[uf-discuss] Re: Precise Expansion Patterns

[Manu Sporny]

Benjamin Hawkes-Lewis wrote:
> Here's another question that needs asking. How much real-world value
> does the use of the ISO standard for date time representations
> actually add in this /particular/ case (hAudio duration)?
> How often do these reasons apply to hAudio duration?

These questions are dangerously short-sighted, focusing on just hAudio
will hurt the Microformats community in the long run... I'll explain why
below.

> The impression I get is that the majority of cases would be served
> fine by:
>
> <span class="duration"><span class="minutes">2</span>:<span
> class="s">25</span></span>

This would be a giant pain to author, IMHO. Then again, it fixes our
accessibility/usability issues - placing much of the load on the
authors. If we make it this verbose to mark up this sort of information,
do we think authors are going to mark up DURATION?

For sites like ours that are automated, we would have no problem
adopting the approach listed above... but think about the blog author
that has to mark up DURATION. Are they going to take the time, or skip it?

What happens when somebody wants to specify a time duration as
"twenty-three minutes"? Or a measurement as "two stones"? Your proposal
above doesn't solve those problems.

Andy Mabbett wrote:
> On Sun, December 16, 2007 19:14, Manu Sporny wrote:
>> Paul Wilkins wrote:
>>
>>> Another possible solution is to provide greater detail for the time
>>> itself.
>>>
>>> <abbr title="00:23:00">2:23</abbr>
> 
>> If you are suggesting that we use the hh:mm:ss time format for
>> expressing duration, we cannot. That would be an abuse of the ISO 8601
>> standard.
> 
> We can, becasue we are not mandated to use the ISO 8601 standard.

So assume that we do that today...

We're locking in DURATION to have a very specific meaning "a length of
time". To denote that length of time, you have hours, minutes and
seconds HH:MM:SS.

Some time further down the line, somebody has another Microformat that
needs to specify a time duration. Their time durations, however, are in
years.

The problem comes in when the second person wants to denote their time
duration in years. We've already said that DURATION is "a length of
time" and specified a format HH:MM:SS. So now, authors have to translate
years into hours... quite a pain, but it gets worse.

Later yet, in a future Microformat far, far away, somebody comes along
and wants to specify time in fractions of a second.

Once again, they can't use DURATION because there is no space for
fractions of a second (which you can specify in ISO8601).

Remember, this same thing happened with the TITLE tag in Microformats.
TITLE is used to specify "a job title" in hCard. This meant that we
couldn't use TITLE in hAudio because it meant "a job title", not "the
title of an object, such as, but not limited to, a book, movie, album,
or person".

By constricting DURATION to have a restrictive format, HH:MM:SS, we are
being short-sighted and are not thinking about the other Microformats
that are still to come that will need to specify DURATION.

"00:02:23" is being shortsighted. Let's learn from our past and not make
the same mistake again... let's not be short-sighted about this decision.

-- manu