- From: Henri Sivonen <hsivonen@iki.fi>
- Date: Mon, 4 Jan 2010 12:10:43 +0200
- To: Larry Masinter <masinter@adobe.com>
- Cc: Jirka Kosek <jirka@kosek.cz>, "Henry S. Thompson" <ht@inf.ed.ac.uk>, Toby Inkster <tai@g5n.co.uk>, "David Carlisle" <davidc@nag.co.uk>, "julian.reschke@gmx.de" <julian.reschke@gmx.de>, "public-html@w3.org" <public-html@w3.org>
On Jan 2, 2010, at 23:49, Larry Masinter wrote: > I understand the XSLT use case. I am asking very specifically > about the XML-editor-being-used-for-XHTML-then-served-as-html > use case, and gave seven URLs for documentation about (different?) > XML editors which allude to using the DOCTYPE declaration to > control the XML editor's behavior. > >> Welcome in a real world ;-) > > My "real world" includes > XML-editor-being-used-for-XHTML-then-served-as-html. Using a generic XML editor for editing content that gets served as text/html is unsafe regardless of doctype issues. An editor needs a non-generic text/html-aware serializer anyway to deal with void vs. non-void elements and <pre> reasonably. When a fully generic XML editor is used to edit text/html content, the user is already in the process of operating a foot gun. Therefore, I think the premise of the use case is faulty and the spec doesn't need to support the use case. (Once the editor is non-generic enough to know *something* about text/html considerations, it might as well allow the use of <!DOCTYPE html> without ill effects. OTOH, an editor that doesn't allow <!DOCTYPE html> isn't a generic XML editor to begin with, because <!DOCTYPE html> is permitted in XML regardless of entity resolver configuration.) -- Henri Sivonen hsivonen@iki.fi http://hsivonen.iki.fi/
Received on Monday, 4 January 2010 10:11:25 UTC