- From: Wayne E Dick <wayneedick@gmail.com>
- Date: Mon, 19 May 2014 09:09:45 -0700
- To: Bim Egan <bim@w3.org>, w3c-wai-eo@w3.org
- Message-ID: <537A2CC9.4040805@gmail.com>
I actually looked in HTML 4 and HTML 5 and the upper left cell is ambigous. Since the next section is irregular tables, a brief transition from simple tables to might simply explain the upper left corner problem: Something like: "You may wonder whether the <th> cell in the upper left corner is interpreted as a row header or a column header. It could be both. The answer is that the situation is ambiguous without using the "scope" attribute of the <th> element. The next section on irregular tables will cover how we resolve the upper left corner problem and other unusual configurations using the attributes of <th> cells. Wayne On 5/19/2014 1:48 AM, Bim Egan wrote: > Hi Wayne, > > No, there's no de facto rule to say what direction the header in the top > left corner covers. It is both directions equally. To my mind this means > that most tables should use the scope attribute to ensure that screen > readers are given the requisite information. However in practice this isn't > going to happen, and while I'm fairly sick of hearing "January December" > when moving into the last column of a calendar, in most cases the human > will either be able to determine or guess why this happens and not believe > that January is the header for the December cell. When headers are less > predictable though, it can be a real issue. > > Cheers, > > Bim > > > > -----Original Message----- > From: Wayne E Dick [mailto:wayneedick@gmail.com] > Sent: 17 May 2014 01:28 > To: w3c-wai-eo@w3.org > Subject: Question about ambiguous scope > > Is the scope of upper left corner of a table ever > unambiguous without a scope attribute? How does a program > determine the scope upper left corner in any simple table > without a scope attribute? Are there an HTML semantic > rules that says it is column or row header? > > Wow, I never thought about this until this morning. What > a nifty paradox. > > >
Received on Monday, 19 May 2014 16:10:20 UTC