W3C home > Mailing lists > Public > w3c-wai-ig@w3.org > April to June 1998

Re: CAST Announces Bobby upgrade & Beta release of Bobby Application

From: Bruce Bailey <bbailey@clark.net>
Date: Fri, 24 Apr 1998 15:45:32 -0400
Message-ID: <01BD6F98.105A2FE0@bbailey.clark.net>
To: "'w3c-wai-ig@w3.org'" <w3c-wai-ig@w3.org>
I have been lurking for a while so I know the debate on the 
merits/limitations of Bobby.

Let us agree that:
A)	Bobby is great (or at least very good) and it is fantastic that it 
exists.
B)	Bobby has it's share of quirks.

Let me pose this question to the group:
Do we, who are in the accessibilty field, dare snub a "Bobby Approved" 
rating because we KNOW our finely crafted pages to be screen reader 
friendly, just because Bobby says otherwise?

I, for one, am commited to the "Bobby Approved" rating and will change my 
work to accomodate the limitations of the program, even when Bobby 
complains about nonsense.

That said, let me publicly vent my two most severe Bobby complaints. 
 (Sorry David Clark, you took too long to answer in private.)

1)	Bobby should reject HREF'ed IMG's with ALT="" unless the link is 
available elsewhere on the same page.  Take a look at url: 
 http://www.dors.state.md.us/test.html
This short page has four invisible buttons, yet Bobby complains not at all!

2)	Bobby discriminates between full-blown text-heavy tables and simple 
tables that are used to align graphics with text.  Bobby should also make 
allowences for tables with only two rows or only two columns and just a few 
words per cell.  Such tables usually manageable with a screen reader 
regardless if they are read line by line, column by column, or row by row.

Before Bobby 2.0, I had taken great pains to code a name/telephone table 
that rendered well in Lynx or a GUI and tested it with a couple different 
screen readers.  The failure of this simple table to pass muster was my 
primary motivation for breaking Bobby with a simple example.   Granted, 
there is no way Bobby could test for _good_ (or even approriate) alt text 
(e.g., a photo.jpeg with alt="yes").  Deliberately empty alt text that 
hides a link deserves special scrutinization.

Bruce Bailey
Maryland Division of Rehabilitation Services
http://www.dors.state.md.us
Received on Friday, 24 April 1998 15:38:45 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Tuesday, 19 July 2011 18:13:39 GMT