> / Cory Snavely <[EMAIL PROTECTED]> was heard to say: > | Keep in mind that a significant portion of XML technology hasn't much to > | do with text processing but with databases and application development. > | In fact it's my impression that's where the bulk of the interest lies, > | which was (and is still) not the case with SGML. SGML was always been > | largely applied to text processing. > > I don't dispute that that's true, but I think it's wrong to > extrapolate that "XML is for data" or "SGML is for text". XML is as > much a marketing victory for SGML as anything else. > > Making SGML a tad simpler made writing parsers easier. Making validity > optional further reduced the "startup cost" of converting to SGML. So > suddenly this 10 year old technology got a new lease on life and > thousands of people discovered something that had been there all the > time. > > Markup is good. Proprietary is bad. Therefore XML.
Yeah a couple of posts seemed to indicate I was somehow implying that SGML was more suitable for text processing than XML. Maybe I wasn't clear. That's not my opinion, nor the intended implication. I was merely pointing out that XML has found broader areas of application, whereas SGML has traditionally been applied mostly to text processing. Right, simpler parsing, schemas as opposed to DTD--that's the lowering-the-bar-of-entry strategy of XML that we all acknowledge. Those are the factors which have made XML a pretty big splash in many fields and applicable for a lot of uses. c
