[rfc-i] Normalized HTML

Iljitsch van Beijnum iljitsch at muada.com
Wed Mar 28 01:52:32 PDT 2012

On 28 Mar 2012, at 10:46 , Peter Saint-Andre wrote:

>> <p>
>> bla blah blah
>> we don't close paragraph tags!

> Please no. Life is hard enough without well-formed XHTML.

So do we impose rules on authors to make life easier for the tools or the other way around?

But I guess the <p> issue isn't important, if they need to be closed that's doable. It's the nested <t>s in the XML that are much more difficult, but these are no longer necessary if we use <h1> / <h2> type headings rather than derive the nesting from the <t>s.

Note that the point here is not to make perfect HTML, but to encode meta data in something that is still mostly plain text, and getting that reflowable version in a browser as a bonus. So for instance we can be more strict than HTML and require that all tags occupy an entire line and start at the first character, so they can be filtered out extremely easily. Or relax HTML if that provides a compelling benefit. (But only then.)

More information about the rfc-interest mailing list