- From: 河内 隆仁 <kochi@google.com>
- Date: Mon, 30 Mar 2015 19:36:48 +0900
- To: "Tab Atkins Jr." <jackalmage@gmail.com>
- Cc: www-style list <www-style@w3.org>
Received on Monday, 30 March 2015 10:37:35 UTC
Thanks Tab, for all the comments. IIUC the 'tab group' thing is a recurring discussion and it may be high time for considering seriously. I must admit that the interaction of tabindex and tabstop is very confusing, even for me, and already found some defects that it cannot explain an <a> behavior (with custom elements + shadow dom) completely, so I would like to explore fleshing out 'tab group' and 'isTabGroup' idea. At the same time, I've almost finished the experimental tabstop implementation for Blink (behind experimental web platforms flag). https://codereview.chromium.org/1046853002/ is the last patch (under review) for attribute <-> property reflection. The basic rule is, unless HTML "tabstop" attribute is there (if tabStop property is set via script, it is reflected in HTML attribute, so there is), tabStop reflects what tabindex for the element means. I'll see how this can or cannot fit for e.g. creating a custom <input-date> etc.
Received on Monday, 30 March 2015 10:37:35 UTC