AttributeUi: Rendering Options

[rpbouman]

Rendering options would allow the user to control how the item values are transformed to output. Concrete examples would be:

• urls. these could be used to render a hyperlink, or an image or some other media (video, audio)
• long text descriptions that might use markdown
• text fields containing programming language code
• binary data, which could be used to render an image or some other media (video, audio) The current rendering would be "text output" (i.e. the value itself is output as text that describes the value directly

It is distinct from formatting, as formatting refers to how the item values themselves are presented in text output.

Concreate example: github-events dataset. It would be nice to be able to produce links for the actor url and an image for the actor avatar url

[rpbouman]

One challenge with this proposal is that it is not clear whether some supported rendering method actually makes sense for the particular dataset.

Even for the simple example of a url:

• we don't know in advance whether some text attribute will hold urls, this would have to be profiled
• if we know it's a url, we can do little more than produce a hyperlink. We'd actually need to access the url to figure out what content types it can deliver