Hacker News Reader

We use JSON for graphs. What sucks about it? In what sense is XML better at representing graphs?

There's no way to point from one part of a JSON doc to another without inventing a terminology or convention for marking the start (anchor) and end of the arc (href). People use 'id' for one end but there's no way to say a json value is actually a reference (href) not just a string. XML has that built in (ID IDREF) and so does HTML, but I didn't say XML was better, I said JSON sucks at markup and graphs. JSON's handy for serializing trees of data with no loops.

What do you mean by "built in"? I don't see how or why an ID couldn't be used in the same manner; the implementation is just a little different because they store data differently.

XML is only a series of nodes and attributes. There isn't really anything else special about it and it's trivial to represent it in JSON so I'm not sure I follow your issue. Could you provide an example?

Reply on news.ycombinator.com