The inch is also based on the SI standard.

btw, since it's a derived unit, that also means by definition it's less accurate.

Latter is an incorrect statement, but a very interesting one, because this leads to some deep thoughts about ratio determination for any of the SI-units.

At first, 'derived units' are all these, which are derived by powers, products or quotients of the 7 base units.

Examples (relevant for EEVBLOG) are the Ohm (kg·m^2·s^-3·A^-2) and the Volt (kg·m^2·s^-3·A^-1).

The inch is a non-SI unit of length, but it's not a derived unit. Though the inch is not recommended for use, it has the same acceptance, as other non-SI units, but which are still in daily use, like: minute, hour, tonne, litre, etc.

Anyhow, NIST defines conversion factors between the SI definitions and many of the U.S. units.

These factors may be exact or an approximation.

https://www.nist.gov/physical-measurement-laboratory/nist-guide-si-appendix-b8The inch / meter conversion factor is exactly 2.54E-2.

Therefore, the

**definition** of the inch is exact, like the meter.

But there's always the

**realization** (mise en pratique) of units.

The meter is realized via the exact definitions (i.e. zero uncertainty) of the speed of light, and of the second, therefore, any fraction or multiple of the meter can be realized with zero uncertainty - within the experiment precision limits, of course.

So the inch is in terms of definition AND realization as accurate as the meter itself.

The length and the time are defined by invariable constants of nature, therefore universally constant and exact.

Volt and Ohm are currently not exact units, neither in definition, nor in realization (~ 1e-7 uncertainty).

This will also change by SI-2018. Their new definition and realization will turn them into exact units, which are also invariable and universal. The new SI - Volt definition by a Josephson Junction Array also has the big advantage, that any fractions and multiples of a Volt can be realized with zero uncertainty (within experiments accuracy), whereas the Klitzing-Ohm has to be divided by methods, which are not exact.

Now up to the mass / kilogram, that's also different. The current definition by means of an artifact is exact.

Anyhow, this unit is not invariant, as the prototype kilogram and its 40 copies all drift over time, obviously.

The realization of the mass, that means any multiples or fractions of the prototype kilogram, also is not exact, because at first, it's only possible to make a direct comparison to the prototype by balances, which have about µg uncertainty, and the creation of fractions like 500g or 100g is even less precise, because there's no such precise 'ratio machine' for masses. Realization of an ounce is even worse, due to the non-exact conversion factor.

Analogously, before the discovery of the Josephson Junction, ratios of the Volt depended on the Kelvin-Varley-divider (Fluke 720A), or the Hamon-divider (Fluke 752A), ratio uncertainty > 1E-7 only.

The new kilogram definition by means of an exact value for Plancks constant, and the Avogadro constant will allow to make exact copies, and exact fractions and multiples, as comparisons to artefacts are no longer necessary.

You may simply design a Si-sphere to let's say 125g, and determine its exact (currently to ~2E-8 uncertainty) mass by counting its atoms.

In a similar manner, the Watt balance also allows exact creation or measurement of random masses by appropriate variation of its measuring voltage.

This is the 2nd reason and advantage of the new mass definition with SI-2018.

Frank