5

When I first heard about the W3C-backed MicroXML, I thought of an idea I read somewhere years ago: a simpler subset of the XML specification without attributes (and without namespaces IIRC). The theory is that such a spec would allow simpler, faster and more memory-efficient parser implementations which can be used in situations where full XML is perceived as too heavy, and most of its features are not needed.

However, it turns out MicroXML is not that simple - while it is simpler, it does retain attributes, for example.

So does something like this exist? A spec? An implementation (preferably in Java)? The closest thing I've found is XML-- (XML minus minus) by Christopher Diggins, but that seems to have never taken off.

Note: I do realize that JSON or several other formats could be used in situations I describe. Out of curiousity, though, I am interested in an XML subset, which is readable by regular XML parsers and can be (even when parsed by the "reduced" special parser) accessed through DOM or SAX APIs.

Paul Sweatte
  • 24,148
  • 7
  • 127
  • 265
Jens Bannmann
  • 4,845
  • 5
  • 49
  • 76
  • It's not correct to describe MicroXML as "W3C-backed". W3C provides a venue for the group of individuals discussing the project, but that doesn't imply any kind of endorsement of the project. It's not a working group with an approved charter, for example. – Michael Kay Jul 17 '12 at 14:32
  • I think the problems with subsetting a spec are always the same: (a) for any feature, there's probably someone who thinks it's essential, and (b) what are the benefits to users of using the subset rather than the full thing? – Michael Kay Jul 17 '12 at 14:34
  • If you feel that the full spec is too heavy, with namespaces etc, simply don't use them. If what you are writing doesn't use them, you don't need to try reading them. – Woody Jul 23 '12 at 12:44

1 Answers1

2

I'm the one who proposed the ideas of XML-- In the end I have found that JSON satisfied exactly my needs and I did not pursue the XML specification. I preferred JSON because it could be rapidly parsed by Javascript implementations, was much less verbose than XML, and could be trivially converted to XML as needed.

cdiggins
  • 17,602
  • 7
  • 105
  • 102