Why should we have to use ten different tools to check the quality of a single web page? We think there should be a tool to gather observations made on a single document by various validators and quality checkers, and summarize all of that neatly for the user.
Do you think we are dreaming? Meet our unicorn.
As a previous entry on this site shows, making sure the whole of a site is using valid markup is not always an easy task, even when one cares very seriously about it. And just like a sentence in english can be grammatically correct but contain spelling mistakes, or just make no sense at all, validity is only a subset of conformance, which itself is only one checkpoint on the way to quality.
Take a single, simple XHTML 1.0 page. Verifying its quality is not limited to validating its markup. There are quite a few things in the XHTML specification that is not included in validation checking, such as the HTML compatibility guidelines (a.k.a “Appendix C”. Some things can’t even be checked by a machine, for instance, whether <address> or <blockquote> elements are used within their actual semantics.
And then, there is also the CSS stylesheets to check, verifying that there is no broken links, find spelling mistakes… And the list could go on forever: for each of these checks, a different tool, a different homepage to visit. And no way to have an overview of the quality checking of the web page.
Checking the quality of a Web page should be simpler.
With this in mind, we have started a new project, code named Unicorn. Based on some reflections by QA-dev participants Björn Höhrmann, Terje Bless and others, the project aims to define a document observation notation, and use it in a centralised framework using RESTful Web services to gather observation by diverse tools on a single document.
W3C interns Jean-Guilhem Rouel and Damien Leroy, working closely with other W3C staff and QA-dev participants, have released documented specifications for the formats and architecture of Unicorn. They are welcoming feedback on that document, preferably sent to the public mailing-list firstname.lastname@example.org. (Since that list is used for several projects, please mention that your feedback is about unicorn in the subject of your mail, e.g. « [Unicorn] suggestion for the contract schema » — thanks)
And coming very soon: a working implementation of the framework, combining several of W3C’s validators and checkers. We are testing a prototype internally, and will make it public in the days or weeks to come.
To be continued…
4 thoughts on “Meet the Unicorn”
That’s a nifty idea, I’ll be all ears in the weeks to come!
I hope Unicorn would come out soon. Many sites, especially sites designed in ancient ages, were designed page by page. Having a systematic tool to validate all pages would be an excellent idea for all webmasters.
We will contribute what we can in supporting this project.
A great idea- it’s always annoying having to validate a page’s HTML, CSS, make sure the links are OK and that it displays OK in every browser: at least Unicorn will remove the need to do some of those things!
A superb idea this should save webmasters alot of time over time, any plans on adding a mobile validator for it? If the new version is public can you also send me the link for it, or possibly let me try out the beta version.
Comments are closed.