W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > October to December 1998

Re: spell-check, yes

From: Al Gilman <asgilman@access.digex.net>
Date: Wed, 11 Nov 1998 12:00:07 -0500 (EST)
Message-Id: <199811111700.MAA28140@access5.digex.net>
To: chisholm@trace.wisc.edu (Wendy A Chisholm)
Cc: w3c-wai-gl@w3.org
Sounds to me like a good place for it.

to follow up on what Wendy A Chisholm said:

> What if we add this as an item in the testing section (Appendix A)?
> Something along the lines of:
> Use a spell checker. A person reading a page with a speech synthesizer may
> not be able to decipher the synthesizer's best guess for a word with a
> spelling errror.  
> However, upon hearing the garbled word, a person could read the word letter
> by letter and hopefully decipher the intended word.  This increases
> cognitive load and time to read the page.  Spell checking is a good
> practice, so including it in Testing keeps it out of the guidelines, but
> points people in the direction we would like them to head.

The "however" paragraph is unnecessary.

Maybe this is outside what you meant as draft text.

> thoughts?
> --w
> At 03:52 AM 10/26/98 , you wrote:
> >OK. Priorety 3.
> >We should allow though to include misspelt words in 
> ><META name="keywords"> for our illetrate readers who look us up
> >in search engines.
> >
> >Best,
> >Nir.
> >
> >> And, by the way, I do believe we _should_ mention spell-checking.
> >> 
> >> This is another case of a general good practice, where failing to
> >> follow the practice has more severe impacts on the text-to-speech
> >> user than on the visual user.  Someone reading with their eyes
> >> can usually pick out the spelling error or will in fact read
> >> right through it without perceiving the error.  But someone
> >> depending on text to speech will get garbage for a word with a
> >> spelling error.  This heightened vulnerability to something that
> >> otherwise is a minor nuisance _should be mentioned_.
> >> 
> >> Al
> >> 
> >> PS: This is a "do as I say, not as I do."  One can find ample
> >> evidence of this effect among the perplexed readers of my
> >> unchecked email <wince>.
> > 
Received on Wednesday, 11 November 1998 12:00:11 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:28 UTC