Re: Irrelevant re-invention
OS coordinates are two letters followed by 4 to 10 digits. An error in the first letter (which is S, T, H, N or very rarely O) will be in the wrong 500km x 500km area of the country; an error in the second (any letter except I) will put you in the wrong 100km x 100km subsquare. In the rare case a 10 digit O/S reference is supplied (to a 1m x 1m square) errors in the 5th or 10th digits will yield a result close enough for most purposes, though perhaps not for parcel deliveries. On the other hand, small errors in the first or (n/2+1)th digit or large errors in the other digits will give you a reference that is hundreds to tens of thousands of metres wrong: easily enough to put you in the wrong place but not far enough out to be obviously wrong before you set off.
This is a* claimed, and - AFAICS - so far not disproved, strength of W3W, that most small errors in reference give (contextually) wildly improbably locations and you are far more likely to know on receipt of the reference (and even if you are offline) that an error has been made.
This is not the case with OS refs, or Lat / Long, either, where errors in the 1st to 3rd decimal places of degrees have similar 'bad but not mad' results. If I were a helicopter pilot looking for, e.g., an injured walker, I'm pretty sure that fruitlessly looking a few km away from my target, mistakenly fairly certain that I were in the right place, would be vastly less preferable to being in a position to say "that reference is in Vietnam, so I think unless you can get us the correct reference, we'll fly round and use the old Mk I eyeball"
* other claims that 3 arbitrary words are easier to remember, or easier to communicate, may well be regarded with more scepticism.
Biting the hand that feeds IT
