[uf-discuss] First version of Currency proposal

[Scott Reynen]

On Oct 11, 2006, at 2:43 PM, Guillaume Lebleu wrote:

>> Can we just treat everything as the default unit
> Not sure what you mean in the context of the 99c example. The  
> default unit for the US currency is the Dollar, not the Cent, but  
> in the context of the 99c context, we need to ensure that "Dollar"  
> is not picked as the unit, but "Cent" instead is.

What I'm suggesting is that everything be treated as dollars in USD  
and everything be treated as Yen in JPY.  Isn't that how most  
applications and people deal with money anyway?  If you want to  
consider your money as cents in your own publishing or your own  
application, you can make the simple conversion (multiply or divide  
by 100) necessary to use the microformat's assumed unit of dollar.   
Is there any currency with multiple units that aren't simple and  
constant fractions of a standard unit?

>> and adjust the machine-readable values accordingly?  E.g.:
>>
>> <span class="money"><abbr class="amount" title="0.99">99</ 
>> abbr><abbr class="currency" title="USD">¢</abbr></span>
>>
> So using this notation would require the parser to interpret the ¢  
> as the cent unit of the USD currency without this information being  
> disambiguated using a microformat.

It seems like you're making this more complicated than it needs to  
be.  The amount is 0.99.  That's an amount in the dollar unit, which  
would be assumed for all USD currency.  There is no cent unit here,  
and no requirement to interpret a cent unit.  When you say "99¢," do  
you mean anything that isn't a direct equivalent of $0.99?  Why  
should a parser care what symbols you use to convey that meaning?  If  
everyone uses the same unit, communication will be much easier, and  
I'm not seeing any reason everyone can't do easily use the same unit  
for machines.  It seems most publishers (well over 80%) already use  
the same unit for people, and the edge cases can convert for machines  
using <abbr> just like we all convert to a standard date format for  
machines.

> It would certainly work, assuming you can deal with the different  
> way of representing the "Cent" information

Doesn't the word "cent" literally mean one perCENT of a "dollar"?   
I'm not seeing the benefit of treating "cent" as a distinct unit.

> but isn't our goal here to make interpretation easier, and not  
> require to have to deal with all these tiny differences?

Yes, and I'd say the concept of a "cent" is a tiny difference that  
parsers shouldn't have to deal with when the concept of a "dollar"  
works just fine to communicate the exact same information.

Peace,
Scott