Can we add metric prefix to dur function?

[Dennis]

On Monday, 21 June 2021 at 10:34:51 UTC, Andrea Fontana wrote:
> Can't we add the metric prefix as alternative?

Please no. I agree that the current list is not the most 
consistent or easiest to remember, but the current error message 
is actually really informative:

```
Error: template instance core.time.dur!"millis" does not match 
template declaration dur(string units)(long length)
   with units = "millis"
   whose parameters have the following constraints:
   ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   > units == "weeks"
     or:
   > units == "days"
     or:
   > units == "hours"
     or:
   > units == "minutes"
     or:
   > units == "seconds"
     or:
   > units == "msecs"
     or:
   > units == "usecs"
     or:
   > units == "hnsecs"
     or:
   > units == "nsecs"
   ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
```

Adding a bunch of synonyms makes things more complex. Whenever 
there are multiple alternatives for something, it spawns 
questions like "are there differences" and "which is the 
preferred one" which you simply don't have with one name per unit.

> 20.dur!"µs"  <-- (altgr+m on my linux keyboard)

Next thing someone wants "us" to be consistent with "usecs". Also 
let's add "secs" and "useconds" and "µseconds". Once there's 
precedent for synonyms, it can easily get out of hand.

> We should add also non-SI abbrevetion for other common accepted 
> units like:
>
> 1.dur!"m"
> 1.dur!"h"
> 1.dur!"w"

m = months and w = weeks? I'm familiar with m = meter or milli, 
but m = month is not obvious to me.