This ain’t your older brother’s web any more
The result of this is that whereas it used to be possible to find the totality of a web site or web application by going to the home page, fetching the HTML content, pulling out the HREF links and then following those until you find no more links that you have not yet seen, that is no longer enough.
The death of the href and the dumb spider
For first generation accessibility tools, this has the following implications:
- Spidering, by following URLs gives you only a part of the picture – because many applications redirect to URLs based on event handler logic, many of the URLs that make up a site or an application are not discoverable by looking at the HTML of the page. A spider therefore gets a partial picture of an application’s URLs when it does this.
- Looking at URLs for testing uniqueness is not enough, the content needs to be evaluated to determine whether it is unique or not.
- Development of a single “web page” application can take months of work by a team of developers. The code that is used by the site might be hundreds of thousands of lines of code. Doing an evaluation of this “page” in the quality assurance phase after it has been implemented is too late to impact the first release. Also web accessibility tools need to support the workflow of multiple developers and testers working on a single page or application.
These implications mean that a second generation of web accessibility tools is needed to help us to develop and evaluate the accessibility of our web sites and web applications. These tools need to do everything that the old automated tools were able to do and more.
- They need to be able to track and evaluate dynamic page updates
- They need to do this for developers while they are developing, for quality assurance engineers while they are doing functional tests and for web site managers while their application is actively deployed.
- They need to support the work of the individual developer or tester as well as the collaborative work of a development team.