RE: Capturing the discussion (was Re: NY Property Tax Explorer)

All,

 

Reading some of this discussion around scoping this work, it sounds a bit like this group is now focusing primarily on Best Practices for Metadata on the Web. 

 

Laufer wrote:

 

For me, the central thing that we can point is about metadata. We can talk about data too, but I think that each one of the cases will be very particular and probably will be treated by a particular WG.

 

and

 

But I think that if we entered in all techniques of consuming/harvesting data and the way people should or must publish to facilitate these things, we will have to take the whole world in our hands.

 

I do agree that trying to come up with best practices for publication of all kinds of data formats and ‘genres’ – text, numbers, sound, image, video, maps, 3D models, online games – is probably not something we can do without involving the experts in those fields, maybe as Laufer writes, in separate WGs. However, is it now established that this group won’t talk about the primary data at all, but just limits its work to metadata?

 

But then again, if we ‘only’ consider metadata, there is the same can of worms waiting. There is often a close connection between the data format or genre and the metadata, both in terms of the metadata model and in how metadata is encoded and embedded, as in ODF, PDF, MP3, JPEG and many, many domain-specific formats. So, I am not sure that, by not talking about how people should publish data in specific formats on the Web but only about how people should publish metadata about data in specific formats, we make our life any easier.

 

What I do think is that if we abstract from data formats and genres, our best practices are going to be on the level of “provide metadata in machine-readable format”.  I am not saying this is not useful advice, but is it really what we would call ‘best practice’?

 

And then, we make statements like “mark as an error instances where vocabularies such as [DC-TERMS] and [VOCAB-DCAT] could have been used but were not” in http://www.w3.org/TR/dwbp/#ProvideMetadataStandardized. Do we really want to say that it is an error to provide metadata using schema.org, XMP, EXIF, CKAN etc. etc., irrespective of what the data is and how it is intended to be used? 

 

Makx.