Re: [xml-dev] XML vs JSON

["Ghislain Fourny" <gfourny@inf.ethz.ch>]

Hi Mike,

Interestingly, the XBRL community had the exact opposite epiphany: when they realized that one can tag a human-readable text (aka a fiscal report, in HTML) with context here and there (such as: this number is our company's revenue of last year in US dollars, or, this paragraph is a policy that applies in this period), they went for it. This is now known as inline XBRL (iXBRL).

It shows how both aspects (data vs. document/narrative) are important and have use cases.

Kind regards,
Ghislain


> On 21 Aug 2017, at 10:44, Michael Kay <mike@saxonica.com> wrote:
> 
>> 
>> A document-oriented XML document looks more like an HTML page with mixed content, XHTML and docbook being two prominent examples:
>> 
>> <paragraph>This is <bold>bold</bold> text and this is <italic>italic</italic> text</paragraph>
>> 
> 
> 
> I like to say that the defining characteristic of pure document-oriented XML (I sometimes call it "narrative XML") is that if you remove the markup, you are left with meaningful human-readable text.
> 
> And I always felt that the reason XML became so popular was not its ability to handle pure narrative XML, but its ability to mix structured data and textual narrative in a single message. So often (consider a CV/resume) you want to handle both at the same time. XML thus brought document processing capability to the data processing masses; and by reverting to JSON, the data processing masses are saying "that's too difficult for us to cope with".
> 
> Michael Kay
> Saxonica