From HTML WG Wiki
Tracking research tasks
This page is intended as a central repository to track research needs for the HTML WG. (DanConnolly, a co-chair of the WG, is wary of attempts to centralize like this; I much prefer to see three examples before instituting a pattern, let alone a list. See ListMode in the original wiki.) Research needs could include any possible methodology. For example, just to name a few, the WG might require:
- library research into a particular relevant topic
- research into behavior of existing UA implementations (including specialized UAs)
- user-testing of experimental implementations or UI approaches
- bot research into the current usage of HTML standards and custom extensions in internet or intranet deployed conten
Again, these are just some broad categories of research the WG might require. Some research approaches may not cleanly fit into any one of these categories. Any research task may be added to this page.
Members may also want to visit the DefinitionCompatibility: Compatibility definition page which discusses some of the parsing behavior of major browsers.
Adding a priority to the task. A priority of "high" indicates the research supports a WG supported feature that could have backwards compatibility or degrades gracefully issues with existing content or existing UAs. A priority of "low" instead indicates a task that constitutes exploratory research into possible changes to HTML. Low priority tasks are worth listing here since other members of the WG might be in a position to easily complete the task listed by other members or have access to research results already performed.
- ResearchNeeds/Item1: Aural-oriented and other UA treatment of LABEL and @for in form controls Prioirty: low
- Usage of longdesc (at the WHATWG wiki) 
HTML5 online tools
- HTML5Lib based parse tree Enter a publicly available URL to see the parse tree.
- Another parse tree viewer using the Validor.nu HTML parser (in Java) Enter a publicly available URL to see the parse tree.
- HTML5Lib Versatile tool for parsing HTML serialization according to HTML5 (draft) parsing rules (i.e., parses well-formed and ill-formed HTML).
- Live DOM Viewer (text/html) Paste, or compose HTML and watch how your current browser parses the source into a DOM tree.
- Live DOM Viewer (XML) Same as prior, but for XML parsing (currently broken, please help)