RNIB redesign does not go far enough
Well, I have to confess that I had high hopes for the new design for the RNIB’s web site. The re-working is not just limited to the structure/design of the site, but also the content of the site itself:
“If you’ve noticed that this site looks different to the last time you visited, that’s because it’s a completely new website. The site has been completely redesigned to make it more accessible, easier to navigate and visually attractive, but the really big difference is that all of the information has been updated and rewritten for the web.”
However, my initial reaction upon seeing the site is that they missed an opportunity. The site looks OK, but is not as graphically pleasing as it could be. You might argue that the audience doesn’t require that - and I understand this completely - but it’s not just blind and visually impaired users who may visit the site. This redesign gave an opportunity to show that an accessible site - and one that itself indirectly promotes the ethic of web accessibility - can be good-looking (but not necessarily in an overtly extravagant way).
Among the crimes that have been committed in this redesign are:
- A table based layout - this was the perfect opportunity to ditch a table for layout and make the site work in a PDA just as well as a fully CSS-able browser viewed on a large screen. Not only that, but the tables are quite heavily nested. Here’s an actual grab of some of the source code:
</td></tr>
</table>
</td>
</tr>
</table>
</td>
</tr>
</table>
</td>
</tr>
</table>
- Invalid or deprecated markup - a casual glance at the markup reveals that a number of attributes used are completely invalid for the doctype. Oops, did I say doctype? There wasn’t one specified. Offending tags and attributes include:
<font>tags- the use of
<b>tag instead of<strong>(find out what’s wrong with the<b>tag here) - bgcolor attribute - background colours specified in table cells (surely one of the most widely supported CSS1 properties - why not use it?
- Incorrect document structure - why could I find a
<h2>in the source code when there is no<h1>preceding it? - Poor use of margins - how can it be that text butting right up against the left margin is good for visually impaired users? Margins are easy to control using CSS. The way that the text sits on the page is just poor design generally, if you ask me, let alone for people who may have difficulty viewing verticals that appear to be merging with the vertical divider immediately to the left:
There may be other faults with the site, but this really is little more than a cursory glance, really.
Some people may disagree with me on this. Blind people may not really give a hoot about some of the points above given that:
- they can jump around the page using hidden skip navigation links (or signposts, as the RNIB calls them) regardless of the look of the site
- it works fine in their screen reader
- the information presented is clear and easy to understand.
However, I feel that the site could have gone further. It should have been possible to make this site smaller in filesize (per page), more flexible in terms of device-independence, comformant to established standards and still serve its purpose in providing information for blind/visually impaired people (or their carers/relatives) in an attractive way. It’s not the radical relaunch I was expecting. Screenshot taken 25 June 2003
Tim Roberts comments: Have to agree with you on the RNIB site. Definite anti-climax.
Tips for screenreaders and many other links returned this:
Content Server Request Failed
Unable to retrieve file. The dynamic application file is not at the
specified location
(’d:/xpedio/weblayout/groups/public/documents/publicwebsite/xpedio/group
s/public/documents/publicwebsite/public_screentips.hcsp’).
You would have thought that the RNIB would have also gone for a CSS
XHTML layout considering the inherent accessibility of such techniques.
They also seem to be using a CMS, but rather than just not serving up
content that is not relevant (eg, for users that are logged in) they
comment the code out, adding useless bandwidth to the document.
Step backwards I think. God knows what they are thinking.