The ability to extend a data format while retaining backwards compatibility is important, and was no doubt one of the objectives of the design of XML.
But I don't think that's primarily what the "extensibility" property is about. The "X" in "XML" is there primarily to contrast with HTML, and it refers to the fact that XML is a syntax without a vocabulary: you roll your own vocabulary rather than using one that comes out-of-the-box with the markup syntax.
Of course that property isn't unique to XML either; it was there in SGML and in ASN.1, for example. For that matter it's there in CSV.
XML doesn't have one unique selling point where it offers something that no other format offers. The reason it was successful was (a) that it offered enough functionality to meet people's needs, (b) combined with enough simplicity to make it inexpensive to implement, (c) at a time when the need for such a format existed and was recognized, and (d) when, for reasons I have never fully understood, there happened to be no serious competitor around.
And no, I don't find it remotely disheartening that there are other formats that can do some of the same things that XML can. If the ideas are technically sound, then in a healthy competitive world, the same ideas will crop up elsewhere.
Michael Kay Saxonica
Hi Folks, Thank you (again) for your outstanding responses. I found your responses surprising. In a nutshell, you told me that many, if not most, data formats – both text and binary – are extensible. XML, it seems to me, has always made it a point to advertise that one of its virtues is that it is an extensible data format. After all, it’s built right into its name: Extensible Markup Language. But now you tell me that, in fact, extensibility
is part of many data formats. XML is not unique in that regard. Isn’t that disheartening? /Roger
|
