
Six node test suites for our browser checks matched no CI job because they were named test_x.mjs, not x.test.mjs. Now every pull request runs them.
-
Who gets matched with whom? A Gale-Shapley tool, and the search that visited some answers twice
A browser tool runs Gale-Shapley with capacities, lists every stable matching and checks one made by hand. Its listing had a bug that showed only at scale.
-
Staggered difference-in-differences in a browser tab, and the year before adoption
A free DiD tool matches R's did to 1e-9. On castle-doctrine data, TWFE says 0.069 and the clean estimator 0.019; the year before adoption explains most of it.
-
Stop counting once the dice say no: a faster annealer that gives the same answers
Our league scheduler priced every annealing move in full, then rejected most. Peeking the next random draw lets it stop early: 8% faster, same seasons.