About then proposal "3*centi(m)", "km._2" or "km.sqrt" I find it confusing too.

I'm reading your old posts to better understand the reason for moving from uunit.pas to dim.pas.

I have found two main reasons:

1) There is always a multiplication by a factor, even when it is not necessary (5*m, 10*s ...)

If this is true only when assign a value the first time then I think that it isn't a important problem.

How many people use m and how many people use others (mm cm, dm, foot, inches...) ?

2) It's needed to define km_h ...

Yes, this is a little cons. km_h isn't intuitive to use.

There are other reasons ?

For my application, an usable dim.pas unit must have capability to manage and combine:

mm, mm2, mm3, mm4

Nm, N/m, N/mm2 (MPa)

kg/dm3

Hz

C°

plus some extra units for calculation (1/mm, 1/mm2, 1/mm3, mm2/s2).

This is minimum requirement, but I want to extend it to english measurement system.

Consider that I have modified some formulas because, for example, in some calculations

the mm4 are exceeded... or there were new units not supported.

The situations can be very different, it' depend on branch, engineering, physics.

I hope that this description can help you to understand which is the better way.

At the moment the game is 1-0 for uunit.pas.

Regards