- From: Andrew Fedoniouk <news@terrainformatica.com>
- Date: Thu, 27 Dec 2007 21:43:40 -0800
- To: "Patrick H. Lauke" <redux@splintered.co.uk>
- CC: public-html@w3.org
Patrick H. Lauke wrote: > > Andrew Fedoniouk wrote: > >> There is one case that we use in UI for <u> - to mark shortcut >> combinations: >> >> <button><u>O</u>pen</button> >> >> We do have special behavior defined for the <u> element that >> synthesizes button click event when user presses Ctrl-O >> (for the button above). >> >> So this brings some semantic meaning to the element >> but this is not what it was intended for of course. > > Yes, you're basically defining some visual presentation of text, and > then redefining how it behaves. Maybe splitting hairs, but it's not > semantic meaning as such...it's presentation that you then hijack with > scripting. > > A possibly more semantic way would be to define an @accesskey for the > button element, and then, since current UAs don't automatically do the > underlining, add that presentational aspect on top (incidentally, when > doing stuff like extensions for Firefox, this is automatically taken > care of - defining an @accesskey attribute on a XUL menuitem of > toolbarbutton results in the relevant letter being underlined in the UI). > For human this markup: <button><u>O</u>pen</button> plus associated behavior has clear semantic value and <button accesskey="^O">Open</button> is just nothing - user simply has no clue about that accesskey. Information that is unknown to particular person has semantic value of zero for that person, isn't it? -- Andrew Fedoniouk. http://terrainformatica.com
Received on Friday, 28 December 2007 05:43:54 UTC