Apologies - that is something that we should have addressed explicitly here. The reason it was missed is because the contrast problem actually arose as a result of the other issues we discussed above - the testing and communications problems. The coder who did the redesign actually did an enormous amount of research and testing for accessibility, and as far as it was possible coded the site to conform to current accessibility standards. As part of that, she implemented a background colour which met current guidelines for default contrasts; however, a lot of our staff had aesthetic issues with it and it displayed very differently in different browsers. We had some miscommunication about the reasoning behind the actual choice and as a result of the other feedback we changed the background to the one we deployed. So, it wasn't that there was a lack of research or initial testing, but the other problems in our process meant that some of that was lost along the way. As you're probably aware, there's no way to guarantee that contrast - or anything else - is accessible for everyone, since both people and browsers vary so much, but what we deployed caused problems for a disproportionate amount of people. We know that the whiter background also causes some accessibility problems for some people, so we may change again in future if we can find a colour that will work better.
We're pretty devastated about this since the fundamental concept behind the redesign was that accessibility should be the default, not the special case (this has always been at the heart of our design philosophy, although in practice we haven't always had the levels of knowledge to make it work). As far as we know, there aren't any other aspects of the redesign which made the site less accessible (except in cases where a bug stopped the design working as it should) - if you do encounter problems then please do let us know about them.
If we keep up our current levels of research and fix all the other things we discussed in this post, then this is an area that should improve immeasurably. However, we definitely should have made this more explicit - apologies. We also welcome any input on how we could build accessible design more firmly into our processes.
Comment on Accessibility, Design and Technology Meeting - 3 December 2011
LucyP Wed 28 Dec 2011 10:42PM UTC
Comment Actions