HTML Unleashed. SGML and the HTML DTD: Other DTDs and Related Resources

ou are probably aware that, besides consecutive HTML versions (the latest being 4.0), there exists a number of HTML "flavors" deviating from the standard in the scope of supported features.  The most notorious of these flavors is "Netscape HTML", a vague term used to circumscribe the suite of HTML extensions ("netscapisms") introduced by different versions of Netscape Navigator browser and now making their way to other browsers and HTML flavors.

Unfortunately, Netscape HTML extensions aren't officially documented in the form of a DTD.  Other companies, as a rule, are more reliable in this respect; for example, the DTD for the version of HTML supported by Microsoft Internet Explorer can be found here.

There were also independent attempts to provide DTDs for various HTML flavors, including Netscape extensions.  The HTML Pro project attempts to combine in one gigantic DTD all HTML variants and extensions proposed by standard-setting organizations and browser manufacturers.

Certainly, the whole wizardry of DTD syntax would be pointless if there were no programs to automatically parse the DTD declarations.  The most sensible purpose of such parsing (as well as of formally defined syntax in general) is to check HTML documents against the DTD to ensure they are valid SGML documents using only declared elements and attributes.

The confusion of HTML flavors notwithstanding, it is always a good idea to make sure that your document is formally correct from the viewpoint of at least one of the DTDs out there, preferably the DTD of the current official HTML version adopted by a standard-setting body.  (To watch for latest developments in this area, visit the W3C page or the home page of IETF.)

Such a validation can also be helpful by ensuring that the document contains no syntax errors such as unclosed tags or delimiters.  Validation packages are sometimes combined with functions to check for broken links, estimate download time, examine images, check spelling, and so on.

Some of these validators are accessible over the Internet.  The WebTechs validation service is a pure SGML validator without any extras, but it offers a big collection of DTDs to choose from and can check not only HTTP- accessible documents but also HTML fragments entered interactively.  The site also offers a handy hypertext version for each of the DTDs it uses.