[uf-discuss] 'currency' microformat straw-man proposal.

[Gazza]

Andy Mabbett mumbled the following on 20/09/2006 23:59:
> In message <4511C248.1000708 at garyjones.co.uk>, Gazza
> <gazza at garyjones.co.uk> writes

>> Usually, when talking about currency, the word 'type' is used (see
>> xe.com)
> 
> It may be used sometimes,; often, even, but is it "usual"? I and people
> I know are far more likely to ask "What currency is use in Albania?"
> than "what type of currency is used in Albania?".

In which case, the pertinent answer for your first question may be 
"money, metal and goats".

>> <span class="currency">
>>  <span class="type">$</span>
>>  <span class="value">5.00</span>
>> </span>
> 
> "Dollars" is a currency. "Five Dollars" is money.

No, money is a currency, metal is another type of currency. Dollars is a 
  /type/ of currency.

>> (which follows the value excerpting model of using type and value
>> classes), or, better:
> 
> Note sure what you mean here.

<http://microformats.org/wiki/hcard#Value_excerpting>

> [...]
>> even ISO4217 has codes for "currencies" that don't use symbols:
>>
>> <span class="currency">
>>  <span class="value">23</span> ounces of
>>  <abbr class="type" title="XAG">Gold</abbr>
>> </span>
> 
> "once" would be the "unit", in that case.

Nope - the ISO4217 specifies that XAG is Gold measured in ounces - 
there's no need to replicate a "unit" of ounces anywhere.
Similarly XAU relates to ounces of silver.

>> Following on from this, the use of a "money" class should not be used;
>> currency does not _have_ to be money
> 
> That's an interesting point.

I agree - once it can be agreed that there is a subtle but important 
difference between the semantic definition of "currency" and "money" 
(even if most people interchange them), then the structure starts to 
form itself.

One fleeting glimpse I had, though I wouldn't be keen on (but bears 
mentioning for completeness of discussion), might be a subclass of money 
/ a method - this could be paper, coins, cheque, etc. Not sure how this 
could be implemented, or even if there's (m)any real-life examples to 
warrant needing this level of detail.

>> If there was one further issue, perhaps an "amount" class could be used
>> instead of type & value:
>>
>> <span class="currency">
>>  <span class="amount">£14 6s 4d</span>
>> </span>
>>
>> <span class="currency">
>>  <span class="amount">50 pence</span>
>> </span>
> 
> Those won't allow a user agent to extract the numeric value, easily.

Agreed, but knew someone might pick up on those examples :o) I suppose 
you could have your "unit" class here:

<span class="currency">
   <span class="amount">
     <abbr class="unit" title="GBP">£</abbr>14
     6<abbr class="unit" title="shillings">s</abbr>
     4<abbr class="unit" title="old pence">d</abbr>
   </span>
</span>

^^^ There's all sorts of problems with that example, but you get the 
general idea.

The fair proportion of anyone using this uf, would be to markup current 
(iso4217) currencies, so these odd examples might be considered trivial.
Time for that current real-life examples page!
-- 
Regards,
Gazza