The contention that XML's core use is for semi-structured data is misleading and, as far as I can tell, an effort to revive XML usage through re-branding. But nobody is seriously suggesting the whole idea behind XML has failed, so I don't think there is any need to come to its defense in the first place. Afterall, the idea behind it gave birth to the data format behind _the entire web_ as we know it today. As the article mentions, there are lots of cool data formats out there, based on XML, like SVG and RSS or Atom.
Those examples do, however, have something in common which is very telling: they are simple formats that can be handled in a text editor and can be interpreted easily. (Sure SVG files tend to get huge, but they're not really complex.) These areas are not where XML has failed. It failed to get traction because of scenarios where big "tools" vendors made it their personal mission to deliver bloated datasets for bloated enterprise applications in an effort to sell unnecessary software, supported by a legion of consultants who made a lot of money trying (and often failing) to make this mutant juggernaut work as intended. Much like Enterprise Java, contact with these wasteful projects scarred reasonable developers for life who desperately wanted to get their stuff working with as little fuss as possible. The result after these experiences was that almost no productive developer wanted to touch this stuff if he had a choice.
(I say "productive developer" here to distinguish, somewhat simplistically, people who just want to get their stuff done in as little time as possible from corporate devs who have little stake in the success of their projects and who get paid monthly for sitting at their desks regardless of the eventual outcome, as long as they keep looking really busy when it's time for their performance review where the total number of lines coded is still considered a valid benchmark.)
So, no, I don't think XML is misunderstood. We appreciate it, and it still is valuable in areas where JSON will never be the right choice. At the same time, polemically declaring JSON as a simple key-value format one step above the flat file is frankly disingenuous. JSON can (and often does in practice) contain the same hierarchical data representation as XML.
Comments
The contention that XML's core use is for semi-structured data is misleading and, as far as I can tell, an effort to revive XML usage through re-branding. But nobody is seriously suggesting the whole idea behind XML has failed, so I don't think there is any need to come to its defense in the first place. Afterall, the idea behind it gave birth to the data format behind _the entire web_ as we know it today. As the article mentions, there are lots of cool data formats out there, based on XML, like SVG and RSS or Atom.
Those examples do, however, have something in common which is very telling: they are simple formats that can be handled in a text editor and can be interpreted easily. (Sure SVG files tend to get huge, but they're not really complex.) These areas are not where XML has failed. It failed to get traction because of scenarios where big "tools" vendors made it their personal mission to deliver bloated datasets for bloated enterprise applications in an effort to sell unnecessary software, supported by a legion of consultants who made a lot of money trying (and often failing) to make this mutant juggernaut work as intended. Much like Enterprise Java, contact with these wasteful projects scarred reasonable developers for life who desperately wanted to get their stuff working with as little fuss as possible. The result after these experiences was that almost no productive developer wanted to touch this stuff if he had a choice.
(I say "productive developer" here to distinguish, somewhat simplistically, people who just want to get their stuff done in as little time as possible from corporate devs who have little stake in the success of their projects and who get paid monthly for sitting at their desks regardless of the eventual outcome, as long as they keep looking really busy when it's time for their performance review where the total number of lines coded is still considered a valid benchmark.)
So, no, I don't think XML is misunderstood. We appreciate it, and it still is valuable in areas where JSON will never be the right choice. At the same time, polemically declaring JSON as a simple key-value format one step above the flat file is frankly disingenuous. JSON can (and often does in practice) contain the same hierarchical data representation as XML.