Thread Subject: Re: Definition Consensus Decision: Text

Note

This archival content is maintained by WebAIM and NCDAE on behalf of TEITAC and the U.S. Access Board . Additional details on the updates to section 508 and section 255 can be found at the Access Board web site.

From: Gregg Vanderheiden
Date: Thu, Mar 13 2008 11:40 AM

Hi Allen,

I made the same observation a long time ago in WCAG - that when we say that
all information should be in text we really mean etext or some other word.
But it gets very complicated and awkward to say etext file, non-etext
content etc. You can substitute anything you want for etext and you get
the same problem (or worse if it is two words long). So we went with text
and images of text in WCAG.


In a standard many of the words used are narrower than public use. So you
define them in the standard so people know what they mean if you don't mean
the general meaning. But this one is hard to do right - and needs to be
because having text be readable by AT is central to so many of the
provisions in TEITAC. If the text can't be seen or read by the AT then all
the provisions requiring that information be in text have no meaning. So I
think following WCAG lead as was suggested is the way to do this.


Gregg
-- ------------------------------
Gregg C Vanderheiden Ph.D.



> -----Original Message-----
> From: = EMAIL ADDRESS REMOVED =
> [mailto: = EMAIL ADDRESS REMOVED = ] On Behalf
> Of Hoffman, Allen
> Sent: Thursday, March 13, 2008 7:02 AM
> To: TEITAC Committee
> Subject: Re: [teitac-committee] Definition Consensus Decision: Text
>
> Andi wrote:
> Text is a binary encoded representation of a human readable
> sequence of characters and the words they form. It can be
> determined exactly from the pattern of 1s and 0s in the
> electronic storage media, assuming you know the encoding
> format, and does not depend on any recognition algortithms to
> extract it.
>
> This isn't "text", it is electronic text, or electronically
> encoded text. Should "electronically encoded or stored text"
> be the actual definition then? Or, do we just state for
> purposes of these recommendations "text" is...?
>
>
>
>
>
> Allen Hoffman -- = EMAIL ADDRESS REMOVED = ; v: 202-447-0303
>
> -----Original Message-----
> From: = EMAIL ADDRESS REMOVED =
> [mailto: = EMAIL ADDRESS REMOVED = ] On Behalf
> Of Andi Snow-Weaver
> Sent: Wednesday, March 12, 2008 11:27 PM
> To: TEITAC Committee
> Subject: Re: [teitac-committee] Definition Consensus Decision: Text
>
> Gregg,
>
> Sorry that I have not weighed in on this. I don't have an
> Internet connection during the day here at CSUN.
>
> With regard to this comment:
>
> "... TODAY a bitmap image of text does not qualify as a TEXT
> alternative".
> This will likely be true for the life of this revision of the
> standard - but I do expect that someday a clear image of text
> would qualify as 'programmatically determinable' but a
> highly stylized bitmap image of text would not (think CAPTCHA
> or the Illustrated first letter of a chapter in a
> formal religious book or manuscript"). The trick at that point will
> be
> having a bright line between what AT can handle or not. Til then -
> 'programmatically determinable text' is likely to not be
> satisfied by 'images of text'."
>
> I don't think a bitmap of text will ever be a "text
> alternative". When OCR advances to the state that "text
> alternatives" are no longer needed for clear images of text,
> then we will change the provisions in the standard that
> require text alternatives but we don't change the meaning of
> what text is.
>
> Text is a binary encoded representation of a human readable
> sequence of characters and the words they form. It can be
> determined exactly from the pattern of 1s and 0s in the
> electronic storage media, assuming you know the encoding
> format, and does not depend on any recognition algortithms to
> extract it.
>
> Most IT developers would interpret that as being
> "programmatically determinable" but I share Peter's concern
> about trying to consense on a definition of text that uses
> "programmatically determinable" is impossible when we have
> such fundamental issues with the definition of
> "programmatically determinable".
>
> I do want to find a way to move forward on this and agree
> with what you and Peter have come to regarding having two
> versions of both "text" and "programmatically determinable"
> as long as we are clear that a particular version of each is
> tied to a particular version of the other.
> We don't want the AB picking one version of "text" and the
> wrong version of "programmatically determinable".
>
> Andi
>

WebAIM is an initiative of:
Center for Persons with Disabilities (CPD) Utah State University