[uf-discuss] haudio contributor

[Andy Mabbett]

In message <47A73A66.90508 at digitalbazaar.com>, Manu Sporny
<msporny at digitalbazaar.com> writes

>Andy Mabbett wrote:
>> In message <47A60E74.5080209 at digitalbazaar.com>, Manu Sporny
>> <msporny at digitalbazaar.com> writes
>>
>>> If only one contributor is listed, it is assumed that he/she/it is also
>>> the creator of the hAudio. If multiple contributors are listed, it is
>>> assumed that the first contributor is the creator, and all subsequent
>>> contributors played supporting roles in the creation of the hAudio.
>>
>> That fails as soon as we want to mark up something like:
>>
>>         Simon Rattle conducted the CBSO in a marvellous rendition of
>>         Beethoven's Fifth
>
>Yes, the use of 'contributor' falls apart completely when we have markup
>like that... which is uncommon.

For some value of "uncommon".

>You should have noted that markup such as that is an edge case. Look at
>the audio-info-examples and you will be lucky if you find 1 or 2
>instances of the markup you describe above.

Perhaps, but why assume that the limited number and types of examples
are representative of all references to audio recordings?

>My point is that it is easy to manufacture words that break hAudio - but
>much harder to find actual examples online that break it.

It's not necessary to "manufacture" examples - I've already provided
real evidence, some weeks ago.

>If you have issues with this approach, you can always use hAudio RDFa,
>which does make the distinction between "dc:creator" and
>"dc:contributor". If you wanted to be even more specific, just include
>the Music Ontology vocabulary and mark it up using that.

I could do, but I'd rather help to get hAudio right; and further
microformats in general.

>>> Thus, it can be said:
>>>
>>> Not all contributors are creators.
>>> Not all contributors are artists.
>>
>> That can certainly be said. However, it cannot be expressed in hAudio
>> without requiring the publishers of such examples to re-order their
>> content. It is a microformats "principle" to not do so.
>
>For the publishers that need to re-order their content to mark up
>hAudio, they are obviously stretching what hAudio uF can do
[...]

Again, not according to the example I gave a while ago.


>We don't have enough examples to split "contributor" into "label",
>"publisher", "creator", and "artist" - which is what the examples showed
>to be the most prominently displayed contributors across the 93+ sites
>that we analyzed for hAudio.

Then we need more examples and realising that such examples may never be
sufficiently exhaustive, should treat them accordingly.

To paraphrase an old maxim:

        the process should be for the guidance of wise men and the blind
        obedience of fools

>>>> It doesn't seem to be based on established practice, as from the
>>>> overview it looks like existing markup overwhelming uses 'artist'.
>>>> http://microformats.org/wiki/audio-info-brainstorming#artist
>>> If we used artist, we would not have been able to mark up publishers,
>>> composers, audio technicians, etc.
>>
>> If we used *only* 'artist', perhaps, but not if we used 'artist' *AND*
>> 'composer' + 'technician'.
>
>There aren't enough examples that list the composer or the technician to
>make the argument for adding those into hAudio.

If there are insufficient examples of "composer" being listed, that
itself is evidence that the examples are inadequate:

        <http://www.google.co.uk/search?q=beethoven+mp3>

        <http://www.google.co.uk/search?q=mozart+mp3>

        <http://www.google.co.uk/search?q=holland+dozier+mp3>

likewise for conductors:

        <http://www.google.co.uk/search?q=%22von+karajan%22+mp3>

        <http://www.google.co.uk/search?q=%22simon+rattle%22+mp3>

and, (though I grant that technicians generally get less attention):

        <http://www.google.co.uk/search?q=%22george+martin%22+mp3>


Consider also how to indicate a reference to a song performed by its
composer:

        <foo class="artist composer">Bob Dylan</foo>

>You're more than welcome
>to go back through the audio-info-examples and re-analyze all of the
>sites to prove your point, though.

Why would I, when they're clearly inadequate?

-- 
Andy Mabbett