Out of the way

Microsoft recently announced some stuff about how they would handle accesskey and keyboard shortcuts in IE7 (so being opinionated in this fieldI commented - apparently before my good mate John Foliot).

What they choose to do is sometimes important. The fact that they aren't considering site-specific preferences except for "some future version, perhaps" only matters to their users - it is a convenience or accessibility functionality.

But their implementation of accesskey, which was also followed by Mozilla, causes problems for everyone. Because IE has such a large market share, things that cause sites to break in IE are not acceptable for many site designers. Making a page that causes the browser to behave in unpredicted ways and breaks the predicted functionality isn't helpful. So people are going to avoid using accesskeys that conflict with normal IE behaviour.

This is good of them, unless you are one of the people whose life would be made a lot easier by good shortcut navigation around websites. (On average people are not, like me their life will be unaffected or made just a little bit easier, but nobody who reads this blog is average, right? ). Being an international product with various localisations means that a lot of the keyboard gets used up. For many years now, authors have been trying to find the keys that cause the least conflict with bad browser implementations, instead of suggesting something that is actually memorable.

(Of course authors making better use of the rel attribute to support very highly predictable navigation would be nice, too. The fact that not all browsers have good support for it isn't much of a reason not to use it, since unlike accesskey it doesn't do funny things even to browsers with poor or no implementation).

So I would love to believe that Microsoft are going to improve their implementation in IE7, and Mozilla their next version, to something that reduces the damage to the accessibility of the Web. (Yes, I do mean something like what Opera does for accesskey). It should not be difficult for them to move to the approach of having a pass-through key instead of just changing expected browser behaviour (without any warning, at the moment).

This implies changing the way part of the user interface works for people with disabilities. In general this isn't considered a brilliant idea. On the other hand the people who want to use accesskey are not getting any support, because authors avoid using it. At least some of those people would be heavy keyboard users in general, so having their browser functions vanish on them will be a frustration too. It will cause a few authors a little pain. Those who have been helpful enough to have implemented access keys and then realised that they needed to do more work to explain what is happening in IE and Mozilla/Firefox will have to change their text that says

press alt+key to go to this link

to take account of the fact that there are different implementations out there. It seems a relatively small price (if somewhat unfair) for a relatively big improvement.