User talk:RexxS/List of symptoms of diving disorders

Introduction
Imagine that this is just one section of a list article compiling the signs and symptoms of various diving disorders – decompression sickness, oxygen toxicity, nitrogen narcosis, and high pressure nervous syndrome are the obvious ones.

The table in the first version has no scopes, and the table in the second version has scopes. Both of them have row headers. They are at present visually identical (unless the stylesheet is modified) – this is to show that "scope=" has no visual effect in itself. The issue is always going to be that the first column contains row headers.

Questions: If we can experiment here, perhaps we can get a clearer idea of what common ground can be found. --RexxS (talk) 23:25, 9 November 2010 (UTC)
 * 1) What faults would a developer find with each of the two versions?
 * 2) What faults would a reviewer at FLC find with each of the two versions?
 * 3) What faults would a blind viewer find with each of the two versions?
 * Sorry, I didn't have time to comment at the time. Do you still consider the case open, or do you consider the current situation with "plainrowheaders" to be satisfactory? Yours, Dodoïste (talk) 15:49, 21 November 2010 (UTC)
 * Yes, I'm afraid it's still a live issue. At least this page shows that adding 'scope' alone does not affect the visual display. I guess I'll have to take a test case to FLC. Cheers, --RexxS (talk) 22:43, 21 November 2010 (UTC)

Comments

 * Comment from Jack Merridew : First-off, obviously, the first table is missing the scope attributes, which ill-serves anyone using a screen-reader. This omission also impedes proper access to the structure of the information in the table by software agents, such as Googlebot, and any piece of software attempting to parse the table. Ideally, the scope attributes would be generated per a few rules; should be feasible for a large number of typical tables. This would be at the MediaWiki-level.
 * Both versions are currently centering the row-headers, and I think this a mere artifact of a mindset that was focused on styling the column-headers, and I have no issue with site css that tweaks this to the default align-left for text. The bolding of these row-headers is part of the long-standing site styling, as is the background-color; together they are the key indicator that the cell *is* a row-header, which I feel is an important piece of information to communicate to all readers (screen reader users will have this information told to them from the '!' and scope). The left-align tweak does have the advantage that it communicates the row-vs-col header status to the sighted, too, by the difference in text alignment. Given that typical row-headers will often be longish and at odds with some-other styling conventions, I'm agreeable to their being styled as non-bold. It was suggested by Edokter that an intermediate #f5f5f5 be used and I could warm to that as an additional distinction from the col-headers. I'd be concerned that some might simply favour it as closer to no-colour ;) Jack Merridew 00:42, 10 November 2010 (UTC)


 * Comment from xxx : yyyyy