Yes, this document has been making me laugh for 20 years!
Probably because I recognize myself a little too much in it!
Well, more or less to a pinch of iotta, or to a cow's tail!
Even if it won't help our friend @gaithbenhamouda much, I couldn't miss the opportunity to pull this pifometric standard out of my hat!
The improved French version here:
https://abcelectronique.net/forum/attachment.php?attachmentid=84368
Chances are it's not that far away.
Hello
Having worked with an American company, the ANSI and ISO ratings are quite different.
Quite often in the USA, the number of digits gives the applicable tolerance.
So if there is 3.0" on the plan you will have a wide tolerance but if it is 3.000" it will be a tight tolerance.
In my opinion, the excerpt of the map you shared is incomplete: there must be missing information to the left or right of this table.
Like this:
If the table at the bottom of the plan is not more explicit than what is displayed, I would opt for:
3.000 + 5 → 3" +0 / +0.005"
0.375 + 3 → 0.375" +0 / +0.003"
If A is given with 3 digits after the decimal point +/-0.002 if it is given with 2 digits: +/- 0.02
It immediately makes more sense.
I wasn't that far away.