A physics engine, and a verification tool enforced on top of it. PhysWall inverts a closed, published, non-linear physical law at a single measurement point — and refuses when the inverse is not unique. Seven laws, one engine, and the same refusal in all of them.
Twenty-eight domains. 16 of them can tell you a number is impossible, 9 can tell you this law does not govern your case — and four of them are not physics at all.
This page is a list, and it is also the argument. The claim is not that we know twenty-eight laws. It is that twenty-eight different things fit the same shape, so a reader gets the same three answers whichever one fires.
Every figure below is counted from the extension's own files when this page is built. None of them is typed. A page about counts is the worst place to keep a second copy of them.
Filled — can say a number is impossible. Ringed — can say so with nothing asked of you. Plain — can only say what is missing. Dashed — no closed law yet.
carry a ceiling, so they can contradict the answer. 2 of those need nothing from you at all — the limit is a published constant and the number either clears it or does not.
have no ceiling to compare against, so they name what is missing instead. This is the common case and it is the useful one. Most numbers a model gives you are not impossible. They are unverifiable, which is a different thing and nobody says it.
This is the part worth stopping on. It reads like a caveat and it is the opposite.
| Law | Domain | Field |
|---|---|---|
| Amdahl 1967 | Parallel speed-up against the serial fraction | computing |
| Miller & Sanjurjo 2018 | Streak, or sample size | statistics |
| Nyquist 1928, Shannon 1949 | Sampling rate against bandwidth | information theory |
| Wilson 1927 | Is this percentage distinguishable | statistics |
A confidence interval and a heat engine have nothing in common. They sit in the same row, with the same fields, and produce the same three verdicts. Nothing in the scanner knows which is which.
That is why it works, and it is the only interesting claim on this page. Adding a domain is a row, not a project. The twenty-ninth will behave like the first twenty-eight because nothing about the first twenty-eight is special-cased.
One of the twenty-eight has no published law behind it. The basketball release angle tells you what a shot angle would need before it can be judged, and there it stops. It is in the list because it does useful work; it is named here because a list that hides its weakest entry is not a list, it is a brochure.
| Domain | What it can say | Source |
|---|---|---|
| Wind turbine power coefficient | says it is impossible, with nothing asked of you | Betz 1919 |
| Single-junction solar cell efficiency | says it is impossible, with nothing asked of you | Shockley & Queisser 1961 |
| Diffraction limit on resolution | says it is impossible | Abbe 1873 |
| Parallel speed-up against the serial fraction | says it is impossible | Amdahl 1967 |
| Battery specific energy against its chemistry | says it is impossible | Abraham, J. Phys. Chem. Lett. 6:830, 2015, Eq. 5 and Tables 2–3 · Li–S 2,600 Wh/kg of sulfur from McCreary et al., Batteries 7:82, 2021 |
| Heat-engine efficiency | says it is impossible | Carnot 1824 |
| Heat pump coefficient of performance | says it is impossible | Carnot 1824 |
| Trace length against a loss budget | says it is impossible | conductor and dielectric loss, standard form |
| Dielectric loss factor | says it is impossible | dielectric loss, standard form |
| Radiation Q for a small antenna | says it is impossible | Wheeler 1947, Chu 1948, Pascale 2021 |
| Minimum energy to erase a bit | says it is impossible | Landauer 1961 |
| Channel discharge from stage | says it is impossible | Manning 1889 |
| Sampling rate against bandwidth | says it is impossible | Nyquist 1928, Shannon 1949 |
| Photosynthetic solar-to-biomass efficiency | says it is impossible | Zhu, Long & Ort, Annu. Rev. Plant Biol. 61:235, 2010 · 4.6% C3 and 6.0% C4 on incident solar at 30 °C and 387 ppm CO2 |
| Shannon-Hartley channel capacity | says it is impossible | Shannon, Proc. IRE 37(1):10, 1949, Theorem 2 — "it is not possible by any encoding method to send at a higher rate and have an arbitrarily low frequency of errors" |
| Rocket delta-v against mass ratio | says it is impossible | Tsiolkovsky 1903 |
| Distance from round-trip light time | says what is missing | c is exact by definition since 1983 |
| Activation energy from a rate | says what is missing | Arrhenius 1889 |
| Speed climbing power | says what is missing | work against gravity, closed form |
| Return loss against load Q | says what is missing | Bode 1945, Fano 1950 |
| Anapole disk radius | says what is missing | fitted to seven published disks |
| Is this percentage distinguishable | says what is missing | Wilson 1927 |
| Radiated power from temperature | says what is missing | Stefan 1879, Boltzmann 1884 |
| Streak, or sample size | says what is missing | Miller & Sanjurjo 2018 |
| Thermal noise frequency | says what is missing | kT/h |
| Which wave approximation applies | says what is missing | Airy 1845 |
| Peak wavelength from temperature | says what is missing | Wien 1893, constant from CODATA 2018 |
| Release angle and consistency | says what is missing | no closed law here yet |
24 physics, 4 not. 27 rest on a law somebody else published and named; one does not, and says so.